Removed the Requirement to Install Python and NodeJS (Now Bundled with Borealis)

Dependencies/Python/include/Python.h

@@ -0,0 +1,139 @@
+// Entry point of the Python C API.
+// C extensions should only #include <Python.h>, and not include directly
+// the other Python header files included by <Python.h>.
+
+#ifndef Py_PYTHON_H
+#define Py_PYTHON_H
+
+// Since this is a "meta-include" file, "#ifdef __cplusplus / extern "C" {"
+// is not needed.
+
+
+// Include Python header files
+#include "patchlevel.h"
+#include "pyconfig.h"
+#include "pymacconfig.h"
+
+
+// Include standard header files
+#include <assert.h>               // assert()
+#include <inttypes.h>             // uintptr_t
+#include <limits.h>               // INT_MAX
+#include <math.h>                 // HUGE_VAL
+#include <stdarg.h>               // va_list
+#include <wchar.h>                // wchar_t
+#ifdef HAVE_SYS_TYPES_H
+#  include <sys/types.h>          // ssize_t
+#endif
+
+// <errno.h>, <stdio.h>, <stdlib.h> and <string.h> headers are no longer used
+// by Python, but kept for the backward compatibility of existing third party C
+// extensions. They are not included by limited C API version 3.11 and newer.
+//
+// The <ctype.h> and <unistd.h> headers are not included by limited C API
+// version 3.13 and newer.
+#if !defined(Py_LIMITED_API) || Py_LIMITED_API+0 < 0x030b0000
+#  include <errno.h>              // errno
+#  include <stdio.h>              // FILE*
+#  include <stdlib.h>             // getenv()
+#  include <string.h>             // memcpy()
+#endif
+#if !defined(Py_LIMITED_API) || Py_LIMITED_API+0 < 0x030d0000
+#  include <ctype.h>              // tolower()
+#  ifndef MS_WINDOWS
+#    include <unistd.h>           // close()
+#  endif
+#endif
+
+// gh-111506: The free-threaded build is not compatible with the limited API
+// or the stable ABI.
+#if defined(Py_LIMITED_API) && defined(Py_GIL_DISABLED)
+#  error "The limited API is not currently supported in the free-threaded build"
+#endif
+
+#if defined(Py_GIL_DISABLED) && defined(_MSC_VER)
+#  include <intrin.h>             // __readgsqword()
+#endif
+
+#if defined(Py_GIL_DISABLED) && defined(__MINGW32__)
+#  include <intrin.h>             // __readgsqword()
+#endif
+
+// Include Python header files
+#include "pyport.h"
+#include "pymacro.h"
+#include "pymath.h"
+#include "pymem.h"
+#include "pytypedefs.h"
+#include "pybuffer.h"
+#include "pystats.h"
+#include "pyatomic.h"
+#include "lock.h"
+#include "object.h"
+#include "objimpl.h"
+#include "typeslots.h"
+#include "pyhash.h"
+#include "cpython/pydebug.h"
+#include "bytearrayobject.h"
+#include "bytesobject.h"
+#include "unicodeobject.h"
+#include "pyerrors.h"
+#include "longobject.h"
+#include "cpython/longintrepr.h"
+#include "boolobject.h"
+#include "floatobject.h"
+#include "complexobject.h"
+#include "rangeobject.h"
+#include "memoryobject.h"
+#include "tupleobject.h"
+#include "listobject.h"
+#include "dictobject.h"
+#include "cpython/odictobject.h"
+#include "enumobject.h"
+#include "setobject.h"
+#include "methodobject.h"
+#include "moduleobject.h"
+#include "monitoring.h"
+#include "cpython/funcobject.h"
+#include "cpython/classobject.h"
+#include "fileobject.h"
+#include "pycapsule.h"
+#include "cpython/code.h"
+#include "pyframe.h"
+#include "traceback.h"
+#include "sliceobject.h"
+#include "cpython/cellobject.h"
+#include "iterobject.h"
+#include "cpython/initconfig.h"
+#include "pystate.h"
+#include "cpython/genobject.h"
+#include "descrobject.h"
+#include "genericaliasobject.h"
+#include "warnings.h"
+#include "weakrefobject.h"
+#include "structseq.h"
+#include "cpython/picklebufobject.h"
+#include "cpython/pytime.h"
+#include "codecs.h"
+#include "pythread.h"
+#include "cpython/context.h"
+#include "modsupport.h"
+#include "compile.h"
+#include "pythonrun.h"
+#include "pylifecycle.h"
+#include "ceval.h"
+#include "sysmodule.h"
+#include "osmodule.h"
+#include "intrcheck.h"
+#include "import.h"
+#include "abstract.h"
+#include "bltinmodule.h"
+#include "critical_section.h"
+#include "cpython/pyctype.h"
+#include "pystrtod.h"
+#include "pystrcmp.h"
+#include "fileutils.h"
+#include "cpython/pyfpe.h"
+#include "cpython/tracemalloc.h"
+
+#endif /* !Py_PYTHON_H */

Dependencies/Python/include/abstract.h

@@ -0,0 +1,911 @@
+/* Abstract Object Interface (many thanks to Jim Fulton) */
+
+#ifndef Py_ABSTRACTOBJECT_H
+#define Py_ABSTRACTOBJECT_H
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/* === Object Protocol ================================================== */
+
+/* Implemented elsewhere:
+
+   int PyObject_Print(PyObject *o, FILE *fp, int flags);
+
+   Print an object 'o' on file 'fp'.  Returns -1 on error. The flags argument
+   is used to enable certain printing options. The only option currently
+   supported is Py_PRINT_RAW. By default (flags=0), PyObject_Print() formats
+   the object by calling PyObject_Repr(). If flags equals to Py_PRINT_RAW, it
+   formats the object by calling PyObject_Str(). */
+
+
+/* Implemented elsewhere:
+
+   int PyObject_HasAttrString(PyObject *o, const char *attr_name);
+
+   Returns 1 if object 'o' has the attribute attr_name, and 0 otherwise.
+
+   This is equivalent to the Python expression: hasattr(o,attr_name).
+
+   This function always succeeds. */
+
+
+/* Implemented elsewhere:
+
+   PyObject* PyObject_GetAttrString(PyObject *o, const char *attr_name);
+
+   Retrieve an attributed named attr_name form object o.
+   Returns the attribute value on success, or NULL on failure.
+
+   This is the equivalent of the Python expression: o.attr_name. */
+
+
+/* Implemented elsewhere:
+
+   int PyObject_HasAttr(PyObject *o, PyObject *attr_name);
+
+   Returns 1 if o has the attribute attr_name, and 0 otherwise.
+
+   This is equivalent to the Python expression: hasattr(o,attr_name).
+
+   This function always succeeds. */
+
+
+/* Implemented elsewhere:
+
+   int PyObject_HasAttrStringWithError(PyObject *o, const char *attr_name);
+
+   Returns 1 if object 'o' has the attribute attr_name, and 0 otherwise.
+   This is equivalent to the Python expression: hasattr(o,attr_name).
+   Returns -1 on failure. */
+
+
+/* Implemented elsewhere:
+
+   int PyObject_HasAttrWithError(PyObject *o, PyObject *attr_name);
+
+   Returns 1 if o has the attribute attr_name, and 0 otherwise.
+   This is equivalent to the Python expression: hasattr(o,attr_name).
+   Returns -1 on failure. */
+
+
+/* Implemented elsewhere:
+
+   PyObject* PyObject_GetAttr(PyObject *o, PyObject *attr_name);
+
+   Retrieve an attributed named 'attr_name' form object 'o'.
+   Returns the attribute value on success, or NULL on failure.
+
+   This is the equivalent of the Python expression: o.attr_name. */
+
+
+/* Implemented elsewhere:
+
+   int PyObject_GetOptionalAttr(PyObject *obj, PyObject *attr_name, PyObject **result);
+
+   Variant of PyObject_GetAttr() which doesn't raise AttributeError
+   if the attribute is not found.
+
+   If the attribute is found, return 1 and set *result to a new strong
+   reference to the attribute.
+   If the attribute is not found, return 0 and set *result to NULL;
+   the AttributeError is silenced.
+   If an error other than AttributeError is raised, return -1 and
+   set *result to NULL.
+*/
+
+
+/* Implemented elsewhere:
+
+   int PyObject_GetOptionalAttrString(PyObject *obj, const char *attr_name, PyObject **result);
+
+   Variant of PyObject_GetAttrString() which doesn't raise AttributeError
+   if the attribute is not found.
+
+   If the attribute is found, return 1 and set *result to a new strong
+   reference to the attribute.
+   If the attribute is not found, return 0 and set *result to NULL;
+   the AttributeError is silenced.
+   If an error other than AttributeError is raised, return -1 and
+   set *result to NULL.
+*/
+
+
+/* Implemented elsewhere:
+
+   int PyObject_SetAttrString(PyObject *o, const char *attr_name, PyObject *v);
+
+   Set the value of the attribute named attr_name, for object 'o',
+   to the value 'v'. Raise an exception and return -1 on failure; return 0 on
+   success.
+
+   This is the equivalent of the Python statement o.attr_name=v. */
+
+
+/* Implemented elsewhere:
+
+   int PyObject_SetAttr(PyObject *o, PyObject *attr_name, PyObject *v);
+
+   Set the value of the attribute named attr_name, for object 'o', to the value
+   'v'. an exception and return -1 on failure; return 0 on success.
+
+   This is the equivalent of the Python statement o.attr_name=v. */
+
+/* Implemented elsewhere:
+
+   int PyObject_DelAttrString(PyObject *o, const char *attr_name);
+
+   Delete attribute named attr_name, for object o. Returns
+   -1 on failure.
+
+   This is the equivalent of the Python statement: del o.attr_name. */
+
+
+/* Implemented elsewhere:
+
+   int PyObject_DelAttr(PyObject *o, PyObject *attr_name);
+
+   Delete attribute named attr_name, for object o. Returns -1
+   on failure.  This is the equivalent of the Python
+   statement: del o.attr_name. */
+
+
+/* Implemented elsewhere:
+
+   PyObject *PyObject_Repr(PyObject *o);
+
+   Compute the string representation of object 'o'.  Returns the
+   string representation on success, NULL on failure.
+
+   This is the equivalent of the Python expression: repr(o).
+
+   Called by the repr() built-in function. */
+
+
+/* Implemented elsewhere:
+
+   PyObject *PyObject_Str(PyObject *o);
+
+   Compute the string representation of object, o.  Returns the
+   string representation on success, NULL on failure.
+
+   This is the equivalent of the Python expression: str(o).
+
+   Called by the str() and print() built-in functions. */
+
+
+/* Declared elsewhere
+
+   PyAPI_FUNC(int) PyCallable_Check(PyObject *o);
+
+   Determine if the object, o, is callable.  Return 1 if the object is callable
+   and 0 otherwise.
+
+   This function always succeeds. */
+
+
+#if !defined(Py_LIMITED_API) || Py_LIMITED_API+0 >= 0x03090000
+/* Call a callable Python object without any arguments */
+PyAPI_FUNC(PyObject *) PyObject_CallNoArgs(PyObject *func);
+#endif
+
+
+/* Call a callable Python object 'callable' with arguments given by the
+   tuple 'args' and keywords arguments given by the dictionary 'kwargs'.
+
+   'args' must not be NULL, use an empty tuple if no arguments are
+   needed. If no named arguments are needed, 'kwargs' can be NULL.
+
+   This is the equivalent of the Python expression:
+   callable(*args, **kwargs). */
+PyAPI_FUNC(PyObject *) PyObject_Call(PyObject *callable,
+                                     PyObject *args, PyObject *kwargs);
+
+
+/* Call a callable Python object 'callable', with arguments given by the
+   tuple 'args'.  If no arguments are needed, then 'args' can be NULL.
+
+   Returns the result of the call on success, or NULL on failure.
+
+   This is the equivalent of the Python expression:
+   callable(*args). */
+PyAPI_FUNC(PyObject *) PyObject_CallObject(PyObject *callable,
+                                           PyObject *args);
+
+/* Call a callable Python object, callable, with a variable number of C
+   arguments. The C arguments are described using a mkvalue-style format
+   string.
+
+   The format may be NULL, indicating that no arguments are provided.
+
+   Returns the result of the call on success, or NULL on failure.
+
+   This is the equivalent of the Python expression:
+   callable(arg1, arg2, ...). */
+PyAPI_FUNC(PyObject *) PyObject_CallFunction(PyObject *callable,
+                                             const char *format, ...);
+
+/* Call the method named 'name' of object 'obj' with a variable number of
+   C arguments.  The C arguments are described by a mkvalue format string.
+
+   The format can be NULL, indicating that no arguments are provided.
+
+   Returns the result of the call on success, or NULL on failure.
+
+   This is the equivalent of the Python expression:
+   obj.name(arg1, arg2, ...). */
+PyAPI_FUNC(PyObject *) PyObject_CallMethod(PyObject *obj,
+                                           const char *name,
+                                           const char *format, ...);
+
+/* Call a callable Python object 'callable' with a variable number of C
+   arguments. The C arguments are provided as PyObject* values, terminated
+   by a NULL.
+
+   Returns the result of the call on success, or NULL on failure.
+
+   This is the equivalent of the Python expression:
+   callable(arg1, arg2, ...). */
+PyAPI_FUNC(PyObject *) PyObject_CallFunctionObjArgs(PyObject *callable,
+                                                    ...);
+
+/* Call the method named 'name' of object 'obj' with a variable number of
+   C arguments.  The C arguments are provided as PyObject* values, terminated
+   by NULL.
+
+   Returns the result of the call on success, or NULL on failure.
+
+   This is the equivalent of the Python expression: obj.name(*args). */
+
+PyAPI_FUNC(PyObject *) PyObject_CallMethodObjArgs(
+    PyObject *obj,
+    PyObject *name,
+    ...);
+
+/* Given a vectorcall nargsf argument, return the actual number of arguments.
+ * (For use outside the limited API, this is re-defined as a static inline
+ * function in cpython/abstract.h)
+ */
+PyAPI_FUNC(Py_ssize_t) PyVectorcall_NARGS(size_t nargsf);
+
+/* Call "callable" (which must support vectorcall) with positional arguments
+   "tuple" and keyword arguments "dict". "dict" may also be NULL */
+PyAPI_FUNC(PyObject *) PyVectorcall_Call(PyObject *callable, PyObject *tuple, PyObject *dict);
+
+#if !defined(Py_LIMITED_API) || Py_LIMITED_API+0 >= 0x030C0000
+#define PY_VECTORCALL_ARGUMENTS_OFFSET \
+    (_Py_STATIC_CAST(size_t, 1) << (8 * sizeof(size_t) - 1))
+
+/* Perform a PEP 590-style vector call on 'callable' */
+PyAPI_FUNC(PyObject *) PyObject_Vectorcall(
+    PyObject *callable,
+    PyObject *const *args,
+    size_t nargsf,
+    PyObject *kwnames);
+
+/* Call the method 'name' on args[0] with arguments in args[1..nargsf-1]. */
+PyAPI_FUNC(PyObject *) PyObject_VectorcallMethod(
+    PyObject *name, PyObject *const *args,
+    size_t nargsf, PyObject *kwnames);
+#endif
+
+/* Implemented elsewhere:
+
+   Py_hash_t PyObject_Hash(PyObject *o);
+
+   Compute and return the hash, hash_value, of an object, o.  On
+   failure, return -1.
+
+   This is the equivalent of the Python expression: hash(o). */
+
+
+/* Implemented elsewhere:
+
+   int PyObject_IsTrue(PyObject *o);
+
+   Returns 1 if the object, o, is considered to be true, 0 if o is
+   considered to be false and -1 on failure.
+
+   This is equivalent to the Python expression: not not o. */
+
+
+/* Implemented elsewhere:
+
+   int PyObject_Not(PyObject *o);
+
+   Returns 0 if the object, o, is considered to be true, 1 if o is
+   considered to be false and -1 on failure.
+
+   This is equivalent to the Python expression: not o. */
+
+
+/* Get the type of an object.
+
+   On success, returns a type object corresponding to the object type of object
+   'o'. On failure, returns NULL.
+
+   This is equivalent to the Python expression: type(o) */
+PyAPI_FUNC(PyObject *) PyObject_Type(PyObject *o);
+
+
+/* Return the size of object 'o'.  If the object 'o' provides both sequence and
+   mapping protocols, the sequence size is returned.
+
+   On error, -1 is returned.
+
+   This is the equivalent to the Python expression: len(o) */
+PyAPI_FUNC(Py_ssize_t) PyObject_Size(PyObject *o);
+
+
+/* For DLL compatibility */
+#undef PyObject_Length
+PyAPI_FUNC(Py_ssize_t) PyObject_Length(PyObject *o);
+#define PyObject_Length PyObject_Size
+
+/* Return element of 'o' corresponding to the object 'key'. Return NULL
+  on failure.
+
+  This is the equivalent of the Python expression: o[key] */
+PyAPI_FUNC(PyObject *) PyObject_GetItem(PyObject *o, PyObject *key);
+
+
+/* Map the object 'key' to the value 'v' into 'o'.
+
+   Raise an exception and return -1 on failure; return 0 on success.
+
+   This is the equivalent of the Python statement: o[key]=v. */
+PyAPI_FUNC(int) PyObject_SetItem(PyObject *o, PyObject *key, PyObject *v);
+
+/* Remove the mapping for the string 'key' from the object 'o'.
+   Returns -1 on failure.
+
+   This is equivalent to the Python statement: del o[key]. */
+PyAPI_FUNC(int) PyObject_DelItemString(PyObject *o, const char *key);
+
+/* Delete the mapping for the object 'key' from the object 'o'.
+   Returns -1 on failure.
+
+   This is the equivalent of the Python statement: del o[key]. */
+PyAPI_FUNC(int) PyObject_DelItem(PyObject *o, PyObject *key);
+
+
+/* Takes an arbitrary object and returns the result of calling
+   obj.__format__(format_spec). */
+PyAPI_FUNC(PyObject *) PyObject_Format(PyObject *obj,
+                                       PyObject *format_spec);
+
+
+/* ==== Iterators ================================================ */
+
+/* Takes an object and returns an iterator for it.
+   This is typically a new iterator but if the argument is an iterator, this
+   returns itself. */
+PyAPI_FUNC(PyObject *) PyObject_GetIter(PyObject *);
+
+/* Takes an AsyncIterable object and returns an AsyncIterator for it.
+   This is typically a new iterator but if the argument is an AsyncIterator,
+   this returns itself. */
+PyAPI_FUNC(PyObject *) PyObject_GetAIter(PyObject *);
+
+/* Returns non-zero if the object 'obj' provides iterator protocols, and 0 otherwise.
+
+   This function always succeeds. */
+PyAPI_FUNC(int) PyIter_Check(PyObject *);
+
+/* Returns non-zero if the object 'obj' provides AsyncIterator protocols, and 0 otherwise.
+
+   This function always succeeds. */
+PyAPI_FUNC(int) PyAIter_Check(PyObject *);
+
+/* Takes an iterator object and calls its tp_iternext slot,
+   returning the next value.
+
+   If the iterator is exhausted, this returns NULL without setting an
+   exception.
+
+   NULL with an exception means an error occurred. */
+PyAPI_FUNC(PyObject *) PyIter_Next(PyObject *);
+
+#if !defined(Py_LIMITED_API) || Py_LIMITED_API+0 >= 0x030A0000
+
+/* Takes generator, coroutine or iterator object and sends the value into it.
+   Returns:
+   - PYGEN_RETURN (0) if generator has returned.
+     'result' parameter is filled with return value
+   - PYGEN_ERROR (-1) if exception was raised.
+     'result' parameter is NULL
+   - PYGEN_NEXT (1) if generator has yielded.
+     'result' parameter is filled with yielded value. */
+PyAPI_FUNC(PySendResult) PyIter_Send(PyObject *, PyObject *, PyObject **);
+#endif
+
+
+/* === Number Protocol ================================================== */
+
+/* Returns 1 if the object 'o' provides numeric protocols, and 0 otherwise.
+
+   This function always succeeds. */
+PyAPI_FUNC(int) PyNumber_Check(PyObject *o);
+
+/* Returns the result of adding o1 and o2, or NULL on failure.
+
+   This is the equivalent of the Python expression: o1 + o2. */
+PyAPI_FUNC(PyObject *) PyNumber_Add(PyObject *o1, PyObject *o2);
+
+/* Returns the result of subtracting o2 from o1, or NULL on failure.
+
+   This is the equivalent of the Python expression: o1 - o2. */
+PyAPI_FUNC(PyObject *) PyNumber_Subtract(PyObject *o1, PyObject *o2);
+
+/* Returns the result of multiplying o1 and o2, or NULL on failure.
+
+   This is the equivalent of the Python expression: o1 * o2. */
+PyAPI_FUNC(PyObject *) PyNumber_Multiply(PyObject *o1, PyObject *o2);
+
+#if !defined(Py_LIMITED_API) || Py_LIMITED_API+0 >= 0x03050000
+/* This is the equivalent of the Python expression: o1 @ o2. */
+PyAPI_FUNC(PyObject *) PyNumber_MatrixMultiply(PyObject *o1, PyObject *o2);
+#endif
+
+/* Returns the result of dividing o1 by o2 giving an integral result,
+   or NULL on failure.
+
+   This is the equivalent of the Python expression: o1 // o2. */
+PyAPI_FUNC(PyObject *) PyNumber_FloorDivide(PyObject *o1, PyObject *o2);
+
+/* Returns the result of dividing o1 by o2 giving a float result, or NULL on
+   failure.
+
+   This is the equivalent of the Python expression: o1 / o2. */
+PyAPI_FUNC(PyObject *) PyNumber_TrueDivide(PyObject *o1, PyObject *o2);
+
+/* Returns the remainder of dividing o1 by o2, or NULL on failure.
+
+   This is the equivalent of the Python expression: o1 % o2. */
+PyAPI_FUNC(PyObject *) PyNumber_Remainder(PyObject *o1, PyObject *o2);
+
+/* See the built-in function divmod.
+
+   Returns NULL on failure.
+
+   This is the equivalent of the Python expression: divmod(o1, o2). */
+PyAPI_FUNC(PyObject *) PyNumber_Divmod(PyObject *o1, PyObject *o2);
+
+/* See the built-in function pow. Returns NULL on failure.
+
+   This is the equivalent of the Python expression: pow(o1, o2, o3),
+   where o3 is optional. */
+PyAPI_FUNC(PyObject *) PyNumber_Power(PyObject *o1, PyObject *o2,
+                                      PyObject *o3);
+
+/* Returns the negation of o on success, or NULL on failure.
+
+ This is the equivalent of the Python expression: -o. */
+PyAPI_FUNC(PyObject *) PyNumber_Negative(PyObject *o);
+
+/* Returns the positive of o on success, or NULL on failure.
+
+   This is the equivalent of the Python expression: +o. */
+PyAPI_FUNC(PyObject *) PyNumber_Positive(PyObject *o);
+
+/* Returns the absolute value of 'o', or NULL on failure.
+
+   This is the equivalent of the Python expression: abs(o). */
+PyAPI_FUNC(PyObject *) PyNumber_Absolute(PyObject *o);
+
+/* Returns the bitwise negation of 'o' on success, or NULL on failure.
+
+   This is the equivalent of the Python expression: ~o. */
+PyAPI_FUNC(PyObject *) PyNumber_Invert(PyObject *o);
+
+/* Returns the result of left shifting o1 by o2 on success, or NULL on failure.
+
+   This is the equivalent of the Python expression: o1 << o2. */
+PyAPI_FUNC(PyObject *) PyNumber_Lshift(PyObject *o1, PyObject *o2);
+
+/* Returns the result of right shifting o1 by o2 on success, or NULL on
+   failure.
+
+   This is the equivalent of the Python expression: o1 >> o2. */
+PyAPI_FUNC(PyObject *) PyNumber_Rshift(PyObject *o1, PyObject *o2);
+
+/* Returns the result of bitwise and of o1 and o2 on success, or NULL on
+   failure.
+
+   This is the equivalent of the Python expression: o1 & o2. */
+PyAPI_FUNC(PyObject *) PyNumber_And(PyObject *o1, PyObject *o2);
+
+/* Returns the bitwise exclusive or of o1 by o2 on success, or NULL on failure.
+
+   This is the equivalent of the Python expression: o1 ^ o2. */
+PyAPI_FUNC(PyObject *) PyNumber_Xor(PyObject *o1, PyObject *o2);
+
+/* Returns the result of bitwise or on o1 and o2 on success, or NULL on
+   failure.
+
+   This is the equivalent of the Python expression: o1 | o2. */
+PyAPI_FUNC(PyObject *) PyNumber_Or(PyObject *o1, PyObject *o2);
+
+/* Returns 1 if obj is an index integer (has the nb_index slot of the
+   tp_as_number structure filled in), and 0 otherwise. */
+PyAPI_FUNC(int) PyIndex_Check(PyObject *);
+
+/* Returns the object 'o' converted to a Python int, or NULL with an exception
+   raised on failure. */
+PyAPI_FUNC(PyObject *) PyNumber_Index(PyObject *o);
+
+/* Returns the object 'o' converted to Py_ssize_t by going through
+   PyNumber_Index() first.
+
+   If an overflow error occurs while converting the int to Py_ssize_t, then the
+   second argument 'exc' is the error-type to return.  If it is NULL, then the
+   overflow error is cleared and the value is clipped. */
+PyAPI_FUNC(Py_ssize_t) PyNumber_AsSsize_t(PyObject *o, PyObject *exc);
+
+/* Returns the object 'o' converted to an integer object on success, or NULL
+   on failure.
+
+   This is the equivalent of the Python expression: int(o). */
+PyAPI_FUNC(PyObject *) PyNumber_Long(PyObject *o);
+
+/* Returns the object 'o' converted to a float object on success, or NULL
+  on failure.
+
+  This is the equivalent of the Python expression: float(o). */
+PyAPI_FUNC(PyObject *) PyNumber_Float(PyObject *o);
+
+
+/* --- In-place variants of (some of) the above number protocol functions -- */
+
+/* Returns the result of adding o2 to o1, possibly in-place, or NULL
+   on failure.
+
+   This is the equivalent of the Python expression: o1 += o2. */
+PyAPI_FUNC(PyObject *) PyNumber_InPlaceAdd(PyObject *o1, PyObject *o2);
+
+/* Returns the result of subtracting o2 from o1, possibly in-place or
+   NULL on failure.
+
+   This is the equivalent of the Python expression: o1 -= o2. */
+PyAPI_FUNC(PyObject *) PyNumber_InPlaceSubtract(PyObject *o1, PyObject *o2);
+
+/* Returns the result of multiplying o1 by o2, possibly in-place, or NULL on
+   failure.
+
+   This is the equivalent of the Python expression: o1 *= o2. */
+PyAPI_FUNC(PyObject *) PyNumber_InPlaceMultiply(PyObject *o1, PyObject *o2);
+
+#if !defined(Py_LIMITED_API) || Py_LIMITED_API+0 >= 0x03050000
+/* This is the equivalent of the Python expression: o1 @= o2. */
+PyAPI_FUNC(PyObject *) PyNumber_InPlaceMatrixMultiply(PyObject *o1, PyObject *o2);
+#endif
+
+/* Returns the result of dividing o1 by o2 giving an integral result, possibly
+   in-place, or NULL on failure.
+
+   This is the equivalent of the Python expression: o1 /= o2. */
+PyAPI_FUNC(PyObject *) PyNumber_InPlaceFloorDivide(PyObject *o1,
+                                                   PyObject *o2);
+
+/* Returns the result of dividing o1 by o2 giving a float result, possibly
+   in-place, or null on failure.
+
+   This is the equivalent of the Python expression: o1 /= o2. */
+PyAPI_FUNC(PyObject *) PyNumber_InPlaceTrueDivide(PyObject *o1,
+                                                  PyObject *o2);
+
+/* Returns the remainder of dividing o1 by o2, possibly in-place, or NULL on
+   failure.
+
+   This is the equivalent of the Python expression: o1 %= o2. */
+PyAPI_FUNC(PyObject *) PyNumber_InPlaceRemainder(PyObject *o1, PyObject *o2);
+
+/* Returns the result of raising o1 to the power of o2, possibly in-place,
+   or NULL on failure.
+
+   This is the equivalent of the Python expression: o1 **= o2,
+   or o1 = pow(o1, o2, o3) if o3 is present. */
+PyAPI_FUNC(PyObject *) PyNumber_InPlacePower(PyObject *o1, PyObject *o2,
+                                             PyObject *o3);
+
+/* Returns the result of left shifting o1 by o2, possibly in-place, or NULL
+   on failure.
+
+   This is the equivalent of the Python expression: o1 <<= o2. */
+PyAPI_FUNC(PyObject *) PyNumber_InPlaceLshift(PyObject *o1, PyObject *o2);
+
+/* Returns the result of right shifting o1 by o2, possibly in-place or NULL
+   on failure.
+
+   This is the equivalent of the Python expression: o1 >>= o2. */
+PyAPI_FUNC(PyObject *) PyNumber_InPlaceRshift(PyObject *o1, PyObject *o2);
+
+/* Returns the result of bitwise and of o1 and o2, possibly in-place, or NULL
+   on failure.
+
+   This is the equivalent of the Python expression: o1 &= o2. */
+PyAPI_FUNC(PyObject *) PyNumber_InPlaceAnd(PyObject *o1, PyObject *o2);
+
+/* Returns the bitwise exclusive or of o1 by o2, possibly in-place, or NULL
+   on failure.
+
+   This is the equivalent of the Python expression: o1 ^= o2. */
+PyAPI_FUNC(PyObject *) PyNumber_InPlaceXor(PyObject *o1, PyObject *o2);
+
+/* Returns the result of bitwise or of o1 and o2, possibly in-place,
+   or NULL on failure.
+
+   This is the equivalent of the Python expression: o1 |= o2. */
+PyAPI_FUNC(PyObject *) PyNumber_InPlaceOr(PyObject *o1, PyObject *o2);
+
+/* Returns the integer n converted to a string with a base, with a base
+   marker of 0b, 0o or 0x prefixed if applicable.
+
+   If n is not an int object, it is converted with PyNumber_Index first. */
+PyAPI_FUNC(PyObject *) PyNumber_ToBase(PyObject *n, int base);
+
+
+/* === Sequence protocol ================================================ */
+
+/* Return 1 if the object provides sequence protocol, and zero
+   otherwise.
+
+   This function always succeeds. */
+PyAPI_FUNC(int) PySequence_Check(PyObject *o);
+
+/* Return the size of sequence object o, or -1 on failure. */
+PyAPI_FUNC(Py_ssize_t) PySequence_Size(PyObject *o);
+
+/* For DLL compatibility */
+#undef PySequence_Length
+PyAPI_FUNC(Py_ssize_t) PySequence_Length(PyObject *o);
+#define PySequence_Length PySequence_Size
+
+
+/* Return the concatenation of o1 and o2 on success, and NULL on failure.
+
+   This is the equivalent of the Python expression: o1 + o2. */
+PyAPI_FUNC(PyObject *) PySequence_Concat(PyObject *o1, PyObject *o2);
+
+/* Return the result of repeating sequence object 'o' 'count' times,
+  or NULL on failure.
+
+  This is the equivalent of the Python expression: o * count. */
+PyAPI_FUNC(PyObject *) PySequence_Repeat(PyObject *o, Py_ssize_t count);
+
+/* Return the ith element of o, or NULL on failure.
+
+   This is the equivalent of the Python expression: o[i]. */
+PyAPI_FUNC(PyObject *) PySequence_GetItem(PyObject *o, Py_ssize_t i);
+
+/* Return the slice of sequence object o between i1 and i2, or NULL on failure.
+
+   This is the equivalent of the Python expression: o[i1:i2]. */
+PyAPI_FUNC(PyObject *) PySequence_GetSlice(PyObject *o, Py_ssize_t i1, Py_ssize_t i2);
+
+/* Assign object 'v' to the ith element of the sequence 'o'. Raise an exception
+   and return -1 on failure; return 0 on success.
+
+   This is the equivalent of the Python statement o[i] = v. */
+PyAPI_FUNC(int) PySequence_SetItem(PyObject *o, Py_ssize_t i, PyObject *v);
+
+/* Delete the 'i'-th element of the sequence 'v'. Returns -1 on failure.
+
+   This is the equivalent of the Python statement: del o[i]. */
+PyAPI_FUNC(int) PySequence_DelItem(PyObject *o, Py_ssize_t i);
+
+/* Assign the sequence object 'v' to the slice in sequence object 'o',
+   from 'i1' to 'i2'. Returns -1 on failure.
+
+   This is the equivalent of the Python statement: o[i1:i2] = v. */
+PyAPI_FUNC(int) PySequence_SetSlice(PyObject *o, Py_ssize_t i1, Py_ssize_t i2,
+                                    PyObject *v);
+
+/* Delete the slice in sequence object 'o' from 'i1' to 'i2'.
+   Returns -1 on failure.
+
+   This is the equivalent of the Python statement: del o[i1:i2]. */
+PyAPI_FUNC(int) PySequence_DelSlice(PyObject *o, Py_ssize_t i1, Py_ssize_t i2);
+
+/* Returns the sequence 'o' as a tuple on success, and NULL on failure.
+
+   This is equivalent to the Python expression: tuple(o). */
+PyAPI_FUNC(PyObject *) PySequence_Tuple(PyObject *o);
+
+/* Returns the sequence 'o' as a list on success, and NULL on failure.
+   This is equivalent to the Python expression: list(o) */
+PyAPI_FUNC(PyObject *) PySequence_List(PyObject *o);
+
+/* Return the sequence 'o' as a list, unless it's already a tuple or list.
+
+   Use PySequence_Fast_GET_ITEM to access the members of this list, and
+   PySequence_Fast_GET_SIZE to get its length.
+
+   Returns NULL on failure.  If the object does not support iteration, raises a
+   TypeError exception with 'm' as the message text. */
+PyAPI_FUNC(PyObject *) PySequence_Fast(PyObject *o, const char* m);
+
+/* Return the size of the sequence 'o', assuming that 'o' was returned by
+   PySequence_Fast and is not NULL. */
+#define PySequence_Fast_GET_SIZE(o) \
+    (PyList_Check(o) ? PyList_GET_SIZE(o) : PyTuple_GET_SIZE(o))
+
+/* Return the 'i'-th element of the sequence 'o', assuming that o was returned
+   by PySequence_Fast, and that i is within bounds. */
+#define PySequence_Fast_GET_ITEM(o, i)\
+     (PyList_Check(o) ? PyList_GET_ITEM((o), (i)) : PyTuple_GET_ITEM((o), (i)))
+
+/* Return a pointer to the underlying item array for
+   an object returned by PySequence_Fast */
+#define PySequence_Fast_ITEMS(sf) \
+    (PyList_Check(sf) ? ((PyListObject *)(sf))->ob_item \
+                      : ((PyTupleObject *)(sf))->ob_item)
+
+/* Return the number of occurrences on value on 'o', that is, return
+   the number of keys for which o[key] == value.
+
+   On failure, return -1.  This is equivalent to the Python expression:
+   o.count(value). */
+PyAPI_FUNC(Py_ssize_t) PySequence_Count(PyObject *o, PyObject *value);
+
+/* Return 1 if 'ob' is in the sequence 'seq'; 0 if 'ob' is not in the sequence
+   'seq'; -1 on error.
+
+   Use __contains__ if possible, else _PySequence_IterSearch(). */
+PyAPI_FUNC(int) PySequence_Contains(PyObject *seq, PyObject *ob);
+
+/* For DLL-level backwards compatibility */
+#undef PySequence_In
+/* Determine if the sequence 'o' contains 'value'. If an item in 'o' is equal
+   to 'value', return 1, otherwise return 0. On error, return -1.
+
+   This is equivalent to the Python expression: value in o. */
+PyAPI_FUNC(int) PySequence_In(PyObject *o, PyObject *value);
+
+/* For source-level backwards compatibility */
+#define PySequence_In PySequence_Contains
+
+
+/* Return the first index for which o[i] == value.
+   On error, return -1.
+
+   This is equivalent to the Python expression: o.index(value). */
+PyAPI_FUNC(Py_ssize_t) PySequence_Index(PyObject *o, PyObject *value);
+
+
+/* --- In-place versions of some of the above Sequence functions --- */
+
+/* Append sequence 'o2' to sequence 'o1', in-place when possible. Return the
+   resulting object, which could be 'o1', or NULL on failure.
+
+  This is the equivalent of the Python expression: o1 += o2. */
+PyAPI_FUNC(PyObject *) PySequence_InPlaceConcat(PyObject *o1, PyObject *o2);
+
+/* Repeat sequence 'o' by 'count', in-place when possible. Return the resulting
+   object, which could be 'o', or NULL on failure.
+
+   This is the equivalent of the Python expression: o1 *= count.  */
+PyAPI_FUNC(PyObject *) PySequence_InPlaceRepeat(PyObject *o, Py_ssize_t count);
+
+
+/* === Mapping protocol ================================================= */
+
+/* Return 1 if the object provides mapping protocol, and 0 otherwise.
+
+   This function always succeeds. */
+PyAPI_FUNC(int) PyMapping_Check(PyObject *o);
+
+/* Returns the number of keys in mapping object 'o' on success, and -1 on
+  failure. This is equivalent to the Python expression: len(o). */
+PyAPI_FUNC(Py_ssize_t) PyMapping_Size(PyObject *o);
+
+/* For DLL compatibility */
+#undef PyMapping_Length
+PyAPI_FUNC(Py_ssize_t) PyMapping_Length(PyObject *o);
+#define PyMapping_Length PyMapping_Size
+
+
+/* Implemented as a macro:
+
+   int PyMapping_DelItemString(PyObject *o, const char *key);
+
+   Remove the mapping for the string 'key' from the mapping 'o'. Returns -1 on
+   failure.
+
+   This is equivalent to the Python statement: del o[key]. */
+#define PyMapping_DelItemString(O, K) PyObject_DelItemString((O), (K))
+
+/* Implemented as a macro:
+
+   int PyMapping_DelItem(PyObject *o, PyObject *key);
+
+   Remove the mapping for the object 'key' from the mapping object 'o'.
+   Returns -1 on failure.
+
+   This is equivalent to the Python statement: del o[key]. */
+#define PyMapping_DelItem(O, K) PyObject_DelItem((O), (K))
+
+/* On success, return 1 if the mapping object 'o' has the key 'key',
+   and 0 otherwise.
+
+   This is equivalent to the Python expression: key in o.
+
+   This function always succeeds. */
+PyAPI_FUNC(int) PyMapping_HasKeyString(PyObject *o, const char *key);
+
+/* Return 1 if the mapping object has the key 'key', and 0 otherwise.
+
+   This is equivalent to the Python expression: key in o.
+
+   This function always succeeds. */
+PyAPI_FUNC(int) PyMapping_HasKey(PyObject *o, PyObject *key);
+
+/* Return 1 if the mapping object has the key 'key', and 0 otherwise.
+   This is equivalent to the Python expression: key in o.
+   On failure, return -1. */
+
+PyAPI_FUNC(int) PyMapping_HasKeyWithError(PyObject *o, PyObject *key);
+
+/* Return 1 if the mapping object has the key 'key', and 0 otherwise.
+   This is equivalent to the Python expression: key in o.
+   On failure, return -1. */
+
+PyAPI_FUNC(int) PyMapping_HasKeyStringWithError(PyObject *o, const char *key);
+
+/* On success, return a list or tuple of the keys in mapping object 'o'.
+   On failure, return NULL. */
+PyAPI_FUNC(PyObject *) PyMapping_Keys(PyObject *o);
+
+/* On success, return a list or tuple of the values in mapping object 'o'.
+   On failure, return NULL. */
+PyAPI_FUNC(PyObject *) PyMapping_Values(PyObject *o);
+
+/* On success, return a list or tuple of the items in mapping object 'o',
+   where each item is a tuple containing a key-value pair. On failure, return
+   NULL. */
+PyAPI_FUNC(PyObject *) PyMapping_Items(PyObject *o);
+
+/* Return element of 'o' corresponding to the string 'key' or NULL on failure.
+
+   This is the equivalent of the Python expression: o[key]. */
+PyAPI_FUNC(PyObject *) PyMapping_GetItemString(PyObject *o,
+                                               const char *key);
+
+/* Variants of PyObject_GetItem() and PyMapping_GetItemString() which don't
+   raise KeyError if the key is not found.
+
+   If the key is found, return 1 and set *result to a new strong
+   reference to the corresponding value.
+   If the key is not found, return 0 and set *result to NULL;
+   the KeyError is silenced.
+   If an error other than KeyError is raised, return -1 and
+   set *result to NULL.
+*/
+#if !defined(Py_LIMITED_API) || Py_LIMITED_API+0 >= 0x030d0000
+PyAPI_FUNC(int) PyMapping_GetOptionalItem(PyObject *, PyObject *, PyObject **);
+PyAPI_FUNC(int) PyMapping_GetOptionalItemString(PyObject *, const char *, PyObject **);
+#endif
+
+/* Map the string 'key' to the value 'v' in the mapping 'o'.
+   Returns -1 on failure.
+
+   This is the equivalent of the Python statement: o[key]=v. */
+PyAPI_FUNC(int) PyMapping_SetItemString(PyObject *o, const char *key,
+                                        PyObject *value);
+
+/* isinstance(object, typeorclass) */
+PyAPI_FUNC(int) PyObject_IsInstance(PyObject *object, PyObject *typeorclass);
+
+/* issubclass(object, typeorclass) */
+PyAPI_FUNC(int) PyObject_IsSubclass(PyObject *object, PyObject *typeorclass);
+
+#ifndef Py_LIMITED_API
+#  define Py_CPYTHON_ABSTRACTOBJECT_H
+#  include "cpython/abstract.h"
+#  undef Py_CPYTHON_ABSTRACTOBJECT_H
+#endif
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* Py_ABSTRACTOBJECT_H */

Dependencies/Python/include/bltinmodule.h

@@ -0,0 +1,14 @@
+#ifndef Py_BLTINMODULE_H
+#define Py_BLTINMODULE_H
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+PyAPI_DATA(PyTypeObject) PyFilter_Type;
+PyAPI_DATA(PyTypeObject) PyMap_Type;
+PyAPI_DATA(PyTypeObject) PyZip_Type;
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* !Py_BLTINMODULE_H */

Dependencies/Python/include/boolobject.h

@@ -0,0 +1,47 @@
+/* Boolean object interface */
+
+#ifndef Py_BOOLOBJECT_H
+#define Py_BOOLOBJECT_H
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+
+// PyBool_Type is declared by object.h
+
+#define PyBool_Check(x) Py_IS_TYPE((x), &PyBool_Type)
+
+/* Py_False and Py_True are the only two bools in existence. */
+
+/* Don't use these directly */
+PyAPI_DATA(PyLongObject) _Py_FalseStruct;
+PyAPI_DATA(PyLongObject) _Py_TrueStruct;
+
+/* Use these macros */
+#if defined(Py_LIMITED_API) && Py_LIMITED_API+0 >= 0x030D0000
+#  define Py_False Py_GetConstantBorrowed(Py_CONSTANT_FALSE)
+#  define Py_True Py_GetConstantBorrowed(Py_CONSTANT_TRUE)
+#else
+#  define Py_False _PyObject_CAST(&_Py_FalseStruct)
+#  define Py_True _PyObject_CAST(&_Py_TrueStruct)
+#endif
+
+// Test if an object is the True singleton, the same as "x is True" in Python.
+PyAPI_FUNC(int) Py_IsTrue(PyObject *x);
+#define Py_IsTrue(x) Py_Is((x), Py_True)
+
+// Test if an object is the False singleton, the same as "x is False" in Python.
+PyAPI_FUNC(int) Py_IsFalse(PyObject *x);
+#define Py_IsFalse(x) Py_Is((x), Py_False)
+
+/* Macros for returning Py_True or Py_False, respectively */
+#define Py_RETURN_TRUE return Py_True
+#define Py_RETURN_FALSE return Py_False
+
+/* Function to return a bool from a C long */
+PyAPI_FUNC(PyObject *) PyBool_FromLong(long);
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* !Py_BOOLOBJECT_H */

Dependencies/Python/include/bytearrayobject.h

@@ -0,0 +1,44 @@
+/* ByteArray object interface */
+
+#ifndef Py_BYTEARRAYOBJECT_H
+#define Py_BYTEARRAYOBJECT_H
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/* Type PyByteArrayObject represents a mutable array of bytes.
+ * The Python API is that of a sequence;
+ * the bytes are mapped to ints in [0, 256).
+ * Bytes are not characters; they may be used to encode characters.
+ * The only way to go between bytes and str/unicode is via encoding
+ * and decoding.
+ * For the convenience of C programmers, the bytes type is considered
+ * to contain a char pointer, not an unsigned char pointer.
+ */
+
+/* Type object */
+PyAPI_DATA(PyTypeObject) PyByteArray_Type;
+PyAPI_DATA(PyTypeObject) PyByteArrayIter_Type;
+
+/* Type check macros */
+#define PyByteArray_Check(self) PyObject_TypeCheck((self), &PyByteArray_Type)
+#define PyByteArray_CheckExact(self) Py_IS_TYPE((self), &PyByteArray_Type)
+
+/* Direct API functions */
+PyAPI_FUNC(PyObject *) PyByteArray_FromObject(PyObject *);
+PyAPI_FUNC(PyObject *) PyByteArray_Concat(PyObject *, PyObject *);
+PyAPI_FUNC(PyObject *) PyByteArray_FromStringAndSize(const char *, Py_ssize_t);
+PyAPI_FUNC(Py_ssize_t) PyByteArray_Size(PyObject *);
+PyAPI_FUNC(char *) PyByteArray_AsString(PyObject *);
+PyAPI_FUNC(int) PyByteArray_Resize(PyObject *, Py_ssize_t);
+
+#ifndef Py_LIMITED_API
+#  define Py_CPYTHON_BYTEARRAYOBJECT_H
+#  include "cpython/bytearrayobject.h"
+#  undef Py_CPYTHON_BYTEARRAYOBJECT_H
+#endif
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* !Py_BYTEARRAYOBJECT_H */

Dependencies/Python/include/bytesobject.h

@@ -0,0 +1,66 @@
+// Bytes object interface
+
+#ifndef Py_BYTESOBJECT_H
+#define Py_BYTESOBJECT_H
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/*
+Type PyBytesObject represents a byte string.  An extra zero byte is
+reserved at the end to ensure it is zero-terminated, but a size is
+present so strings with null bytes in them can be represented.  This
+is an immutable object type.
+
+There are functions to create new bytes objects, to test
+an object for bytes-ness, and to get the
+byte string value.  The latter function returns a null pointer
+if the object is not of the proper type.
+There is a variant that takes an explicit size as well as a
+variant that assumes a zero-terminated string.  Note that none of the
+functions should be applied to NULL pointer.
+*/
+
+PyAPI_DATA(PyTypeObject) PyBytes_Type;
+PyAPI_DATA(PyTypeObject) PyBytesIter_Type;
+
+#define PyBytes_Check(op) \
+                 PyType_FastSubclass(Py_TYPE(op), Py_TPFLAGS_BYTES_SUBCLASS)
+#define PyBytes_CheckExact(op) Py_IS_TYPE((op), &PyBytes_Type)
+
+PyAPI_FUNC(PyObject *) PyBytes_FromStringAndSize(const char *, Py_ssize_t);
+PyAPI_FUNC(PyObject *) PyBytes_FromString(const char *);
+PyAPI_FUNC(PyObject *) PyBytes_FromObject(PyObject *);
+PyAPI_FUNC(PyObject *) PyBytes_FromFormatV(const char*, va_list)
+                                Py_GCC_ATTRIBUTE((format(printf, 1, 0)));
+PyAPI_FUNC(PyObject *) PyBytes_FromFormat(const char*, ...)
+                                Py_GCC_ATTRIBUTE((format(printf, 1, 2)));
+PyAPI_FUNC(Py_ssize_t) PyBytes_Size(PyObject *);
+PyAPI_FUNC(char *) PyBytes_AsString(PyObject *);
+PyAPI_FUNC(PyObject *) PyBytes_Repr(PyObject *, int);
+PyAPI_FUNC(void) PyBytes_Concat(PyObject **, PyObject *);
+PyAPI_FUNC(void) PyBytes_ConcatAndDel(PyObject **, PyObject *);
+PyAPI_FUNC(PyObject *) PyBytes_DecodeEscape(const char *, Py_ssize_t,
+                                            const char *, Py_ssize_t,
+                                            const char *);
+
+/* Provides access to the internal data buffer and size of a bytes object.
+   Passing NULL as len parameter will force the string buffer to be
+   0-terminated (passing a string with embedded NUL characters will
+   cause an exception).  */
+PyAPI_FUNC(int) PyBytes_AsStringAndSize(
+    PyObject *obj,      /* bytes object */
+    char **s,           /* pointer to buffer variable */
+    Py_ssize_t *len     /* pointer to length variable or NULL */
+    );
+
+#ifndef Py_LIMITED_API
+#  define Py_CPYTHON_BYTESOBJECT_H
+#  include "cpython/bytesobject.h"
+#  undef Py_CPYTHON_BYTESOBJECT_H
+#endif
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* !Py_BYTESOBJECT_H */

Dependencies/Python/include/ceval.h

@@ -0,0 +1,145 @@
+/* Interface to random parts in ceval.c */
+
+#ifndef Py_CEVAL_H
+#define Py_CEVAL_H
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+
+PyAPI_FUNC(PyObject *) PyEval_EvalCode(PyObject *, PyObject *, PyObject *);
+
+PyAPI_FUNC(PyObject *) PyEval_EvalCodeEx(PyObject *co,
+                                         PyObject *globals,
+                                         PyObject *locals,
+                                         PyObject *const *args, int argc,
+                                         PyObject *const *kwds, int kwdc,
+                                         PyObject *const *defs, int defc,
+                                         PyObject *kwdefs, PyObject *closure);
+
+PyAPI_FUNC(PyObject *) PyEval_GetBuiltins(void);
+PyAPI_FUNC(PyObject *) PyEval_GetGlobals(void);
+PyAPI_FUNC(PyObject *) PyEval_GetLocals(void);
+PyAPI_FUNC(PyFrameObject *) PyEval_GetFrame(void);
+
+PyAPI_FUNC(PyObject *) PyEval_GetFrameBuiltins(void);
+PyAPI_FUNC(PyObject *) PyEval_GetFrameGlobals(void);
+PyAPI_FUNC(PyObject *) PyEval_GetFrameLocals(void);
+
+PyAPI_FUNC(int) Py_AddPendingCall(int (*func)(void *), void *arg);
+PyAPI_FUNC(int) Py_MakePendingCalls(void);
+
+/* Protection against deeply nested recursive calls
+
+   In Python 3.0, this protection has two levels:
+   * normal anti-recursion protection is triggered when the recursion level
+     exceeds the current recursion limit. It raises a RecursionError, and sets
+     the "overflowed" flag in the thread state structure. This flag
+     temporarily *disables* the normal protection; this allows cleanup code
+     to potentially outgrow the recursion limit while processing the
+     RecursionError.
+   * "last chance" anti-recursion protection is triggered when the recursion
+     level exceeds "current recursion limit + 50". By construction, this
+     protection can only be triggered when the "overflowed" flag is set. It
+     means the cleanup code has itself gone into an infinite loop, or the
+     RecursionError has been mistakingly ignored. When this protection is
+     triggered, the interpreter aborts with a Fatal Error.
+
+   In addition, the "overflowed" flag is automatically reset when the
+   recursion level drops below "current recursion limit - 50". This heuristic
+   is meant to ensure that the normal anti-recursion protection doesn't get
+   disabled too long.
+
+   Please note: this scheme has its own limitations. See:
+   http://mail.python.org/pipermail/python-dev/2008-August/082106.html
+   for some observations.
+*/
+PyAPI_FUNC(void) Py_SetRecursionLimit(int);
+PyAPI_FUNC(int) Py_GetRecursionLimit(void);
+
+PyAPI_FUNC(int) Py_EnterRecursiveCall(const char *where);
+PyAPI_FUNC(void) Py_LeaveRecursiveCall(void);
+
+PyAPI_FUNC(const char *) PyEval_GetFuncName(PyObject *);
+PyAPI_FUNC(const char *) PyEval_GetFuncDesc(PyObject *);
+
+PyAPI_FUNC(PyObject *) PyEval_EvalFrame(PyFrameObject *);
+PyAPI_FUNC(PyObject *) PyEval_EvalFrameEx(PyFrameObject *f, int exc);
+
+/* Interface for threads.
+
+   A module that plans to do a blocking system call (or something else
+   that lasts a long time and doesn't touch Python data) can allow other
+   threads to run as follows:
+
+    ...preparations here...
+    Py_BEGIN_ALLOW_THREADS
+    ...blocking system call here...
+    Py_END_ALLOW_THREADS
+    ...interpret result here...
+
+   The Py_BEGIN_ALLOW_THREADS/Py_END_ALLOW_THREADS pair expands to a
+   {}-surrounded block.
+   To leave the block in the middle (e.g., with return), you must insert
+   a line containing Py_BLOCK_THREADS before the return, e.g.
+
+    if (...premature_exit...) {
+        Py_BLOCK_THREADS
+        PyErr_SetFromErrno(PyExc_OSError);
+        return NULL;
+    }
+
+   An alternative is:
+
+    Py_BLOCK_THREADS
+    if (...premature_exit...) {
+        PyErr_SetFromErrno(PyExc_OSError);
+        return NULL;
+    }
+    Py_UNBLOCK_THREADS
+
+   For convenience, that the value of 'errno' is restored across
+   Py_END_ALLOW_THREADS and Py_BLOCK_THREADS.
+
+   WARNING: NEVER NEST CALLS TO Py_BEGIN_ALLOW_THREADS AND
+   Py_END_ALLOW_THREADS!!!
+
+   Note that not yet all candidates have been converted to use this
+   mechanism!
+*/
+
+PyAPI_FUNC(PyThreadState *) PyEval_SaveThread(void);
+PyAPI_FUNC(void) PyEval_RestoreThread(PyThreadState *);
+
+Py_DEPRECATED(3.9) PyAPI_FUNC(void) PyEval_InitThreads(void);
+
+PyAPI_FUNC(void) PyEval_AcquireThread(PyThreadState *tstate);
+PyAPI_FUNC(void) PyEval_ReleaseThread(PyThreadState *tstate);
+
+#define Py_BEGIN_ALLOW_THREADS { \
+                        PyThreadState *_save; \
+                        _save = PyEval_SaveThread();
+#define Py_BLOCK_THREADS        PyEval_RestoreThread(_save);
+#define Py_UNBLOCK_THREADS      _save = PyEval_SaveThread();
+#define Py_END_ALLOW_THREADS    PyEval_RestoreThread(_save); \
+                 }
+
+/* Masks and values used by FORMAT_VALUE opcode. */
+#define FVC_MASK      0x3
+#define FVC_NONE      0x0
+#define FVC_STR       0x1
+#define FVC_REPR      0x2
+#define FVC_ASCII     0x3
+#define FVS_MASK      0x4
+#define FVS_HAVE_SPEC 0x4
+
+#ifndef Py_LIMITED_API
+#  define Py_CPYTHON_CEVAL_H
+#  include "cpython/ceval.h"
+#  undef Py_CPYTHON_CEVAL_H
+#endif
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* !Py_CEVAL_H */

Dependencies/Python/include/codecs.h

@@ -0,0 +1,176 @@
+#ifndef Py_CODECREGISTRY_H
+#define Py_CODECREGISTRY_H
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/* ------------------------------------------------------------------------
+
+   Python Codec Registry and support functions
+
+
+Written by Marc-Andre Lemburg (mal@lemburg.com).
+
+Copyright (c) Corporation for National Research Initiatives.
+
+   ------------------------------------------------------------------------ */
+
+/* Register a new codec search function.
+
+   As side effect, this tries to load the encodings package, if not
+   yet done, to make sure that it is always first in the list of
+   search functions.
+
+   The search_function's refcount is incremented by this function. */
+
+PyAPI_FUNC(int) PyCodec_Register(
+       PyObject *search_function
+       );
+
+/* Unregister a codec search function and clear the registry's cache.
+   If the search function is not registered, do nothing.
+   Return 0 on success. Raise an exception and return -1 on error. */
+
+PyAPI_FUNC(int) PyCodec_Unregister(
+       PyObject *search_function
+       );
+
+/* Codec registry encoding check API.
+
+   Returns 1/0 depending on whether there is a registered codec for
+   the given encoding.
+
+*/
+
+PyAPI_FUNC(int) PyCodec_KnownEncoding(
+       const char *encoding
+       );
+
+/* Generic codec based encoding API.
+
+   object is passed through the encoder function found for the given
+   encoding using the error handling method defined by errors. errors
+   may be NULL to use the default method defined for the codec.
+
+   Raises a LookupError in case no encoder can be found.
+
+ */
+
+PyAPI_FUNC(PyObject *) PyCodec_Encode(
+       PyObject *object,
+       const char *encoding,
+       const char *errors
+       );
+
+/* Generic codec based decoding API.
+
+   object is passed through the decoder function found for the given
+   encoding using the error handling method defined by errors. errors
+   may be NULL to use the default method defined for the codec.
+
+   Raises a LookupError in case no encoder can be found.
+
+ */
+
+PyAPI_FUNC(PyObject *) PyCodec_Decode(
+       PyObject *object,
+       const char *encoding,
+       const char *errors
+       );
+
+// --- Codec Lookup APIs --------------------------------------------------
+
+/* Codec registry lookup API.
+
+   Looks up the given encoding and returns a CodecInfo object with
+   function attributes which implement the different aspects of
+   processing the encoding.
+
+   The encoding string is looked up converted to all lower-case
+   characters. This makes encodings looked up through this mechanism
+   effectively case-insensitive.
+
+   If no codec is found, a KeyError is set and NULL returned.
+
+   As side effect, this tries to load the encodings package, if not
+   yet done. This is part of the lazy load strategy for the encodings
+   package.
+ */
+
+/* Get an encoder function for the given encoding. */
+
+PyAPI_FUNC(PyObject *) PyCodec_Encoder(const char *encoding);
+
+/* Get a decoder function for the given encoding. */
+
+PyAPI_FUNC(PyObject *) PyCodec_Decoder(const char *encoding);
+
+/* Get an IncrementalEncoder object for the given encoding. */
+
+PyAPI_FUNC(PyObject *) PyCodec_IncrementalEncoder(
+   const char *encoding,
+   const char *errors);
+
+/* Get an IncrementalDecoder object function for the given encoding. */
+
+PyAPI_FUNC(PyObject *) PyCodec_IncrementalDecoder(
+   const char *encoding,
+   const char *errors);
+
+/* Get a StreamReader factory function for the given encoding. */
+
+PyAPI_FUNC(PyObject *) PyCodec_StreamReader(
+   const char *encoding,
+   PyObject *stream,
+   const char *errors);
+
+/* Get a StreamWriter factory function for the given encoding. */
+
+PyAPI_FUNC(PyObject *) PyCodec_StreamWriter(
+   const char *encoding,
+   PyObject *stream,
+   const char *errors);
+
+/* Unicode encoding error handling callback registry API */
+
+/* Register the error handling callback function error under the given
+   name. This function will be called by the codec when it encounters
+   unencodable characters/undecodable bytes and doesn't know the
+   callback name, when name is specified as the error parameter
+   in the call to the encode/decode function.
+   Return 0 on success, -1 on error */
+PyAPI_FUNC(int) PyCodec_RegisterError(const char *name, PyObject *error);
+
+/* Lookup the error handling callback function registered under the given
+   name. As a special case NULL can be passed, in which case
+   the error handling callback for "strict" will be returned. */
+PyAPI_FUNC(PyObject *) PyCodec_LookupError(const char *name);
+
+/* raise exc as an exception */
+PyAPI_FUNC(PyObject *) PyCodec_StrictErrors(PyObject *exc);
+
+/* ignore the unicode error, skipping the faulty input */
+PyAPI_FUNC(PyObject *) PyCodec_IgnoreErrors(PyObject *exc);
+
+/* replace the unicode encode error with ? or U+FFFD */
+PyAPI_FUNC(PyObject *) PyCodec_ReplaceErrors(PyObject *exc);
+
+/* replace the unicode encode error with XML character references */
+PyAPI_FUNC(PyObject *) PyCodec_XMLCharRefReplaceErrors(PyObject *exc);
+
+/* replace the unicode encode error with backslash escapes (\x, \u and \U) */
+PyAPI_FUNC(PyObject *) PyCodec_BackslashReplaceErrors(PyObject *exc);
+
+#if !defined(Py_LIMITED_API) || Py_LIMITED_API+0 >= 0x03050000
+/* replace the unicode encode error with backslash escapes (\N, \x, \u and \U) */
+PyAPI_FUNC(PyObject *) PyCodec_NameReplaceErrors(PyObject *exc);
+#endif
+
+#ifndef Py_LIMITED_API
+PyAPI_DATA(const char *) Py_hexdigits;
+#endif
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* !Py_CODECREGISTRY_H */

Dependencies/Python/include/compile.h

@@ -0,0 +1,22 @@
+#ifndef Py_COMPILE_H
+#define Py_COMPILE_H
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/* These definitions must match corresponding definitions in graminit.h. */
+#define Py_single_input 256
+#define Py_file_input 257
+#define Py_eval_input 258
+#define Py_func_type_input 345
+
+#ifndef Py_LIMITED_API
+#  define Py_CPYTHON_COMPILE_H
+#  include "cpython/compile.h"
+#  undef Py_CPYTHON_COMPILE_H
+#endif
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* !Py_COMPILE_H */

Dependencies/Python/include/complexobject.h

@@ -0,0 +1,30 @@
+/* Complex number structure */
+
+#ifndef Py_COMPLEXOBJECT_H
+#define Py_COMPLEXOBJECT_H
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/* Complex object interface */
+
+PyAPI_DATA(PyTypeObject) PyComplex_Type;
+
+#define PyComplex_Check(op) PyObject_TypeCheck((op), &PyComplex_Type)
+#define PyComplex_CheckExact(op) Py_IS_TYPE((op), &PyComplex_Type)
+
+PyAPI_FUNC(PyObject *) PyComplex_FromDoubles(double real, double imag);
+
+PyAPI_FUNC(double) PyComplex_RealAsDouble(PyObject *op);
+PyAPI_FUNC(double) PyComplex_ImagAsDouble(PyObject *op);
+
+#ifndef Py_LIMITED_API
+#  define Py_CPYTHON_COMPLEXOBJECT_H
+#  include "cpython/complexobject.h"
+#  undef Py_CPYTHON_COMPLEXOBJECT_H
+#endif
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* !Py_COMPLEXOBJECT_H */

Dependencies/Python/include/cpython/abstract.h

@@ -0,0 +1,87 @@
+#ifndef Py_CPYTHON_ABSTRACTOBJECT_H
+#  error "this header file must not be included directly"
+#endif
+
+/* === Object Protocol ================================================== */
+
+/* Like PyObject_CallMethod(), but expect a _Py_Identifier*
+   as the method name. */
+PyAPI_FUNC(PyObject*) _PyObject_CallMethodId(
+    PyObject *obj,
+    _Py_Identifier *name,
+    const char *format, ...);
+
+/* Convert keyword arguments from the FASTCALL (stack: C array, kwnames: tuple)
+   format to a Python dictionary ("kwargs" dict).
+
+   The type of kwnames keys is not checked. The final function getting
+   arguments is responsible to check if all keys are strings, for example using
+   PyArg_ParseTupleAndKeywords() or PyArg_ValidateKeywordArguments().
+
+   Duplicate keys are merged using the last value. If duplicate keys must raise
+   an exception, the caller is responsible to implement an explicit keys on
+   kwnames. */
+PyAPI_FUNC(PyObject*) _PyStack_AsDict(PyObject *const *values, PyObject *kwnames);
+
+
+/* === Vectorcall protocol (PEP 590) ============================= */
+
+// PyVectorcall_NARGS() is exported as a function for the stable ABI.
+// Here (when we are not using the stable ABI), the name is overridden to
+// call a static inline function for best performance.
+static inline Py_ssize_t
+_PyVectorcall_NARGS(size_t n)
+{
+    return n & ~PY_VECTORCALL_ARGUMENTS_OFFSET;
+}
+#define PyVectorcall_NARGS(n) _PyVectorcall_NARGS(n)
+
+PyAPI_FUNC(vectorcallfunc) PyVectorcall_Function(PyObject *callable);
+
+// Backwards compatibility aliases (PEP 590) for API that was provisional
+// in Python 3.8
+#define _PyObject_Vectorcall PyObject_Vectorcall
+#define _PyObject_VectorcallMethod PyObject_VectorcallMethod
+#define _PyObject_FastCallDict PyObject_VectorcallDict
+#define _PyVectorcall_Function PyVectorcall_Function
+#define _PyObject_CallOneArg PyObject_CallOneArg
+#define _PyObject_CallMethodNoArgs PyObject_CallMethodNoArgs
+#define _PyObject_CallMethodOneArg PyObject_CallMethodOneArg
+
+/* Same as PyObject_Vectorcall except that keyword arguments are passed as
+   dict, which may be NULL if there are no keyword arguments. */
+PyAPI_FUNC(PyObject *) PyObject_VectorcallDict(
+    PyObject *callable,
+    PyObject *const *args,
+    size_t nargsf,
+    PyObject *kwargs);
+
+PyAPI_FUNC(PyObject *) PyObject_CallOneArg(PyObject *func, PyObject *arg);
+
+static inline PyObject *
+PyObject_CallMethodNoArgs(PyObject *self, PyObject *name)
+{
+    size_t nargsf = 1 | PY_VECTORCALL_ARGUMENTS_OFFSET;
+    return PyObject_VectorcallMethod(name, &self, nargsf, _Py_NULL);
+}
+
+static inline PyObject *
+PyObject_CallMethodOneArg(PyObject *self, PyObject *name, PyObject *arg)
+{
+    PyObject *args[2] = {self, arg};
+    size_t nargsf = 2 | PY_VECTORCALL_ARGUMENTS_OFFSET;
+    assert(arg != NULL);
+    return PyObject_VectorcallMethod(name, args, nargsf, _Py_NULL);
+}
+
+/* Guess the size of object 'o' using len(o) or o.__length_hint__().
+   If neither of those return a non-negative value, then return the default
+   value.  If one of the calls fails, this function returns -1. */
+PyAPI_FUNC(Py_ssize_t) PyObject_LengthHint(PyObject *o, Py_ssize_t);
+
+/* === Sequence protocol ================================================ */
+
+/* Assume tp_as_sequence and sq_item exist and that 'i' does not
+   need to be corrected for a negative index. */
+#define PySequence_ITEM(o, i)\
+    ( Py_TYPE(o)->tp_as_sequence->sq_item((o), (i)) )

Dependencies/Python/include/cpython/bytearrayobject.h

@@ -0,0 +1,34 @@
+#ifndef Py_CPYTHON_BYTEARRAYOBJECT_H
+#  error "this header file must not be included directly"
+#endif
+
+/* Object layout */
+typedef struct {
+    PyObject_VAR_HEAD
+    Py_ssize_t ob_alloc;   /* How many bytes allocated in ob_bytes */
+    char *ob_bytes;        /* Physical backing buffer */
+    char *ob_start;        /* Logical start inside ob_bytes */
+    Py_ssize_t ob_exports; /* How many buffer exports */
+} PyByteArrayObject;
+
+PyAPI_DATA(char) _PyByteArray_empty_string[];
+
+/* Macros and static inline functions, trading safety for speed */
+#define _PyByteArray_CAST(op) \
+    (assert(PyByteArray_Check(op)), _Py_CAST(PyByteArrayObject*, op))
+
+static inline char* PyByteArray_AS_STRING(PyObject *op)
+{
+    PyByteArrayObject *self = _PyByteArray_CAST(op);
+    if (Py_SIZE(self)) {
+        return self->ob_start;
+    }
+    return _PyByteArray_empty_string;
+}
+#define PyByteArray_AS_STRING(self) PyByteArray_AS_STRING(_PyObject_CAST(self))
+
+static inline Py_ssize_t PyByteArray_GET_SIZE(PyObject *op) {
+    PyByteArrayObject *self = _PyByteArray_CAST(op);
+    return Py_SIZE(self);
+}
+#define PyByteArray_GET_SIZE(self) PyByteArray_GET_SIZE(_PyObject_CAST(self))

Dependencies/Python/include/cpython/bytesobject.h

@@ -0,0 +1,37 @@
+#ifndef Py_CPYTHON_BYTESOBJECT_H
+#  error "this header file must not be included directly"
+#endif
+
+typedef struct {
+    PyObject_VAR_HEAD
+    Py_DEPRECATED(3.11) Py_hash_t ob_shash;
+    char ob_sval[1];
+
+    /* Invariants:
+     *     ob_sval contains space for 'ob_size+1' elements.
+     *     ob_sval[ob_size] == 0.
+     *     ob_shash is the hash of the byte string or -1 if not computed yet.
+     */
+} PyBytesObject;
+
+PyAPI_FUNC(int) _PyBytes_Resize(PyObject **, Py_ssize_t);
+
+/* Macros and static inline functions, trading safety for speed */
+#define _PyBytes_CAST(op) \
+    (assert(PyBytes_Check(op)), _Py_CAST(PyBytesObject*, op))
+
+static inline char* PyBytes_AS_STRING(PyObject *op)
+{
+    return _PyBytes_CAST(op)->ob_sval;
+}
+#define PyBytes_AS_STRING(op) PyBytes_AS_STRING(_PyObject_CAST(op))
+
+static inline Py_ssize_t PyBytes_GET_SIZE(PyObject *op) {
+    PyBytesObject *self = _PyBytes_CAST(op);
+    return Py_SIZE(self);
+}
+#define PyBytes_GET_SIZE(self) PyBytes_GET_SIZE(_PyObject_CAST(self))
+
+/* _PyBytes_Join(sep, x) is like sep.join(x).  sep must be PyBytesObject*,
+   x must be an iterable object. */
+PyAPI_FUNC(PyObject*) _PyBytes_Join(PyObject *sep, PyObject *x);

Dependencies/Python/include/cpython/cellobject.h

@@ -0,0 +1,44 @@
+/* Cell object interface */
+
+#ifndef Py_LIMITED_API
+#ifndef Py_CELLOBJECT_H
+#define Py_CELLOBJECT_H
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+typedef struct {
+    PyObject_HEAD
+    /* Content of the cell or NULL when empty */
+    PyObject *ob_ref;
+} PyCellObject;
+
+PyAPI_DATA(PyTypeObject) PyCell_Type;
+
+#define PyCell_Check(op) Py_IS_TYPE((op), &PyCell_Type)
+
+PyAPI_FUNC(PyObject *) PyCell_New(PyObject *);
+PyAPI_FUNC(PyObject *) PyCell_Get(PyObject *);
+PyAPI_FUNC(int) PyCell_Set(PyObject *, PyObject *);
+
+static inline PyObject* PyCell_GET(PyObject *op) {
+    PyCellObject *cell;
+    assert(PyCell_Check(op));
+    cell = _Py_CAST(PyCellObject*, op);
+    return cell->ob_ref;
+}
+#define PyCell_GET(op) PyCell_GET(_PyObject_CAST(op))
+
+static inline void PyCell_SET(PyObject *op, PyObject *value) {
+    PyCellObject *cell;
+    assert(PyCell_Check(op));
+    cell = _Py_CAST(PyCellObject*, op);
+    cell->ob_ref = value;
+}
+#define PyCell_SET(op, value) PyCell_SET(_PyObject_CAST(op), (value))
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* !Py_TUPLEOBJECT_H */
+#endif /* Py_LIMITED_API */

Dependencies/Python/include/cpython/ceval.h

@@ -0,0 +1,25 @@
+#ifndef Py_CPYTHON_CEVAL_H
+#  error "this header file must not be included directly"
+#endif
+
+PyAPI_FUNC(void) PyEval_SetProfile(Py_tracefunc, PyObject *);
+PyAPI_FUNC(void) PyEval_SetProfileAllThreads(Py_tracefunc, PyObject *);
+PyAPI_FUNC(void) PyEval_SetTrace(Py_tracefunc, PyObject *);
+PyAPI_FUNC(void) PyEval_SetTraceAllThreads(Py_tracefunc, PyObject *);
+
+/* Look at the current frame's (if any) code's co_flags, and turn on
+   the corresponding compiler flags in cf->cf_flags.  Return 1 if any
+   flag was set, else return 0. */
+PyAPI_FUNC(int) PyEval_MergeCompilerFlags(PyCompilerFlags *cf);
+
+PyAPI_FUNC(PyObject *) _PyEval_EvalFrameDefault(PyThreadState *tstate, struct _PyInterpreterFrame *f, int exc);
+
+PyAPI_FUNC(Py_ssize_t) PyUnstable_Eval_RequestCodeExtraIndex(freefunc);
+// Old name -- remove when this API changes:
+_Py_DEPRECATED_EXTERNALLY(3.12) static inline Py_ssize_t
+_PyEval_RequestCodeExtraIndex(freefunc f) {
+    return PyUnstable_Eval_RequestCodeExtraIndex(f);
+}
+
+PyAPI_FUNC(int) _PyEval_SliceIndex(PyObject *, Py_ssize_t *);
+PyAPI_FUNC(int) _PyEval_SliceIndexNotNone(PyObject *, Py_ssize_t *);

Dependencies/Python/include/cpython/classobject.h

@@ -0,0 +1,71 @@
+/* Former class object interface -- now only bound methods are here  */
+
+/* Revealing some structures (not for general use) */
+
+#ifndef Py_LIMITED_API
+#ifndef Py_CLASSOBJECT_H
+#define Py_CLASSOBJECT_H
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+typedef struct {
+    PyObject_HEAD
+    PyObject *im_func;   /* The callable object implementing the method */
+    PyObject *im_self;   /* The instance it is bound to */
+    PyObject *im_weakreflist; /* List of weak references */
+    vectorcallfunc vectorcall;
+} PyMethodObject;
+
+PyAPI_DATA(PyTypeObject) PyMethod_Type;
+
+#define PyMethod_Check(op) Py_IS_TYPE((op), &PyMethod_Type)
+
+PyAPI_FUNC(PyObject *) PyMethod_New(PyObject *, PyObject *);
+
+PyAPI_FUNC(PyObject *) PyMethod_Function(PyObject *);
+PyAPI_FUNC(PyObject *) PyMethod_Self(PyObject *);
+
+#define _PyMethod_CAST(meth) \
+    (assert(PyMethod_Check(meth)), _Py_CAST(PyMethodObject*, meth))
+
+/* Static inline functions for direct access to these values.
+   Type checks are *not* done, so use with care. */
+static inline PyObject* PyMethod_GET_FUNCTION(PyObject *meth) {
+    return _PyMethod_CAST(meth)->im_func;
+}
+#define PyMethod_GET_FUNCTION(meth) PyMethod_GET_FUNCTION(_PyObject_CAST(meth))
+
+static inline PyObject* PyMethod_GET_SELF(PyObject *meth) {
+    return _PyMethod_CAST(meth)->im_self;
+}
+#define PyMethod_GET_SELF(meth) PyMethod_GET_SELF(_PyObject_CAST(meth))
+
+typedef struct {
+    PyObject_HEAD
+    PyObject *func;
+} PyInstanceMethodObject;
+
+PyAPI_DATA(PyTypeObject) PyInstanceMethod_Type;
+
+#define PyInstanceMethod_Check(op) Py_IS_TYPE((op), &PyInstanceMethod_Type)
+
+PyAPI_FUNC(PyObject *) PyInstanceMethod_New(PyObject *);
+PyAPI_FUNC(PyObject *) PyInstanceMethod_Function(PyObject *);
+
+#define _PyInstanceMethod_CAST(meth) \
+    (assert(PyInstanceMethod_Check(meth)), \
+     _Py_CAST(PyInstanceMethodObject*, meth))
+
+/* Static inline function for direct access to these values.
+   Type checks are *not* done, so use with care. */
+static inline PyObject* PyInstanceMethod_GET_FUNCTION(PyObject *meth) {
+    return _PyInstanceMethod_CAST(meth)->func;
+}
+#define PyInstanceMethod_GET_FUNCTION(meth) PyInstanceMethod_GET_FUNCTION(_PyObject_CAST(meth))
+
+#ifdef __cplusplus
+}
+#endif
+#endif   // !Py_CLASSOBJECT_H
+#endif   // !Py_LIMITED_API

Dependencies/Python/include/cpython/code.h

@@ -0,0 +1,358 @@
+/* Definitions for bytecode */
+
+#ifndef Py_LIMITED_API
+#ifndef Py_CODE_H
+#define Py_CODE_H
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/* Count of all local monitoring events */
+#define  _PY_MONITORING_LOCAL_EVENTS 10
+/* Count of all "real" monitoring events (not derived from other events) */
+#define _PY_MONITORING_UNGROUPED_EVENTS 15
+/* Count of all  monitoring events */
+#define _PY_MONITORING_EVENTS 17
+
+/* Tables of which tools are active for each monitored event. */
+typedef struct _Py_LocalMonitors {
+    uint8_t tools[_PY_MONITORING_LOCAL_EVENTS];
+} _Py_LocalMonitors;
+
+typedef struct _Py_GlobalMonitors {
+    uint8_t tools[_PY_MONITORING_UNGROUPED_EVENTS];
+} _Py_GlobalMonitors;
+
+
+typedef struct {
+    PyObject *_co_code;
+    PyObject *_co_varnames;
+    PyObject *_co_cellvars;
+    PyObject *_co_freevars;
+} _PyCoCached;
+
+/* Ancillary data structure used for instrumentation.
+   Line instrumentation creates this with sufficient
+   space for one entry per code unit. The total size
+   of the data will be `bytes_per_entry * Py_SIZE(code)` */
+typedef struct {
+    uint8_t bytes_per_entry;
+    uint8_t data[1];
+} _PyCoLineInstrumentationData;
+
+
+typedef struct {
+    int size;
+    int capacity;
+    struct _PyExecutorObject *executors[1];
+} _PyExecutorArray;
+
+/* Main data structure used for instrumentation.
+ * This is allocated when needed for instrumentation
+ */
+typedef struct {
+    /* Monitoring specific to this code object */
+    _Py_LocalMonitors local_monitors;
+    /* Monitoring that is active on this code object */
+    _Py_LocalMonitors active_monitors;
+    /* The tools that are to be notified for events for the matching code unit */
+    uint8_t *tools;
+    /* Information to support line events */
+    _PyCoLineInstrumentationData *lines;
+    /* The tools that are to be notified for line events for the matching code unit */
+    uint8_t *line_tools;
+    /* Information to support instruction events */
+    /* The underlying instructions, which can themselves be instrumented */
+    uint8_t *per_instruction_opcodes;
+    /* The tools that are to be notified for instruction events for the matching code unit */
+    uint8_t *per_instruction_tools;
+} _PyCoMonitoringData;
+
+// To avoid repeating ourselves in deepfreeze.py, all PyCodeObject members are
+// defined in this macro:
+#define _PyCode_DEF(SIZE) {                                                    \
+    PyObject_VAR_HEAD                                                          \
+                                                                               \
+    /* Note only the following fields are used in hash and/or comparisons      \
+     *                                                                         \
+     * - co_name                                                               \
+     * - co_argcount                                                           \
+     * - co_posonlyargcount                                                    \
+     * - co_kwonlyargcount                                                     \
+     * - co_nlocals                                                            \
+     * - co_stacksize                                                          \
+     * - co_flags                                                              \
+     * - co_firstlineno                                                        \
+     * - co_consts                                                             \
+     * - co_names                                                              \
+     * - co_localsplusnames                                                    \
+     * This is done to preserve the name and line number for tracebacks        \
+     * and debuggers; otherwise, constant de-duplication would collapse        \
+     * identical functions/lambdas defined on different lines.                 \
+     */                                                                        \
+                                                                               \
+    /* These fields are set with provided values on new code objects. */       \
+                                                                               \
+    /* The hottest fields (in the eval loop) are grouped here at the top. */   \
+    PyObject *co_consts;           /* list (constants used) */                 \
+    PyObject *co_names;            /* list of strings (names used) */          \
+    PyObject *co_exceptiontable;   /* Byte string encoding exception handling  \
+                                      table */                                 \
+    int co_flags;                  /* CO_..., see below */                     \
+                                                                               \
+    /* The rest are not so impactful on performance. */                        \
+    int co_argcount;              /* #arguments, except *args */               \
+    int co_posonlyargcount;       /* #positional only arguments */             \
+    int co_kwonlyargcount;        /* #keyword only arguments */                \
+    int co_stacksize;             /* #entries needed for evaluation stack */   \
+    int co_firstlineno;           /* first source line number */               \
+                                                                               \
+    /* redundant values (derived from co_localsplusnames and                   \
+       co_localspluskinds) */                                                  \
+    int co_nlocalsplus;           /* number of local + cell + free variables */ \
+    int co_framesize;             /* Size of frame in words */                 \
+    int co_nlocals;               /* number of local variables */              \
+    int co_ncellvars;             /* total number of cell variables */         \
+    int co_nfreevars;             /* number of free variables */               \
+    uint32_t co_version;          /* version number */                         \
+                                                                               \
+    PyObject *co_localsplusnames; /* tuple mapping offsets to names */         \
+    PyObject *co_localspluskinds; /* Bytes mapping to local kinds (one byte    \
+                                     per variable) */                          \
+    PyObject *co_filename;        /* unicode (where it was loaded from) */     \
+    PyObject *co_name;            /* unicode (name, for reference) */          \
+    PyObject *co_qualname;        /* unicode (qualname, for reference) */      \
+    PyObject *co_linetable;       /* bytes object that holds location info */  \
+    PyObject *co_weakreflist;     /* to support weakrefs to code objects */    \
+    _PyExecutorArray *co_executors;      /* executors from optimizer */        \
+    _PyCoCached *_co_cached;      /* cached co_* attributes */                 \
+    uintptr_t _co_instrumentation_version; /* current instrumentation version */ \
+    _PyCoMonitoringData *_co_monitoring; /* Monitoring data */                 \
+    int _co_firsttraceable;       /* index of first traceable instruction */   \
+    /* Scratch space for extra data relating to the code object.               \
+       Type is a void* to keep the format private in codeobject.c to force     \
+       people to go through the proper APIs. */                                \
+    void *co_extra;                                                            \
+    char co_code_adaptive[(SIZE)];                                             \
+}
+
+/* Bytecode object */
+struct PyCodeObject _PyCode_DEF(1);
+
+/* Masks for co_flags above */
+#define CO_OPTIMIZED    0x0001
+#define CO_NEWLOCALS    0x0002
+#define CO_VARARGS      0x0004
+#define CO_VARKEYWORDS  0x0008
+#define CO_NESTED       0x0010
+#define CO_GENERATOR    0x0020
+
+/* The CO_COROUTINE flag is set for coroutine functions (defined with
+   ``async def`` keywords) */
+#define CO_COROUTINE            0x0080
+#define CO_ITERABLE_COROUTINE   0x0100
+#define CO_ASYNC_GENERATOR      0x0200
+
+/* bpo-39562: These constant values are changed in Python 3.9
+   to prevent collision with compiler flags. CO_FUTURE_ and PyCF_
+   constants must be kept unique. PyCF_ constants can use bits from
+   0x0100 to 0x10000. CO_FUTURE_ constants use bits starting at 0x20000. */
+#define CO_FUTURE_DIVISION      0x20000
+#define CO_FUTURE_ABSOLUTE_IMPORT 0x40000 /* do absolute imports by default */
+#define CO_FUTURE_WITH_STATEMENT  0x80000
+#define CO_FUTURE_PRINT_FUNCTION  0x100000
+#define CO_FUTURE_UNICODE_LITERALS 0x200000
+
+#define CO_FUTURE_BARRY_AS_BDFL  0x400000
+#define CO_FUTURE_GENERATOR_STOP  0x800000
+#define CO_FUTURE_ANNOTATIONS    0x1000000
+
+#define CO_NO_MONITORING_EVENTS 0x2000000
+
+/* This should be defined if a future statement modifies the syntax.
+   For example, when a keyword is added.
+*/
+#define PY_PARSER_REQUIRES_FUTURE_KEYWORD
+
+#define CO_MAXBLOCKS 21 /* Max static block nesting within a function */
+
+PyAPI_DATA(PyTypeObject) PyCode_Type;
+
+#define PyCode_Check(op) Py_IS_TYPE((op), &PyCode_Type)
+
+static inline Py_ssize_t PyCode_GetNumFree(PyCodeObject *op) {
+    assert(PyCode_Check(op));
+    return op->co_nfreevars;
+}
+
+static inline int PyUnstable_Code_GetFirstFree(PyCodeObject *op) {
+    assert(PyCode_Check(op));
+    return op->co_nlocalsplus - op->co_nfreevars;
+}
+
+Py_DEPRECATED(3.13) static inline int PyCode_GetFirstFree(PyCodeObject *op) {
+    return PyUnstable_Code_GetFirstFree(op);
+}
+
+/* Unstable public interface */
+PyAPI_FUNC(PyCodeObject *) PyUnstable_Code_New(
+        int, int, int, int, int, PyObject *, PyObject *,
+        PyObject *, PyObject *, PyObject *, PyObject *,
+        PyObject *, PyObject *, PyObject *, int, PyObject *,
+        PyObject *);
+
+PyAPI_FUNC(PyCodeObject *) PyUnstable_Code_NewWithPosOnlyArgs(
+        int, int, int, int, int, int, PyObject *, PyObject *,
+        PyObject *, PyObject *, PyObject *, PyObject *,
+        PyObject *, PyObject *, PyObject *, int, PyObject *,
+        PyObject *);
+        /* same as struct above */
+// Old names -- remove when this API changes:
+_Py_DEPRECATED_EXTERNALLY(3.12) static inline PyCodeObject *
+PyCode_New(
+        int a, int b, int c, int d, int e, PyObject *f, PyObject *g,
+        PyObject *h, PyObject *i, PyObject *j, PyObject *k,
+        PyObject *l, PyObject *m, PyObject *n, int o, PyObject *p,
+        PyObject *q)
+{
+    return PyUnstable_Code_New(
+        a, b, c, d, e, f, g, h, i, j, k, l, m, n, o, p, q);
+}
+_Py_DEPRECATED_EXTERNALLY(3.12) static inline PyCodeObject *
+PyCode_NewWithPosOnlyArgs(
+        int a, int poac, int b, int c, int d, int e, PyObject *f, PyObject *g,
+        PyObject *h, PyObject *i, PyObject *j, PyObject *k,
+        PyObject *l, PyObject *m, PyObject *n, int o, PyObject *p,
+        PyObject *q)
+{
+    return PyUnstable_Code_NewWithPosOnlyArgs(
+        a, poac, b, c, d, e, f, g, h, i, j, k, l, m, n, o, p, q);
+}
+
+/* Creates a new empty code object with the specified source location. */
+PyAPI_FUNC(PyCodeObject *)
+PyCode_NewEmpty(const char *filename, const char *funcname, int firstlineno);
+
+/* Return the line number associated with the specified bytecode index
+   in this code object.  If you just need the line number of a frame,
+   use PyFrame_GetLineNumber() instead. */
+PyAPI_FUNC(int) PyCode_Addr2Line(PyCodeObject *, int);
+
+PyAPI_FUNC(int) PyCode_Addr2Location(PyCodeObject *, int, int *, int *, int *, int *);
+
+#define PY_FOREACH_CODE_EVENT(V) \
+    V(CREATE)                 \
+    V(DESTROY)
+
+typedef enum {
+    #define PY_DEF_EVENT(op) PY_CODE_EVENT_##op,
+    PY_FOREACH_CODE_EVENT(PY_DEF_EVENT)
+    #undef PY_DEF_EVENT
+} PyCodeEvent;
+
+
+/*
+ * A callback that is invoked for different events in a code object's lifecycle.
+ *
+ * The callback is invoked with a borrowed reference to co, after it is
+ * created and before it is destroyed.
+ *
+ * If the callback sets an exception, it must return -1. Otherwise
+ * it should return 0.
+ */
+typedef int (*PyCode_WatchCallback)(
+  PyCodeEvent event,
+  PyCodeObject* co);
+
+/*
+ * Register a per-interpreter callback that will be invoked for code object
+ * lifecycle events.
+ *
+ * Returns a handle that may be passed to PyCode_ClearWatcher on success,
+ * or -1 and sets an error if no more handles are available.
+ */
+PyAPI_FUNC(int) PyCode_AddWatcher(PyCode_WatchCallback callback);
+
+/*
+ * Clear the watcher associated with the watcher_id handle.
+ *
+ * Returns 0 on success or -1 if no watcher exists for the provided id.
+ */
+PyAPI_FUNC(int) PyCode_ClearWatcher(int watcher_id);
+
+/* for internal use only */
+struct _opaque {
+    int computed_line;
+    const uint8_t *lo_next;
+    const uint8_t *limit;
+};
+
+typedef struct _line_offsets {
+    int ar_start;
+    int ar_end;
+    int ar_line;
+    struct _opaque opaque;
+} PyCodeAddressRange;
+
+/* Update *bounds to describe the first and one-past-the-last instructions in the
+   same line as lasti.  Return the number of that line.
+*/
+PyAPI_FUNC(int) _PyCode_CheckLineNumber(int lasti, PyCodeAddressRange *bounds);
+
+/* Create a comparable key used to compare constants taking in account the
+ * object type. It is used to make sure types are not coerced (e.g., float and
+ * complex) _and_ to distinguish 0.0 from -0.0 e.g. on IEEE platforms
+ *
+ * Return (type(obj), obj, ...): a tuple with variable size (at least 2 items)
+ * depending on the type and the value. The type is the first item to not
+ * compare bytes and str which can raise a BytesWarning exception. */
+PyAPI_FUNC(PyObject*) _PyCode_ConstantKey(PyObject *obj);
+
+PyAPI_FUNC(PyObject*) PyCode_Optimize(PyObject *code, PyObject* consts,
+                                      PyObject *names, PyObject *lnotab);
+
+PyAPI_FUNC(int) PyUnstable_Code_GetExtra(
+    PyObject *code, Py_ssize_t index, void **extra);
+PyAPI_FUNC(int) PyUnstable_Code_SetExtra(
+    PyObject *code, Py_ssize_t index, void *extra);
+// Old names -- remove when this API changes:
+_Py_DEPRECATED_EXTERNALLY(3.12) static inline int
+_PyCode_GetExtra(PyObject *code, Py_ssize_t index, void **extra)
+{
+    return PyUnstable_Code_GetExtra(code, index, extra);
+}
+_Py_DEPRECATED_EXTERNALLY(3.12) static inline int
+_PyCode_SetExtra(PyObject *code, Py_ssize_t index, void *extra)
+{
+    return PyUnstable_Code_SetExtra(code, index, extra);
+}
+
+/* Equivalent to getattr(code, 'co_code') in Python.
+   Returns a strong reference to a bytes object. */
+PyAPI_FUNC(PyObject *) PyCode_GetCode(PyCodeObject *code);
+/* Equivalent to getattr(code, 'co_varnames') in Python. */
+PyAPI_FUNC(PyObject *) PyCode_GetVarnames(PyCodeObject *code);
+/* Equivalent to getattr(code, 'co_cellvars') in Python. */
+PyAPI_FUNC(PyObject *) PyCode_GetCellvars(PyCodeObject *code);
+/* Equivalent to getattr(code, 'co_freevars') in Python. */
+PyAPI_FUNC(PyObject *) PyCode_GetFreevars(PyCodeObject *code);
+
+typedef enum _PyCodeLocationInfoKind {
+    /* short forms are 0 to 9 */
+    PY_CODE_LOCATION_INFO_SHORT0 = 0,
+    /* one lineforms are 10 to 12 */
+    PY_CODE_LOCATION_INFO_ONE_LINE0 = 10,
+    PY_CODE_LOCATION_INFO_ONE_LINE1 = 11,
+    PY_CODE_LOCATION_INFO_ONE_LINE2 = 12,
+
+    PY_CODE_LOCATION_INFO_NO_COLUMNS = 13,
+    PY_CODE_LOCATION_INFO_LONG = 14,
+    PY_CODE_LOCATION_INFO_NONE = 15
+} _PyCodeLocationInfoKind;
+
+#ifdef __cplusplus
+}
+#endif
+#endif  // !Py_CODE_H
+#endif  // !Py_LIMITED_API

Dependencies/Python/include/cpython/compile.h

@@ -0,0 +1,50 @@
+#ifndef Py_CPYTHON_COMPILE_H
+#  error "this header file must not be included directly"
+#endif
+
+/* Public interface */
+#define PyCF_MASK (CO_FUTURE_DIVISION | CO_FUTURE_ABSOLUTE_IMPORT | \
+                   CO_FUTURE_WITH_STATEMENT | CO_FUTURE_PRINT_FUNCTION | \
+                   CO_FUTURE_UNICODE_LITERALS | CO_FUTURE_BARRY_AS_BDFL | \
+                   CO_FUTURE_GENERATOR_STOP | CO_FUTURE_ANNOTATIONS)
+#define PyCF_MASK_OBSOLETE (CO_NESTED)
+
+/* bpo-39562: CO_FUTURE_ and PyCF_ constants must be kept unique.
+   PyCF_ constants can use bits from 0x0100 to 0x10000.
+   CO_FUTURE_ constants use bits starting at 0x20000. */
+#define PyCF_SOURCE_IS_UTF8  0x0100
+#define PyCF_DONT_IMPLY_DEDENT 0x0200
+#define PyCF_ONLY_AST 0x0400
+#define PyCF_IGNORE_COOKIE 0x0800
+#define PyCF_TYPE_COMMENTS 0x1000
+#define PyCF_ALLOW_TOP_LEVEL_AWAIT 0x2000
+#define PyCF_ALLOW_INCOMPLETE_INPUT 0x4000
+#define PyCF_OPTIMIZED_AST (0x8000 | PyCF_ONLY_AST)
+#define PyCF_COMPILE_MASK (PyCF_ONLY_AST | PyCF_ALLOW_TOP_LEVEL_AWAIT | \
+                           PyCF_TYPE_COMMENTS | PyCF_DONT_IMPLY_DEDENT | \
+                           PyCF_ALLOW_INCOMPLETE_INPUT | PyCF_OPTIMIZED_AST)
+
+typedef struct {
+    int cf_flags;  /* bitmask of CO_xxx flags relevant to future */
+    int cf_feature_version;  /* minor Python version (PyCF_ONLY_AST) */
+} PyCompilerFlags;
+
+#define _PyCompilerFlags_INIT \
+    (PyCompilerFlags){.cf_flags = 0, .cf_feature_version = PY_MINOR_VERSION}
+
+/* Future feature support */
+
+#define FUTURE_NESTED_SCOPES "nested_scopes"
+#define FUTURE_GENERATORS "generators"
+#define FUTURE_DIVISION "division"
+#define FUTURE_ABSOLUTE_IMPORT "absolute_import"
+#define FUTURE_WITH_STATEMENT "with_statement"
+#define FUTURE_PRINT_FUNCTION "print_function"
+#define FUTURE_UNICODE_LITERALS "unicode_literals"
+#define FUTURE_BARRY_AS_BDFL "barry_as_FLUFL"
+#define FUTURE_GENERATOR_STOP "generator_stop"
+#define FUTURE_ANNOTATIONS "annotations"
+
+#define PY_INVALID_STACK_EFFECT INT_MAX
+PyAPI_FUNC(int) PyCompile_OpcodeStackEffect(int opcode, int oparg);
+PyAPI_FUNC(int) PyCompile_OpcodeStackEffectWithJump(int opcode, int oparg, int jump);

Dependencies/Python/include/cpython/complexobject.h

@@ -0,0 +1,33 @@
+#ifndef Py_CPYTHON_COMPLEXOBJECT_H
+#  error "this header file must not be included directly"
+#endif
+
+typedef struct {
+    double real;
+    double imag;
+} Py_complex;
+
+// Operations on complex numbers.
+PyAPI_FUNC(Py_complex) _Py_c_sum(Py_complex, Py_complex);
+PyAPI_FUNC(Py_complex) _Py_c_diff(Py_complex, Py_complex);
+PyAPI_FUNC(Py_complex) _Py_c_neg(Py_complex);
+PyAPI_FUNC(Py_complex) _Py_c_prod(Py_complex, Py_complex);
+PyAPI_FUNC(Py_complex) _Py_c_quot(Py_complex, Py_complex);
+PyAPI_FUNC(Py_complex) _Py_c_pow(Py_complex, Py_complex);
+PyAPI_FUNC(double) _Py_c_abs(Py_complex);
+
+
+/* Complex object interface */
+
+/*
+PyComplexObject represents a complex number with double-precision
+real and imaginary parts.
+*/
+typedef struct {
+    PyObject_HEAD
+    Py_complex cval;
+} PyComplexObject;
+
+PyAPI_FUNC(PyObject *) PyComplex_FromCComplex(Py_complex);
+
+PyAPI_FUNC(Py_complex) PyComplex_AsCComplex(PyObject *op);

Dependencies/Python/include/cpython/context.h

@@ -0,0 +1,74 @@
+#ifndef Py_LIMITED_API
+#ifndef Py_CONTEXT_H
+#define Py_CONTEXT_H
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+PyAPI_DATA(PyTypeObject) PyContext_Type;
+typedef struct _pycontextobject PyContext;
+
+PyAPI_DATA(PyTypeObject) PyContextVar_Type;
+typedef struct _pycontextvarobject PyContextVar;
+
+PyAPI_DATA(PyTypeObject) PyContextToken_Type;
+typedef struct _pycontexttokenobject PyContextToken;
+
+
+#define PyContext_CheckExact(o) Py_IS_TYPE((o), &PyContext_Type)
+#define PyContextVar_CheckExact(o) Py_IS_TYPE((o), &PyContextVar_Type)
+#define PyContextToken_CheckExact(o) Py_IS_TYPE((o), &PyContextToken_Type)
+
+
+PyAPI_FUNC(PyObject *) PyContext_New(void);
+PyAPI_FUNC(PyObject *) PyContext_Copy(PyObject *);
+PyAPI_FUNC(PyObject *) PyContext_CopyCurrent(void);
+
+PyAPI_FUNC(int) PyContext_Enter(PyObject *);
+PyAPI_FUNC(int) PyContext_Exit(PyObject *);
+
+
+/* Create a new context variable.
+
+   default_value can be NULL.
+*/
+PyAPI_FUNC(PyObject *) PyContextVar_New(
+    const char *name, PyObject *default_value);
+
+
+/* Get a value for the variable.
+
+   Returns -1 if an error occurred during lookup.
+
+   Returns 0 if value either was or was not found.
+
+   If value was found, *value will point to it.
+   If not, it will point to:
+
+   - default_value, if not NULL;
+   - the default value of "var", if not NULL;
+   - NULL.
+
+   '*value' will be a new ref, if not NULL.
+*/
+PyAPI_FUNC(int) PyContextVar_Get(
+    PyObject *var, PyObject *default_value, PyObject **value);
+
+
+/* Set a new value for the variable.
+   Returns NULL if an error occurs.
+*/
+PyAPI_FUNC(PyObject *) PyContextVar_Set(PyObject *var, PyObject *value);
+
+
+/* Reset a variable to its previous value.
+   Returns 0 on success, -1 on error.
+*/
+PyAPI_FUNC(int) PyContextVar_Reset(PyObject *var, PyObject *token);
+
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* !Py_CONTEXT_H */
+#endif /* !Py_LIMITED_API */

Dependencies/Python/include/cpython/critical_section.h

@@ -0,0 +1,134 @@
+#ifndef Py_CPYTHON_CRITICAL_SECTION_H
+#  error "this header file must not be included directly"
+#endif
+
+// Python critical sections
+//
+// Conceptually, critical sections are a deadlock avoidance layer on top of
+// per-object locks. These helpers, in combination with those locks, replace
+// our usage of the global interpreter lock to provide thread-safety for
+// otherwise thread-unsafe objects, such as dict.
+//
+// NOTE: These APIs are no-ops in non-free-threaded builds.
+//
+// Straightforward per-object locking could introduce deadlocks that were not
+// present when running with the GIL. Threads may hold locks for multiple
+// objects simultaneously because Python operations can nest. If threads were
+// to acquire the same locks in different orders, they would deadlock.
+//
+// One way to avoid deadlocks is to allow threads to hold only the lock (or
+// locks) for a single operation at a time (typically a single lock, but some
+// operations involve two locks). When a thread begins a nested operation it
+// could suspend the locks for any outer operation: before beginning the nested
+// operation, the locks for the outer operation are released and when the
+// nested operation completes, the locks for the outer operation are
+// reacquired.
+//
+// To improve performance, this API uses a variation of the above scheme.
+// Instead of immediately suspending locks any time a nested operation begins,
+// locks are only suspended if the thread would block. This reduces the number
+// of lock acquisitions and releases for nested operations, while still
+// avoiding deadlocks.
+//
+// Additionally, the locks for any active operation are suspended around
+// other potentially blocking operations, such as I/O. This is because the
+// interaction between locks and blocking operations can lead to deadlocks in
+// the same way as the interaction between multiple locks.
+//
+// Each thread's critical sections and their corresponding locks are tracked in
+// a stack in `PyThreadState.critical_section`. When a thread calls
+// `_PyThreadState_Detach()`, such as before a blocking I/O operation or when
+// waiting to acquire a lock, the thread suspends all of its active critical
+// sections, temporarily releasing the associated locks. When the thread calls
+// `_PyThreadState_Attach()`, it resumes the top-most (i.e., most recent)
+// critical section by reacquiring the associated lock or locks.  See
+// `_PyCriticalSection_Resume()`.
+//
+// NOTE: Only the top-most critical section is guaranteed to be active.
+// Operations that need to lock two objects at once must use
+// `Py_BEGIN_CRITICAL_SECTION2()`. You *CANNOT* use nested critical sections
+// to lock more than one object at once, because the inner critical section
+// may  suspend the outer critical sections. This API does not provide a way
+// to lock more than two objects at once (though it could be added later
+// if actually needed).
+//
+// NOTE: Critical sections implicitly behave like reentrant locks because
+// attempting to acquire the same lock will suspend any outer (earlier)
+// critical sections. However, they are less efficient for this use case than
+// purposefully designed reentrant locks.
+//
+// Example usage:
+//  Py_BEGIN_CRITICAL_SECTION(op);
+//  ...
+//  Py_END_CRITICAL_SECTION();
+//
+// To lock two objects at once:
+//  Py_BEGIN_CRITICAL_SECTION2(op1, op2);
+//  ...
+//  Py_END_CRITICAL_SECTION2();
+
+typedef struct PyCriticalSection PyCriticalSection;
+typedef struct PyCriticalSection2 PyCriticalSection2;
+
+PyAPI_FUNC(void)
+PyCriticalSection_Begin(PyCriticalSection *c, PyObject *op);
+
+PyAPI_FUNC(void)
+PyCriticalSection_End(PyCriticalSection *c);
+
+PyAPI_FUNC(void)
+PyCriticalSection2_Begin(PyCriticalSection2 *c, PyObject *a, PyObject *b);
+
+PyAPI_FUNC(void)
+PyCriticalSection2_End(PyCriticalSection2 *c);
+
+#ifndef Py_GIL_DISABLED
+# define Py_BEGIN_CRITICAL_SECTION(op)      \
+    {
+# define Py_END_CRITICAL_SECTION()          \
+    }
+# define Py_BEGIN_CRITICAL_SECTION2(a, b)   \
+    {
+# define Py_END_CRITICAL_SECTION2()         \
+    }
+#else /* !Py_GIL_DISABLED */
+
+// NOTE: the contents of this struct are private and may change betweeen
+// Python releases without a deprecation period.
+struct PyCriticalSection {
+    // Tagged pointer to an outer active critical section (or 0).
+    uintptr_t _cs_prev;
+
+    // Mutex used to protect critical section
+    PyMutex *_cs_mutex;
+};
+
+// A critical section protected by two mutexes. Use
+// Py_BEGIN_CRITICAL_SECTION2 and Py_END_CRITICAL_SECTION2.
+// NOTE: the contents of this struct are private and may change betweeen
+// Python releases without a deprecation period.
+struct PyCriticalSection2 {
+    PyCriticalSection _cs_base;
+
+    PyMutex *_cs_mutex2;
+};
+
+# define Py_BEGIN_CRITICAL_SECTION(op)                                  \
+    {                                                                   \
+        PyCriticalSection _py_cs;                                       \
+        PyCriticalSection_Begin(&_py_cs, _PyObject_CAST(op))
+
+# define Py_END_CRITICAL_SECTION()                                      \
+        PyCriticalSection_End(&_py_cs);                                 \
+    }
+
+# define Py_BEGIN_CRITICAL_SECTION2(a, b)                               \
+    {                                                                   \
+        PyCriticalSection2 _py_cs2;                                     \
+        PyCriticalSection2_Begin(&_py_cs2, _PyObject_CAST(a), _PyObject_CAST(b))
+
+# define Py_END_CRITICAL_SECTION2()                                     \
+        PyCriticalSection2_End(&_py_cs2);                               \
+    }
+
+#endif

Dependencies/Python/include/cpython/descrobject.h

@@ -0,0 +1,62 @@
+#ifndef Py_CPYTHON_DESCROBJECT_H
+#  error "this header file must not be included directly"
+#endif
+
+typedef PyObject *(*wrapperfunc)(PyObject *self, PyObject *args,
+                                 void *wrapped);
+
+typedef PyObject *(*wrapperfunc_kwds)(PyObject *self, PyObject *args,
+                                      void *wrapped, PyObject *kwds);
+
+struct wrapperbase {
+    const char *name;
+    int offset;
+    void *function;
+    wrapperfunc wrapper;
+    const char *doc;
+    int flags;
+    PyObject *name_strobj;
+};
+
+/* Flags for above struct */
+#define PyWrapperFlag_KEYWORDS 1 /* wrapper function takes keyword args */
+
+/* Various kinds of descriptor objects */
+
+typedef struct {
+    PyObject_HEAD
+    PyTypeObject *d_type;
+    PyObject *d_name;
+    PyObject *d_qualname;
+} PyDescrObject;
+
+#define PyDescr_COMMON PyDescrObject d_common
+
+#define PyDescr_TYPE(x) (((PyDescrObject *)(x))->d_type)
+#define PyDescr_NAME(x) (((PyDescrObject *)(x))->d_name)
+
+typedef struct {
+    PyDescr_COMMON;
+    PyMethodDef *d_method;
+    vectorcallfunc vectorcall;
+} PyMethodDescrObject;
+
+typedef struct {
+    PyDescr_COMMON;
+    PyMemberDef *d_member;
+} PyMemberDescrObject;
+
+typedef struct {
+    PyDescr_COMMON;
+    PyGetSetDef *d_getset;
+} PyGetSetDescrObject;
+
+typedef struct {
+    PyDescr_COMMON;
+    struct wrapperbase *d_base;
+    void *d_wrapped; /* This can be any function pointer */
+} PyWrapperDescrObject;
+
+PyAPI_FUNC(PyObject *) PyDescr_NewWrapper(PyTypeObject *,
+                                                struct wrapperbase *, void *);
+PyAPI_FUNC(int) PyDescr_IsData(PyObject *);

Dependencies/Python/include/cpython/dictobject.h

@@ -0,0 +1,102 @@
+#ifndef Py_CPYTHON_DICTOBJECT_H
+#  error "this header file must not be included directly"
+#endif
+
+typedef struct _dictkeysobject PyDictKeysObject;
+typedef struct _dictvalues PyDictValues;
+
+/* The ma_values pointer is NULL for a combined table
+ * or points to an array of PyObject* for a split table
+ */
+typedef struct {
+    PyObject_HEAD
+
+    /* Number of items in the dictionary */
+    Py_ssize_t ma_used;
+
+    /* Dictionary version: globally unique, value change each time
+       the dictionary is modified */
+#ifdef Py_BUILD_CORE
+    /* Bits 0-7 are for dict watchers.
+     * Bits 8-11 are for the watched mutation counter (used by tier2 optimization)
+     * The remaining bits (12-63) are the actual version tag. */
+    uint64_t ma_version_tag;
+#else
+    Py_DEPRECATED(3.12) uint64_t ma_version_tag;
+#endif
+
+    PyDictKeysObject *ma_keys;
+
+    /* If ma_values is NULL, the table is "combined": keys and values
+       are stored in ma_keys.
+
+       If ma_values is not NULL, the table is split:
+       keys are stored in ma_keys and values are stored in ma_values */
+    PyDictValues *ma_values;
+} PyDictObject;
+
+PyAPI_FUNC(PyObject *) _PyDict_GetItem_KnownHash(PyObject *mp, PyObject *key,
+                                                 Py_hash_t hash);
+PyAPI_FUNC(PyObject *) _PyDict_GetItemStringWithError(PyObject *, const char *);
+PyAPI_FUNC(PyObject *) PyDict_SetDefault(
+    PyObject *mp, PyObject *key, PyObject *defaultobj);
+
+// Inserts `key` with a value `default_value`, if `key` is not already present
+// in the dictionary.  If `result` is not NULL, then the value associated
+// with `key` is returned in `*result` (either the existing value, or the now
+// inserted `default_value`).
+// Returns:
+//   -1 on error
+//    0 if `key` was not present and `default_value` was inserted
+//    1 if `key` was present and `default_value` was not inserted
+PyAPI_FUNC(int) PyDict_SetDefaultRef(PyObject *mp, PyObject *key, PyObject *default_value, PyObject **result);
+
+/* Get the number of items of a dictionary. */
+static inline Py_ssize_t PyDict_GET_SIZE(PyObject *op) {
+    PyDictObject *mp;
+    assert(PyDict_Check(op));
+    mp = _Py_CAST(PyDictObject*, op);
+#ifdef Py_GIL_DISABLED
+    return _Py_atomic_load_ssize_relaxed(&mp->ma_used);
+#else
+    return mp->ma_used;
+#endif
+}
+#define PyDict_GET_SIZE(op) PyDict_GET_SIZE(_PyObject_CAST(op))
+
+PyAPI_FUNC(int) PyDict_ContainsString(PyObject *mp, const char *key);
+
+PyAPI_FUNC(PyObject *) _PyDict_NewPresized(Py_ssize_t minused);
+
+PyAPI_FUNC(int) PyDict_Pop(PyObject *dict, PyObject *key, PyObject **result);
+PyAPI_FUNC(int) PyDict_PopString(PyObject *dict, const char *key, PyObject **result);
+PyAPI_FUNC(PyObject *) _PyDict_Pop(PyObject *dict, PyObject *key, PyObject *default_value);
+
+/* Dictionary watchers */
+
+#define PY_FOREACH_DICT_EVENT(V) \
+    V(ADDED)                     \
+    V(MODIFIED)                  \
+    V(DELETED)                   \
+    V(CLONED)                    \
+    V(CLEARED)                   \
+    V(DEALLOCATED)
+
+typedef enum {
+    #define PY_DEF_EVENT(EVENT) PyDict_EVENT_##EVENT,
+    PY_FOREACH_DICT_EVENT(PY_DEF_EVENT)
+    #undef PY_DEF_EVENT
+} PyDict_WatchEvent;
+
+// Callback to be invoked when a watched dict is cleared, dealloced, or modified.
+// In clear/dealloc case, key and new_value will be NULL. Otherwise, new_value will be the
+// new value for key, NULL if key is being deleted.
+typedef int(*PyDict_WatchCallback)(PyDict_WatchEvent event, PyObject* dict, PyObject* key, PyObject* new_value);
+
+// Register/unregister a dict-watcher callback
+PyAPI_FUNC(int) PyDict_AddWatcher(PyDict_WatchCallback callback);
+PyAPI_FUNC(int) PyDict_ClearWatcher(int watcher_id);
+
+// Mark given dictionary as "watched" (callback will be called if it is modified)
+PyAPI_FUNC(int) PyDict_Watch(int watcher_id, PyObject* dict);
+PyAPI_FUNC(int) PyDict_Unwatch(int watcher_id, PyObject* dict);

Dependencies/Python/include/cpython/fileobject.h

@@ -0,0 +1,16 @@
+#ifndef Py_CPYTHON_FILEOBJECT_H
+#  error "this header file must not be included directly"
+#endif
+
+PyAPI_FUNC(char *) Py_UniversalNewlineFgets(char *, int, FILE*, PyObject *);
+
+/* The std printer acts as a preliminary sys.stderr until the new io
+   infrastructure is in place. */
+PyAPI_FUNC(PyObject *) PyFile_NewStdPrinter(int);
+PyAPI_DATA(PyTypeObject) PyStdPrinter_Type;
+
+typedef PyObject * (*Py_OpenCodeHookFunction)(PyObject *, void *);
+
+PyAPI_FUNC(PyObject *) PyFile_OpenCode(const char *utf8path);
+PyAPI_FUNC(PyObject *) PyFile_OpenCodeObject(PyObject *path);
+PyAPI_FUNC(int) PyFile_SetOpenCodeHook(Py_OpenCodeHookFunction hook, void *userData);

Dependencies/Python/include/cpython/fileutils.h

@@ -0,0 +1,8 @@
+#ifndef Py_CPYTHON_FILEUTILS_H
+#  error "this header file must not be included directly"
+#endif
+
+// Used by _testcapi which must not use the internal C API
+PyAPI_FUNC(FILE*) _Py_fopen_obj(
+    PyObject *path,
+    const char *mode);

Dependencies/Python/include/cpython/floatobject.h

@@ -0,0 +1,27 @@
+#ifndef Py_CPYTHON_FLOATOBJECT_H
+#  error "this header file must not be included directly"
+#endif
+
+typedef struct {
+    PyObject_HEAD
+    double ob_fval;
+} PyFloatObject;
+
+#define _PyFloat_CAST(op) \
+    (assert(PyFloat_Check(op)), _Py_CAST(PyFloatObject*, op))
+
+// Static inline version of PyFloat_AsDouble() trading safety for speed.
+// It doesn't check if op is a double object.
+static inline double PyFloat_AS_DOUBLE(PyObject *op) {
+    return _PyFloat_CAST(op)->ob_fval;
+}
+#define PyFloat_AS_DOUBLE(op) PyFloat_AS_DOUBLE(_PyObject_CAST(op))
+
+
+PyAPI_FUNC(int) PyFloat_Pack2(double x, char *p, int le);
+PyAPI_FUNC(int) PyFloat_Pack4(double x, char *p, int le);
+PyAPI_FUNC(int) PyFloat_Pack8(double x, char *p, int le);
+
+PyAPI_FUNC(double) PyFloat_Unpack2(const char *p, int le);
+PyAPI_FUNC(double) PyFloat_Unpack4(const char *p, int le);
+PyAPI_FUNC(double) PyFloat_Unpack8(const char *p, int le);

Dependencies/Python/include/cpython/frameobject.h

@@ -0,0 +1,35 @@
+/* Frame object interface */
+
+#ifndef Py_CPYTHON_FRAMEOBJECT_H
+#  error "this header file must not be included directly"
+#endif
+
+/* Standard object interface */
+
+PyAPI_FUNC(PyFrameObject *) PyFrame_New(PyThreadState *, PyCodeObject *,
+                                        PyObject *, PyObject *);
+
+/* The rest of the interface is specific for frame objects */
+
+/* Conversions between "fast locals" and locals in dictionary */
+
+PyAPI_FUNC(void) PyFrame_LocalsToFast(PyFrameObject *, int);
+
+/* -- Caveat emptor --
+ * The concept of entry frames is an implementation detail of the CPython
+ * interpreter. This API is considered unstable and is provided for the
+ * convenience of debuggers, profilers and state-inspecting tools. Notice that
+ * this API can be changed in future minor versions if the underlying frame
+ * mechanism change or the concept of an 'entry frame' or its semantics becomes
+ * obsolete or outdated. */
+
+PyAPI_FUNC(int) _PyFrame_IsEntryFrame(PyFrameObject *frame);
+
+PyAPI_FUNC(int) PyFrame_FastToLocalsWithError(PyFrameObject *f);
+PyAPI_FUNC(void) PyFrame_FastToLocals(PyFrameObject *);
+
+
+typedef struct {
+    PyObject_HEAD
+    PyFrameObject* frame;
+} PyFrameLocalsProxyObject;

Dependencies/Python/include/cpython/funcobject.h

@@ -0,0 +1,184 @@
+/* Function object interface */
+
+#ifndef Py_LIMITED_API
+#ifndef Py_FUNCOBJECT_H
+#define Py_FUNCOBJECT_H
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+
+#define _Py_COMMON_FIELDS(PREFIX) \
+    PyObject *PREFIX ## globals; \
+    PyObject *PREFIX ## builtins; \
+    PyObject *PREFIX ## name; \
+    PyObject *PREFIX ## qualname; \
+    PyObject *PREFIX ## code;        /* A code object, the __code__ attribute */ \
+    PyObject *PREFIX ## defaults;    /* NULL or a tuple */ \
+    PyObject *PREFIX ## kwdefaults;  /* NULL or a dict */ \
+    PyObject *PREFIX ## closure;     /* NULL or a tuple of cell objects */
+
+typedef struct {
+    _Py_COMMON_FIELDS(fc_)
+} PyFrameConstructor;
+
+/* Function objects and code objects should not be confused with each other:
+ *
+ * Function objects are created by the execution of the 'def' statement.
+ * They reference a code object in their __code__ attribute, which is a
+ * purely syntactic object, i.e. nothing more than a compiled version of some
+ * source code lines.  There is one code object per source code "fragment",
+ * but each code object can be referenced by zero or many function objects
+ * depending only on how many times the 'def' statement in the source was
+ * executed so far.
+ */
+
+typedef struct {
+    PyObject_HEAD
+    _Py_COMMON_FIELDS(func_)
+    PyObject *func_doc;         /* The __doc__ attribute, can be anything */
+    PyObject *func_dict;        /* The __dict__ attribute, a dict or NULL */
+    PyObject *func_weakreflist; /* List of weak references */
+    PyObject *func_module;      /* The __module__ attribute, can be anything */
+    PyObject *func_annotations; /* Annotations, a dict or NULL */
+    PyObject *func_typeparams;  /* Tuple of active type variables or NULL */
+    vectorcallfunc vectorcall;
+    /* Version number for use by specializer.
+     * Can set to non-zero when we want to specialize.
+     * Will be set to zero if any of these change:
+     *     defaults
+     *     kwdefaults (only if the object changes, not the contents of the dict)
+     *     code
+     *     annotations
+     *     vectorcall function pointer */
+    uint32_t func_version;
+
+    /* Invariant:
+     *     func_closure contains the bindings for func_code->co_freevars, so
+     *     PyTuple_Size(func_closure) == PyCode_GetNumFree(func_code)
+     *     (func_closure may be NULL if PyCode_GetNumFree(func_code) == 0).
+     */
+} PyFunctionObject;
+
+#undef _Py_COMMON_FIELDS
+
+PyAPI_DATA(PyTypeObject) PyFunction_Type;
+
+#define PyFunction_Check(op) Py_IS_TYPE((op), &PyFunction_Type)
+
+PyAPI_FUNC(PyObject *) PyFunction_New(PyObject *, PyObject *);
+PyAPI_FUNC(PyObject *) PyFunction_NewWithQualName(PyObject *, PyObject *, PyObject *);
+PyAPI_FUNC(PyObject *) PyFunction_GetCode(PyObject *);
+PyAPI_FUNC(PyObject *) PyFunction_GetGlobals(PyObject *);
+PyAPI_FUNC(PyObject *) PyFunction_GetModule(PyObject *);
+PyAPI_FUNC(PyObject *) PyFunction_GetDefaults(PyObject *);
+PyAPI_FUNC(int) PyFunction_SetDefaults(PyObject *, PyObject *);
+PyAPI_FUNC(void) PyFunction_SetVectorcall(PyFunctionObject *, vectorcallfunc);
+PyAPI_FUNC(PyObject *) PyFunction_GetKwDefaults(PyObject *);
+PyAPI_FUNC(int) PyFunction_SetKwDefaults(PyObject *, PyObject *);
+PyAPI_FUNC(PyObject *) PyFunction_GetClosure(PyObject *);
+PyAPI_FUNC(int) PyFunction_SetClosure(PyObject *, PyObject *);
+PyAPI_FUNC(PyObject *) PyFunction_GetAnnotations(PyObject *);
+PyAPI_FUNC(int) PyFunction_SetAnnotations(PyObject *, PyObject *);
+
+#define _PyFunction_CAST(func) \
+    (assert(PyFunction_Check(func)), _Py_CAST(PyFunctionObject*, func))
+
+/* Static inline functions for direct access to these values.
+   Type checks are *not* done, so use with care. */
+static inline PyObject* PyFunction_GET_CODE(PyObject *func) {
+    return _PyFunction_CAST(func)->func_code;
+}
+#define PyFunction_GET_CODE(func) PyFunction_GET_CODE(_PyObject_CAST(func))
+
+static inline PyObject* PyFunction_GET_GLOBALS(PyObject *func) {
+    return _PyFunction_CAST(func)->func_globals;
+}
+#define PyFunction_GET_GLOBALS(func) PyFunction_GET_GLOBALS(_PyObject_CAST(func))
+
+static inline PyObject* PyFunction_GET_MODULE(PyObject *func) {
+    return _PyFunction_CAST(func)->func_module;
+}
+#define PyFunction_GET_MODULE(func) PyFunction_GET_MODULE(_PyObject_CAST(func))
+
+static inline PyObject* PyFunction_GET_DEFAULTS(PyObject *func) {
+    return _PyFunction_CAST(func)->func_defaults;
+}
+#define PyFunction_GET_DEFAULTS(func) PyFunction_GET_DEFAULTS(_PyObject_CAST(func))
+
+static inline PyObject* PyFunction_GET_KW_DEFAULTS(PyObject *func) {
+    return _PyFunction_CAST(func)->func_kwdefaults;
+}
+#define PyFunction_GET_KW_DEFAULTS(func) PyFunction_GET_KW_DEFAULTS(_PyObject_CAST(func))
+
+static inline PyObject* PyFunction_GET_CLOSURE(PyObject *func) {
+    return _PyFunction_CAST(func)->func_closure;
+}
+#define PyFunction_GET_CLOSURE(func) PyFunction_GET_CLOSURE(_PyObject_CAST(func))
+
+static inline PyObject* PyFunction_GET_ANNOTATIONS(PyObject *func) {
+    return _PyFunction_CAST(func)->func_annotations;
+}
+#define PyFunction_GET_ANNOTATIONS(func) PyFunction_GET_ANNOTATIONS(_PyObject_CAST(func))
+
+/* The classmethod and staticmethod types lives here, too */
+PyAPI_DATA(PyTypeObject) PyClassMethod_Type;
+PyAPI_DATA(PyTypeObject) PyStaticMethod_Type;
+
+PyAPI_FUNC(PyObject *) PyClassMethod_New(PyObject *);
+PyAPI_FUNC(PyObject *) PyStaticMethod_New(PyObject *);
+
+#define PY_FOREACH_FUNC_EVENT(V) \
+    V(CREATE)                    \
+    V(DESTROY)                   \
+    V(MODIFY_CODE)               \
+    V(MODIFY_DEFAULTS)           \
+    V(MODIFY_KWDEFAULTS)
+
+typedef enum {
+    #define PY_DEF_EVENT(EVENT) PyFunction_EVENT_##EVENT,
+    PY_FOREACH_FUNC_EVENT(PY_DEF_EVENT)
+    #undef PY_DEF_EVENT
+} PyFunction_WatchEvent;
+
+/*
+ * A callback that is invoked for different events in a function's lifecycle.
+ *
+ * The callback is invoked with a borrowed reference to func, after it is
+ * created and before it is modified or destroyed. The callback should not
+ * modify func.
+ *
+ * When a function's code object, defaults, or kwdefaults are modified the
+ * callback will be invoked with the respective event and new_value will
+ * contain a borrowed reference to the new value that is about to be stored in
+ * the function. Otherwise the third argument is NULL.
+ *
+ * If the callback returns with an exception set, it must return -1. Otherwise
+ * it should return 0.
+ */
+typedef int (*PyFunction_WatchCallback)(
+  PyFunction_WatchEvent event,
+  PyFunctionObject *func,
+  PyObject *new_value);
+
+/*
+ * Register a per-interpreter callback that will be invoked for function lifecycle
+ * events.
+ *
+ * Returns a handle that may be passed to PyFunction_ClearWatcher on success,
+ * or -1 and sets an error if no more handles are available.
+ */
+PyAPI_FUNC(int) PyFunction_AddWatcher(PyFunction_WatchCallback callback);
+
+/*
+ * Clear the watcher associated with the watcher_id handle.
+ *
+ * Returns 0 on success or -1 if no watcher exists for the supplied id.
+ */
+PyAPI_FUNC(int) PyFunction_ClearWatcher(int watcher_id);
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* !Py_FUNCOBJECT_H */
+#endif /* Py_LIMITED_API */

Dependencies/Python/include/cpython/genobject.h

@@ -0,0 +1,83 @@
+/* Generator object interface */
+
+#ifndef Py_LIMITED_API
+#ifndef Py_GENOBJECT_H
+#define Py_GENOBJECT_H
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/* --- Generators --------------------------------------------------------- */
+
+/* _PyGenObject_HEAD defines the initial segment of generator
+   and coroutine objects. */
+#define _PyGenObject_HEAD(prefix)                                           \
+    PyObject_HEAD                                                           \
+    /* List of weak reference. */                                           \
+    PyObject *prefix##_weakreflist;                                         \
+    /* Name of the generator. */                                            \
+    PyObject *prefix##_name;                                                \
+    /* Qualified name of the generator. */                                  \
+    PyObject *prefix##_qualname;                                            \
+    _PyErr_StackItem prefix##_exc_state;                                    \
+    PyObject *prefix##_origin_or_finalizer;                                 \
+    char prefix##_hooks_inited;                                             \
+    char prefix##_closed;                                                   \
+    char prefix##_running_async;                                            \
+    /* The frame */                                                         \
+    int8_t prefix##_frame_state;                                            \
+    PyObject *prefix##_iframe[1];                                           \
+
+typedef struct {
+    /* The gi_ prefix is intended to remind of generator-iterator. */
+    _PyGenObject_HEAD(gi)
+} PyGenObject;
+
+PyAPI_DATA(PyTypeObject) PyGen_Type;
+
+#define PyGen_Check(op) PyObject_TypeCheck((op), &PyGen_Type)
+#define PyGen_CheckExact(op) Py_IS_TYPE((op), &PyGen_Type)
+
+PyAPI_FUNC(PyObject *) PyGen_New(PyFrameObject *);
+PyAPI_FUNC(PyObject *) PyGen_NewWithQualName(PyFrameObject *,
+    PyObject *name, PyObject *qualname);
+PyAPI_FUNC(PyCodeObject *) PyGen_GetCode(PyGenObject *gen);
+
+
+/* --- PyCoroObject ------------------------------------------------------- */
+
+typedef struct {
+    _PyGenObject_HEAD(cr)
+} PyCoroObject;
+
+PyAPI_DATA(PyTypeObject) PyCoro_Type;
+
+#define PyCoro_CheckExact(op) Py_IS_TYPE((op), &PyCoro_Type)
+PyAPI_FUNC(PyObject *) PyCoro_New(PyFrameObject *,
+    PyObject *name, PyObject *qualname);
+
+
+/* --- Asynchronous Generators -------------------------------------------- */
+
+typedef struct {
+    _PyGenObject_HEAD(ag)
+} PyAsyncGenObject;
+
+PyAPI_DATA(PyTypeObject) PyAsyncGen_Type;
+PyAPI_DATA(PyTypeObject) _PyAsyncGenASend_Type;
+
+PyAPI_FUNC(PyObject *) PyAsyncGen_New(PyFrameObject *,
+    PyObject *name, PyObject *qualname);
+
+#define PyAsyncGen_CheckExact(op) Py_IS_TYPE((op), &PyAsyncGen_Type)
+
+#define PyAsyncGenASend_CheckExact(op) Py_IS_TYPE((op), &_PyAsyncGenASend_Type)
+
+
+#undef _PyGenObject_HEAD
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* !Py_GENOBJECT_H */
+#endif /* Py_LIMITED_API */

Dependencies/Python/include/cpython/import.h

@@ -0,0 +1,25 @@
+#ifndef Py_CPYTHON_IMPORT_H
+#  error "this header file must not be included directly"
+#endif
+
+PyMODINIT_FUNC PyInit__imp(void);
+
+struct _inittab {
+    const char *name;           /* ASCII encoded string */
+    PyObject* (*initfunc)(void);
+};
+// This is not used after Py_Initialize() is called.
+PyAPI_DATA(struct _inittab *) PyImport_Inittab;
+PyAPI_FUNC(int) PyImport_ExtendInittab(struct _inittab *newtab);
+
+struct _frozen {
+    const char *name;                 /* ASCII encoded string */
+    const unsigned char *code;
+    int size;
+    int is_package;
+};
+
+/* Embedding apps may change this pointer to point to their favorite
+   collection of frozen modules: */
+
+PyAPI_DATA(const struct _frozen *) PyImport_FrozenModules;

Dependencies/Python/include/cpython/initconfig.h

@@ -0,0 +1,274 @@
+#ifndef Py_PYCORECONFIG_H
+#define Py_PYCORECONFIG_H
+#ifndef Py_LIMITED_API
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/* --- PyStatus ----------------------------------------------- */
+
+typedef struct {
+    enum {
+        _PyStatus_TYPE_OK=0,
+        _PyStatus_TYPE_ERROR=1,
+        _PyStatus_TYPE_EXIT=2
+    } _type;
+    const char *func;
+    const char *err_msg;
+    int exitcode;
+} PyStatus;
+
+PyAPI_FUNC(PyStatus) PyStatus_Ok(void);
+PyAPI_FUNC(PyStatus) PyStatus_Error(const char *err_msg);
+PyAPI_FUNC(PyStatus) PyStatus_NoMemory(void);
+PyAPI_FUNC(PyStatus) PyStatus_Exit(int exitcode);
+PyAPI_FUNC(int) PyStatus_IsError(PyStatus err);
+PyAPI_FUNC(int) PyStatus_IsExit(PyStatus err);
+PyAPI_FUNC(int) PyStatus_Exception(PyStatus err);
+
+/* --- PyWideStringList ------------------------------------------------ */
+
+typedef struct {
+    /* If length is greater than zero, items must be non-NULL
+       and all items strings must be non-NULL */
+    Py_ssize_t length;
+    wchar_t **items;
+} PyWideStringList;
+
+PyAPI_FUNC(PyStatus) PyWideStringList_Append(PyWideStringList *list,
+    const wchar_t *item);
+PyAPI_FUNC(PyStatus) PyWideStringList_Insert(PyWideStringList *list,
+    Py_ssize_t index,
+    const wchar_t *item);
+
+
+/* --- PyPreConfig ----------------------------------------------- */
+
+typedef struct PyPreConfig {
+    int _config_init;     /* _PyConfigInitEnum value */
+
+    /* Parse Py_PreInitializeFromBytesArgs() arguments?
+       See PyConfig.parse_argv */
+    int parse_argv;
+
+    /* If greater than 0, enable isolated mode: sys.path contains
+       neither the script's directory nor the user's site-packages directory.
+
+       Set to 1 by the -I command line option. If set to -1 (default), inherit
+       Py_IsolatedFlag value. */
+    int isolated;
+
+    /* If greater than 0: use environment variables.
+       Set to 0 by -E command line option. If set to -1 (default), it is
+       set to !Py_IgnoreEnvironmentFlag. */
+    int use_environment;
+
+    /* Set the LC_CTYPE locale to the user preferred locale? If equals to 0,
+       set coerce_c_locale and coerce_c_locale_warn to 0. */
+    int configure_locale;
+
+    /* Coerce the LC_CTYPE locale if it's equal to "C"? (PEP 538)
+
+       Set to 0 by PYTHONCOERCECLOCALE=0. Set to 1 by PYTHONCOERCECLOCALE=1.
+       Set to 2 if the user preferred LC_CTYPE locale is "C".
+
+       If it is equal to 1, LC_CTYPE locale is read to decide if it should be
+       coerced or not (ex: PYTHONCOERCECLOCALE=1). Internally, it is set to 2
+       if the LC_CTYPE locale must be coerced.
+
+       Disable by default (set to 0). Set it to -1 to let Python decide if it
+       should be enabled or not. */
+    int coerce_c_locale;
+
+    /* Emit a warning if the LC_CTYPE locale is coerced?
+
+       Set to 1 by PYTHONCOERCECLOCALE=warn.
+
+       Disable by default (set to 0). Set it to -1 to let Python decide if it
+       should be enabled or not. */
+    int coerce_c_locale_warn;
+
+#ifdef MS_WINDOWS
+    /* If greater than 1, use the "mbcs" encoding instead of the UTF-8
+       encoding for the filesystem encoding.
+
+       Set to 1 if the PYTHONLEGACYWINDOWSFSENCODING environment variable is
+       set to a non-empty string. If set to -1 (default), inherit
+       Py_LegacyWindowsFSEncodingFlag value.
+
+       See PEP 529 for more details. */
+    int legacy_windows_fs_encoding;
+#endif
+
+    /* Enable UTF-8 mode? (PEP 540)
+
+       Disabled by default (equals to 0).
+
+       Set to 1 by "-X utf8" and "-X utf8=1" command line options.
+       Set to 1 by PYTHONUTF8=1 environment variable.
+
+       Set to 0 by "-X utf8=0" and PYTHONUTF8=0.
+
+       If equals to -1, it is set to 1 if the LC_CTYPE locale is "C" or
+       "POSIX", otherwise it is set to 0. Inherit Py_UTF8Mode value value. */
+    int utf8_mode;
+
+    /* If non-zero, enable the Python Development Mode.
+
+       Set to 1 by the -X dev command line option. Set by the PYTHONDEVMODE
+       environment variable. */
+    int dev_mode;
+
+    /* Memory allocator: PYTHONMALLOC env var.
+       See PyMemAllocatorName for valid values. */
+    int allocator;
+} PyPreConfig;
+
+PyAPI_FUNC(void) PyPreConfig_InitPythonConfig(PyPreConfig *config);
+PyAPI_FUNC(void) PyPreConfig_InitIsolatedConfig(PyPreConfig *config);
+
+
+/* --- PyConfig ---------------------------------------------- */
+
+/* This structure is best documented in the Doc/c-api/init_config.rst file. */
+typedef struct PyConfig {
+    int _config_init;     /* _PyConfigInitEnum value */
+
+    int isolated;
+    int use_environment;
+    int dev_mode;
+    int install_signal_handlers;
+    int use_hash_seed;
+    unsigned long hash_seed;
+    int faulthandler;
+    int tracemalloc;
+    int perf_profiling;
+    int import_time;
+    int code_debug_ranges;
+    int show_ref_count;
+    int dump_refs;
+    wchar_t *dump_refs_file;
+    int malloc_stats;
+    wchar_t *filesystem_encoding;
+    wchar_t *filesystem_errors;
+    wchar_t *pycache_prefix;
+    int parse_argv;
+    PyWideStringList orig_argv;
+    PyWideStringList argv;
+    PyWideStringList xoptions;
+    PyWideStringList warnoptions;
+    int site_import;
+    int bytes_warning;
+    int warn_default_encoding;
+    int inspect;
+    int interactive;
+    int optimization_level;
+    int parser_debug;
+    int write_bytecode;
+    int verbose;
+    int quiet;
+    int user_site_directory;
+    int configure_c_stdio;
+    int buffered_stdio;
+    wchar_t *stdio_encoding;
+    wchar_t *stdio_errors;
+#ifdef MS_WINDOWS
+    int legacy_windows_stdio;
+#endif
+    wchar_t *check_hash_pycs_mode;
+    int use_frozen_modules;
+    int safe_path;
+    int int_max_str_digits;
+
+    int cpu_count;
+#ifdef Py_GIL_DISABLED
+    int enable_gil;
+#endif
+
+    /* --- Path configuration inputs ------------ */
+    int pathconfig_warnings;
+    wchar_t *program_name;
+    wchar_t *pythonpath_env;
+    wchar_t *home;
+    wchar_t *platlibdir;
+
+    /* --- Path configuration outputs ----------- */
+    int module_search_paths_set;
+    PyWideStringList module_search_paths;
+    wchar_t *stdlib_dir;
+    wchar_t *executable;
+    wchar_t *base_executable;
+    wchar_t *prefix;
+    wchar_t *base_prefix;
+    wchar_t *exec_prefix;
+    wchar_t *base_exec_prefix;
+
+    /* --- Parameter only used by Py_Main() ---------- */
+    int skip_source_first_line;
+    wchar_t *run_command;
+    wchar_t *run_module;
+    wchar_t *run_filename;
+
+    /* --- Set by Py_Main() -------------------------- */
+    wchar_t *sys_path_0;
+
+    /* --- Private fields ---------------------------- */
+
+    // Install importlib? If equals to 0, importlib is not initialized at all.
+    // Needed by freeze_importlib.
+    int _install_importlib;
+
+    // If equal to 0, stop Python initialization before the "main" phase.
+    int _init_main;
+
+    // If non-zero, we believe we're running from a source tree.
+    int _is_python_build;
+
+#ifdef Py_STATS
+    // If non-zero, turns on statistics gathering.
+    int _pystats;
+#endif
+
+#ifdef Py_DEBUG
+    // If not empty, import a non-__main__ module before site.py is executed.
+    // PYTHON_PRESITE=package.module or -X presite=package.module
+    wchar_t *run_presite;
+#endif
+} PyConfig;
+
+PyAPI_FUNC(void) PyConfig_InitPythonConfig(PyConfig *config);
+PyAPI_FUNC(void) PyConfig_InitIsolatedConfig(PyConfig *config);
+PyAPI_FUNC(void) PyConfig_Clear(PyConfig *);
+PyAPI_FUNC(PyStatus) PyConfig_SetString(
+    PyConfig *config,
+    wchar_t **config_str,
+    const wchar_t *str);
+PyAPI_FUNC(PyStatus) PyConfig_SetBytesString(
+    PyConfig *config,
+    wchar_t **config_str,
+    const char *str);
+PyAPI_FUNC(PyStatus) PyConfig_Read(PyConfig *config);
+PyAPI_FUNC(PyStatus) PyConfig_SetBytesArgv(
+    PyConfig *config,
+    Py_ssize_t argc,
+    char * const *argv);
+PyAPI_FUNC(PyStatus) PyConfig_SetArgv(PyConfig *config,
+    Py_ssize_t argc,
+    wchar_t * const *argv);
+PyAPI_FUNC(PyStatus) PyConfig_SetWideStringList(PyConfig *config,
+    PyWideStringList *list,
+    Py_ssize_t length, wchar_t **items);
+
+
+/* --- Helper functions --------------------------------------- */
+
+/* Get the original command line arguments, before Python modified them.
+
+   See also PyConfig.orig_argv. */
+PyAPI_FUNC(void) Py_GetArgcArgv(int *argc, wchar_t ***argv);
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* !Py_LIMITED_API */
+#endif /* !Py_PYCORECONFIG_H */

Dependencies/Python/include/cpython/listobject.h

@@ -0,0 +1,53 @@
+#ifndef Py_CPYTHON_LISTOBJECT_H
+#  error "this header file must not be included directly"
+#endif
+
+typedef struct {
+    PyObject_VAR_HEAD
+    /* Vector of pointers to list elements.  list[0] is ob_item[0], etc. */
+    PyObject **ob_item;
+
+    /* ob_item contains space for 'allocated' elements.  The number
+     * currently in use is ob_size.
+     * Invariants:
+     *     0 <= ob_size <= allocated
+     *     len(list) == ob_size
+     *     ob_item == NULL implies ob_size == allocated == 0
+     * list.sort() temporarily sets allocated to -1 to detect mutations.
+     *
+     * Items must normally not be NULL, except during construction when
+     * the list is not yet visible outside the function that builds it.
+     */
+    Py_ssize_t allocated;
+} PyListObject;
+
+/* Cast argument to PyListObject* type. */
+#define _PyList_CAST(op) \
+    (assert(PyList_Check(op)), _Py_CAST(PyListObject*, (op)))
+
+// Macros and static inline functions, trading safety for speed
+
+static inline Py_ssize_t PyList_GET_SIZE(PyObject *op) {
+    PyListObject *list = _PyList_CAST(op);
+#ifdef Py_GIL_DISABLED
+    return _Py_atomic_load_ssize_relaxed(&(_PyVarObject_CAST(list)->ob_size));
+#else
+    return Py_SIZE(list);
+#endif
+}
+#define PyList_GET_SIZE(op) PyList_GET_SIZE(_PyObject_CAST(op))
+
+#define PyList_GET_ITEM(op, index) (_PyList_CAST(op)->ob_item[(index)])
+
+static inline void
+PyList_SET_ITEM(PyObject *op, Py_ssize_t index, PyObject *value) {
+    PyListObject *list = _PyList_CAST(op);
+    assert(0 <= index);
+    assert(index < list->allocated);
+    list->ob_item[index] = value;
+}
+#define PyList_SET_ITEM(op, index, value) \
+    PyList_SET_ITEM(_PyObject_CAST(op), (index), _PyObject_CAST(value))
+
+PyAPI_FUNC(int) PyList_Extend(PyObject *self, PyObject *iterable);
+PyAPI_FUNC(int) PyList_Clear(PyObject *self);

Dependencies/Python/include/cpython/lock.h

@@ -0,0 +1,63 @@
+#ifndef Py_CPYTHON_LOCK_H
+#  error "this header file must not be included directly"
+#endif
+
+#define _Py_UNLOCKED    0
+#define _Py_LOCKED      1
+
+// A mutex that occupies one byte. The lock can be zero initialized to
+// represent the unlocked state.
+//
+// Typical initialization:
+//   PyMutex m = (PyMutex){0};
+//
+// Or initialize as global variables:
+//   static PyMutex m;
+//
+// Typical usage:
+//   PyMutex_Lock(&m);
+//   ...
+//   PyMutex_Unlock(&m);
+//
+// The contents of the PyMutex are not part of the public API, but are
+// described to aid in understanding the implementation and debugging. Only
+// the two least significant bits are used. The remaining bits are always zero:
+// 0b00: unlocked
+// 0b01: locked
+// 0b10: unlocked and has parked threads
+// 0b11: locked and has parked threads
+typedef struct PyMutex {
+    uint8_t _bits;  // (private)
+} PyMutex;
+
+// exported function for locking the mutex
+PyAPI_FUNC(void) PyMutex_Lock(PyMutex *m);
+
+// exported function for unlocking the mutex
+PyAPI_FUNC(void) PyMutex_Unlock(PyMutex *m);
+
+// Locks the mutex.
+//
+// If the mutex is currently locked, the calling thread will be parked until
+// the mutex is unlocked. If the current thread holds the GIL, then the GIL
+// will be released while the thread is parked.
+static inline void
+_PyMutex_Lock(PyMutex *m)
+{
+    uint8_t expected = _Py_UNLOCKED;
+    if (!_Py_atomic_compare_exchange_uint8(&m->_bits, &expected, _Py_LOCKED)) {
+        PyMutex_Lock(m);
+    }
+}
+#define PyMutex_Lock _PyMutex_Lock
+
+// Unlocks the mutex.
+static inline void
+_PyMutex_Unlock(PyMutex *m)
+{
+    uint8_t expected = _Py_LOCKED;
+    if (!_Py_atomic_compare_exchange_uint8(&m->_bits, &expected, _Py_UNLOCKED)) {
+        PyMutex_Unlock(m);
+    }
+}
+#define PyMutex_Unlock _PyMutex_Unlock

Dependencies/Python/include/cpython/longintrepr.h

@@ -0,0 +1,146 @@
+#ifndef Py_LIMITED_API
+#ifndef Py_LONGINTREPR_H
+#define Py_LONGINTREPR_H
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+
+/* This is published for the benefit of "friends" marshal.c and _decimal.c. */
+
+/* Parameters of the integer representation.  There are two different
+   sets of parameters: one set for 30-bit digits, stored in an unsigned 32-bit
+   integer type, and one set for 15-bit digits with each digit stored in an
+   unsigned short.  The value of PYLONG_BITS_IN_DIGIT, defined either at
+   configure time or in pyport.h, is used to decide which digit size to use.
+
+   Type 'digit' should be able to hold 2*PyLong_BASE-1, and type 'twodigits'
+   should be an unsigned integer type able to hold all integers up to
+   PyLong_BASE*PyLong_BASE-1.  x_sub assumes that 'digit' is an unsigned type,
+   and that overflow is handled by taking the result modulo 2**N for some N >
+   PyLong_SHIFT.  The majority of the code doesn't care about the precise
+   value of PyLong_SHIFT, but there are some notable exceptions:
+
+   - PyLong_{As,From}ByteArray require that PyLong_SHIFT be at least 8
+
+   - long_hash() requires that PyLong_SHIFT is *strictly* less than the number
+     of bits in an unsigned long, as do the PyLong <-> long (or unsigned long)
+     conversion functions
+
+   - the Python int <-> size_t/Py_ssize_t conversion functions expect that
+     PyLong_SHIFT is strictly less than the number of bits in a size_t
+
+   - the marshal code currently expects that PyLong_SHIFT is a multiple of 15
+
+   - NSMALLNEGINTS and NSMALLPOSINTS should be small enough to fit in a single
+     digit; with the current values this forces PyLong_SHIFT >= 9
+
+  The values 15 and 30 should fit all of the above requirements, on any
+  platform.
+*/
+
+#if PYLONG_BITS_IN_DIGIT == 30
+typedef uint32_t digit;
+typedef int32_t sdigit; /* signed variant of digit */
+typedef uint64_t twodigits;
+typedef int64_t stwodigits; /* signed variant of twodigits */
+#define PyLong_SHIFT    30
+#define _PyLong_DECIMAL_SHIFT   9 /* max(e such that 10**e fits in a digit) */
+#define _PyLong_DECIMAL_BASE    ((digit)1000000000) /* 10 ** DECIMAL_SHIFT */
+#elif PYLONG_BITS_IN_DIGIT == 15
+typedef unsigned short digit;
+typedef short sdigit; /* signed variant of digit */
+typedef unsigned long twodigits;
+typedef long stwodigits; /* signed variant of twodigits */
+#define PyLong_SHIFT    15
+#define _PyLong_DECIMAL_SHIFT   4 /* max(e such that 10**e fits in a digit) */
+#define _PyLong_DECIMAL_BASE    ((digit)10000) /* 10 ** DECIMAL_SHIFT */
+#else
+#error "PYLONG_BITS_IN_DIGIT should be 15 or 30"
+#endif
+#define PyLong_BASE     ((digit)1 << PyLong_SHIFT)
+#define PyLong_MASK     ((digit)(PyLong_BASE - 1))
+
+/* Long integer representation.
+
+   Long integers are made up of a number of 30- or 15-bit digits, depending on
+   the platform. The number of digits (ndigits) is stored in the high bits of
+   the lv_tag field (lvtag >> _PyLong_NON_SIZE_BITS).
+
+   The absolute value of a number is equal to
+        SUM(for i=0 through ndigits-1) ob_digit[i] * 2**(PyLong_SHIFT*i)
+
+   The sign of the value is stored in the lower 2 bits of lv_tag.
+
+    - 0: Positive
+    - 1: Zero
+    - 2: Negative
+
+   The third lowest bit of lv_tag is reserved for an immortality flag, but is
+   not currently used.
+
+   In a normalized number, ob_digit[ndigits-1] (the most significant
+   digit) is never zero.  Also, in all cases, for all valid i,
+        0 <= ob_digit[i] <= PyLong_MASK.
+
+   The allocation function takes care of allocating extra memory
+   so that ob_digit[0] ... ob_digit[ndigits-1] are actually available.
+   We always allocate memory for at least one digit, so accessing ob_digit[0]
+   is always safe. However, in the case ndigits == 0, the contents of
+   ob_digit[0] may be undefined.
+*/
+
+typedef struct _PyLongValue {
+    uintptr_t lv_tag; /* Number of digits, sign and flags */
+    digit ob_digit[1];
+} _PyLongValue;
+
+struct _longobject {
+    PyObject_HEAD
+    _PyLongValue long_value;
+};
+
+PyAPI_FUNC(PyLongObject*) _PyLong_New(Py_ssize_t);
+
+// Return a copy of src.
+PyAPI_FUNC(PyObject*) _PyLong_Copy(PyLongObject *src);
+
+PyAPI_FUNC(PyLongObject*) _PyLong_FromDigits(
+    int negative,
+    Py_ssize_t digit_count,
+    digit *digits);
+
+
+/* Inline some internals for speed. These should be in pycore_long.h
+ * if user code didn't need them inlined. */
+
+#define _PyLong_SIGN_MASK 3
+#define _PyLong_NON_SIZE_BITS 3
+
+
+static inline int
+_PyLong_IsCompact(const PyLongObject* op) {
+    assert(PyType_HasFeature((op)->ob_base.ob_type, Py_TPFLAGS_LONG_SUBCLASS));
+    return op->long_value.lv_tag < (2 << _PyLong_NON_SIZE_BITS);
+}
+
+#define PyUnstable_Long_IsCompact _PyLong_IsCompact
+
+static inline Py_ssize_t
+_PyLong_CompactValue(const PyLongObject *op)
+{
+    Py_ssize_t sign;
+    assert(PyType_HasFeature((op)->ob_base.ob_type, Py_TPFLAGS_LONG_SUBCLASS));
+    assert(PyUnstable_Long_IsCompact(op));
+    sign = 1 - (op->long_value.lv_tag & _PyLong_SIGN_MASK);
+    return sign * (Py_ssize_t)op->long_value.ob_digit[0];
+}
+
+#define PyUnstable_Long_CompactValue _PyLong_CompactValue
+
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* !Py_LONGINTREPR_H */
+#endif /* Py_LIMITED_API */

Dependencies/Python/include/cpython/longobject.h

@@ -0,0 +1,116 @@
+#ifndef Py_CPYTHON_LONGOBJECT_H
+#  error "this header file must not be included directly"
+#endif
+
+PyAPI_FUNC(PyObject*) PyLong_FromUnicodeObject(PyObject *u, int base);
+
+#define Py_ASNATIVEBYTES_DEFAULTS -1
+#define Py_ASNATIVEBYTES_BIG_ENDIAN 0
+#define Py_ASNATIVEBYTES_LITTLE_ENDIAN 1
+#define Py_ASNATIVEBYTES_NATIVE_ENDIAN 3
+#define Py_ASNATIVEBYTES_UNSIGNED_BUFFER 4
+#define Py_ASNATIVEBYTES_REJECT_NEGATIVE 8
+#define Py_ASNATIVEBYTES_ALLOW_INDEX 16
+
+/* PyLong_AsNativeBytes: Copy the integer value to a native variable.
+   buffer points to the first byte of the variable.
+   n_bytes is the number of bytes available in the buffer. Pass 0 to request
+   the required size for the value.
+   flags is a bitfield of the following flags:
+   * 1 - little endian
+   * 2 - native endian
+   * 4 - unsigned destination (e.g. don't reject copying 255 into one byte)
+   * 8 - raise an exception for negative inputs
+   * 16 - call __index__ on non-int types
+   If flags is -1 (all bits set), native endian is used, value truncation
+   behaves most like C (allows negative inputs and allow MSB set), and non-int
+   objects will raise a TypeError.
+   Big endian mode will write the most significant byte into the address
+   directly referenced by buffer; little endian will write the least significant
+   byte into that address.
+
+   If an exception is raised, returns a negative value.
+   Otherwise, returns the number of bytes that are required to store the value.
+   To check that the full value is represented, ensure that the return value is
+   equal or less than n_bytes.
+   All n_bytes are guaranteed to be written (unless an exception occurs), and
+   so ignoring a positive return value is the equivalent of a downcast in C.
+   In cases where the full value could not be represented, the returned value
+   may be larger than necessary - this function is not an accurate way to
+   calculate the bit length of an integer object.
+   */
+PyAPI_FUNC(Py_ssize_t) PyLong_AsNativeBytes(PyObject* v, void* buffer,
+    Py_ssize_t n_bytes, int flags);
+
+/* PyLong_FromNativeBytes: Create an int value from a native integer
+   n_bytes is the number of bytes to read from the buffer. Passing 0 will
+   always produce the zero int.
+   PyLong_FromUnsignedNativeBytes always produces a non-negative int.
+   flags is the same as for PyLong_AsNativeBytes, but only supports selecting
+   the endianness or forcing an unsigned buffer.
+
+   Returns the int object, or NULL with an exception set. */
+PyAPI_FUNC(PyObject*) PyLong_FromNativeBytes(const void* buffer, size_t n_bytes,
+    int flags);
+PyAPI_FUNC(PyObject*) PyLong_FromUnsignedNativeBytes(const void* buffer,
+    size_t n_bytes, int flags);
+
+PyAPI_FUNC(int) PyUnstable_Long_IsCompact(const PyLongObject* op);
+PyAPI_FUNC(Py_ssize_t) PyUnstable_Long_CompactValue(const PyLongObject* op);
+
+// _PyLong_Sign.  Return 0 if v is 0, -1 if v < 0, +1 if v > 0.
+// v must not be NULL, and must be a normalized long.
+// There are no error cases.
+PyAPI_FUNC(int) _PyLong_Sign(PyObject *v);
+
+/* _PyLong_NumBits.  Return the number of bits needed to represent the
+   absolute value of a long.  For example, this returns 1 for 1 and -1, 2
+   for 2 and -2, and 2 for 3 and -3.  It returns 0 for 0.
+   v must not be NULL, and must be a normalized long.
+   (size_t)-1 is returned and OverflowError set if the true result doesn't
+   fit in a size_t.
+*/
+PyAPI_FUNC(size_t) _PyLong_NumBits(PyObject *v);
+
+/* _PyLong_FromByteArray:  View the n unsigned bytes as a binary integer in
+   base 256, and return a Python int with the same numeric value.
+   If n is 0, the integer is 0.  Else:
+   If little_endian is 1/true, bytes[n-1] is the MSB and bytes[0] the LSB;
+   else (little_endian is 0/false) bytes[0] is the MSB and bytes[n-1] the
+   LSB.
+   If is_signed is 0/false, view the bytes as a non-negative integer.
+   If is_signed is 1/true, view the bytes as a 2's-complement integer,
+   non-negative if bit 0x80 of the MSB is clear, negative if set.
+   Error returns:
+   + Return NULL with the appropriate exception set if there's not
+     enough memory to create the Python int.
+*/
+PyAPI_FUNC(PyObject *) _PyLong_FromByteArray(
+    const unsigned char* bytes, size_t n,
+    int little_endian, int is_signed);
+
+/* _PyLong_AsByteArray: Convert the least-significant 8*n bits of long
+   v to a base-256 integer, stored in array bytes.  Normally return 0,
+   return -1 on error.
+   If little_endian is 1/true, store the MSB at bytes[n-1] and the LSB at
+   bytes[0]; else (little_endian is 0/false) store the MSB at bytes[0] and
+   the LSB at bytes[n-1].
+   If is_signed is 0/false, it's an error if v < 0; else (v >= 0) n bytes
+   are filled and there's nothing special about bit 0x80 of the MSB.
+   If is_signed is 1/true, bytes is filled with the 2's-complement
+   representation of v's value.  Bit 0x80 of the MSB is the sign bit.
+   Error returns (-1):
+   + is_signed is 0 and v < 0.  TypeError is set in this case, and bytes
+     isn't altered.
+   + n isn't big enough to hold the full mathematical value of v.  For
+     example, if is_signed is 0 and there are more digits in the v than
+     fit in n; or if is_signed is 1, v < 0, and n is just 1 bit shy of
+     being large enough to hold a sign bit.  OverflowError is set in this
+     case, but bytes holds the least-significant n bytes of the true value.
+*/
+PyAPI_FUNC(int) _PyLong_AsByteArray(PyLongObject* v,
+    unsigned char* bytes, size_t n,
+    int little_endian, int is_signed, int with_exceptions);
+
+/* For use by the gcd function in mathmodule.c */
+PyAPI_FUNC(PyObject *) _PyLong_GCD(PyObject *, PyObject *);

Dependencies/Python/include/cpython/memoryobject.h

@@ -0,0 +1,50 @@
+#ifndef Py_CPYTHON_MEMORYOBJECT_H
+#  error "this header file must not be included directly"
+#endif
+
+/* The structs are declared here so that macros can work, but they shouldn't
+   be considered public. Don't access their fields directly, use the macros
+   and functions instead! */
+#define _Py_MANAGED_BUFFER_RELEASED    0x001  /* access to exporter blocked */
+#define _Py_MANAGED_BUFFER_FREE_FORMAT 0x002  /* free format */
+
+typedef struct {
+    PyObject_HEAD
+    int flags;          /* state flags */
+    Py_ssize_t exports; /* number of direct memoryview exports */
+    Py_buffer master; /* snapshot buffer obtained from the original exporter */
+} _PyManagedBufferObject;
+
+
+/* memoryview state flags */
+#define _Py_MEMORYVIEW_RELEASED    0x001  /* access to master buffer blocked */
+#define _Py_MEMORYVIEW_C           0x002  /* C-contiguous layout */
+#define _Py_MEMORYVIEW_FORTRAN     0x004  /* Fortran contiguous layout */
+#define _Py_MEMORYVIEW_SCALAR      0x008  /* scalar: ndim = 0 */
+#define _Py_MEMORYVIEW_PIL         0x010  /* PIL-style layout */
+#define _Py_MEMORYVIEW_RESTRICTED  0x020  /* Disallow new references to the memoryview's buffer */
+
+typedef struct {
+    PyObject_VAR_HEAD
+    _PyManagedBufferObject *mbuf; /* managed buffer */
+    Py_hash_t hash;               /* hash value for read-only views */
+    int flags;                    /* state flags */
+    Py_ssize_t exports;           /* number of buffer re-exports */
+    Py_buffer view;               /* private copy of the exporter's view */
+    PyObject *weakreflist;
+    Py_ssize_t ob_array[1];       /* shape, strides, suboffsets */
+} PyMemoryViewObject;
+
+#define _PyMemoryView_CAST(op) _Py_CAST(PyMemoryViewObject*, op)
+
+/* Get a pointer to the memoryview's private copy of the exporter's buffer. */
+static inline Py_buffer* PyMemoryView_GET_BUFFER(PyObject *op) {
+    return (&_PyMemoryView_CAST(op)->view);
+}
+#define PyMemoryView_GET_BUFFER(op) PyMemoryView_GET_BUFFER(_PyObject_CAST(op))
+
+/* Get a pointer to the exporting object (this may be NULL!). */
+static inline PyObject* PyMemoryView_GET_BASE(PyObject *op) {
+    return _PyMemoryView_CAST(op)->view.obj;
+}
+#define PyMemoryView_GET_BASE(op) PyMemoryView_GET_BASE(_PyObject_CAST(op))

Dependencies/Python/include/cpython/methodobject.h

@@ -0,0 +1,66 @@
+#ifndef Py_CPYTHON_METHODOBJECT_H
+#  error "this header file must not be included directly"
+#endif
+
+// PyCFunctionObject structure
+
+typedef struct {
+    PyObject_HEAD
+    PyMethodDef *m_ml; /* Description of the C function to call */
+    PyObject    *m_self; /* Passed as 'self' arg to the C func, can be NULL */
+    PyObject    *m_module; /* The __module__ attribute, can be anything */
+    PyObject    *m_weakreflist; /* List of weak references */
+    vectorcallfunc vectorcall;
+} PyCFunctionObject;
+
+#define _PyCFunctionObject_CAST(func) \
+    (assert(PyCFunction_Check(func)), \
+     _Py_CAST(PyCFunctionObject*, (func)))
+
+
+// PyCMethodObject structure
+
+typedef struct {
+    PyCFunctionObject func;
+    PyTypeObject *mm_class; /* Class that defines this method */
+} PyCMethodObject;
+
+#define _PyCMethodObject_CAST(func) \
+    (assert(PyCMethod_Check(func)), \
+     _Py_CAST(PyCMethodObject*, (func)))
+
+PyAPI_DATA(PyTypeObject) PyCMethod_Type;
+
+#define PyCMethod_CheckExact(op) Py_IS_TYPE((op), &PyCMethod_Type)
+#define PyCMethod_Check(op) PyObject_TypeCheck((op), &PyCMethod_Type)
+
+
+/* Static inline functions for direct access to these values.
+   Type checks are *not* done, so use with care. */
+static inline PyCFunction PyCFunction_GET_FUNCTION(PyObject *func) {
+    return _PyCFunctionObject_CAST(func)->m_ml->ml_meth;
+}
+#define PyCFunction_GET_FUNCTION(func) PyCFunction_GET_FUNCTION(_PyObject_CAST(func))
+
+static inline PyObject* PyCFunction_GET_SELF(PyObject *func_obj) {
+    PyCFunctionObject *func = _PyCFunctionObject_CAST(func_obj);
+    if (func->m_ml->ml_flags & METH_STATIC) {
+        return _Py_NULL;
+    }
+    return func->m_self;
+}
+#define PyCFunction_GET_SELF(func) PyCFunction_GET_SELF(_PyObject_CAST(func))
+
+static inline int PyCFunction_GET_FLAGS(PyObject *func) {
+    return _PyCFunctionObject_CAST(func)->m_ml->ml_flags;
+}
+#define PyCFunction_GET_FLAGS(func) PyCFunction_GET_FLAGS(_PyObject_CAST(func))
+
+static inline PyTypeObject* PyCFunction_GET_CLASS(PyObject *func_obj) {
+    PyCFunctionObject *func = _PyCFunctionObject_CAST(func_obj);
+    if (func->m_ml->ml_flags & METH_METHOD) {
+        return _PyCMethodObject_CAST(func)->mm_class;
+    }
+    return _Py_NULL;
+}
+#define PyCFunction_GET_CLASS(func) PyCFunction_GET_CLASS(_PyObject_CAST(func))

Dependencies/Python/include/cpython/modsupport.h

@@ -0,0 +1,26 @@
+#ifndef Py_CPYTHON_MODSUPPORT_H
+#  error "this header file must not be included directly"
+#endif
+
+// A data structure that can be used to run initialization code once in a
+// thread-safe manner. The C++11 equivalent is std::call_once.
+typedef struct {
+    uint8_t v;
+} _PyOnceFlag;
+
+typedef struct _PyArg_Parser {
+    const char *format;
+    const char * const *keywords;
+    const char *fname;
+    const char *custom_msg;
+    _PyOnceFlag once;       /* atomic one-time initialization flag */
+    int is_kwtuple_owned;   /* does this parser own the kwtuple object? */
+    int pos;                /* number of positional-only arguments */
+    int min;                /* minimal number of arguments */
+    int max;                /* maximal number of positional arguments */
+    PyObject *kwtuple;      /* tuple of keyword parameter names */
+    struct _PyArg_Parser *next;
+} _PyArg_Parser;
+
+PyAPI_FUNC(int) _PyArg_ParseTupleAndKeywordsFast(PyObject *, PyObject *,
+                                                 struct _PyArg_Parser *, ...);

Dependencies/Python/include/cpython/monitoring.h

@@ -0,0 +1,250 @@
+#ifndef Py_CPYTHON_MONITORING_H
+#  error "this header file must not be included directly"
+#endif
+
+/* Local events.
+ * These require bytecode instrumentation */
+
+#define PY_MONITORING_EVENT_PY_START 0
+#define PY_MONITORING_EVENT_PY_RESUME 1
+#define PY_MONITORING_EVENT_PY_RETURN 2
+#define PY_MONITORING_EVENT_PY_YIELD 3
+#define PY_MONITORING_EVENT_CALL 4
+#define PY_MONITORING_EVENT_LINE 5
+#define PY_MONITORING_EVENT_INSTRUCTION 6
+#define PY_MONITORING_EVENT_JUMP 7
+#define PY_MONITORING_EVENT_BRANCH 8
+#define PY_MONITORING_EVENT_STOP_ITERATION 9
+
+#define PY_MONITORING_IS_INSTRUMENTED_EVENT(ev) \
+    ((ev) < _PY_MONITORING_LOCAL_EVENTS)
+
+/* Other events, mainly exceptions */
+
+#define PY_MONITORING_EVENT_RAISE 10
+#define PY_MONITORING_EVENT_EXCEPTION_HANDLED 11
+#define PY_MONITORING_EVENT_PY_UNWIND 12
+#define PY_MONITORING_EVENT_PY_THROW 13
+#define PY_MONITORING_EVENT_RERAISE 14
+
+
+/* Ancillary events */
+
+#define PY_MONITORING_EVENT_C_RETURN 15
+#define PY_MONITORING_EVENT_C_RAISE 16
+
+
+typedef struct _PyMonitoringState {
+    uint8_t active;
+    uint8_t opaque;
+} PyMonitoringState;
+
+
+PyAPI_FUNC(int)
+PyMonitoring_EnterScope(PyMonitoringState *state_array, uint64_t *version,
+                         const uint8_t *event_types, Py_ssize_t length);
+
+PyAPI_FUNC(int)
+PyMonitoring_ExitScope(void);
+
+
+PyAPI_FUNC(int)
+_PyMonitoring_FirePyStartEvent(PyMonitoringState *state, PyObject *codelike, int32_t offset);
+
+PyAPI_FUNC(int)
+_PyMonitoring_FirePyResumeEvent(PyMonitoringState *state, PyObject *codelike, int32_t offset);
+
+PyAPI_FUNC(int)
+_PyMonitoring_FirePyReturnEvent(PyMonitoringState *state, PyObject *codelike, int32_t offset,
+                                PyObject *retval);
+
+PyAPI_FUNC(int)
+_PyMonitoring_FirePyYieldEvent(PyMonitoringState *state, PyObject *codelike, int32_t offset,
+                               PyObject *retval);
+
+PyAPI_FUNC(int)
+_PyMonitoring_FireCallEvent(PyMonitoringState *state, PyObject *codelike, int32_t offset,
+                            PyObject* callable, PyObject *arg0);
+
+PyAPI_FUNC(int)
+_PyMonitoring_FireLineEvent(PyMonitoringState *state, PyObject *codelike, int32_t offset,
+                            int lineno);
+
+PyAPI_FUNC(int)
+_PyMonitoring_FireJumpEvent(PyMonitoringState *state, PyObject *codelike, int32_t offset,
+                            PyObject *target_offset);
+
+PyAPI_FUNC(int)
+_PyMonitoring_FireBranchEvent(PyMonitoringState *state, PyObject *codelike, int32_t offset,
+                              PyObject *target_offset);
+
+PyAPI_FUNC(int)
+_PyMonitoring_FireCReturnEvent(PyMonitoringState *state, PyObject *codelike, int32_t offset,
+                               PyObject *retval);
+
+PyAPI_FUNC(int)
+_PyMonitoring_FirePyThrowEvent(PyMonitoringState *state, PyObject *codelike, int32_t offset);
+
+PyAPI_FUNC(int)
+_PyMonitoring_FireRaiseEvent(PyMonitoringState *state, PyObject *codelike, int32_t offset);
+
+PyAPI_FUNC(int)
+_PyMonitoring_FireReraiseEvent(PyMonitoringState *state, PyObject *codelike, int32_t offset);
+
+PyAPI_FUNC(int)
+_PyMonitoring_FireExceptionHandledEvent(PyMonitoringState *state, PyObject *codelike, int32_t offset);
+
+PyAPI_FUNC(int)
+_PyMonitoring_FireCRaiseEvent(PyMonitoringState *state, PyObject *codelike, int32_t offset);
+
+PyAPI_FUNC(int)
+_PyMonitoring_FirePyUnwindEvent(PyMonitoringState *state, PyObject *codelike, int32_t offset);
+
+PyAPI_FUNC(int)
+_PyMonitoring_FireStopIterationEvent(PyMonitoringState *state, PyObject *codelike, int32_t offset, PyObject *value);
+
+
+#define _PYMONITORING_IF_ACTIVE(STATE, X)  \
+    if ((STATE)->active) { \
+        return (X); \
+    } \
+    else { \
+        return 0; \
+    }
+
+static inline int
+PyMonitoring_FirePyStartEvent(PyMonitoringState *state, PyObject *codelike, int32_t offset)
+{
+    _PYMONITORING_IF_ACTIVE(
+        state,
+        _PyMonitoring_FirePyStartEvent(state, codelike, offset));
+}
+
+static inline int
+PyMonitoring_FirePyResumeEvent(PyMonitoringState *state, PyObject *codelike, int32_t offset)
+{
+    _PYMONITORING_IF_ACTIVE(
+        state,
+        _PyMonitoring_FirePyResumeEvent(state, codelike, offset));
+}
+
+static inline int
+PyMonitoring_FirePyReturnEvent(PyMonitoringState *state, PyObject *codelike, int32_t offset,
+                               PyObject *retval)
+{
+    _PYMONITORING_IF_ACTIVE(
+        state,
+        _PyMonitoring_FirePyReturnEvent(state, codelike, offset, retval));
+}
+
+static inline int
+PyMonitoring_FirePyYieldEvent(PyMonitoringState *state, PyObject *codelike, int32_t offset,
+                              PyObject *retval)
+{
+    _PYMONITORING_IF_ACTIVE(
+        state,
+        _PyMonitoring_FirePyYieldEvent(state, codelike, offset, retval));
+}
+
+static inline int
+PyMonitoring_FireCallEvent(PyMonitoringState *state, PyObject *codelike, int32_t offset,
+                           PyObject* callable, PyObject *arg0)
+{
+    _PYMONITORING_IF_ACTIVE(
+        state,
+        _PyMonitoring_FireCallEvent(state, codelike, offset, callable, arg0));
+}
+
+static inline int
+PyMonitoring_FireLineEvent(PyMonitoringState *state, PyObject *codelike, int32_t offset,
+                           int lineno)
+{
+    _PYMONITORING_IF_ACTIVE(
+        state,
+        _PyMonitoring_FireLineEvent(state, codelike, offset, lineno));
+}
+
+static inline int
+PyMonitoring_FireJumpEvent(PyMonitoringState *state, PyObject *codelike, int32_t offset,
+                           PyObject *target_offset)
+{
+    _PYMONITORING_IF_ACTIVE(
+        state,
+        _PyMonitoring_FireJumpEvent(state, codelike, offset, target_offset));
+}
+
+static inline int
+PyMonitoring_FireBranchEvent(PyMonitoringState *state, PyObject *codelike, int32_t offset,
+                             PyObject *target_offset)
+{
+    _PYMONITORING_IF_ACTIVE(
+        state,
+        _PyMonitoring_FireBranchEvent(state, codelike, offset, target_offset));
+}
+
+static inline int
+PyMonitoring_FireCReturnEvent(PyMonitoringState *state, PyObject *codelike, int32_t offset,
+                              PyObject *retval)
+{
+    _PYMONITORING_IF_ACTIVE(
+        state,
+        _PyMonitoring_FireCReturnEvent(state, codelike, offset, retval));
+}
+
+static inline int
+PyMonitoring_FirePyThrowEvent(PyMonitoringState *state, PyObject *codelike, int32_t offset)
+{
+    _PYMONITORING_IF_ACTIVE(
+        state,
+        _PyMonitoring_FirePyThrowEvent(state, codelike, offset));
+}
+
+static inline int
+PyMonitoring_FireRaiseEvent(PyMonitoringState *state, PyObject *codelike, int32_t offset)
+{
+    _PYMONITORING_IF_ACTIVE(
+        state,
+        _PyMonitoring_FireRaiseEvent(state, codelike, offset));
+}
+
+static inline int
+PyMonitoring_FireReraiseEvent(PyMonitoringState *state, PyObject *codelike, int32_t offset)
+{
+    _PYMONITORING_IF_ACTIVE(
+        state,
+        _PyMonitoring_FireReraiseEvent(state, codelike, offset));
+}
+
+static inline int
+PyMonitoring_FireExceptionHandledEvent(PyMonitoringState *state, PyObject *codelike, int32_t offset)
+{
+    _PYMONITORING_IF_ACTIVE(
+        state,
+        _PyMonitoring_FireExceptionHandledEvent(state, codelike, offset));
+}
+
+static inline int
+PyMonitoring_FireCRaiseEvent(PyMonitoringState *state, PyObject *codelike, int32_t offset)
+{
+    _PYMONITORING_IF_ACTIVE(
+        state,
+        _PyMonitoring_FireCRaiseEvent(state, codelike, offset));
+}
+
+static inline int
+PyMonitoring_FirePyUnwindEvent(PyMonitoringState *state, PyObject *codelike, int32_t offset)
+{
+    _PYMONITORING_IF_ACTIVE(
+        state,
+        _PyMonitoring_FirePyUnwindEvent(state, codelike, offset));
+}
+
+static inline int
+PyMonitoring_FireStopIterationEvent(PyMonitoringState *state, PyObject *codelike, int32_t offset, PyObject *value)
+{
+    _PYMONITORING_IF_ACTIVE(
+        state,
+        _PyMonitoring_FireStopIterationEvent(state, codelike, offset, value));
+}
+
+#undef _PYMONITORING_IF_ACTIVE

Dependencies/Python/include/cpython/object.h

@@ -0,0 +1,525 @@
+#ifndef Py_CPYTHON_OBJECT_H
+#  error "this header file must not be included directly"
+#endif
+
+PyAPI_FUNC(void) _Py_NewReference(PyObject *op);
+PyAPI_FUNC(void) _Py_NewReferenceNoTotal(PyObject *op);
+PyAPI_FUNC(void) _Py_ResurrectReference(PyObject *op);
+
+#ifdef Py_REF_DEBUG
+/* These are useful as debugging aids when chasing down refleaks. */
+PyAPI_FUNC(Py_ssize_t) _Py_GetGlobalRefTotal(void);
+#  define _Py_GetRefTotal() _Py_GetGlobalRefTotal()
+PyAPI_FUNC(Py_ssize_t) _Py_GetLegacyRefTotal(void);
+PyAPI_FUNC(Py_ssize_t) _PyInterpreterState_GetRefTotal(PyInterpreterState *);
+#endif
+
+
+/********************* String Literals ****************************************/
+/* This structure helps managing static strings. The basic usage goes like this:
+   Instead of doing
+
+       r = PyObject_CallMethod(o, "foo", "args", ...);
+
+   do
+
+       _Py_IDENTIFIER(foo);
+       ...
+       r = _PyObject_CallMethodId(o, &PyId_foo, "args", ...);
+
+   PyId_foo is a static variable, either on block level or file level. On first
+   usage, the string "foo" is interned, and the structures are linked. On interpreter
+   shutdown, all strings are released.
+
+   Alternatively, _Py_static_string allows choosing the variable name.
+   _PyUnicode_FromId returns a borrowed reference to the interned string.
+   _PyObject_{Get,Set,Has}AttrId are __getattr__ versions using _Py_Identifier*.
+*/
+typedef struct _Py_Identifier {
+    const char* string;
+    // Index in PyInterpreterState.unicode.ids.array. It is process-wide
+    // unique and must be initialized to -1.
+    Py_ssize_t index;
+    // Hidden PyMutex struct for non free-threaded build.
+    struct {
+        uint8_t v;
+    } mutex;
+} _Py_Identifier;
+
+#ifndef Py_BUILD_CORE
+// For now we are keeping _Py_IDENTIFIER for continued use
+// in non-builtin extensions (and naughty PyPI modules).
+
+#define _Py_static_string_init(value) { .string = (value), .index = -1 }
+#define _Py_static_string(varname, value)  static _Py_Identifier varname = _Py_static_string_init(value)
+#define _Py_IDENTIFIER(varname) _Py_static_string(PyId_##varname, #varname)
+
+#endif /* !Py_BUILD_CORE */
+
+
+typedef struct {
+    /* Number implementations must check *both*
+       arguments for proper type and implement the necessary conversions
+       in the slot functions themselves. */
+
+    binaryfunc nb_add;
+    binaryfunc nb_subtract;
+    binaryfunc nb_multiply;
+    binaryfunc nb_remainder;
+    binaryfunc nb_divmod;
+    ternaryfunc nb_power;
+    unaryfunc nb_negative;
+    unaryfunc nb_positive;
+    unaryfunc nb_absolute;
+    inquiry nb_bool;
+    unaryfunc nb_invert;
+    binaryfunc nb_lshift;
+    binaryfunc nb_rshift;
+    binaryfunc nb_and;
+    binaryfunc nb_xor;
+    binaryfunc nb_or;
+    unaryfunc nb_int;
+    void *nb_reserved;  /* the slot formerly known as nb_long */
+    unaryfunc nb_float;
+
+    binaryfunc nb_inplace_add;
+    binaryfunc nb_inplace_subtract;
+    binaryfunc nb_inplace_multiply;
+    binaryfunc nb_inplace_remainder;
+    ternaryfunc nb_inplace_power;
+    binaryfunc nb_inplace_lshift;
+    binaryfunc nb_inplace_rshift;
+    binaryfunc nb_inplace_and;
+    binaryfunc nb_inplace_xor;
+    binaryfunc nb_inplace_or;
+
+    binaryfunc nb_floor_divide;
+    binaryfunc nb_true_divide;
+    binaryfunc nb_inplace_floor_divide;
+    binaryfunc nb_inplace_true_divide;
+
+    unaryfunc nb_index;
+
+    binaryfunc nb_matrix_multiply;
+    binaryfunc nb_inplace_matrix_multiply;
+} PyNumberMethods;
+
+typedef struct {
+    lenfunc sq_length;
+    binaryfunc sq_concat;
+    ssizeargfunc sq_repeat;
+    ssizeargfunc sq_item;
+    void *was_sq_slice;
+    ssizeobjargproc sq_ass_item;
+    void *was_sq_ass_slice;
+    objobjproc sq_contains;
+
+    binaryfunc sq_inplace_concat;
+    ssizeargfunc sq_inplace_repeat;
+} PySequenceMethods;
+
+typedef struct {
+    lenfunc mp_length;
+    binaryfunc mp_subscript;
+    objobjargproc mp_ass_subscript;
+} PyMappingMethods;
+
+typedef PySendResult (*sendfunc)(PyObject *iter, PyObject *value, PyObject **result);
+
+typedef struct {
+    unaryfunc am_await;
+    unaryfunc am_aiter;
+    unaryfunc am_anext;
+    sendfunc am_send;
+} PyAsyncMethods;
+
+typedef struct {
+     getbufferproc bf_getbuffer;
+     releasebufferproc bf_releasebuffer;
+} PyBufferProcs;
+
+/* Allow printfunc in the tp_vectorcall_offset slot for
+ * backwards-compatibility */
+typedef Py_ssize_t printfunc;
+
+// If this structure is modified, Doc/includes/typestruct.h should be updated
+// as well.
+struct _typeobject {
+    PyObject_VAR_HEAD
+    const char *tp_name; /* For printing, in format "<module>.<name>" */
+    Py_ssize_t tp_basicsize, tp_itemsize; /* For allocation */
+
+    /* Methods to implement standard operations */
+
+    destructor tp_dealloc;
+    Py_ssize_t tp_vectorcall_offset;
+    getattrfunc tp_getattr;
+    setattrfunc tp_setattr;
+    PyAsyncMethods *tp_as_async; /* formerly known as tp_compare (Python 2)
+                                    or tp_reserved (Python 3) */
+    reprfunc tp_repr;
+
+    /* Method suites for standard classes */
+
+    PyNumberMethods *tp_as_number;
+    PySequenceMethods *tp_as_sequence;
+    PyMappingMethods *tp_as_mapping;
+
+    /* More standard operations (here for binary compatibility) */
+
+    hashfunc tp_hash;
+    ternaryfunc tp_call;
+    reprfunc tp_str;
+    getattrofunc tp_getattro;
+    setattrofunc tp_setattro;
+
+    /* Functions to access object as input/output buffer */
+    PyBufferProcs *tp_as_buffer;
+
+    /* Flags to define presence of optional/expanded features */
+    unsigned long tp_flags;
+
+    const char *tp_doc; /* Documentation string */
+
+    /* Assigned meaning in release 2.0 */
+    /* call function for all accessible objects */
+    traverseproc tp_traverse;
+
+    /* delete references to contained objects */
+    inquiry tp_clear;
+
+    /* Assigned meaning in release 2.1 */
+    /* rich comparisons */
+    richcmpfunc tp_richcompare;
+
+    /* weak reference enabler */
+    Py_ssize_t tp_weaklistoffset;
+
+    /* Iterators */
+    getiterfunc tp_iter;
+    iternextfunc tp_iternext;
+
+    /* Attribute descriptor and subclassing stuff */
+    PyMethodDef *tp_methods;
+    PyMemberDef *tp_members;
+    PyGetSetDef *tp_getset;
+    // Strong reference on a heap type, borrowed reference on a static type
+    PyTypeObject *tp_base;
+    PyObject *tp_dict;
+    descrgetfunc tp_descr_get;
+    descrsetfunc tp_descr_set;
+    Py_ssize_t tp_dictoffset;
+    initproc tp_init;
+    allocfunc tp_alloc;
+    newfunc tp_new;
+    freefunc tp_free; /* Low-level free-memory routine */
+    inquiry tp_is_gc; /* For PyObject_IS_GC */
+    PyObject *tp_bases;
+    PyObject *tp_mro; /* method resolution order */
+    PyObject *tp_cache; /* no longer used */
+    void *tp_subclasses;  /* for static builtin types this is an index */
+    PyObject *tp_weaklist; /* not used for static builtin types */
+    destructor tp_del;
+
+    /* Type attribute cache version tag. Added in version 2.6 */
+    unsigned int tp_version_tag;
+
+    destructor tp_finalize;
+    vectorcallfunc tp_vectorcall;
+
+    /* bitset of which type-watchers care about this type */
+    unsigned char tp_watched;
+    uint16_t tp_versions_used;
+};
+
+/* This struct is used by the specializer
+ * It should be treated as an opaque blob
+ * by code other than the specializer and interpreter. */
+struct _specialization_cache {
+    // In order to avoid bloating the bytecode with lots of inline caches, the
+    // members of this structure have a somewhat unique contract. They are set
+    // by the specialization machinery, and are invalidated by PyType_Modified.
+    // The rules for using them are as follows:
+    // - If getitem is non-NULL, then it is the same Python function that
+    //   PyType_Lookup(cls, "__getitem__") would return.
+    // - If getitem is NULL, then getitem_version is meaningless.
+    // - If getitem->func_version == getitem_version, then getitem can be called
+    //   with two positional arguments and no keyword arguments, and has neither
+    //   *args nor **kwargs (as required by BINARY_SUBSCR_GETITEM):
+    PyObject *getitem;
+    uint32_t getitem_version;
+    PyObject *init;
+};
+
+/* The *real* layout of a type object when allocated on the heap */
+typedef struct _heaptypeobject {
+    /* Note: there's a dependency on the order of these members
+       in slotptr() in typeobject.c . */
+    PyTypeObject ht_type;
+    PyAsyncMethods as_async;
+    PyNumberMethods as_number;
+    PyMappingMethods as_mapping;
+    PySequenceMethods as_sequence; /* as_sequence comes after as_mapping,
+                                      so that the mapping wins when both
+                                      the mapping and the sequence define
+                                      a given operator (e.g. __getitem__).
+                                      see add_operators() in typeobject.c . */
+    PyBufferProcs as_buffer;
+    PyObject *ht_name, *ht_slots, *ht_qualname;
+    struct _dictkeysobject *ht_cached_keys;
+    PyObject *ht_module;
+    char *_ht_tpname;  // Storage for "tp_name"; see PyType_FromModuleAndSpec
+    struct _specialization_cache _spec_cache; // For use by the specializer.
+    /* here are optional user slots, followed by the members. */
+} PyHeapTypeObject;
+
+PyAPI_FUNC(const char *) _PyType_Name(PyTypeObject *);
+PyAPI_FUNC(PyObject *) _PyType_Lookup(PyTypeObject *, PyObject *);
+PyAPI_FUNC(PyObject *) _PyType_LookupRef(PyTypeObject *, PyObject *);
+PyAPI_FUNC(PyObject *) PyType_GetDict(PyTypeObject *);
+
+PyAPI_FUNC(int) PyObject_Print(PyObject *, FILE *, int);
+PyAPI_FUNC(void) _Py_BreakPoint(void);
+PyAPI_FUNC(void) _PyObject_Dump(PyObject *);
+
+PyAPI_FUNC(PyObject*) _PyObject_GetAttrId(PyObject *, _Py_Identifier *);
+
+PyAPI_FUNC(PyObject **) _PyObject_GetDictPtr(PyObject *);
+PyAPI_FUNC(void) PyObject_CallFinalizer(PyObject *);
+PyAPI_FUNC(int) PyObject_CallFinalizerFromDealloc(PyObject *);
+
+PyAPI_FUNC(void) PyUnstable_Object_ClearWeakRefsNoCallbacks(PyObject *);
+
+/* Same as PyObject_Generic{Get,Set}Attr, but passing the attributes
+   dict as the last parameter. */
+PyAPI_FUNC(PyObject *)
+_PyObject_GenericGetAttrWithDict(PyObject *, PyObject *, PyObject *, int);
+PyAPI_FUNC(int)
+_PyObject_GenericSetAttrWithDict(PyObject *, PyObject *,
+                                 PyObject *, PyObject *);
+
+PyAPI_FUNC(PyObject *) _PyObject_FunctionStr(PyObject *);
+
+/* Safely decref `dst` and set `dst` to `src`.
+ *
+ * As in case of Py_CLEAR "the obvious" code can be deadly:
+ *
+ *     Py_DECREF(dst);
+ *     dst = src;
+ *
+ * The safe way is:
+ *
+ *      Py_SETREF(dst, src);
+ *
+ * That arranges to set `dst` to `src` _before_ decref'ing, so that any code
+ * triggered as a side-effect of `dst` getting torn down no longer believes
+ * `dst` points to a valid object.
+ *
+ * Temporary variables are used to only evalutate macro arguments once and so
+ * avoid the duplication of side effects. _Py_TYPEOF() or memcpy() is used to
+ * avoid a miscompilation caused by type punning. See Py_CLEAR() comment for
+ * implementation details about type punning.
+ *
+ * The memcpy() implementation does not emit a compiler warning if 'src' has
+ * not the same type than 'src': any pointer type is accepted for 'src'.
+ */
+#ifdef _Py_TYPEOF
+#define Py_SETREF(dst, src) \
+    do { \
+        _Py_TYPEOF(dst)* _tmp_dst_ptr = &(dst); \
+        _Py_TYPEOF(dst) _tmp_old_dst = (*_tmp_dst_ptr); \
+        *_tmp_dst_ptr = (src); \
+        Py_DECREF(_tmp_old_dst); \
+    } while (0)
+#else
+#define Py_SETREF(dst, src) \
+    do { \
+        PyObject **_tmp_dst_ptr = _Py_CAST(PyObject**, &(dst)); \
+        PyObject *_tmp_old_dst = (*_tmp_dst_ptr); \
+        PyObject *_tmp_src = _PyObject_CAST(src); \
+        memcpy(_tmp_dst_ptr, &_tmp_src, sizeof(PyObject*)); \
+        Py_DECREF(_tmp_old_dst); \
+    } while (0)
+#endif
+
+/* Py_XSETREF() is a variant of Py_SETREF() that uses Py_XDECREF() instead of
+ * Py_DECREF().
+ */
+#ifdef _Py_TYPEOF
+#define Py_XSETREF(dst, src) \
+    do { \
+        _Py_TYPEOF(dst)* _tmp_dst_ptr = &(dst); \
+        _Py_TYPEOF(dst) _tmp_old_dst = (*_tmp_dst_ptr); \
+        *_tmp_dst_ptr = (src); \
+        Py_XDECREF(_tmp_old_dst); \
+    } while (0)
+#else
+#define Py_XSETREF(dst, src) \
+    do { \
+        PyObject **_tmp_dst_ptr = _Py_CAST(PyObject**, &(dst)); \
+        PyObject *_tmp_old_dst = (*_tmp_dst_ptr); \
+        PyObject *_tmp_src = _PyObject_CAST(src); \
+        memcpy(_tmp_dst_ptr, &_tmp_src, sizeof(PyObject*)); \
+        Py_XDECREF(_tmp_old_dst); \
+    } while (0)
+#endif
+
+
+/* Define a pair of assertion macros:
+   _PyObject_ASSERT_FROM(), _PyObject_ASSERT_WITH_MSG() and _PyObject_ASSERT().
+
+   These work like the regular C assert(), in that they will abort the
+   process with a message on stderr if the given condition fails to hold,
+   but compile away to nothing if NDEBUG is defined.
+
+   However, before aborting, Python will also try to call _PyObject_Dump() on
+   the given object.  This may be of use when investigating bugs in which a
+   particular object is corrupt (e.g. buggy a tp_visit method in an extension
+   module breaking the garbage collector), to help locate the broken objects.
+
+   The WITH_MSG variant allows you to supply an additional message that Python
+   will attempt to print to stderr, after the object dump. */
+#ifdef NDEBUG
+   /* No debugging: compile away the assertions: */
+#  define _PyObject_ASSERT_FROM(obj, expr, msg, filename, lineno, func) \
+    ((void)0)
+#else
+   /* With debugging: generate checks: */
+#  define _PyObject_ASSERT_FROM(obj, expr, msg, filename, lineno, func) \
+    ((expr) \
+      ? (void)(0) \
+      : _PyObject_AssertFailed((obj), Py_STRINGIFY(expr), \
+                               (msg), (filename), (lineno), (func)))
+#endif
+
+#define _PyObject_ASSERT_WITH_MSG(obj, expr, msg) \
+    _PyObject_ASSERT_FROM((obj), expr, (msg), __FILE__, __LINE__, __func__)
+#define _PyObject_ASSERT(obj, expr) \
+    _PyObject_ASSERT_WITH_MSG((obj), expr, NULL)
+
+#define _PyObject_ASSERT_FAILED_MSG(obj, msg) \
+    _PyObject_AssertFailed((obj), NULL, (msg), __FILE__, __LINE__, __func__)
+
+/* Declare and define _PyObject_AssertFailed() even when NDEBUG is defined,
+   to avoid causing compiler/linker errors when building extensions without
+   NDEBUG against a Python built with NDEBUG defined.
+
+   msg, expr and function can be NULL. */
+PyAPI_FUNC(void) _Py_NO_RETURN _PyObject_AssertFailed(
+    PyObject *obj,
+    const char *expr,
+    const char *msg,
+    const char *file,
+    int line,
+    const char *function);
+
+
+/* Trashcan mechanism, thanks to Christian Tismer.
+
+When deallocating a container object, it's possible to trigger an unbounded
+chain of deallocations, as each Py_DECREF in turn drops the refcount on "the
+next" object in the chain to 0.  This can easily lead to stack overflows,
+especially in threads (which typically have less stack space to work with).
+
+A container object can avoid this by bracketing the body of its tp_dealloc
+function with a pair of macros:
+
+static void
+mytype_dealloc(mytype *p)
+{
+    ... declarations go here ...
+
+    PyObject_GC_UnTrack(p);        // must untrack first
+    Py_TRASHCAN_BEGIN(p, mytype_dealloc)
+    ... The body of the deallocator goes here, including all calls ...
+    ... to Py_DECREF on contained objects.                         ...
+    Py_TRASHCAN_END                // there should be no code after this
+}
+
+CAUTION:  Never return from the middle of the body!  If the body needs to
+"get out early", put a label immediately before the Py_TRASHCAN_END
+call, and goto it.  Else the call-depth counter (see below) will stay
+above 0 forever, and the trashcan will never get emptied.
+
+How it works:  The BEGIN macro increments a call-depth counter.  So long
+as this counter is small, the body of the deallocator is run directly without
+further ado.  But if the counter gets large, it instead adds p to a list of
+objects to be deallocated later, skips the body of the deallocator, and
+resumes execution after the END macro.  The tp_dealloc routine then returns
+without deallocating anything (and so unbounded call-stack depth is avoided).
+
+When the call stack finishes unwinding again, code generated by the END macro
+notices this, and calls another routine to deallocate all the objects that
+may have been added to the list of deferred deallocations.  In effect, a
+chain of N deallocations is broken into (N-1)/(Py_TRASHCAN_HEADROOM-1) pieces,
+with the call stack never exceeding a depth of Py_TRASHCAN_HEADROOM.
+
+Since the tp_dealloc of a subclass typically calls the tp_dealloc of the base
+class, we need to ensure that the trashcan is only triggered on the tp_dealloc
+of the actual class being deallocated. Otherwise we might end up with a
+partially-deallocated object. To check this, the tp_dealloc function must be
+passed as second argument to Py_TRASHCAN_BEGIN().
+*/
+
+/* Python 3.9 private API, invoked by the macros below. */
+PyAPI_FUNC(int) _PyTrash_begin(PyThreadState *tstate, PyObject *op);
+PyAPI_FUNC(void) _PyTrash_end(PyThreadState *tstate);
+
+PyAPI_FUNC(void) _PyTrash_thread_deposit_object(PyThreadState *tstate, PyObject *op);
+PyAPI_FUNC(void) _PyTrash_thread_destroy_chain(PyThreadState *tstate);
+
+
+/* Python 3.10 private API, invoked by the Py_TRASHCAN_BEGIN(). */
+
+/* To avoid raising recursion errors during dealloc trigger trashcan before we reach
+ * recursion limit. To avoid trashing, we don't attempt to empty the trashcan until
+ * we have headroom above the trigger limit */
+#define Py_TRASHCAN_HEADROOM 50
+
+#define Py_TRASHCAN_BEGIN(op, dealloc) \
+do { \
+    PyThreadState *tstate = PyThreadState_Get(); \
+    if (tstate->c_recursion_remaining <= Py_TRASHCAN_HEADROOM && Py_TYPE(op)->tp_dealloc == (destructor)dealloc) { \
+        _PyTrash_thread_deposit_object(tstate, (PyObject *)op); \
+        break; \
+    } \
+    tstate->c_recursion_remaining--;
+    /* The body of the deallocator is here. */
+#define Py_TRASHCAN_END \
+    tstate->c_recursion_remaining++; \
+    if (tstate->delete_later && tstate->c_recursion_remaining > (Py_TRASHCAN_HEADROOM*2)) { \
+        _PyTrash_thread_destroy_chain(tstate); \
+    } \
+} while (0);
+
+
+PyAPI_FUNC(void *) PyObject_GetItemData(PyObject *obj);
+
+PyAPI_FUNC(int) PyObject_VisitManagedDict(PyObject *obj, visitproc visit, void *arg);
+PyAPI_FUNC(int) _PyObject_SetManagedDict(PyObject *obj, PyObject *new_dict);
+PyAPI_FUNC(void) PyObject_ClearManagedDict(PyObject *obj);
+
+#define TYPE_MAX_WATCHERS 8
+
+typedef int(*PyType_WatchCallback)(PyTypeObject *);
+PyAPI_FUNC(int) PyType_AddWatcher(PyType_WatchCallback callback);
+PyAPI_FUNC(int) PyType_ClearWatcher(int watcher_id);
+PyAPI_FUNC(int) PyType_Watch(int watcher_id, PyObject *type);
+PyAPI_FUNC(int) PyType_Unwatch(int watcher_id, PyObject *type);
+
+/* Attempt to assign a version tag to the given type.
+ *
+ * Returns 1 if the type already had a valid version tag or a new one was
+ * assigned, or 0 if a new tag could not be assigned.
+ */
+PyAPI_FUNC(int) PyUnstable_Type_AssignVersionTag(PyTypeObject *type);
+
+
+typedef enum {
+    PyRefTracer_CREATE = 0,
+    PyRefTracer_DESTROY = 1,
+} PyRefTracerEvent;
+
+typedef int (*PyRefTracer)(PyObject *, PyRefTracerEvent event, void *);
+PyAPI_FUNC(int) PyRefTracer_SetTracer(PyRefTracer tracer, void *data);
+PyAPI_FUNC(PyRefTracer) PyRefTracer_GetTracer(void**);

Dependencies/Python/include/cpython/objimpl.h

@@ -0,0 +1,104 @@
+#ifndef Py_CPYTHON_OBJIMPL_H
+#  error "this header file must not be included directly"
+#endif
+
+static inline size_t _PyObject_SIZE(PyTypeObject *type) {
+    return _Py_STATIC_CAST(size_t, type->tp_basicsize);
+}
+
+/* _PyObject_VAR_SIZE returns the number of bytes (as size_t) allocated for a
+   vrbl-size object with nitems items, exclusive of gc overhead (if any).  The
+   value is rounded up to the closest multiple of sizeof(void *), in order to
+   ensure that pointer fields at the end of the object are correctly aligned
+   for the platform (this is of special importance for subclasses of, e.g.,
+   str or int, so that pointers can be stored after the embedded data).
+
+   Note that there's no memory wastage in doing this, as malloc has to
+   return (at worst) pointer-aligned memory anyway.
+*/
+#if ((SIZEOF_VOID_P - 1) & SIZEOF_VOID_P) != 0
+#   error "_PyObject_VAR_SIZE requires SIZEOF_VOID_P be a power of 2"
+#endif
+
+static inline size_t _PyObject_VAR_SIZE(PyTypeObject *type, Py_ssize_t nitems) {
+    size_t size = _Py_STATIC_CAST(size_t, type->tp_basicsize);
+    size += _Py_STATIC_CAST(size_t, nitems) * _Py_STATIC_CAST(size_t, type->tp_itemsize);
+    return _Py_SIZE_ROUND_UP(size, SIZEOF_VOID_P);
+}
+
+
+/* This example code implements an object constructor with a custom
+   allocator, where PyObject_New is inlined, and shows the important
+   distinction between two steps (at least):
+       1) the actual allocation of the object storage;
+       2) the initialization of the Python specific fields
+      in this storage with PyObject_{Init, InitVar}.
+
+   PyObject *
+   YourObject_New(...)
+   {
+       PyObject *op;
+
+       op = (PyObject *) Your_Allocator(_PyObject_SIZE(YourTypeStruct));
+       if (op == NULL) {
+           return PyErr_NoMemory();
+       }
+
+       PyObject_Init(op, &YourTypeStruct);
+
+       op->ob_field = value;
+       ...
+       return op;
+   }
+
+   Note that in C++, the use of the new operator usually implies that
+   the 1st step is performed automatically for you, so in a C++ class
+   constructor you would start directly with PyObject_Init/InitVar. */
+
+
+typedef struct {
+    /* user context passed as the first argument to the 2 functions */
+    void *ctx;
+
+    /* allocate an arena of size bytes */
+    void* (*alloc) (void *ctx, size_t size);
+
+    /* free an arena */
+    void (*free) (void *ctx, void *ptr, size_t size);
+} PyObjectArenaAllocator;
+
+/* Get the arena allocator. */
+PyAPI_FUNC(void) PyObject_GetArenaAllocator(PyObjectArenaAllocator *allocator);
+
+/* Set the arena allocator. */
+PyAPI_FUNC(void) PyObject_SetArenaAllocator(PyObjectArenaAllocator *allocator);
+
+
+/* Test if an object implements the garbage collector protocol */
+PyAPI_FUNC(int) PyObject_IS_GC(PyObject *obj);
+
+
+// Test if a type supports weak references
+PyAPI_FUNC(int) PyType_SUPPORTS_WEAKREFS(PyTypeObject *type);
+
+PyAPI_FUNC(PyObject **) PyObject_GET_WEAKREFS_LISTPTR(PyObject *op);
+
+PyAPI_FUNC(PyObject *) PyUnstable_Object_GC_NewWithExtraData(PyTypeObject *,
+                                                             size_t);
+
+
+/* Visit all live GC-capable objects, similar to gc.get_objects(None). The
+ * supplied callback is called on every such object with the void* arg set
+ * to the supplied arg. Returning 0 from the callback ends iteration, returning
+ * 1 allows iteration to continue. Returning any other value may result in
+ * undefined behaviour.
+ *
+ * If new objects are (de)allocated by the callback it is undefined if they
+ * will be visited.
+
+ * Garbage collection is disabled during operation. Explicitly running a
+ * collection in the callback may lead to undefined behaviour e.g. visiting the
+ * same objects multiple times or not at all.
+ */
+typedef int (*gcvisitobjects_t)(PyObject*, void*);
+PyAPI_FUNC(void) PyUnstable_GC_VisitObjects(gcvisitobjects_t callback, void* arg);

Dependencies/Python/include/cpython/odictobject.h

@@ -0,0 +1,43 @@
+#ifndef Py_ODICTOBJECT_H
+#define Py_ODICTOBJECT_H
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+
+/* OrderedDict */
+/* This API is optional and mostly redundant. */
+
+#ifndef Py_LIMITED_API
+
+typedef struct _odictobject PyODictObject;
+
+PyAPI_DATA(PyTypeObject) PyODict_Type;
+PyAPI_DATA(PyTypeObject) PyODictIter_Type;
+PyAPI_DATA(PyTypeObject) PyODictKeys_Type;
+PyAPI_DATA(PyTypeObject) PyODictItems_Type;
+PyAPI_DATA(PyTypeObject) PyODictValues_Type;
+
+#define PyODict_Check(op) PyObject_TypeCheck((op), &PyODict_Type)
+#define PyODict_CheckExact(op) Py_IS_TYPE((op), &PyODict_Type)
+#define PyODict_SIZE(op) PyDict_GET_SIZE((op))
+
+PyAPI_FUNC(PyObject *) PyODict_New(void);
+PyAPI_FUNC(int) PyODict_SetItem(PyObject *od, PyObject *key, PyObject *item);
+PyAPI_FUNC(int) PyODict_DelItem(PyObject *od, PyObject *key);
+
+/* wrappers around PyDict* functions */
+#define PyODict_GetItem(od, key) PyDict_GetItem(_PyObject_CAST(od), (key))
+#define PyODict_GetItemWithError(od, key) \
+    PyDict_GetItemWithError(_PyObject_CAST(od), (key))
+#define PyODict_Contains(od, key) PyDict_Contains(_PyObject_CAST(od), (key))
+#define PyODict_Size(od) PyDict_Size(_PyObject_CAST(od))
+#define PyODict_GetItemString(od, key) \
+    PyDict_GetItemString(_PyObject_CAST(od), (key))
+
+#endif
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* !Py_ODICTOBJECT_H */

Dependencies/Python/include/cpython/picklebufobject.h

@@ -0,0 +1,31 @@
+/* PickleBuffer object. This is built-in for ease of use from third-party
+ * C extensions.
+ */
+
+#ifndef Py_PICKLEBUFOBJECT_H
+#define Py_PICKLEBUFOBJECT_H
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+#ifndef Py_LIMITED_API
+
+PyAPI_DATA(PyTypeObject) PyPickleBuffer_Type;
+
+#define PyPickleBuffer_Check(op) Py_IS_TYPE((op), &PyPickleBuffer_Type)
+
+/* Create a PickleBuffer redirecting to the given buffer-enabled object */
+PyAPI_FUNC(PyObject *) PyPickleBuffer_FromObject(PyObject *);
+/* Get the PickleBuffer's underlying view to the original object
+ * (NULL if released)
+ */
+PyAPI_FUNC(const Py_buffer *) PyPickleBuffer_GetBuffer(PyObject *);
+/* Release the PickleBuffer.  Returns 0 on success, -1 on error. */
+PyAPI_FUNC(int) PyPickleBuffer_Release(PyObject *);
+
+#endif /* !Py_LIMITED_API */
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* !Py_PICKLEBUFOBJECT_H */

Dependencies/Python/include/cpython/pthread_stubs.h

@@ -0,0 +1,105 @@
+#ifndef Py_CPYTHON_PTRHEAD_STUBS_H
+#define Py_CPYTHON_PTRHEAD_STUBS_H
+
+#if !defined(HAVE_PTHREAD_STUBS)
+#  error "this header file requires stubbed pthreads."
+#endif
+
+#ifndef _POSIX_THREADS
+#  define _POSIX_THREADS 1
+#endif
+
+/* Minimal pthread stubs for CPython.
+ *
+ * The stubs implement the minimum pthread API for CPython.
+ * - pthread_create() fails.
+ * - pthread_exit() calls exit(0).
+ * - pthread_key_*() functions implement minimal TSS without destructor.
+ * - all other functions do nothing and return 0.
+ */
+
+#ifdef __wasi__
+// WASI's bits/alltypes.h provides type definitions when __NEED_ is set.
+// The header file can be included multiple times.
+//
+// <sys/types.h> may also define these macros.
+#  ifndef __NEED_pthread_cond_t
+#    define __NEED_pthread_cond_t 1
+#  endif
+#  ifndef __NEED_pthread_condattr_t
+#    define __NEED_pthread_condattr_t 1
+#  endif
+#  ifndef __NEED_pthread_mutex_t
+#    define __NEED_pthread_mutex_t 1
+#  endif
+#  ifndef __NEED_pthread_mutexattr_t
+#    define __NEED_pthread_mutexattr_t 1
+#  endif
+#  ifndef __NEED_pthread_key_t
+#    define __NEED_pthread_key_t 1
+#  endif
+#  ifndef __NEED_pthread_t
+#    define __NEED_pthread_t 1
+#  endif
+#  ifndef __NEED_pthread_attr_t
+#    define __NEED_pthread_attr_t 1
+#  endif
+#  include <bits/alltypes.h>
+#else
+typedef struct { void *__x; } pthread_cond_t;
+typedef struct { unsigned __attr; } pthread_condattr_t;
+typedef struct { void *__x; } pthread_mutex_t;
+typedef struct { unsigned __attr; } pthread_mutexattr_t;
+typedef unsigned pthread_key_t;
+typedef unsigned pthread_t;
+typedef struct { unsigned __attr; } pthread_attr_t;
+#endif
+
+// mutex
+PyAPI_FUNC(int) pthread_mutex_init(pthread_mutex_t *restrict mutex,
+                                   const pthread_mutexattr_t *restrict attr);
+PyAPI_FUNC(int) pthread_mutex_destroy(pthread_mutex_t *mutex);
+PyAPI_FUNC(int) pthread_mutex_trylock(pthread_mutex_t *mutex);
+PyAPI_FUNC(int) pthread_mutex_lock(pthread_mutex_t *mutex);
+PyAPI_FUNC(int) pthread_mutex_unlock(pthread_mutex_t *mutex);
+
+// condition
+PyAPI_FUNC(int) pthread_cond_init(pthread_cond_t *restrict cond,
+                                  const pthread_condattr_t *restrict attr);
+PyAPI_FUNC(int) pthread_cond_destroy(pthread_cond_t *cond);
+PyAPI_FUNC(int) pthread_cond_wait(pthread_cond_t *restrict cond,
+                                  pthread_mutex_t *restrict mutex);
+PyAPI_FUNC(int) pthread_cond_timedwait(pthread_cond_t *restrict cond,
+                                       pthread_mutex_t *restrict mutex,
+                                       const struct timespec *restrict abstime);
+PyAPI_FUNC(int) pthread_cond_signal(pthread_cond_t *cond);
+PyAPI_FUNC(int) pthread_condattr_init(pthread_condattr_t *attr);
+PyAPI_FUNC(int) pthread_condattr_setclock(
+    pthread_condattr_t *attr, clockid_t clock_id);
+
+// pthread
+PyAPI_FUNC(int) pthread_create(pthread_t *restrict thread,
+                               const pthread_attr_t *restrict attr,
+                               void *(*start_routine)(void *),
+                               void *restrict arg);
+PyAPI_FUNC(int) pthread_detach(pthread_t thread);
+PyAPI_FUNC(int) pthread_join(pthread_t thread, void** value_ptr);
+PyAPI_FUNC(pthread_t) pthread_self(void);
+PyAPI_FUNC(int) pthread_exit(void *retval) __attribute__ ((__noreturn__));
+PyAPI_FUNC(int) pthread_attr_init(pthread_attr_t *attr);
+PyAPI_FUNC(int) pthread_attr_setstacksize(pthread_attr_t *attr, size_t stacksize);
+PyAPI_FUNC(int) pthread_attr_destroy(pthread_attr_t *attr);
+
+
+// pthread_key
+#ifndef PTHREAD_KEYS_MAX
+#  define PTHREAD_KEYS_MAX 128
+#endif
+
+PyAPI_FUNC(int) pthread_key_create(pthread_key_t *key,
+                                   void (*destr_function)(void *));
+PyAPI_FUNC(int) pthread_key_delete(pthread_key_t key);
+PyAPI_FUNC(void *) pthread_getspecific(pthread_key_t key);
+PyAPI_FUNC(int) pthread_setspecific(pthread_key_t key, const void *value);
+
+#endif // Py_CPYTHON_PTRHEAD_STUBS_H

Dependencies/Python/include/cpython/pyatomic.h

@@ -0,0 +1,569 @@
+// This header provides cross-platform low-level atomic operations
+// similar to C11 atomics.
+//
+// Operations are sequentially consistent unless they have a suffix indicating
+// otherwise. If in doubt, prefer the sequentially consistent operations.
+//
+// The "_relaxed" suffix for load and store operations indicates the "relaxed"
+// memory order. They don't provide synchronization, but (roughly speaking)
+// guarantee somewhat sane behavior for races instead of undefined behavior.
+// In practice, they correspond to "normal" hardware load and store
+// instructions, so they are almost as inexpensive as plain loads and stores
+// in C.
+//
+// Note that atomic read-modify-write operations like _Py_atomic_add_* return
+// the previous value of the atomic variable, not the new value.
+//
+// See https://en.cppreference.com/w/c/atomic for more information on C11
+// atomics.
+// See https://www.open-std.org/jtc1/sc22/wg21/docs/papers/2020/p2055r0.pdf
+// "A Relaxed Guide to memory_order_relaxed" for discussion of and common usage
+// or relaxed atomics.
+//
+// Functions with pseudo Python code:
+//
+//   def _Py_atomic_load(obj):
+//        return obj  # sequential consistency
+//
+//   def _Py_atomic_load_relaxed(obj):
+//       return obj  # relaxed consistency
+//
+//   def _Py_atomic_store(obj, value):
+//       obj = value  # sequential consistency
+//
+//   def _Py_atomic_store_relaxed(obj, value):
+//       obj = value  # relaxed consistency
+//
+//   def _Py_atomic_exchange(obj, value):
+//       # sequential consistency
+//       old_obj = obj
+//       obj = value
+//       return old_obj
+//
+//   def _Py_atomic_compare_exchange(obj, expected, desired):
+//       # sequential consistency
+//       if obj == expected:
+//           obj = desired
+//           return True
+//       else:
+//           expected = obj
+//           return False
+//
+//   def _Py_atomic_add(obj, value):
+//       # sequential consistency
+//       old_obj = obj
+//       obj += value
+//       return old_obj
+//
+//   def _Py_atomic_and(obj, value):
+//       # sequential consistency
+//       old_obj = obj
+//       obj &= value
+//       return old_obj
+//
+//   def _Py_atomic_or(obj, value):
+//       # sequential consistency
+//       old_obj = obj
+//       obj |= value
+//       return old_obj
+//
+// Other functions:
+//
+//   def _Py_atomic_load_ptr_acquire(obj):
+//       return obj  # acquire
+//
+//   def _Py_atomic_store_ptr_release(obj):
+//       return obj  # release
+//
+//   def _Py_atomic_fence_seq_cst():
+//       # sequential consistency
+//       ...
+//
+//   def _Py_atomic_fence_release():
+//       # release
+//       ...
+
+#ifndef Py_CPYTHON_ATOMIC_H
+#  error "this header file must not be included directly"
+#endif
+
+// --- _Py_atomic_add --------------------------------------------------------
+// Atomically adds `value` to `obj` and returns the previous value
+
+static inline int
+_Py_atomic_add_int(int *obj, int value);
+
+static inline int8_t
+_Py_atomic_add_int8(int8_t *obj, int8_t value);
+
+static inline int16_t
+_Py_atomic_add_int16(int16_t *obj, int16_t value);
+
+static inline int32_t
+_Py_atomic_add_int32(int32_t *obj, int32_t value);
+
+static inline int64_t
+_Py_atomic_add_int64(int64_t *obj, int64_t value);
+
+static inline intptr_t
+_Py_atomic_add_intptr(intptr_t *obj, intptr_t value);
+
+static inline unsigned int
+_Py_atomic_add_uint(unsigned int *obj, unsigned int value);
+
+static inline uint8_t
+_Py_atomic_add_uint8(uint8_t *obj, uint8_t value);
+
+static inline uint16_t
+_Py_atomic_add_uint16(uint16_t *obj, uint16_t value);
+
+static inline uint32_t
+_Py_atomic_add_uint32(uint32_t *obj, uint32_t value);
+
+static inline uint64_t
+_Py_atomic_add_uint64(uint64_t *obj, uint64_t value);
+
+static inline uintptr_t
+_Py_atomic_add_uintptr(uintptr_t *obj, uintptr_t value);
+
+static inline Py_ssize_t
+_Py_atomic_add_ssize(Py_ssize_t *obj, Py_ssize_t value);
+
+
+// --- _Py_atomic_compare_exchange -------------------------------------------
+// Performs an atomic compare-and-exchange.
+//
+// - If `*obj` and `*expected` are equal, store `desired` into `*obj`
+//   and return 1 (success).
+// - Otherwise, store the `*obj` current value into `*expected`
+//   and return 0 (failure).
+//
+// These correspond to the C11 atomic_compare_exchange_strong() function.
+
+static inline int
+_Py_atomic_compare_exchange_int(int *obj, int *expected, int desired);
+
+static inline int
+_Py_atomic_compare_exchange_int8(int8_t *obj, int8_t *expected, int8_t desired);
+
+static inline int
+_Py_atomic_compare_exchange_int16(int16_t *obj, int16_t *expected, int16_t desired);
+
+static inline int
+_Py_atomic_compare_exchange_int32(int32_t *obj, int32_t *expected, int32_t desired);
+
+static inline int
+_Py_atomic_compare_exchange_int64(int64_t *obj, int64_t *expected, int64_t desired);
+
+static inline int
+_Py_atomic_compare_exchange_intptr(intptr_t *obj, intptr_t *expected, intptr_t desired);
+
+static inline int
+_Py_atomic_compare_exchange_uint(unsigned int *obj, unsigned int *expected, unsigned int desired);
+
+static inline int
+_Py_atomic_compare_exchange_uint8(uint8_t *obj, uint8_t *expected, uint8_t desired);
+
+static inline int
+_Py_atomic_compare_exchange_uint16(uint16_t *obj, uint16_t *expected, uint16_t desired);
+
+static inline int
+_Py_atomic_compare_exchange_uint32(uint32_t *obj, uint32_t *expected, uint32_t desired);
+
+static inline int
+_Py_atomic_compare_exchange_uint64(uint64_t *obj, uint64_t *expected, uint64_t desired);
+
+static inline int
+_Py_atomic_compare_exchange_uintptr(uintptr_t *obj, uintptr_t *expected, uintptr_t desired);
+
+static inline int
+_Py_atomic_compare_exchange_ssize(Py_ssize_t *obj, Py_ssize_t *expected, Py_ssize_t desired);
+
+// NOTE: `obj` and `expected` are logically `void**` types, but we use `void*`
+// so that we can pass types like `PyObject**` without a cast.
+static inline int
+_Py_atomic_compare_exchange_ptr(void *obj, void *expected, void *value);
+
+
+// --- _Py_atomic_exchange ---------------------------------------------------
+// Atomically replaces `*obj` with `value` and returns the previous value of `*obj`.
+
+static inline int
+_Py_atomic_exchange_int(int *obj, int value);
+
+static inline int8_t
+_Py_atomic_exchange_int8(int8_t *obj, int8_t value);
+
+static inline int16_t
+_Py_atomic_exchange_int16(int16_t *obj, int16_t value);
+
+static inline int32_t
+_Py_atomic_exchange_int32(int32_t *obj, int32_t value);
+
+static inline int64_t
+_Py_atomic_exchange_int64(int64_t *obj, int64_t value);
+
+static inline intptr_t
+_Py_atomic_exchange_intptr(intptr_t *obj, intptr_t value);
+
+static inline unsigned int
+_Py_atomic_exchange_uint(unsigned int *obj, unsigned int value);
+
+static inline uint8_t
+_Py_atomic_exchange_uint8(uint8_t *obj, uint8_t value);
+
+static inline uint16_t
+_Py_atomic_exchange_uint16(uint16_t *obj, uint16_t value);
+
+static inline uint32_t
+_Py_atomic_exchange_uint32(uint32_t *obj, uint32_t value);
+
+static inline uint64_t
+_Py_atomic_exchange_uint64(uint64_t *obj, uint64_t value);
+
+static inline uintptr_t
+_Py_atomic_exchange_uintptr(uintptr_t *obj, uintptr_t value);
+
+static inline Py_ssize_t
+_Py_atomic_exchange_ssize(Py_ssize_t *obj, Py_ssize_t value);
+
+static inline void *
+_Py_atomic_exchange_ptr(void *obj, void *value);
+
+
+// --- _Py_atomic_and --------------------------------------------------------
+// Performs `*obj &= value` atomically and returns the previous value of `*obj`.
+
+static inline uint8_t
+_Py_atomic_and_uint8(uint8_t *obj, uint8_t value);
+
+static inline uint16_t
+_Py_atomic_and_uint16(uint16_t *obj, uint16_t value);
+
+static inline uint32_t
+_Py_atomic_and_uint32(uint32_t *obj, uint32_t value);
+
+static inline uint64_t
+_Py_atomic_and_uint64(uint64_t *obj, uint64_t value);
+
+static inline uintptr_t
+_Py_atomic_and_uintptr(uintptr_t *obj, uintptr_t value);
+
+
+// --- _Py_atomic_or ---------------------------------------------------------
+// Performs `*obj |= value` atomically and returns the previous value of `*obj`.
+
+static inline uint8_t
+_Py_atomic_or_uint8(uint8_t *obj, uint8_t value);
+
+static inline uint16_t
+_Py_atomic_or_uint16(uint16_t *obj, uint16_t value);
+
+static inline uint32_t
+_Py_atomic_or_uint32(uint32_t *obj, uint32_t value);
+
+static inline uint64_t
+_Py_atomic_or_uint64(uint64_t *obj, uint64_t value);
+
+static inline uintptr_t
+_Py_atomic_or_uintptr(uintptr_t *obj, uintptr_t value);
+
+
+// --- _Py_atomic_load -------------------------------------------------------
+// Atomically loads `*obj` (sequential consistency)
+
+static inline int
+_Py_atomic_load_int(const int *obj);
+
+static inline int8_t
+_Py_atomic_load_int8(const int8_t *obj);
+
+static inline int16_t
+_Py_atomic_load_int16(const int16_t *obj);
+
+static inline int32_t
+_Py_atomic_load_int32(const int32_t *obj);
+
+static inline int64_t
+_Py_atomic_load_int64(const int64_t *obj);
+
+static inline intptr_t
+_Py_atomic_load_intptr(const intptr_t *obj);
+
+static inline uint8_t
+_Py_atomic_load_uint8(const uint8_t *obj);
+
+static inline uint16_t
+_Py_atomic_load_uint16(const uint16_t *obj);
+
+static inline uint32_t
+_Py_atomic_load_uint32(const uint32_t *obj);
+
+static inline uint64_t
+_Py_atomic_load_uint64(const uint64_t *obj);
+
+static inline uintptr_t
+_Py_atomic_load_uintptr(const uintptr_t *obj);
+
+static inline unsigned int
+_Py_atomic_load_uint(const unsigned int *obj);
+
+static inline Py_ssize_t
+_Py_atomic_load_ssize(const Py_ssize_t *obj);
+
+static inline void *
+_Py_atomic_load_ptr(const void *obj);
+
+
+// --- _Py_atomic_load_relaxed -----------------------------------------------
+// Loads `*obj` (relaxed consistency, i.e., no ordering)
+
+static inline int
+_Py_atomic_load_int_relaxed(const int *obj);
+
+static inline int8_t
+_Py_atomic_load_int8_relaxed(const int8_t *obj);
+
+static inline int16_t
+_Py_atomic_load_int16_relaxed(const int16_t *obj);
+
+static inline int32_t
+_Py_atomic_load_int32_relaxed(const int32_t *obj);
+
+static inline int64_t
+_Py_atomic_load_int64_relaxed(const int64_t *obj);
+
+static inline intptr_t
+_Py_atomic_load_intptr_relaxed(const intptr_t *obj);
+
+static inline uint8_t
+_Py_atomic_load_uint8_relaxed(const uint8_t *obj);
+
+static inline uint16_t
+_Py_atomic_load_uint16_relaxed(const uint16_t *obj);
+
+static inline uint32_t
+_Py_atomic_load_uint32_relaxed(const uint32_t *obj);
+
+static inline uint64_t
+_Py_atomic_load_uint64_relaxed(const uint64_t *obj);
+
+static inline uintptr_t
+_Py_atomic_load_uintptr_relaxed(const uintptr_t *obj);
+
+static inline unsigned int
+_Py_atomic_load_uint_relaxed(const unsigned int *obj);
+
+static inline Py_ssize_t
+_Py_atomic_load_ssize_relaxed(const Py_ssize_t *obj);
+
+static inline void *
+_Py_atomic_load_ptr_relaxed(const void *obj);
+
+static inline unsigned long long
+_Py_atomic_load_ullong_relaxed(const unsigned long long *obj);
+
+// --- _Py_atomic_store ------------------------------------------------------
+// Atomically performs `*obj = value` (sequential consistency)
+
+static inline void
+_Py_atomic_store_int(int *obj, int value);
+
+static inline void
+_Py_atomic_store_int8(int8_t *obj, int8_t value);
+
+static inline void
+_Py_atomic_store_int16(int16_t *obj, int16_t value);
+
+static inline void
+_Py_atomic_store_int32(int32_t *obj, int32_t value);
+
+static inline void
+_Py_atomic_store_int64(int64_t *obj, int64_t value);
+
+static inline void
+_Py_atomic_store_intptr(intptr_t *obj, intptr_t value);
+
+static inline void
+_Py_atomic_store_uint8(uint8_t *obj, uint8_t value);
+
+static inline void
+_Py_atomic_store_uint16(uint16_t *obj, uint16_t value);
+
+static inline void
+_Py_atomic_store_uint32(uint32_t *obj, uint32_t value);
+
+static inline void
+_Py_atomic_store_uint64(uint64_t *obj, uint64_t value);
+
+static inline void
+_Py_atomic_store_uintptr(uintptr_t *obj, uintptr_t value);
+
+static inline void
+_Py_atomic_store_uint(unsigned int *obj, unsigned int value);
+
+static inline void
+_Py_atomic_store_ptr(void *obj, void *value);
+
+static inline void
+_Py_atomic_store_ssize(Py_ssize_t* obj, Py_ssize_t value);
+
+
+// --- _Py_atomic_store_relaxed ----------------------------------------------
+// Stores `*obj = value` (relaxed consistency, i.e., no ordering)
+
+static inline void
+_Py_atomic_store_int_relaxed(int *obj, int value);
+
+static inline void
+_Py_atomic_store_int8_relaxed(int8_t *obj, int8_t value);
+
+static inline void
+_Py_atomic_store_int16_relaxed(int16_t *obj, int16_t value);
+
+static inline void
+_Py_atomic_store_int32_relaxed(int32_t *obj, int32_t value);
+
+static inline void
+_Py_atomic_store_int64_relaxed(int64_t *obj, int64_t value);
+
+static inline void
+_Py_atomic_store_intptr_relaxed(intptr_t *obj, intptr_t value);
+
+static inline void
+_Py_atomic_store_uint8_relaxed(uint8_t* obj, uint8_t value);
+
+static inline void
+_Py_atomic_store_uint16_relaxed(uint16_t *obj, uint16_t value);
+
+static inline void
+_Py_atomic_store_uint32_relaxed(uint32_t *obj, uint32_t value);
+
+static inline void
+_Py_atomic_store_uint64_relaxed(uint64_t *obj, uint64_t value);
+
+static inline void
+_Py_atomic_store_uintptr_relaxed(uintptr_t *obj, uintptr_t value);
+
+static inline void
+_Py_atomic_store_uint_relaxed(unsigned int *obj, unsigned int value);
+
+static inline void
+_Py_atomic_store_ptr_relaxed(void *obj, void *value);
+
+static inline void
+_Py_atomic_store_ssize_relaxed(Py_ssize_t *obj, Py_ssize_t value);
+
+static inline void
+_Py_atomic_store_ullong_relaxed(unsigned long long *obj,
+                                unsigned long long value);
+
+
+// --- _Py_atomic_load_ptr_acquire / _Py_atomic_store_ptr_release ------------
+
+// Loads `*obj` (acquire operation)
+static inline void *
+_Py_atomic_load_ptr_acquire(const void *obj);
+
+static inline uintptr_t
+_Py_atomic_load_uintptr_acquire(const uintptr_t *obj);
+
+// Stores `*obj = value` (release operation)
+static inline void
+_Py_atomic_store_ptr_release(void *obj, void *value);
+
+static inline void
+_Py_atomic_store_uintptr_release(uintptr_t *obj, uintptr_t value);
+
+static inline void
+_Py_atomic_store_ssize_release(Py_ssize_t *obj, Py_ssize_t value);
+
+static inline void
+_Py_atomic_store_int_release(int *obj, int value);
+
+static inline int
+_Py_atomic_load_int_acquire(const int *obj);
+
+static inline void
+_Py_atomic_store_uint32_release(uint32_t *obj, uint32_t value);
+
+static inline void
+_Py_atomic_store_uint64_release(uint64_t *obj, uint64_t value);
+
+static inline uint64_t
+_Py_atomic_load_uint64_acquire(const uint64_t *obj);
+
+static inline uint32_t
+_Py_atomic_load_uint32_acquire(const uint32_t *obj);
+
+static inline Py_ssize_t
+_Py_atomic_load_ssize_acquire(const Py_ssize_t *obj);
+
+
+
+
+// --- _Py_atomic_fence ------------------------------------------------------
+
+// Sequential consistency fence. C11 fences have complex semantics. When
+// possible, use the atomic operations on variables defined above, which
+// generally do not require explicit use of a fence.
+// See https://en.cppreference.com/w/cpp/atomic/atomic_thread_fence
+static inline void _Py_atomic_fence_seq_cst(void);
+
+// Acquire fence
+static inline void _Py_atomic_fence_acquire(void);
+
+// Release fence
+static inline void _Py_atomic_fence_release(void);
+
+
+#ifndef _Py_USE_GCC_BUILTIN_ATOMICS
+#  if defined(__GNUC__) && (__GNUC__ > 4 || (__GNUC__ == 4 && __GNUC_MINOR__ >= 8))
+#    define _Py_USE_GCC_BUILTIN_ATOMICS 1
+#  elif defined(__clang__)
+#    if __has_builtin(__atomic_load)
+#      define _Py_USE_GCC_BUILTIN_ATOMICS 1
+#    endif
+#  endif
+#endif
+
+#if _Py_USE_GCC_BUILTIN_ATOMICS
+#  define Py_ATOMIC_GCC_H
+#  include "pyatomic_gcc.h"
+#  undef Py_ATOMIC_GCC_H
+#elif __STDC_VERSION__ >= 201112L && !defined(__STDC_NO_ATOMICS__)
+#  define Py_ATOMIC_STD_H
+#  include "pyatomic_std.h"
+#  undef Py_ATOMIC_STD_H
+#elif defined(_MSC_VER)
+#  define Py_ATOMIC_MSC_H
+#  include "pyatomic_msc.h"
+#  undef Py_ATOMIC_MSC_H
+#else
+#  error "no available pyatomic implementation for this platform/compiler"
+#endif
+
+
+// --- aliases ---------------------------------------------------------------
+
+#if SIZEOF_LONG == 8
+# define _Py_atomic_load_ulong(p) \
+    _Py_atomic_load_uint64((uint64_t *)p)
+# define _Py_atomic_load_ulong_relaxed(p) \
+    _Py_atomic_load_uint64_relaxed((uint64_t *)p)
+# define _Py_atomic_store_ulong(p, v) \
+    _Py_atomic_store_uint64((uint64_t *)p, v)
+# define _Py_atomic_store_ulong_relaxed(p, v) \
+    _Py_atomic_store_uint64_relaxed((uint64_t *)p, v)
+#elif SIZEOF_LONG == 4
+# define _Py_atomic_load_ulong(p) \
+    _Py_atomic_load_uint32((uint32_t *)p)
+# define _Py_atomic_load_ulong_relaxed(p) \
+    _Py_atomic_load_uint32_relaxed((uint32_t *)p)
+# define _Py_atomic_store_ulong(p, v) \
+    _Py_atomic_store_uint32((uint32_t *)p, v)
+# define _Py_atomic_store_ulong_relaxed(p, v) \
+    _Py_atomic_store_uint32_relaxed((uint32_t *)p, v)
+#else
+# error "long must be 4 or 8 bytes in size"
+#endif  // SIZEOF_LONG

Dependencies/Python/include/cpython/pyatomic_gcc.h

@@ -0,0 +1,551 @@
+// This is the implementation of Python atomic operations using GCC's built-in
+// functions that match the C+11 memory model. This implementation is preferred
+// for GCC compatible compilers, such as Clang. These functions are available
+// in GCC 4.8+ without needing to compile with --std=c11 or --std=gnu11.
+
+#ifndef Py_ATOMIC_GCC_H
+#  error "this header file must not be included directly"
+#endif
+
+
+// --- _Py_atomic_add --------------------------------------------------------
+
+static inline int
+_Py_atomic_add_int(int *obj, int value)
+{ return __atomic_fetch_add(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline int8_t
+_Py_atomic_add_int8(int8_t *obj, int8_t value)
+{ return __atomic_fetch_add(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline int16_t
+_Py_atomic_add_int16(int16_t *obj, int16_t value)
+{ return __atomic_fetch_add(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline int32_t
+_Py_atomic_add_int32(int32_t *obj, int32_t value)
+{ return __atomic_fetch_add(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline int64_t
+_Py_atomic_add_int64(int64_t *obj, int64_t value)
+{ return __atomic_fetch_add(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline intptr_t
+_Py_atomic_add_intptr(intptr_t *obj, intptr_t value)
+{ return __atomic_fetch_add(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline unsigned int
+_Py_atomic_add_uint(unsigned int *obj, unsigned int value)
+{ return __atomic_fetch_add(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline uint8_t
+_Py_atomic_add_uint8(uint8_t *obj, uint8_t value)
+{ return __atomic_fetch_add(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline uint16_t
+_Py_atomic_add_uint16(uint16_t *obj, uint16_t value)
+{ return __atomic_fetch_add(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline uint32_t
+_Py_atomic_add_uint32(uint32_t *obj, uint32_t value)
+{ return __atomic_fetch_add(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline uint64_t
+_Py_atomic_add_uint64(uint64_t *obj, uint64_t value)
+{ return __atomic_fetch_add(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline uintptr_t
+_Py_atomic_add_uintptr(uintptr_t *obj, uintptr_t value)
+{ return __atomic_fetch_add(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline Py_ssize_t
+_Py_atomic_add_ssize(Py_ssize_t *obj, Py_ssize_t value)
+{ return __atomic_fetch_add(obj, value, __ATOMIC_SEQ_CST); }
+
+
+// --- _Py_atomic_compare_exchange -------------------------------------------
+
+static inline int
+_Py_atomic_compare_exchange_int(int *obj, int *expected, int desired)
+{ return __atomic_compare_exchange_n(obj, expected, desired, 0,
+                                     __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+
+static inline int
+_Py_atomic_compare_exchange_int8(int8_t *obj, int8_t *expected, int8_t desired)
+{ return __atomic_compare_exchange_n(obj, expected, desired, 0,
+                                     __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+
+static inline int
+_Py_atomic_compare_exchange_int16(int16_t *obj, int16_t *expected, int16_t desired)
+{ return __atomic_compare_exchange_n(obj, expected, desired, 0,
+                                     __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+
+static inline int
+_Py_atomic_compare_exchange_int32(int32_t *obj, int32_t *expected, int32_t desired)
+{ return __atomic_compare_exchange_n(obj, expected, desired, 0,
+                                     __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+
+static inline int
+_Py_atomic_compare_exchange_int64(int64_t *obj, int64_t *expected, int64_t desired)
+{ return __atomic_compare_exchange_n(obj, expected, desired, 0,
+                                     __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+
+static inline int
+_Py_atomic_compare_exchange_intptr(intptr_t *obj, intptr_t *expected, intptr_t desired)
+{ return __atomic_compare_exchange_n(obj, expected, desired, 0,
+                                     __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+
+static inline int
+_Py_atomic_compare_exchange_uint(unsigned int *obj, unsigned int *expected, unsigned int desired)
+{ return __atomic_compare_exchange_n(obj, expected, desired, 0,
+                                     __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+
+static inline int
+_Py_atomic_compare_exchange_uint8(uint8_t *obj, uint8_t *expected, uint8_t desired)
+{ return __atomic_compare_exchange_n(obj, expected, desired, 0,
+                                     __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+
+static inline int
+_Py_atomic_compare_exchange_uint16(uint16_t *obj, uint16_t *expected, uint16_t desired)
+{ return __atomic_compare_exchange_n(obj, expected, desired, 0,
+                                     __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+
+static inline int
+_Py_atomic_compare_exchange_uint32(uint32_t *obj, uint32_t *expected, uint32_t desired)
+{ return __atomic_compare_exchange_n(obj, expected, desired, 0,
+                                     __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+
+static inline int
+_Py_atomic_compare_exchange_uint64(uint64_t *obj, uint64_t *expected, uint64_t desired)
+{ return __atomic_compare_exchange_n(obj, expected, desired, 0,
+                                     __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+
+static inline int
+_Py_atomic_compare_exchange_uintptr(uintptr_t *obj, uintptr_t *expected, uintptr_t desired)
+{ return __atomic_compare_exchange_n(obj, expected, desired, 0,
+                                     __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+
+static inline int
+_Py_atomic_compare_exchange_ssize(Py_ssize_t *obj, Py_ssize_t *expected, Py_ssize_t desired)
+{ return __atomic_compare_exchange_n(obj, expected, desired, 0,
+                                     __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+
+static inline int
+_Py_atomic_compare_exchange_ptr(void *obj, void *expected, void *desired)
+{ return __atomic_compare_exchange_n((void **)obj, (void **)expected, desired, 0,
+                                     __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+
+
+// --- _Py_atomic_exchange ---------------------------------------------------
+
+static inline int
+_Py_atomic_exchange_int(int *obj, int value)
+{ return __atomic_exchange_n(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline int8_t
+_Py_atomic_exchange_int8(int8_t *obj, int8_t value)
+{ return __atomic_exchange_n(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline int16_t
+_Py_atomic_exchange_int16(int16_t *obj, int16_t value)
+{ return __atomic_exchange_n(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline int32_t
+_Py_atomic_exchange_int32(int32_t *obj, int32_t value)
+{ return __atomic_exchange_n(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline int64_t
+_Py_atomic_exchange_int64(int64_t *obj, int64_t value)
+{ return __atomic_exchange_n(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline intptr_t
+_Py_atomic_exchange_intptr(intptr_t *obj, intptr_t value)
+{ return __atomic_exchange_n(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline unsigned int
+_Py_atomic_exchange_uint(unsigned int *obj, unsigned int value)
+{ return __atomic_exchange_n(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline uint8_t
+_Py_atomic_exchange_uint8(uint8_t *obj, uint8_t value)
+{ return __atomic_exchange_n(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline uint16_t
+_Py_atomic_exchange_uint16(uint16_t *obj, uint16_t value)
+{ return __atomic_exchange_n(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline uint32_t
+_Py_atomic_exchange_uint32(uint32_t *obj, uint32_t value)
+{ return __atomic_exchange_n(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline uint64_t
+_Py_atomic_exchange_uint64(uint64_t *obj, uint64_t value)
+{ return __atomic_exchange_n(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline uintptr_t
+_Py_atomic_exchange_uintptr(uintptr_t *obj, uintptr_t value)
+{ return __atomic_exchange_n(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline Py_ssize_t
+_Py_atomic_exchange_ssize(Py_ssize_t *obj, Py_ssize_t value)
+{ return __atomic_exchange_n(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline void *
+_Py_atomic_exchange_ptr(void *obj, void *value)
+{ return __atomic_exchange_n((void **)obj, value, __ATOMIC_SEQ_CST); }
+
+
+// --- _Py_atomic_and --------------------------------------------------------
+
+static inline uint8_t
+_Py_atomic_and_uint8(uint8_t *obj, uint8_t value)
+{ return __atomic_fetch_and(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline uint16_t
+_Py_atomic_and_uint16(uint16_t *obj, uint16_t value)
+{ return __atomic_fetch_and(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline uint32_t
+_Py_atomic_and_uint32(uint32_t *obj, uint32_t value)
+{ return __atomic_fetch_and(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline uint64_t
+_Py_atomic_and_uint64(uint64_t *obj, uint64_t value)
+{ return __atomic_fetch_and(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline uintptr_t
+_Py_atomic_and_uintptr(uintptr_t *obj, uintptr_t value)
+{ return __atomic_fetch_and(obj, value, __ATOMIC_SEQ_CST); }
+
+
+// --- _Py_atomic_or ---------------------------------------------------------
+
+static inline uint8_t
+_Py_atomic_or_uint8(uint8_t *obj, uint8_t value)
+{ return __atomic_fetch_or(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline uint16_t
+_Py_atomic_or_uint16(uint16_t *obj, uint16_t value)
+{ return __atomic_fetch_or(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline uint32_t
+_Py_atomic_or_uint32(uint32_t *obj, uint32_t value)
+{ return __atomic_fetch_or(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline uint64_t
+_Py_atomic_or_uint64(uint64_t *obj, uint64_t value)
+{ return __atomic_fetch_or(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline uintptr_t
+_Py_atomic_or_uintptr(uintptr_t *obj, uintptr_t value)
+{ return __atomic_fetch_or(obj, value, __ATOMIC_SEQ_CST); }
+
+
+// --- _Py_atomic_load -------------------------------------------------------
+
+static inline int
+_Py_atomic_load_int(const int *obj)
+{ return __atomic_load_n(obj, __ATOMIC_SEQ_CST); }
+
+static inline int8_t
+_Py_atomic_load_int8(const int8_t *obj)
+{ return __atomic_load_n(obj, __ATOMIC_SEQ_CST); }
+
+static inline int16_t
+_Py_atomic_load_int16(const int16_t *obj)
+{ return __atomic_load_n(obj, __ATOMIC_SEQ_CST); }
+
+static inline int32_t
+_Py_atomic_load_int32(const int32_t *obj)
+{ return __atomic_load_n(obj, __ATOMIC_SEQ_CST); }
+
+static inline int64_t
+_Py_atomic_load_int64(const int64_t *obj)
+{ return __atomic_load_n(obj, __ATOMIC_SEQ_CST); }
+
+static inline intptr_t
+_Py_atomic_load_intptr(const intptr_t *obj)
+{ return __atomic_load_n(obj, __ATOMIC_SEQ_CST); }
+
+static inline uint8_t
+_Py_atomic_load_uint8(const uint8_t *obj)
+{ return __atomic_load_n(obj, __ATOMIC_SEQ_CST); }
+
+static inline uint16_t
+_Py_atomic_load_uint16(const uint16_t *obj)
+{ return __atomic_load_n(obj, __ATOMIC_SEQ_CST); }
+
+static inline uint32_t
+_Py_atomic_load_uint32(const uint32_t *obj)
+{ return __atomic_load_n(obj, __ATOMIC_SEQ_CST); }
+
+static inline uint64_t
+_Py_atomic_load_uint64(const uint64_t *obj)
+{ return __atomic_load_n(obj, __ATOMIC_SEQ_CST); }
+
+static inline uintptr_t
+_Py_atomic_load_uintptr(const uintptr_t *obj)
+{ return __atomic_load_n(obj, __ATOMIC_SEQ_CST); }
+
+static inline unsigned int
+_Py_atomic_load_uint(const unsigned int *obj)
+{ return __atomic_load_n(obj, __ATOMIC_SEQ_CST); }
+
+static inline Py_ssize_t
+_Py_atomic_load_ssize(const Py_ssize_t *obj)
+{ return __atomic_load_n(obj, __ATOMIC_SEQ_CST); }
+
+static inline void *
+_Py_atomic_load_ptr(const void *obj)
+{ return (void *)__atomic_load_n((void * const *)obj, __ATOMIC_SEQ_CST); }
+
+
+// --- _Py_atomic_load_relaxed -----------------------------------------------
+
+static inline int
+_Py_atomic_load_int_relaxed(const int *obj)
+{ return __atomic_load_n(obj, __ATOMIC_RELAXED); }
+
+static inline int8_t
+_Py_atomic_load_int8_relaxed(const int8_t *obj)
+{ return __atomic_load_n(obj, __ATOMIC_RELAXED); }
+
+static inline int16_t
+_Py_atomic_load_int16_relaxed(const int16_t *obj)
+{ return __atomic_load_n(obj, __ATOMIC_RELAXED); }
+
+static inline int32_t
+_Py_atomic_load_int32_relaxed(const int32_t *obj)
+{ return __atomic_load_n(obj, __ATOMIC_RELAXED); }
+
+static inline int64_t
+_Py_atomic_load_int64_relaxed(const int64_t *obj)
+{ return __atomic_load_n(obj, __ATOMIC_RELAXED); }
+
+static inline intptr_t
+_Py_atomic_load_intptr_relaxed(const intptr_t *obj)
+{ return __atomic_load_n(obj, __ATOMIC_RELAXED); }
+
+static inline uint8_t
+_Py_atomic_load_uint8_relaxed(const uint8_t *obj)
+{ return __atomic_load_n(obj, __ATOMIC_RELAXED); }
+
+static inline uint16_t
+_Py_atomic_load_uint16_relaxed(const uint16_t *obj)
+{ return __atomic_load_n(obj, __ATOMIC_RELAXED); }
+
+static inline uint32_t
+_Py_atomic_load_uint32_relaxed(const uint32_t *obj)
+{ return __atomic_load_n(obj, __ATOMIC_RELAXED); }
+
+static inline uint64_t
+_Py_atomic_load_uint64_relaxed(const uint64_t *obj)
+{ return __atomic_load_n(obj, __ATOMIC_RELAXED); }
+
+static inline uintptr_t
+_Py_atomic_load_uintptr_relaxed(const uintptr_t *obj)
+{ return __atomic_load_n(obj, __ATOMIC_RELAXED); }
+
+static inline unsigned int
+_Py_atomic_load_uint_relaxed(const unsigned int *obj)
+{ return __atomic_load_n(obj, __ATOMIC_RELAXED); }
+
+static inline Py_ssize_t
+_Py_atomic_load_ssize_relaxed(const Py_ssize_t *obj)
+{ return __atomic_load_n(obj, __ATOMIC_RELAXED); }
+
+static inline void *
+_Py_atomic_load_ptr_relaxed(const void *obj)
+{ return (void *)__atomic_load_n((void * const *)obj, __ATOMIC_RELAXED); }
+
+static inline unsigned long long
+_Py_atomic_load_ullong_relaxed(const unsigned long long *obj)
+{ return __atomic_load_n(obj, __ATOMIC_RELAXED); }
+
+
+// --- _Py_atomic_store ------------------------------------------------------
+
+static inline void
+_Py_atomic_store_int(int *obj, int value)
+{ __atomic_store_n(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline void
+_Py_atomic_store_int8(int8_t *obj, int8_t value)
+{ __atomic_store_n(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline void
+_Py_atomic_store_int16(int16_t *obj, int16_t value)
+{ __atomic_store_n(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline void
+_Py_atomic_store_int32(int32_t *obj, int32_t value)
+{ __atomic_store_n(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline void
+_Py_atomic_store_int64(int64_t *obj, int64_t value)
+{ __atomic_store_n(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline void
+_Py_atomic_store_intptr(intptr_t *obj, intptr_t value)
+{ __atomic_store_n(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline void
+_Py_atomic_store_uint8(uint8_t *obj, uint8_t value)
+{ __atomic_store_n(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline void
+_Py_atomic_store_uint16(uint16_t *obj, uint16_t value)
+{ __atomic_store_n(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline void
+_Py_atomic_store_uint32(uint32_t *obj, uint32_t value)
+{ __atomic_store_n(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline void
+_Py_atomic_store_uint64(uint64_t *obj, uint64_t value)
+{ __atomic_store_n(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline void
+_Py_atomic_store_uintptr(uintptr_t *obj, uintptr_t value)
+{ __atomic_store_n(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline void
+_Py_atomic_store_uint(unsigned int *obj, unsigned int value)
+{ __atomic_store_n(obj, value, __ATOMIC_SEQ_CST); }
+
+static inline void
+_Py_atomic_store_ptr(void *obj, void *value)
+{ __atomic_store_n((void **)obj, value, __ATOMIC_SEQ_CST); }
+
+static inline void
+_Py_atomic_store_ssize(Py_ssize_t *obj, Py_ssize_t value)
+{ __atomic_store_n(obj, value, __ATOMIC_SEQ_CST); }
+
+
+// --- _Py_atomic_store_relaxed ----------------------------------------------
+
+static inline void
+_Py_atomic_store_int_relaxed(int *obj, int value)
+{ __atomic_store_n(obj, value, __ATOMIC_RELAXED); }
+
+static inline void
+_Py_atomic_store_int8_relaxed(int8_t *obj, int8_t value)
+{ __atomic_store_n(obj, value, __ATOMIC_RELAXED); }
+
+static inline void
+_Py_atomic_store_int16_relaxed(int16_t *obj, int16_t value)
+{ __atomic_store_n(obj, value, __ATOMIC_RELAXED); }
+
+static inline void
+_Py_atomic_store_int32_relaxed(int32_t *obj, int32_t value)
+{ __atomic_store_n(obj, value, __ATOMIC_RELAXED); }
+
+static inline void
+_Py_atomic_store_int64_relaxed(int64_t *obj, int64_t value)
+{ __atomic_store_n(obj, value, __ATOMIC_RELAXED); }
+
+static inline void
+_Py_atomic_store_intptr_relaxed(intptr_t *obj, intptr_t value)
+{ __atomic_store_n(obj, value, __ATOMIC_RELAXED); }
+
+static inline void
+_Py_atomic_store_uint8_relaxed(uint8_t *obj, uint8_t value)
+{ __atomic_store_n(obj, value, __ATOMIC_RELAXED); }
+
+static inline void
+_Py_atomic_store_uint16_relaxed(uint16_t *obj, uint16_t value)
+{ __atomic_store_n(obj, value, __ATOMIC_RELAXED); }
+
+static inline void
+_Py_atomic_store_uint32_relaxed(uint32_t *obj, uint32_t value)
+{ __atomic_store_n(obj, value, __ATOMIC_RELAXED); }
+
+static inline void
+_Py_atomic_store_uint64_relaxed(uint64_t *obj, uint64_t value)
+{ __atomic_store_n(obj, value, __ATOMIC_RELAXED); }
+
+static inline void
+_Py_atomic_store_uintptr_relaxed(uintptr_t *obj, uintptr_t value)
+{ __atomic_store_n(obj, value, __ATOMIC_RELAXED); }
+
+static inline void
+_Py_atomic_store_uint_relaxed(unsigned int *obj, unsigned int value)
+{ __atomic_store_n(obj, value, __ATOMIC_RELAXED); }
+
+static inline void
+_Py_atomic_store_ptr_relaxed(void *obj, void *value)
+{ __atomic_store_n((void **)obj, value, __ATOMIC_RELAXED); }
+
+static inline void
+_Py_atomic_store_ssize_relaxed(Py_ssize_t *obj, Py_ssize_t value)
+{ __atomic_store_n(obj, value, __ATOMIC_RELAXED); }
+
+static inline void
+_Py_atomic_store_ullong_relaxed(unsigned long long *obj,
+                                unsigned long long value)
+{ __atomic_store_n(obj, value, __ATOMIC_RELAXED); }
+
+
+// --- _Py_atomic_load_ptr_acquire / _Py_atomic_store_ptr_release ------------
+
+static inline void *
+_Py_atomic_load_ptr_acquire(const void *obj)
+{ return (void *)__atomic_load_n((void * const *)obj, __ATOMIC_ACQUIRE); }
+
+static inline uintptr_t
+_Py_atomic_load_uintptr_acquire(const uintptr_t *obj)
+{ return (uintptr_t)__atomic_load_n(obj, __ATOMIC_ACQUIRE); }
+
+static inline void
+_Py_atomic_store_ptr_release(void *obj, void *value)
+{ __atomic_store_n((void **)obj, value, __ATOMIC_RELEASE); }
+
+static inline void
+_Py_atomic_store_uintptr_release(uintptr_t *obj, uintptr_t value)
+{ __atomic_store_n(obj, value, __ATOMIC_RELEASE); }
+
+static inline void
+_Py_atomic_store_int_release(int *obj, int value)
+{ __atomic_store_n(obj, value, __ATOMIC_RELEASE); }
+
+static inline void
+_Py_atomic_store_ssize_release(Py_ssize_t *obj, Py_ssize_t value)
+{ __atomic_store_n(obj, value, __ATOMIC_RELEASE); }
+
+static inline int
+_Py_atomic_load_int_acquire(const int *obj)
+{ return __atomic_load_n(obj, __ATOMIC_ACQUIRE); }
+
+static inline void
+_Py_atomic_store_uint32_release(uint32_t *obj, uint32_t value)
+{ __atomic_store_n(obj, value, __ATOMIC_RELEASE); }
+
+static inline void
+_Py_atomic_store_uint64_release(uint64_t *obj, uint64_t value)
+{ __atomic_store_n(obj, value, __ATOMIC_RELEASE); }
+
+static inline uint64_t
+_Py_atomic_load_uint64_acquire(const uint64_t *obj)
+{ return __atomic_load_n(obj, __ATOMIC_ACQUIRE); }
+
+static inline uint32_t
+_Py_atomic_load_uint32_acquire(const uint32_t *obj)
+{ return __atomic_load_n(obj, __ATOMIC_ACQUIRE); }
+
+static inline Py_ssize_t
+_Py_atomic_load_ssize_acquire(const Py_ssize_t *obj)
+{ return __atomic_load_n(obj, __ATOMIC_ACQUIRE); }
+
+// --- _Py_atomic_fence ------------------------------------------------------
+
+static inline void
+_Py_atomic_fence_seq_cst(void)
+{ __atomic_thread_fence(__ATOMIC_SEQ_CST); }
+
+ static inline void
+_Py_atomic_fence_acquire(void)
+{ __atomic_thread_fence(__ATOMIC_ACQUIRE); }
+
+ static inline void
+_Py_atomic_fence_release(void)
+{ __atomic_thread_fence(__ATOMIC_RELEASE); }

Dependencies/Python/include/cpython/pyatomic_std.h

@@ -0,0 +1,976 @@
+// This is the implementation of Python atomic operations using C++11 or C11
+// atomics. Note that the pyatomic_gcc.h implementation is preferred for GCC
+// compatible compilers, even if they support C++11 atomics.
+
+#ifndef Py_ATOMIC_STD_H
+#  error "this header file must not be included directly"
+#endif
+
+#ifdef __cplusplus
+extern "C++" {
+#  include <atomic>
+}
+#  define _Py_USING_STD using namespace std
+#  define _Atomic(tp) atomic<tp>
+#else
+#  define  _Py_USING_STD
+#  include <stdatomic.h>
+#endif
+
+
+// --- _Py_atomic_add --------------------------------------------------------
+
+static inline int
+_Py_atomic_add_int(int *obj, int value)
+{
+    _Py_USING_STD;
+    return atomic_fetch_add((_Atomic(int)*)obj, value);
+}
+
+static inline int8_t
+_Py_atomic_add_int8(int8_t *obj, int8_t value)
+{
+    _Py_USING_STD;
+    return atomic_fetch_add((_Atomic(int8_t)*)obj, value);
+}
+
+static inline int16_t
+_Py_atomic_add_int16(int16_t *obj, int16_t value)
+{
+    _Py_USING_STD;
+    return atomic_fetch_add((_Atomic(int16_t)*)obj, value);
+}
+
+static inline int32_t
+_Py_atomic_add_int32(int32_t *obj, int32_t value)
+{
+    _Py_USING_STD;
+    return atomic_fetch_add((_Atomic(int32_t)*)obj, value);
+}
+
+static inline int64_t
+_Py_atomic_add_int64(int64_t *obj, int64_t value)
+{
+    _Py_USING_STD;
+    return atomic_fetch_add((_Atomic(int64_t)*)obj, value);
+}
+
+static inline intptr_t
+_Py_atomic_add_intptr(intptr_t *obj, intptr_t value)
+{
+    _Py_USING_STD;
+    return atomic_fetch_add((_Atomic(intptr_t)*)obj, value);
+}
+
+static inline unsigned int
+_Py_atomic_add_uint(unsigned int *obj, unsigned int value)
+{
+    _Py_USING_STD;
+    return atomic_fetch_add((_Atomic(unsigned int)*)obj, value);
+}
+
+static inline uint8_t
+_Py_atomic_add_uint8(uint8_t *obj, uint8_t value)
+{
+    _Py_USING_STD;
+    return atomic_fetch_add((_Atomic(uint8_t)*)obj, value);
+}
+
+static inline uint16_t
+_Py_atomic_add_uint16(uint16_t *obj, uint16_t value)
+{
+    _Py_USING_STD;
+    return atomic_fetch_add((_Atomic(uint16_t)*)obj, value);
+}
+
+static inline uint32_t
+_Py_atomic_add_uint32(uint32_t *obj, uint32_t value)
+{
+    _Py_USING_STD;
+    return atomic_fetch_add((_Atomic(uint32_t)*)obj, value);
+}
+
+static inline uint64_t
+_Py_atomic_add_uint64(uint64_t *obj, uint64_t value)
+{
+    _Py_USING_STD;
+    return atomic_fetch_add((_Atomic(uint64_t)*)obj, value);
+}
+
+static inline uintptr_t
+_Py_atomic_add_uintptr(uintptr_t *obj, uintptr_t value)
+{
+    _Py_USING_STD;
+    return atomic_fetch_add((_Atomic(uintptr_t)*)obj, value);
+}
+
+static inline Py_ssize_t
+_Py_atomic_add_ssize(Py_ssize_t *obj, Py_ssize_t value)
+{
+    _Py_USING_STD;
+    return atomic_fetch_add((_Atomic(Py_ssize_t)*)obj, value);
+}
+
+
+// --- _Py_atomic_compare_exchange -------------------------------------------
+
+static inline int
+_Py_atomic_compare_exchange_int(int *obj, int *expected, int desired)
+{
+    _Py_USING_STD;
+    return atomic_compare_exchange_strong((_Atomic(int)*)obj,
+                                          expected, desired);
+}
+
+static inline int
+_Py_atomic_compare_exchange_int8(int8_t *obj, int8_t *expected, int8_t desired)
+{
+    _Py_USING_STD;
+    return atomic_compare_exchange_strong((_Atomic(int8_t)*)obj,
+                                          expected, desired);
+}
+
+static inline int
+_Py_atomic_compare_exchange_int16(int16_t *obj, int16_t *expected, int16_t desired)
+{
+    _Py_USING_STD;
+    return atomic_compare_exchange_strong((_Atomic(int16_t)*)obj,
+                                          expected, desired);
+}
+
+static inline int
+_Py_atomic_compare_exchange_int32(int32_t *obj, int32_t *expected, int32_t desired)
+{
+    _Py_USING_STD;
+    return atomic_compare_exchange_strong((_Atomic(int32_t)*)obj,
+                                          expected, desired);
+}
+
+static inline int
+_Py_atomic_compare_exchange_int64(int64_t *obj, int64_t *expected, int64_t desired)
+{
+    _Py_USING_STD;
+    return atomic_compare_exchange_strong((_Atomic(int64_t)*)obj,
+                                          expected, desired);
+}
+
+static inline int
+_Py_atomic_compare_exchange_intptr(intptr_t *obj, intptr_t *expected, intptr_t desired)
+{
+    _Py_USING_STD;
+    return atomic_compare_exchange_strong((_Atomic(intptr_t)*)obj,
+                                          expected, desired);
+}
+
+static inline int
+_Py_atomic_compare_exchange_uint(unsigned int *obj, unsigned int *expected, unsigned int desired)
+{
+    _Py_USING_STD;
+    return atomic_compare_exchange_strong((_Atomic(unsigned int)*)obj,
+                                          expected, desired);
+}
+
+static inline int
+_Py_atomic_compare_exchange_uint8(uint8_t *obj, uint8_t *expected, uint8_t desired)
+{
+    _Py_USING_STD;
+    return atomic_compare_exchange_strong((_Atomic(uint8_t)*)obj,
+                                          expected, desired);
+}
+
+static inline int
+_Py_atomic_compare_exchange_uint16(uint16_t *obj, uint16_t *expected, uint16_t desired)
+{
+    _Py_USING_STD;
+    return atomic_compare_exchange_strong((_Atomic(uint16_t)*)obj,
+                                          expected, desired);
+}
+
+static inline int
+_Py_atomic_compare_exchange_uint32(uint32_t *obj, uint32_t *expected, uint32_t desired)
+{
+    _Py_USING_STD;
+    return atomic_compare_exchange_strong((_Atomic(uint32_t)*)obj,
+                                          expected, desired);
+}
+
+static inline int
+_Py_atomic_compare_exchange_uint64(uint64_t *obj, uint64_t *expected, uint64_t desired)
+{
+    _Py_USING_STD;
+    return atomic_compare_exchange_strong((_Atomic(uint64_t)*)obj,
+                                          expected, desired);
+}
+
+static inline int
+_Py_atomic_compare_exchange_uintptr(uintptr_t *obj, uintptr_t *expected, uintptr_t desired)
+{
+    _Py_USING_STD;
+    return atomic_compare_exchange_strong((_Atomic(uintptr_t)*)obj,
+                                          expected, desired);
+}
+
+static inline int
+_Py_atomic_compare_exchange_ssize(Py_ssize_t *obj, Py_ssize_t *expected, Py_ssize_t desired)
+{
+    _Py_USING_STD;
+    return atomic_compare_exchange_strong((_Atomic(Py_ssize_t)*)obj,
+                                          expected, desired);
+}
+
+static inline int
+_Py_atomic_compare_exchange_ptr(void *obj, void *expected, void *desired)
+{
+    _Py_USING_STD;
+    return atomic_compare_exchange_strong((_Atomic(void *)*)obj,
+                                          (void **)expected, desired);
+}
+
+
+// --- _Py_atomic_exchange ---------------------------------------------------
+
+static inline int
+_Py_atomic_exchange_int(int *obj, int value)
+{
+    _Py_USING_STD;
+    return atomic_exchange((_Atomic(int)*)obj, value);
+}
+
+static inline int8_t
+_Py_atomic_exchange_int8(int8_t *obj, int8_t value)
+{
+    _Py_USING_STD;
+    return atomic_exchange((_Atomic(int8_t)*)obj, value);
+}
+
+static inline int16_t
+_Py_atomic_exchange_int16(int16_t *obj, int16_t value)
+{
+    _Py_USING_STD;
+    return atomic_exchange((_Atomic(int16_t)*)obj, value);
+}
+
+static inline int32_t
+_Py_atomic_exchange_int32(int32_t *obj, int32_t value)
+{
+    _Py_USING_STD;
+    return atomic_exchange((_Atomic(int32_t)*)obj, value);
+}
+
+static inline int64_t
+_Py_atomic_exchange_int64(int64_t *obj, int64_t value)
+{
+    _Py_USING_STD;
+    return atomic_exchange((_Atomic(int64_t)*)obj, value);
+}
+
+static inline intptr_t
+_Py_atomic_exchange_intptr(intptr_t *obj, intptr_t value)
+{
+    _Py_USING_STD;
+    return atomic_exchange((_Atomic(intptr_t)*)obj, value);
+}
+
+static inline unsigned int
+_Py_atomic_exchange_uint(unsigned int *obj, unsigned int value)
+{
+    _Py_USING_STD;
+    return atomic_exchange((_Atomic(unsigned int)*)obj, value);
+}
+
+static inline uint8_t
+_Py_atomic_exchange_uint8(uint8_t *obj, uint8_t value)
+{
+    _Py_USING_STD;
+    return atomic_exchange((_Atomic(uint8_t)*)obj, value);
+}
+
+static inline uint16_t
+_Py_atomic_exchange_uint16(uint16_t *obj, uint16_t value)
+{
+    _Py_USING_STD;
+    return atomic_exchange((_Atomic(uint16_t)*)obj, value);
+}
+
+static inline uint32_t
+_Py_atomic_exchange_uint32(uint32_t *obj, uint32_t value)
+{
+    _Py_USING_STD;
+    return atomic_exchange((_Atomic(uint32_t)*)obj, value);
+}
+
+static inline uint64_t
+_Py_atomic_exchange_uint64(uint64_t *obj, uint64_t value)
+{
+    _Py_USING_STD;
+    return atomic_exchange((_Atomic(uint64_t)*)obj, value);
+}
+
+static inline uintptr_t
+_Py_atomic_exchange_uintptr(uintptr_t *obj, uintptr_t value)
+{
+    _Py_USING_STD;
+    return atomic_exchange((_Atomic(uintptr_t)*)obj, value);
+}
+
+static inline Py_ssize_t
+_Py_atomic_exchange_ssize(Py_ssize_t *obj, Py_ssize_t value)
+{
+    _Py_USING_STD;
+    return atomic_exchange((_Atomic(Py_ssize_t)*)obj, value);
+}
+
+static inline void*
+_Py_atomic_exchange_ptr(void *obj, void *value)
+{
+    _Py_USING_STD;
+    return atomic_exchange((_Atomic(void *)*)obj, value);
+}
+
+
+// --- _Py_atomic_and --------------------------------------------------------
+
+static inline uint8_t
+_Py_atomic_and_uint8(uint8_t *obj, uint8_t value)
+{
+    _Py_USING_STD;
+    return atomic_fetch_and((_Atomic(uint8_t)*)obj, value);
+}
+
+static inline uint16_t
+_Py_atomic_and_uint16(uint16_t *obj, uint16_t value)
+{
+    _Py_USING_STD;
+    return atomic_fetch_and((_Atomic(uint16_t)*)obj, value);
+}
+
+static inline uint32_t
+_Py_atomic_and_uint32(uint32_t *obj, uint32_t value)
+{
+    _Py_USING_STD;
+    return atomic_fetch_and((_Atomic(uint32_t)*)obj, value);
+}
+
+static inline uint64_t
+_Py_atomic_and_uint64(uint64_t *obj, uint64_t value)
+{
+    _Py_USING_STD;
+    return atomic_fetch_and((_Atomic(uint64_t)*)obj, value);
+}
+
+static inline uintptr_t
+_Py_atomic_and_uintptr(uintptr_t *obj, uintptr_t value)
+{
+    _Py_USING_STD;
+    return atomic_fetch_and((_Atomic(uintptr_t)*)obj, value);
+}
+
+
+// --- _Py_atomic_or ---------------------------------------------------------
+
+static inline uint8_t
+_Py_atomic_or_uint8(uint8_t *obj, uint8_t value)
+{
+    _Py_USING_STD;
+    return atomic_fetch_or((_Atomic(uint8_t)*)obj, value);
+}
+
+static inline uint16_t
+_Py_atomic_or_uint16(uint16_t *obj, uint16_t value)
+{
+    _Py_USING_STD;
+    return atomic_fetch_or((_Atomic(uint16_t)*)obj, value);
+}
+
+static inline uint32_t
+_Py_atomic_or_uint32(uint32_t *obj, uint32_t value)
+{
+    _Py_USING_STD;
+    return atomic_fetch_or((_Atomic(uint32_t)*)obj, value);
+}
+
+static inline uint64_t
+_Py_atomic_or_uint64(uint64_t *obj, uint64_t value)
+{
+    _Py_USING_STD;
+    return atomic_fetch_or((_Atomic(uint64_t)*)obj, value);
+}
+
+static inline uintptr_t
+_Py_atomic_or_uintptr(uintptr_t *obj, uintptr_t value)
+{
+    _Py_USING_STD;
+    return atomic_fetch_or((_Atomic(uintptr_t)*)obj, value);
+}
+
+
+// --- _Py_atomic_load -------------------------------------------------------
+
+static inline int
+_Py_atomic_load_int(const int *obj)
+{
+    _Py_USING_STD;
+    return atomic_load((const _Atomic(int)*)obj);
+}
+
+static inline int8_t
+_Py_atomic_load_int8(const int8_t *obj)
+{
+    _Py_USING_STD;
+    return atomic_load((const _Atomic(int8_t)*)obj);
+}
+
+static inline int16_t
+_Py_atomic_load_int16(const int16_t *obj)
+{
+    _Py_USING_STD;
+    return atomic_load((const _Atomic(int16_t)*)obj);
+}
+
+static inline int32_t
+_Py_atomic_load_int32(const int32_t *obj)
+{
+    _Py_USING_STD;
+    return atomic_load((const _Atomic(int32_t)*)obj);
+}
+
+static inline int64_t
+_Py_atomic_load_int64(const int64_t *obj)
+{
+    _Py_USING_STD;
+    return atomic_load((const _Atomic(int64_t)*)obj);
+}
+
+static inline intptr_t
+_Py_atomic_load_intptr(const intptr_t *obj)
+{
+    _Py_USING_STD;
+    return atomic_load((const _Atomic(intptr_t)*)obj);
+}
+
+static inline uint8_t
+_Py_atomic_load_uint8(const uint8_t *obj)
+{
+    _Py_USING_STD;
+    return atomic_load((const _Atomic(uint8_t)*)obj);
+}
+
+static inline uint16_t
+_Py_atomic_load_uint16(const uint16_t *obj)
+{
+    _Py_USING_STD;
+    return atomic_load((const _Atomic(uint32_t)*)obj);
+}
+
+static inline uint32_t
+_Py_atomic_load_uint32(const uint32_t *obj)
+{
+    _Py_USING_STD;
+    return atomic_load((const _Atomic(uint32_t)*)obj);
+}
+
+static inline uint64_t
+_Py_atomic_load_uint64(const uint64_t *obj)
+{
+    _Py_USING_STD;
+    return atomic_load((const _Atomic(uint64_t)*)obj);
+}
+
+static inline uintptr_t
+_Py_atomic_load_uintptr(const uintptr_t *obj)
+{
+    _Py_USING_STD;
+    return atomic_load((const _Atomic(uintptr_t)*)obj);
+}
+
+static inline unsigned int
+_Py_atomic_load_uint(const unsigned int *obj)
+{
+    _Py_USING_STD;
+    return atomic_load((const _Atomic(unsigned int)*)obj);
+}
+
+static inline Py_ssize_t
+_Py_atomic_load_ssize(const Py_ssize_t *obj)
+{
+    _Py_USING_STD;
+    return atomic_load((const _Atomic(Py_ssize_t)*)obj);
+}
+
+static inline void*
+_Py_atomic_load_ptr(const void *obj)
+{
+    _Py_USING_STD;
+    return atomic_load((const _Atomic(void*)*)obj);
+}
+
+
+// --- _Py_atomic_load_relaxed -----------------------------------------------
+
+static inline int
+_Py_atomic_load_int_relaxed(const int *obj)
+{
+    _Py_USING_STD;
+    return atomic_load_explicit((const _Atomic(int)*)obj,
+                                memory_order_relaxed);
+}
+
+static inline int8_t
+_Py_atomic_load_int8_relaxed(const int8_t *obj)
+{
+    _Py_USING_STD;
+    return atomic_load_explicit((const _Atomic(int8_t)*)obj,
+                                memory_order_relaxed);
+}
+
+static inline int16_t
+_Py_atomic_load_int16_relaxed(const int16_t *obj)
+{
+    _Py_USING_STD;
+    return atomic_load_explicit((const _Atomic(int16_t)*)obj,
+                                memory_order_relaxed);
+}
+
+static inline int32_t
+_Py_atomic_load_int32_relaxed(const int32_t *obj)
+{
+    _Py_USING_STD;
+    return atomic_load_explicit((const _Atomic(int32_t)*)obj,
+                                memory_order_relaxed);
+}
+
+static inline int64_t
+_Py_atomic_load_int64_relaxed(const int64_t *obj)
+{
+    _Py_USING_STD;
+    return atomic_load_explicit((const _Atomic(int64_t)*)obj,
+                                memory_order_relaxed);
+}
+
+static inline intptr_t
+_Py_atomic_load_intptr_relaxed(const intptr_t *obj)
+{
+    _Py_USING_STD;
+    return atomic_load_explicit((const _Atomic(intptr_t)*)obj,
+                                memory_order_relaxed);
+}
+
+static inline uint8_t
+_Py_atomic_load_uint8_relaxed(const uint8_t *obj)
+{
+    _Py_USING_STD;
+    return atomic_load_explicit((const _Atomic(uint8_t)*)obj,
+                                memory_order_relaxed);
+}
+
+static inline uint16_t
+_Py_atomic_load_uint16_relaxed(const uint16_t *obj)
+{
+    _Py_USING_STD;
+    return atomic_load_explicit((const _Atomic(uint16_t)*)obj,
+                                memory_order_relaxed);
+}
+
+static inline uint32_t
+_Py_atomic_load_uint32_relaxed(const uint32_t *obj)
+{
+    _Py_USING_STD;
+    return atomic_load_explicit((const _Atomic(uint32_t)*)obj,
+                                memory_order_relaxed);
+}
+
+static inline uint64_t
+_Py_atomic_load_uint64_relaxed(const uint64_t *obj)
+{
+    _Py_USING_STD;
+    return atomic_load_explicit((const _Atomic(uint64_t)*)obj,
+                                memory_order_relaxed);
+}
+
+static inline uintptr_t
+_Py_atomic_load_uintptr_relaxed(const uintptr_t *obj)
+{
+    _Py_USING_STD;
+    return atomic_load_explicit((const _Atomic(uintptr_t)*)obj,
+                                memory_order_relaxed);
+}
+
+static inline unsigned int
+_Py_atomic_load_uint_relaxed(const unsigned int *obj)
+{
+    _Py_USING_STD;
+    return atomic_load_explicit((const _Atomic(unsigned int)*)obj,
+                                memory_order_relaxed);
+}
+
+static inline Py_ssize_t
+_Py_atomic_load_ssize_relaxed(const Py_ssize_t *obj)
+{
+    _Py_USING_STD;
+    return atomic_load_explicit((const _Atomic(Py_ssize_t)*)obj,
+                                memory_order_relaxed);
+}
+
+static inline void*
+_Py_atomic_load_ptr_relaxed(const void *obj)
+{
+    _Py_USING_STD;
+    return atomic_load_explicit((const _Atomic(void*)*)obj,
+                                memory_order_relaxed);
+}
+
+static inline unsigned long long
+_Py_atomic_load_ullong_relaxed(const unsigned long long *obj)
+{
+    _Py_USING_STD;
+    return atomic_load_explicit((const _Atomic(unsigned long long)*)obj,
+                                memory_order_relaxed);
+}
+
+
+// --- _Py_atomic_store ------------------------------------------------------
+
+static inline void
+_Py_atomic_store_int(int *obj, int value)
+{
+    _Py_USING_STD;
+    atomic_store((_Atomic(int)*)obj, value);
+}
+
+static inline void
+_Py_atomic_store_int8(int8_t *obj, int8_t value)
+{
+    _Py_USING_STD;
+    atomic_store((_Atomic(int8_t)*)obj, value);
+}
+
+static inline void
+_Py_atomic_store_int16(int16_t *obj, int16_t value)
+{
+    _Py_USING_STD;
+    atomic_store((_Atomic(int16_t)*)obj, value);
+}
+
+static inline void
+_Py_atomic_store_int32(int32_t *obj, int32_t value)
+{
+    _Py_USING_STD;
+    atomic_store((_Atomic(int32_t)*)obj, value);
+}
+
+static inline void
+_Py_atomic_store_int64(int64_t *obj, int64_t value)
+{
+    _Py_USING_STD;
+    atomic_store((_Atomic(int64_t)*)obj, value);
+}
+
+static inline void
+_Py_atomic_store_intptr(intptr_t *obj, intptr_t value)
+{
+    _Py_USING_STD;
+    atomic_store((_Atomic(intptr_t)*)obj, value);
+}
+
+static inline void
+_Py_atomic_store_uint8(uint8_t *obj, uint8_t value)
+{
+    _Py_USING_STD;
+    atomic_store((_Atomic(uint8_t)*)obj, value);
+}
+
+static inline void
+_Py_atomic_store_uint16(uint16_t *obj, uint16_t value)
+{
+    _Py_USING_STD;
+    atomic_store((_Atomic(uint16_t)*)obj, value);
+}
+
+static inline void
+_Py_atomic_store_uint32(uint32_t *obj, uint32_t value)
+{
+    _Py_USING_STD;
+    atomic_store((_Atomic(uint32_t)*)obj, value);
+}
+
+static inline void
+_Py_atomic_store_uint64(uint64_t *obj, uint64_t value)
+{
+    _Py_USING_STD;
+    atomic_store((_Atomic(uint64_t)*)obj, value);
+}
+
+static inline void
+_Py_atomic_store_uintptr(uintptr_t *obj, uintptr_t value)
+{
+    _Py_USING_STD;
+    atomic_store((_Atomic(uintptr_t)*)obj, value);
+}
+
+static inline void
+_Py_atomic_store_uint(unsigned int *obj, unsigned int value)
+{
+    _Py_USING_STD;
+    atomic_store((_Atomic(unsigned int)*)obj, value);
+}
+
+static inline void
+_Py_atomic_store_ptr(void *obj, void *value)
+{
+    _Py_USING_STD;
+    atomic_store((_Atomic(void*)*)obj, value);
+}
+
+static inline void
+_Py_atomic_store_ssize(Py_ssize_t *obj, Py_ssize_t value)
+{
+    _Py_USING_STD;
+    atomic_store((_Atomic(Py_ssize_t)*)obj, value);
+}
+
+
+// --- _Py_atomic_store_relaxed ----------------------------------------------
+
+static inline void
+_Py_atomic_store_int_relaxed(int *obj, int value)
+{
+    _Py_USING_STD;
+    atomic_store_explicit((_Atomic(int)*)obj, value,
+                          memory_order_relaxed);
+}
+
+static inline void
+_Py_atomic_store_int8_relaxed(int8_t *obj, int8_t value)
+{
+    _Py_USING_STD;
+    atomic_store_explicit((_Atomic(int8_t)*)obj, value,
+                          memory_order_relaxed);
+}
+
+static inline void
+_Py_atomic_store_int16_relaxed(int16_t *obj, int16_t value)
+{
+    _Py_USING_STD;
+    atomic_store_explicit((_Atomic(int16_t)*)obj, value,
+                          memory_order_relaxed);
+}
+
+static inline void
+_Py_atomic_store_int32_relaxed(int32_t *obj, int32_t value)
+{
+    _Py_USING_STD;
+    atomic_store_explicit((_Atomic(int32_t)*)obj, value,
+                          memory_order_relaxed);
+}
+
+static inline void
+_Py_atomic_store_int64_relaxed(int64_t *obj, int64_t value)
+{
+    _Py_USING_STD;
+    atomic_store_explicit((_Atomic(int64_t)*)obj, value,
+                          memory_order_relaxed);
+}
+
+static inline void
+_Py_atomic_store_intptr_relaxed(intptr_t *obj, intptr_t value)
+{
+    _Py_USING_STD;
+    atomic_store_explicit((_Atomic(intptr_t)*)obj, value,
+                          memory_order_relaxed);
+}
+
+static inline void
+_Py_atomic_store_uint8_relaxed(uint8_t *obj, uint8_t value)
+{
+    _Py_USING_STD;
+    atomic_store_explicit((_Atomic(uint8_t)*)obj, value,
+                          memory_order_relaxed);
+}
+
+static inline void
+_Py_atomic_store_uint16_relaxed(uint16_t *obj, uint16_t value)
+{
+    _Py_USING_STD;
+    atomic_store_explicit((_Atomic(uint16_t)*)obj, value,
+                          memory_order_relaxed);
+}
+
+static inline void
+_Py_atomic_store_uint32_relaxed(uint32_t *obj, uint32_t value)
+{
+    _Py_USING_STD;
+    atomic_store_explicit((_Atomic(uint32_t)*)obj, value,
+                          memory_order_relaxed);
+}
+
+static inline void
+_Py_atomic_store_uint64_relaxed(uint64_t *obj, uint64_t value)
+{
+    _Py_USING_STD;
+    atomic_store_explicit((_Atomic(uint64_t)*)obj, value,
+                          memory_order_relaxed);
+}
+
+static inline void
+_Py_atomic_store_uintptr_relaxed(uintptr_t *obj, uintptr_t value)
+{
+    _Py_USING_STD;
+    atomic_store_explicit((_Atomic(uintptr_t)*)obj, value,
+                          memory_order_relaxed);
+}
+
+static inline void
+_Py_atomic_store_uint_relaxed(unsigned int *obj, unsigned int value)
+{
+    _Py_USING_STD;
+    atomic_store_explicit((_Atomic(unsigned int)*)obj, value,
+                          memory_order_relaxed);
+}
+
+static inline void
+_Py_atomic_store_ptr_relaxed(void *obj, void *value)
+{
+    _Py_USING_STD;
+    atomic_store_explicit((_Atomic(void*)*)obj, value,
+                          memory_order_relaxed);
+}
+
+static inline void
+_Py_atomic_store_ssize_relaxed(Py_ssize_t *obj, Py_ssize_t value)
+{
+    _Py_USING_STD;
+    atomic_store_explicit((_Atomic(Py_ssize_t)*)obj, value,
+                          memory_order_relaxed);
+}
+
+static inline void
+_Py_atomic_store_ullong_relaxed(unsigned long long *obj,
+                                unsigned long long value)
+{
+    _Py_USING_STD;
+    atomic_store_explicit((_Atomic(unsigned long long)*)obj, value,
+                          memory_order_relaxed);
+}
+
+
+// --- _Py_atomic_load_ptr_acquire / _Py_atomic_store_ptr_release ------------
+
+static inline void *
+_Py_atomic_load_ptr_acquire(const void *obj)
+{
+    _Py_USING_STD;
+    return atomic_load_explicit((const _Atomic(void*)*)obj,
+                                memory_order_acquire);
+}
+
+static inline uintptr_t
+_Py_atomic_load_uintptr_acquire(const uintptr_t *obj)
+{
+    _Py_USING_STD;
+    return atomic_load_explicit((const _Atomic(uintptr_t)*)obj,
+                                memory_order_acquire);
+}
+
+static inline void
+_Py_atomic_store_ptr_release(void *obj, void *value)
+{
+    _Py_USING_STD;
+    atomic_store_explicit((_Atomic(void*)*)obj, value,
+                          memory_order_release);
+}
+
+static inline void
+_Py_atomic_store_uintptr_release(uintptr_t *obj, uintptr_t value)
+{
+    _Py_USING_STD;
+    atomic_store_explicit((_Atomic(uintptr_t)*)obj, value,
+                          memory_order_release);
+}
+
+static inline void
+_Py_atomic_store_int_release(int *obj, int value)
+{
+    _Py_USING_STD;
+    atomic_store_explicit((_Atomic(int)*)obj, value,
+                          memory_order_release);
+}
+
+static inline void
+_Py_atomic_store_ssize_release(Py_ssize_t *obj, Py_ssize_t value)
+{
+    _Py_USING_STD;
+    atomic_store_explicit((_Atomic(Py_ssize_t)*)obj, value,
+                          memory_order_release);
+}
+
+static inline int
+_Py_atomic_load_int_acquire(const int *obj)
+{
+    _Py_USING_STD;
+    return atomic_load_explicit((const _Atomic(int)*)obj,
+                                memory_order_acquire);
+}
+
+static inline void
+_Py_atomic_store_uint32_release(uint32_t *obj, uint32_t value)
+{
+    _Py_USING_STD;
+    atomic_store_explicit((_Atomic(uint32_t)*)obj, value,
+                          memory_order_release);
+}
+
+static inline void
+_Py_atomic_store_uint64_release(uint64_t *obj, uint64_t value)
+{
+    _Py_USING_STD;
+    atomic_store_explicit((_Atomic(uint64_t)*)obj, value,
+                          memory_order_release);
+}
+
+static inline uint64_t
+_Py_atomic_load_uint64_acquire(const uint64_t *obj)
+{
+    _Py_USING_STD;
+    return atomic_load_explicit((const _Atomic(uint64_t)*)obj,
+                                memory_order_acquire);
+}
+
+static inline uint32_t
+_Py_atomic_load_uint32_acquire(const uint32_t *obj)
+{
+    _Py_USING_STD;
+    return atomic_load_explicit((const _Atomic(uint32_t)*)obj,
+                                memory_order_acquire);
+}
+
+static inline Py_ssize_t
+_Py_atomic_load_ssize_acquire(const Py_ssize_t *obj)
+{
+    _Py_USING_STD;
+    return atomic_load_explicit((const _Atomic(Py_ssize_t)*)obj,
+                                memory_order_acquire);
+}
+
+
+// --- _Py_atomic_fence ------------------------------------------------------
+
+ static inline void
+_Py_atomic_fence_seq_cst(void)
+{
+    _Py_USING_STD;
+    atomic_thread_fence(memory_order_seq_cst);
+}
+
+ static inline void
+_Py_atomic_fence_acquire(void)
+{
+    _Py_USING_STD;
+    atomic_thread_fence(memory_order_acquire);
+}
+
+ static inline void
+_Py_atomic_fence_release(void)
+{
+    _Py_USING_STD;
+    atomic_thread_fence(memory_order_release);
+}

Dependencies/Python/include/cpython/pyctype.h

@@ -0,0 +1,39 @@
+#ifndef Py_LIMITED_API
+#ifndef PYCTYPE_H
+#define PYCTYPE_H
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+#define PY_CTF_LOWER  0x01
+#define PY_CTF_UPPER  0x02
+#define PY_CTF_ALPHA  (PY_CTF_LOWER|PY_CTF_UPPER)
+#define PY_CTF_DIGIT  0x04
+#define PY_CTF_ALNUM  (PY_CTF_ALPHA|PY_CTF_DIGIT)
+#define PY_CTF_SPACE  0x08
+#define PY_CTF_XDIGIT 0x10
+
+PyAPI_DATA(const unsigned int) _Py_ctype_table[256];
+
+/* Unlike their C counterparts, the following macros are not meant to
+ * handle an int with any of the values [EOF, 0-UCHAR_MAX]. The argument
+ * must be a signed/unsigned char. */
+#define Py_ISLOWER(c)  (_Py_ctype_table[Py_CHARMASK(c)] & PY_CTF_LOWER)
+#define Py_ISUPPER(c)  (_Py_ctype_table[Py_CHARMASK(c)] & PY_CTF_UPPER)
+#define Py_ISALPHA(c)  (_Py_ctype_table[Py_CHARMASK(c)] & PY_CTF_ALPHA)
+#define Py_ISDIGIT(c)  (_Py_ctype_table[Py_CHARMASK(c)] & PY_CTF_DIGIT)
+#define Py_ISXDIGIT(c) (_Py_ctype_table[Py_CHARMASK(c)] & PY_CTF_XDIGIT)
+#define Py_ISALNUM(c)  (_Py_ctype_table[Py_CHARMASK(c)] & PY_CTF_ALNUM)
+#define Py_ISSPACE(c)  (_Py_ctype_table[Py_CHARMASK(c)] & PY_CTF_SPACE)
+
+PyAPI_DATA(const unsigned char) _Py_ctype_tolower[256];
+PyAPI_DATA(const unsigned char) _Py_ctype_toupper[256];
+
+#define Py_TOLOWER(c) (_Py_ctype_tolower[Py_CHARMASK(c)])
+#define Py_TOUPPER(c) (_Py_ctype_toupper[Py_CHARMASK(c)])
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* !PYCTYPE_H */
+#endif /* !Py_LIMITED_API */

Dependencies/Python/include/cpython/pydebug.h

@@ -0,0 +1,38 @@
+#ifndef Py_LIMITED_API
+#ifndef Py_PYDEBUG_H
+#define Py_PYDEBUG_H
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+Py_DEPRECATED(3.12) PyAPI_DATA(int) Py_DebugFlag;
+Py_DEPRECATED(3.12) PyAPI_DATA(int) Py_VerboseFlag;
+Py_DEPRECATED(3.12) PyAPI_DATA(int) Py_QuietFlag;
+Py_DEPRECATED(3.12) PyAPI_DATA(int) Py_InteractiveFlag;
+Py_DEPRECATED(3.12) PyAPI_DATA(int) Py_InspectFlag;
+Py_DEPRECATED(3.12) PyAPI_DATA(int) Py_OptimizeFlag;
+Py_DEPRECATED(3.12) PyAPI_DATA(int) Py_NoSiteFlag;
+Py_DEPRECATED(3.12) PyAPI_DATA(int) Py_BytesWarningFlag;
+Py_DEPRECATED(3.12) PyAPI_DATA(int) Py_FrozenFlag;
+Py_DEPRECATED(3.12) PyAPI_DATA(int) Py_IgnoreEnvironmentFlag;
+Py_DEPRECATED(3.12) PyAPI_DATA(int) Py_DontWriteBytecodeFlag;
+Py_DEPRECATED(3.12) PyAPI_DATA(int) Py_NoUserSiteDirectory;
+Py_DEPRECATED(3.12) PyAPI_DATA(int) Py_UnbufferedStdioFlag;
+Py_DEPRECATED(3.12) PyAPI_DATA(int) Py_HashRandomizationFlag;
+Py_DEPRECATED(3.12) PyAPI_DATA(int) Py_IsolatedFlag;
+
+#ifdef MS_WINDOWS
+Py_DEPRECATED(3.12) PyAPI_DATA(int) Py_LegacyWindowsFSEncodingFlag;
+Py_DEPRECATED(3.12) PyAPI_DATA(int) Py_LegacyWindowsStdioFlag;
+#endif
+
+/* this is a wrapper around getenv() that pays attention to
+   Py_IgnoreEnvironmentFlag.  It should be used for getting variables like
+   PYTHONPATH and PYTHONHOME from the environment */
+PyAPI_FUNC(char*) Py_GETENV(const char *name);
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* !Py_PYDEBUG_H */
+#endif /* Py_LIMITED_API */

Dependencies/Python/include/cpython/pyerrors.h

@@ -0,0 +1,131 @@
+#ifndef Py_CPYTHON_ERRORS_H
+#  error "this header file must not be included directly"
+#endif
+
+/* Error objects */
+
+/* PyException_HEAD defines the initial segment of every exception class. */
+#define PyException_HEAD PyObject_HEAD PyObject *dict;\
+             PyObject *args; PyObject *notes; PyObject *traceback;\
+             PyObject *context; PyObject *cause;\
+             char suppress_context;
+
+typedef struct {
+    PyException_HEAD
+} PyBaseExceptionObject;
+
+typedef struct {
+    PyException_HEAD
+    PyObject *msg;
+    PyObject *excs;
+} PyBaseExceptionGroupObject;
+
+typedef struct {
+    PyException_HEAD
+    PyObject *msg;
+    PyObject *filename;
+    PyObject *lineno;
+    PyObject *offset;
+    PyObject *end_lineno;
+    PyObject *end_offset;
+    PyObject *text;
+    PyObject *print_file_and_line;
+} PySyntaxErrorObject;
+
+typedef struct {
+    PyException_HEAD
+    PyObject *msg;
+    PyObject *name;
+    PyObject *path;
+    PyObject *name_from;
+} PyImportErrorObject;
+
+typedef struct {
+    PyException_HEAD
+    PyObject *encoding;
+    PyObject *object;
+    Py_ssize_t start;
+    Py_ssize_t end;
+    PyObject *reason;
+} PyUnicodeErrorObject;
+
+typedef struct {
+    PyException_HEAD
+    PyObject *code;
+} PySystemExitObject;
+
+typedef struct {
+    PyException_HEAD
+    PyObject *myerrno;
+    PyObject *strerror;
+    PyObject *filename;
+    PyObject *filename2;
+#ifdef MS_WINDOWS
+    PyObject *winerror;
+#endif
+    Py_ssize_t written;   /* only for BlockingIOError, -1 otherwise */
+} PyOSErrorObject;
+
+typedef struct {
+    PyException_HEAD
+    PyObject *value;
+} PyStopIterationObject;
+
+typedef struct {
+    PyException_HEAD
+    PyObject *name;
+} PyNameErrorObject;
+
+typedef struct {
+    PyException_HEAD
+    PyObject *obj;
+    PyObject *name;
+} PyAttributeErrorObject;
+
+/* Compatibility typedefs */
+typedef PyOSErrorObject PyEnvironmentErrorObject;
+#ifdef MS_WINDOWS
+typedef PyOSErrorObject PyWindowsErrorObject;
+#endif
+
+/* Context manipulation (PEP 3134) */
+
+PyAPI_FUNC(void) _PyErr_ChainExceptions1(PyObject *);
+
+/* In exceptions.c */
+
+PyAPI_FUNC(PyObject*) PyUnstable_Exc_PrepReraiseStar(
+     PyObject *orig,
+     PyObject *excs);
+
+/* In signalmodule.c */
+
+PyAPI_FUNC(int) PySignal_SetWakeupFd(int fd);
+
+/* Support for adding program text to SyntaxErrors */
+
+PyAPI_FUNC(void) PyErr_SyntaxLocationObject(
+    PyObject *filename,
+    int lineno,
+    int col_offset);
+
+PyAPI_FUNC(void) PyErr_RangedSyntaxLocationObject(
+    PyObject *filename,
+    int lineno,
+    int col_offset,
+    int end_lineno,
+    int end_col_offset);
+
+PyAPI_FUNC(PyObject *) PyErr_ProgramTextObject(
+    PyObject *filename,
+    int lineno);
+
+PyAPI_FUNC(void) _Py_NO_RETURN _Py_FatalErrorFunc(
+    const char *func,
+    const char *message);
+
+PyAPI_FUNC(void) PyErr_FormatUnraisable(const char *, ...);
+
+PyAPI_DATA(PyObject *) PyExc_PythonFinalizationError;
+
+#define Py_FatalError(message) _Py_FatalErrorFunc(__func__, (message))

Dependencies/Python/include/cpython/pyfpe.h

@@ -0,0 +1,15 @@
+#ifndef Py_PYFPE_H
+#define Py_PYFPE_H
+/* Header excluded from the stable API */
+#ifndef Py_LIMITED_API
+
+/* These macros used to do something when Python was built with --with-fpectl,
+ * but support for that was dropped in 3.7. We continue to define them though,
+ * to avoid breaking API users.
+ */
+
+#define PyFPE_START_PROTECT(err_string, leave_stmt)
+#define PyFPE_END_PROTECT(v)
+
+#endif /* !defined(Py_LIMITED_API) */
+#endif /* !Py_PYFPE_H */

Dependencies/Python/include/cpython/pyframe.h

@@ -0,0 +1,45 @@
+#ifndef Py_CPYTHON_PYFRAME_H
+#  error "this header file must not be included directly"
+#endif
+
+PyAPI_DATA(PyTypeObject) PyFrame_Type;
+PyAPI_DATA(PyTypeObject) PyFrameLocalsProxy_Type;
+
+#define PyFrame_Check(op) Py_IS_TYPE((op), &PyFrame_Type)
+#define PyFrameLocalsProxy_Check(op) Py_IS_TYPE((op), &PyFrameLocalsProxy_Type)
+
+PyAPI_FUNC(PyFrameObject *) PyFrame_GetBack(PyFrameObject *frame);
+PyAPI_FUNC(PyObject *) PyFrame_GetLocals(PyFrameObject *frame);
+
+PyAPI_FUNC(PyObject *) PyFrame_GetGlobals(PyFrameObject *frame);
+PyAPI_FUNC(PyObject *) PyFrame_GetBuiltins(PyFrameObject *frame);
+
+PyAPI_FUNC(PyObject *) PyFrame_GetGenerator(PyFrameObject *frame);
+PyAPI_FUNC(int) PyFrame_GetLasti(PyFrameObject *frame);
+PyAPI_FUNC(PyObject*) PyFrame_GetVar(PyFrameObject *frame, PyObject *name);
+PyAPI_FUNC(PyObject*) PyFrame_GetVarString(PyFrameObject *frame, const char *name);
+
+/* The following functions are for use by debuggers and other tools
+ * implementing custom frame evaluators with PEP 523. */
+
+struct _PyInterpreterFrame;
+
+/* Returns the code object of the frame (strong reference).
+ * Does not raise an exception. */
+PyAPI_FUNC(PyObject *) PyUnstable_InterpreterFrame_GetCode(struct _PyInterpreterFrame *frame);
+
+/* Returns a byte ofsset into the last executed instruction.
+ * Does not raise an exception. */
+PyAPI_FUNC(int) PyUnstable_InterpreterFrame_GetLasti(struct _PyInterpreterFrame *frame);
+
+/* Returns the currently executing line number, or -1 if there is no line number.
+ * Does not raise an exception. */
+PyAPI_FUNC(int) PyUnstable_InterpreterFrame_GetLine(struct _PyInterpreterFrame *frame);
+
+#define PyUnstable_EXECUTABLE_KIND_SKIP 0
+#define PyUnstable_EXECUTABLE_KIND_PY_FUNCTION 1
+#define PyUnstable_EXECUTABLE_KIND_BUILTIN_FUNCTION 3
+#define PyUnstable_EXECUTABLE_KIND_METHOD_DESCRIPTOR 4
+#define PyUnstable_EXECUTABLE_KINDS 5
+
+PyAPI_DATA(const PyTypeObject *) const PyUnstable_ExecutableKinds[PyUnstable_EXECUTABLE_KINDS+1];

Dependencies/Python/include/cpython/pyhash.h

@@ -0,0 +1,47 @@
+#ifndef Py_CPYTHON_HASH_H
+#  error "this header file must not be included directly"
+#endif
+
+/* Prime multiplier used in string and various other hashes. */
+#define PyHASH_MULTIPLIER 1000003UL  /* 0xf4243 */
+
+/* Parameters used for the numeric hash implementation.  See notes for
+   _Py_HashDouble in Python/pyhash.c.  Numeric hashes are based on
+   reduction modulo the prime 2**_PyHASH_BITS - 1. */
+
+#if SIZEOF_VOID_P >= 8
+#  define PyHASH_BITS 61
+#else
+#  define PyHASH_BITS 31
+#endif
+
+#define PyHASH_MODULUS (((size_t)1 << _PyHASH_BITS) - 1)
+#define PyHASH_INF 314159
+#define PyHASH_IMAG PyHASH_MULTIPLIER
+
+/* Aliases kept for backward compatibility with Python 3.12 */
+#define _PyHASH_MULTIPLIER PyHASH_MULTIPLIER
+#define _PyHASH_BITS PyHASH_BITS
+#define _PyHASH_MODULUS PyHASH_MODULUS
+#define _PyHASH_INF PyHASH_INF
+#define _PyHASH_IMAG PyHASH_IMAG
+
+/* Helpers for hash functions */
+PyAPI_FUNC(Py_hash_t) _Py_HashDouble(PyObject *, double);
+
+// Kept for backward compatibility
+#define _Py_HashPointer Py_HashPointer
+
+
+/* hash function definition */
+typedef struct {
+    Py_hash_t (*const hash)(const void *, Py_ssize_t);
+    const char *name;
+    const int hash_bits;
+    const int seed_bits;
+} PyHash_FuncDef;
+
+PyAPI_FUNC(PyHash_FuncDef*) PyHash_GetFuncDef(void);
+
+PyAPI_FUNC(Py_hash_t) Py_HashPointer(const void *ptr);
+PyAPI_FUNC(Py_hash_t) PyObject_GenericHash(PyObject *);

Dependencies/Python/include/cpython/pylifecycle.h

@@ -0,0 +1,92 @@
+#ifndef Py_CPYTHON_PYLIFECYCLE_H
+#  error "this header file must not be included directly"
+#endif
+
+/* Py_FrozenMain is kept out of the Limited API until documented and present
+   in all builds of Python */
+PyAPI_FUNC(int) Py_FrozenMain(int argc, char **argv);
+
+/* PEP 432 Multi-phase initialization API (Private while provisional!) */
+
+PyAPI_FUNC(PyStatus) Py_PreInitialize(
+    const PyPreConfig *src_config);
+PyAPI_FUNC(PyStatus) Py_PreInitializeFromBytesArgs(
+    const PyPreConfig *src_config,
+    Py_ssize_t argc,
+    char **argv);
+PyAPI_FUNC(PyStatus) Py_PreInitializeFromArgs(
+    const PyPreConfig *src_config,
+    Py_ssize_t argc,
+    wchar_t **argv);
+
+
+/* Initialization and finalization */
+
+PyAPI_FUNC(PyStatus) Py_InitializeFromConfig(
+    const PyConfig *config);
+
+// Python 3.8 provisional API (PEP 587)
+PyAPI_FUNC(PyStatus) _Py_InitializeMain(void);
+
+PyAPI_FUNC(int) Py_RunMain(void);
+
+
+PyAPI_FUNC(void) _Py_NO_RETURN Py_ExitStatusException(PyStatus err);
+
+PyAPI_FUNC(int) Py_FdIsInteractive(FILE *, const char *);
+
+/* --- PyInterpreterConfig ------------------------------------ */
+
+#define PyInterpreterConfig_DEFAULT_GIL (0)
+#define PyInterpreterConfig_SHARED_GIL (1)
+#define PyInterpreterConfig_OWN_GIL (2)
+
+typedef struct {
+    // XXX "allow_object_sharing"?  "own_objects"?
+    int use_main_obmalloc;
+    int allow_fork;
+    int allow_exec;
+    int allow_threads;
+    int allow_daemon_threads;
+    int check_multi_interp_extensions;
+    int gil;
+} PyInterpreterConfig;
+
+#define _PyInterpreterConfig_INIT \
+    { \
+        .use_main_obmalloc = 0, \
+        .allow_fork = 0, \
+        .allow_exec = 0, \
+        .allow_threads = 1, \
+        .allow_daemon_threads = 0, \
+        .check_multi_interp_extensions = 1, \
+        .gil = PyInterpreterConfig_OWN_GIL, \
+    }
+
+// gh-117649: The free-threaded build does not currently support single-phase
+// init extensions in subinterpreters. For now, we ensure that
+// `check_multi_interp_extensions` is always `1`, even in the legacy config.
+#ifdef Py_GIL_DISABLED
+#  define _PyInterpreterConfig_LEGACY_CHECK_MULTI_INTERP_EXTENSIONS 1
+#else
+#  define _PyInterpreterConfig_LEGACY_CHECK_MULTI_INTERP_EXTENSIONS 0
+#endif
+
+#define _PyInterpreterConfig_LEGACY_INIT \
+    { \
+        .use_main_obmalloc = 1, \
+        .allow_fork = 1, \
+        .allow_exec = 1, \
+        .allow_threads = 1, \
+        .allow_daemon_threads = 1, \
+        .check_multi_interp_extensions = _PyInterpreterConfig_LEGACY_CHECK_MULTI_INTERP_EXTENSIONS, \
+        .gil = PyInterpreterConfig_SHARED_GIL, \
+    }
+
+PyAPI_FUNC(PyStatus) Py_NewInterpreterFromConfig(
+    PyThreadState **tstate_p,
+    const PyInterpreterConfig *config);
+
+typedef void (*atexit_datacallbackfunc)(void *);
+PyAPI_FUNC(int) PyUnstable_AtExit(
+        PyInterpreterState *, atexit_datacallbackfunc, void *);

Dependencies/Python/include/cpython/pymem.h

@@ -0,0 +1,84 @@
+#ifndef Py_CPYTHON_PYMEM_H
+#  error "this header file must not be included directly"
+#endif
+
+typedef enum {
+    /* PyMem_RawMalloc(), PyMem_RawRealloc() and PyMem_RawFree() */
+    PYMEM_DOMAIN_RAW,
+
+    /* PyMem_Malloc(), PyMem_Realloc() and PyMem_Free() */
+    PYMEM_DOMAIN_MEM,
+
+    /* PyObject_Malloc(), PyObject_Realloc() and PyObject_Free() */
+    PYMEM_DOMAIN_OBJ
+} PyMemAllocatorDomain;
+
+typedef enum {
+    PYMEM_ALLOCATOR_NOT_SET = 0,
+    PYMEM_ALLOCATOR_DEFAULT = 1,
+    PYMEM_ALLOCATOR_DEBUG = 2,
+    PYMEM_ALLOCATOR_MALLOC = 3,
+    PYMEM_ALLOCATOR_MALLOC_DEBUG = 4,
+#ifdef WITH_PYMALLOC
+    PYMEM_ALLOCATOR_PYMALLOC = 5,
+    PYMEM_ALLOCATOR_PYMALLOC_DEBUG = 6,
+#endif
+#ifdef WITH_MIMALLOC
+    PYMEM_ALLOCATOR_MIMALLOC = 7,
+    PYMEM_ALLOCATOR_MIMALLOC_DEBUG = 8,
+#endif
+} PyMemAllocatorName;
+
+
+typedef struct {
+    /* user context passed as the first argument to the 4 functions */
+    void *ctx;
+
+    /* allocate a memory block */
+    void* (*malloc) (void *ctx, size_t size);
+
+    /* allocate a memory block initialized by zeros */
+    void* (*calloc) (void *ctx, size_t nelem, size_t elsize);
+
+    /* allocate or resize a memory block */
+    void* (*realloc) (void *ctx, void *ptr, size_t new_size);
+
+    /* release a memory block */
+    void (*free) (void *ctx, void *ptr);
+} PyMemAllocatorEx;
+
+/* Get the memory block allocator of the specified domain. */
+PyAPI_FUNC(void) PyMem_GetAllocator(PyMemAllocatorDomain domain,
+                                    PyMemAllocatorEx *allocator);
+
+/* Set the memory block allocator of the specified domain.
+
+   The new allocator must return a distinct non-NULL pointer when requesting
+   zero bytes.
+
+   For the PYMEM_DOMAIN_RAW domain, the allocator must be thread-safe: the GIL
+   is not held when the allocator is called.
+
+   If the new allocator is not a hook (don't call the previous allocator), the
+   PyMem_SetupDebugHooks() function must be called to reinstall the debug hooks
+   on top on the new allocator. */
+PyAPI_FUNC(void) PyMem_SetAllocator(PyMemAllocatorDomain domain,
+                                    PyMemAllocatorEx *allocator);
+
+/* Setup hooks to detect bugs in the following Python memory allocator
+   functions:
+
+   - PyMem_RawMalloc(), PyMem_RawRealloc(), PyMem_RawFree()
+   - PyMem_Malloc(), PyMem_Realloc(), PyMem_Free()
+   - PyObject_Malloc(), PyObject_Realloc() and PyObject_Free()
+
+   Newly allocated memory is filled with the byte 0xCB, freed memory is filled
+   with the byte 0xDB. Additional checks:
+
+   - detect API violations, ex: PyObject_Free() called on a buffer allocated
+     by PyMem_Malloc()
+   - detect write before the start of the buffer (buffer underflow)
+   - detect write after the end of the buffer (buffer overflow)
+
+   The function does nothing if Python is not compiled is debug mode. */
+PyAPI_FUNC(void) PyMem_SetupDebugHooks(void);

Dependencies/Python/include/cpython/pystate.h

@@ -0,0 +1,277 @@
+#ifndef Py_CPYTHON_PYSTATE_H
+#  error "this header file must not be included directly"
+#endif
+
+
+/* private interpreter helpers */
+
+PyAPI_FUNC(int) _PyInterpreterState_RequiresIDRef(PyInterpreterState *);
+PyAPI_FUNC(void) _PyInterpreterState_RequireIDRef(PyInterpreterState *, int);
+
+PyAPI_FUNC(PyObject *) PyUnstable_InterpreterState_GetMainModule(PyInterpreterState *);
+
+/* State unique per thread */
+
+/* Py_tracefunc return -1 when raising an exception, or 0 for success. */
+typedef int (*Py_tracefunc)(PyObject *, PyFrameObject *, int, PyObject *);
+
+/* The following values are used for 'what' for tracefunc functions
+ *
+ * To add a new kind of trace event, also update "trace_init" in
+ * Python/sysmodule.c to define the Python level event name
+ */
+#define PyTrace_CALL 0
+#define PyTrace_EXCEPTION 1
+#define PyTrace_LINE 2
+#define PyTrace_RETURN 3
+#define PyTrace_C_CALL 4
+#define PyTrace_C_EXCEPTION 5
+#define PyTrace_C_RETURN 6
+#define PyTrace_OPCODE 7
+
+typedef struct _err_stackitem {
+    /* This struct represents a single execution context where we might
+     * be currently handling an exception.  It is a per-coroutine state
+     * (coroutine in the computer science sense, including the thread
+     * and generators).
+     *
+     * This is used as an entry on the exception stack, where each
+     * entry indicates if it is currently handling an exception.
+     * This ensures that the exception state is not impacted
+     * by "yields" from an except handler.  The thread
+     * always has an entry (the bottom-most one).
+     */
+
+    /* The exception currently being handled in this context, if any. */
+    PyObject *exc_value;
+
+    struct _err_stackitem *previous_item;
+
+} _PyErr_StackItem;
+
+typedef struct _stack_chunk {
+    struct _stack_chunk *previous;
+    size_t size;
+    size_t top;
+    PyObject * data[1]; /* Variable sized */
+} _PyStackChunk;
+
+struct _ts {
+    /* See Python/ceval.c for comments explaining most fields */
+
+    PyThreadState *prev;
+    PyThreadState *next;
+    PyInterpreterState *interp;
+
+    /* The global instrumentation version in high bits, plus flags indicating
+       when to break out of the interpreter loop in lower bits. See details in
+       pycore_ceval.h. */
+    uintptr_t eval_breaker;
+
+    struct {
+        /* Has been initialized to a safe state.
+
+           In order to be effective, this must be set to 0 during or right
+           after allocation. */
+        unsigned int initialized:1;
+
+        /* Has been bound to an OS thread. */
+        unsigned int bound:1;
+        /* Has been unbound from its OS thread. */
+        unsigned int unbound:1;
+        /* Has been bound aa current for the GILState API. */
+        unsigned int bound_gilstate:1;
+        /* Currently in use (maybe holds the GIL). */
+        unsigned int active:1;
+        /* Currently holds the GIL. */
+        unsigned int holds_gil:1;
+
+        /* various stages of finalization */
+        unsigned int finalizing:1;
+        unsigned int cleared:1;
+        unsigned int finalized:1;
+
+        /* padding to align to 4 bytes */
+        unsigned int :23;
+    } _status;
+#ifdef Py_BUILD_CORE
+#  define _PyThreadState_WHENCE_NOTSET -1
+#  define _PyThreadState_WHENCE_UNKNOWN 0
+#  define _PyThreadState_WHENCE_INIT 1
+#  define _PyThreadState_WHENCE_FINI 2
+#  define _PyThreadState_WHENCE_THREADING 3
+#  define _PyThreadState_WHENCE_GILSTATE 4
+#  define _PyThreadState_WHENCE_EXEC 5
+#endif
+    int _whence;
+
+    /* Thread state (_Py_THREAD_ATTACHED, _Py_THREAD_DETACHED, _Py_THREAD_SUSPENDED).
+       See Include/internal/pycore_pystate.h for more details. */
+    int state;
+
+    int py_recursion_remaining;
+    int py_recursion_limit;
+
+    int c_recursion_remaining;
+    int recursion_headroom; /* Allow 50 more calls to handle any errors. */
+
+    /* 'tracing' keeps track of the execution depth when tracing/profiling.
+       This is to prevent the actual trace/profile code from being recorded in
+       the trace/profile. */
+    int tracing;
+    int what_event; /* The event currently being monitored, if any. */
+
+    /* Pointer to currently executing frame. */
+    struct _PyInterpreterFrame *current_frame;
+
+    Py_tracefunc c_profilefunc;
+    Py_tracefunc c_tracefunc;
+    PyObject *c_profileobj;
+    PyObject *c_traceobj;
+
+    /* The exception currently being raised */
+    PyObject *current_exception;
+
+    /* Pointer to the top of the exception stack for the exceptions
+     * we may be currently handling.  (See _PyErr_StackItem above.)
+     * This is never NULL. */
+    _PyErr_StackItem *exc_info;
+
+    PyObject *dict;  /* Stores per-thread state */
+
+    int gilstate_counter;
+
+    PyObject *async_exc; /* Asynchronous exception to raise */
+    unsigned long thread_id; /* Thread id where this tstate was created */
+
+    /* Native thread id where this tstate was created. This will be 0 except on
+     * those platforms that have the notion of native thread id, for which the
+     * macro PY_HAVE_THREAD_NATIVE_ID is then defined.
+     */
+    unsigned long native_thread_id;
+
+    PyObject *delete_later;
+
+    /* Tagged pointer to top-most critical section, or zero if there is no
+     * active critical section. Critical sections are only used in
+     * `--disable-gil` builds (i.e., when Py_GIL_DISABLED is defined to 1). In the
+     * default build, this field is always zero.
+     */
+    uintptr_t critical_section;
+
+    int coroutine_origin_tracking_depth;
+
+    PyObject *async_gen_firstiter;
+    PyObject *async_gen_finalizer;
+
+    PyObject *context;
+    uint64_t context_ver;
+
+    /* Unique thread state id. */
+    uint64_t id;
+
+    _PyStackChunk *datastack_chunk;
+    PyObject **datastack_top;
+    PyObject **datastack_limit;
+    /* XXX signal handlers should also be here */
+
+    /* The following fields are here to avoid allocation during init.
+       The data is exposed through PyThreadState pointer fields.
+       These fields should not be accessed directly outside of init.
+       This is indicated by an underscore prefix on the field names.
+
+       All other PyInterpreterState pointer fields are populated when
+       needed and default to NULL.
+       */
+       // Note some fields do not have a leading underscore for backward
+       // compatibility.  See https://bugs.python.org/issue45953#msg412046.
+
+    /* The thread's exception stack entry.  (Always the last entry.) */
+    _PyErr_StackItem exc_state;
+
+    PyObject *previous_executor;
+
+    uint64_t dict_global_version;
+
+    /* Used to store/retrieve `threading.local` keys/values for this thread */
+    PyObject *threading_local_key;
+
+    /* Used by `threading.local`s to be remove keys/values for dying threads.
+       The PyThreadObject must hold the only reference to this value.
+    */
+    PyObject *threading_local_sentinel;
+};
+
+#ifdef Py_DEBUG
+   // A debug build is likely built with low optimization level which implies
+   // higher stack memory usage than a release build: use a lower limit.
+#  define Py_C_RECURSION_LIMIT 500
+#elif defined(__s390x__)
+#  define Py_C_RECURSION_LIMIT 800
+#elif defined(_WIN32) && defined(_M_ARM64)
+#  define Py_C_RECURSION_LIMIT 1000
+#elif defined(_WIN32)
+#  define Py_C_RECURSION_LIMIT 3000
+#elif defined(__ANDROID__)
+   // On an ARM64 emulator, API level 34 was OK with 10000, but API level 21
+   // crashed in test_compiler_recursion_limit.
+#  define Py_C_RECURSION_LIMIT 3000
+#elif defined(_Py_ADDRESS_SANITIZER)
+#  define Py_C_RECURSION_LIMIT 4000
+#elif defined(__wasi__)
+   // Based on wasmtime 16.
+#  define Py_C_RECURSION_LIMIT 5000
+#else
+   // This value is duplicated in Lib/test/support/__init__.py
+#  define Py_C_RECURSION_LIMIT 10000
+#endif
+
+
+/* other API */
+
+/* Similar to PyThreadState_Get(), but don't issue a fatal error
+ * if it is NULL. */
+PyAPI_FUNC(PyThreadState *) PyThreadState_GetUnchecked(void);
+
+// Alias kept for backward compatibility
+#define _PyThreadState_UncheckedGet PyThreadState_GetUnchecked
+
+
+// Disable tracing and profiling.
+PyAPI_FUNC(void) PyThreadState_EnterTracing(PyThreadState *tstate);
+
+// Reset tracing and profiling: enable them if a trace function or a profile
+// function is set, otherwise disable them.
+PyAPI_FUNC(void) PyThreadState_LeaveTracing(PyThreadState *tstate);
+
+/* PyGILState */
+
+/* Helper/diagnostic function - return 1 if the current thread
+   currently holds the GIL, 0 otherwise.
+
+   The function returns 1 if _PyGILState_check_enabled is non-zero. */
+PyAPI_FUNC(int) PyGILState_Check(void);
+
+/* The implementation of sys._current_frames()  Returns a dict mapping
+   thread id to that thread's current frame.
+*/
+PyAPI_FUNC(PyObject*) _PyThread_CurrentFrames(void);
+
+/* Routines for advanced debuggers, requested by David Beazley.
+   Don't use unless you know what you are doing! */
+PyAPI_FUNC(PyInterpreterState *) PyInterpreterState_Main(void);
+PyAPI_FUNC(PyInterpreterState *) PyInterpreterState_Head(void);
+PyAPI_FUNC(PyInterpreterState *) PyInterpreterState_Next(PyInterpreterState *);
+PyAPI_FUNC(PyThreadState *) PyInterpreterState_ThreadHead(PyInterpreterState *);
+PyAPI_FUNC(PyThreadState *) PyThreadState_Next(PyThreadState *);
+PyAPI_FUNC(void) PyThreadState_DeleteCurrent(void);
+
+/* Frame evaluation API */
+
+typedef PyObject* (*_PyFrameEvalFunction)(PyThreadState *tstate, struct _PyInterpreterFrame *, int);
+
+PyAPI_FUNC(_PyFrameEvalFunction) _PyInterpreterState_GetEvalFrameFunc(
+    PyInterpreterState *interp);
+PyAPI_FUNC(void) _PyInterpreterState_SetEvalFrameFunc(
+    PyInterpreterState *interp,
+    _PyFrameEvalFunction eval_frame);

Dependencies/Python/include/cpython/pystats.h

@@ -0,0 +1,175 @@
+// Statistics on Python performance.
+//
+// API:
+//
+// - _Py_INCREF_STAT_INC() and _Py_DECREF_STAT_INC() used by Py_INCREF()
+//   and Py_DECREF().
+// - _Py_stats variable
+//
+// Functions of the sys module:
+//
+// - sys._stats_on()
+// - sys._stats_off()
+// - sys._stats_clear()
+// - sys._stats_dump()
+//
+// Python must be built with ./configure --enable-pystats to define the
+// Py_STATS macro.
+//
+// Define _PY_INTERPRETER macro to increment interpreter_increfs and
+// interpreter_decrefs. Otherwise, increment increfs and decrefs.
+//
+// The number of incref operations counted by `incref` and
+// `interpreter_incref` is the number of increment operations, which is
+// not equal to the total of all reference counts. A single increment
+// operation may increase the reference count of an object by more than
+// one. For example, see `_Py_RefcntAdd`.
+
+#ifndef Py_CPYTHON_PYSTATS_H
+#  error "this header file must not be included directly"
+#endif
+
+#define PYSTATS_MAX_UOP_ID 512
+
+#define SPECIALIZATION_FAILURE_KINDS 36
+
+/* Stats for determining who is calling PyEval_EvalFrame */
+#define EVAL_CALL_TOTAL 0
+#define EVAL_CALL_VECTOR 1
+#define EVAL_CALL_GENERATOR 2
+#define EVAL_CALL_LEGACY 3
+#define EVAL_CALL_FUNCTION_VECTORCALL 4
+#define EVAL_CALL_BUILD_CLASS 5
+#define EVAL_CALL_SLOT 6
+#define EVAL_CALL_FUNCTION_EX 7
+#define EVAL_CALL_API 8
+#define EVAL_CALL_METHOD 9
+
+#define EVAL_CALL_KINDS 10
+
+typedef struct _specialization_stats {
+    uint64_t success;
+    uint64_t failure;
+    uint64_t hit;
+    uint64_t deferred;
+    uint64_t miss;
+    uint64_t deopt;
+    uint64_t failure_kinds[SPECIALIZATION_FAILURE_KINDS];
+} SpecializationStats;
+
+typedef struct _opcode_stats {
+    SpecializationStats specialization;
+    uint64_t execution_count;
+    uint64_t pair_count[256];
+} OpcodeStats;
+
+typedef struct _call_stats {
+    uint64_t inlined_py_calls;
+    uint64_t pyeval_calls;
+    uint64_t frames_pushed;
+    uint64_t frame_objects_created;
+    uint64_t eval_calls[EVAL_CALL_KINDS];
+} CallStats;
+
+typedef struct _object_stats {
+    uint64_t increfs;
+    uint64_t decrefs;
+    uint64_t interpreter_increfs;
+    uint64_t interpreter_decrefs;
+    uint64_t allocations;
+    uint64_t allocations512;
+    uint64_t allocations4k;
+    uint64_t allocations_big;
+    uint64_t frees;
+    uint64_t to_freelist;
+    uint64_t from_freelist;
+    uint64_t inline_values;
+    uint64_t dict_materialized_on_request;
+    uint64_t dict_materialized_new_key;
+    uint64_t dict_materialized_too_big;
+    uint64_t dict_materialized_str_subclass;
+    uint64_t type_cache_hits;
+    uint64_t type_cache_misses;
+    uint64_t type_cache_dunder_hits;
+    uint64_t type_cache_dunder_misses;
+    uint64_t type_cache_collisions;
+    /* Temporary value used during GC */
+    uint64_t object_visits;
+} ObjectStats;
+
+typedef struct _gc_stats {
+    uint64_t collections;
+    uint64_t object_visits;
+    uint64_t objects_collected;
+} GCStats;
+
+typedef struct _uop_stats {
+    uint64_t execution_count;
+    uint64_t miss;
+    uint64_t pair_count[PYSTATS_MAX_UOP_ID + 1];
+} UOpStats;
+
+#define _Py_UOP_HIST_SIZE 32
+
+typedef struct _optimization_stats {
+    uint64_t attempts;
+    uint64_t traces_created;
+    uint64_t traces_executed;
+    uint64_t uops_executed;
+    uint64_t trace_stack_overflow;
+    uint64_t trace_stack_underflow;
+    uint64_t trace_too_long;
+    uint64_t trace_too_short;
+    uint64_t inner_loop;
+    uint64_t recursive_call;
+    uint64_t low_confidence;
+    uint64_t executors_invalidated;
+    UOpStats opcode[PYSTATS_MAX_UOP_ID + 1];
+    uint64_t unsupported_opcode[256];
+    uint64_t trace_length_hist[_Py_UOP_HIST_SIZE];
+    uint64_t trace_run_length_hist[_Py_UOP_HIST_SIZE];
+    uint64_t optimized_trace_length_hist[_Py_UOP_HIST_SIZE];
+    uint64_t optimizer_attempts;
+    uint64_t optimizer_successes;
+    uint64_t optimizer_failure_reason_no_memory;
+    uint64_t remove_globals_builtins_changed;
+    uint64_t remove_globals_incorrect_keys;
+    uint64_t error_in_opcode[PYSTATS_MAX_UOP_ID + 1];
+} OptimizationStats;
+
+typedef struct _rare_event_stats {
+    /* Setting an object's class, obj.__class__ = ... */
+    uint64_t set_class;
+    /* Setting the bases of a class, cls.__bases__ = ... */
+    uint64_t set_bases;
+    /* Setting the PEP 523 frame eval function, _PyInterpreterState_SetFrameEvalFunc() */
+    uint64_t set_eval_frame_func;
+    /* Modifying the builtins,  __builtins__.__dict__[var] = ... */
+    uint64_t builtin_dict;
+    /* Modifying a function, e.g. func.__defaults__ = ..., etc. */
+    uint64_t func_modification;
+    /* Modifying a dict that is being watched */
+    uint64_t watched_dict_modification;
+    uint64_t watched_globals_modification;
+} RareEventStats;
+
+typedef struct _stats {
+    OpcodeStats opcode_stats[256];
+    CallStats call_stats;
+    ObjectStats object_stats;
+    OptimizationStats optimization_stats;
+    RareEventStats rare_event_stats;
+    GCStats *gc_stats;
+} PyStats;
+
+
+// Export for shared extensions like 'math'
+PyAPI_DATA(PyStats*) _Py_stats;
+
+#ifdef _PY_INTERPRETER
+#  define _Py_INCREF_STAT_INC() do { if (_Py_stats) _Py_stats->object_stats.interpreter_increfs++; } while (0)
+#  define _Py_DECREF_STAT_INC() do { if (_Py_stats) _Py_stats->object_stats.interpreter_decrefs++; } while (0)
+#else
+#  define _Py_INCREF_STAT_INC() do { if (_Py_stats) _Py_stats->object_stats.increfs++; } while (0)
+#  define _Py_DECREF_STAT_INC() do { if (_Py_stats) _Py_stats->object_stats.decrefs++; } while (0)
+#endif

Dependencies/Python/include/cpython/pythonrun.h

@@ -0,0 +1,96 @@
+#ifndef Py_CPYTHON_PYTHONRUN_H
+#  error "this header file must not be included directly"
+#endif
+
+PyAPI_FUNC(int) PyRun_SimpleStringFlags(const char *, PyCompilerFlags *);
+PyAPI_FUNC(int) PyRun_AnyFileExFlags(
+    FILE *fp,
+    const char *filename,       /* decoded from the filesystem encoding */
+    int closeit,
+    PyCompilerFlags *flags);
+PyAPI_FUNC(int) PyRun_SimpleFileExFlags(
+    FILE *fp,
+    const char *filename,       /* decoded from the filesystem encoding */
+    int closeit,
+    PyCompilerFlags *flags);
+PyAPI_FUNC(int) PyRun_InteractiveOneFlags(
+    FILE *fp,
+    const char *filename,       /* decoded from the filesystem encoding */
+    PyCompilerFlags *flags);
+PyAPI_FUNC(int) PyRun_InteractiveOneObject(
+    FILE *fp,
+    PyObject *filename,
+    PyCompilerFlags *flags);
+PyAPI_FUNC(int) PyRun_InteractiveLoopFlags(
+    FILE *fp,
+    const char *filename,       /* decoded from the filesystem encoding */
+    PyCompilerFlags *flags);
+
+
+PyAPI_FUNC(PyObject *) PyRun_StringFlags(const char *, int, PyObject *,
+                                         PyObject *, PyCompilerFlags *);
+
+PyAPI_FUNC(PyObject *) PyRun_FileExFlags(
+    FILE *fp,
+    const char *filename,       /* decoded from the filesystem encoding */
+    int start,
+    PyObject *globals,
+    PyObject *locals,
+    int closeit,
+    PyCompilerFlags *flags);
+
+
+PyAPI_FUNC(PyObject *) Py_CompileStringExFlags(
+    const char *str,
+    const char *filename,       /* decoded from the filesystem encoding */
+    int start,
+    PyCompilerFlags *flags,
+    int optimize);
+PyAPI_FUNC(PyObject *) Py_CompileStringObject(
+    const char *str,
+    PyObject *filename, int start,
+    PyCompilerFlags *flags,
+    int optimize);
+
+#define Py_CompileString(str, p, s) Py_CompileStringExFlags((str), (p), (s), NULL, -1)
+#define Py_CompileStringFlags(str, p, s, f) Py_CompileStringExFlags((str), (p), (s), (f), -1)
+
+/* A function flavor is also exported by libpython. It is required when
+    libpython is accessed directly rather than using header files which defines
+    macros below. On Windows, for example, PyAPI_FUNC() uses dllexport to
+    export functions in pythonXX.dll. */
+PyAPI_FUNC(PyObject *) PyRun_String(const char *str, int s, PyObject *g, PyObject *l);
+PyAPI_FUNC(int) PyRun_AnyFile(FILE *fp, const char *name);
+PyAPI_FUNC(int) PyRun_AnyFileEx(FILE *fp, const char *name, int closeit);
+PyAPI_FUNC(int) PyRun_AnyFileFlags(FILE *, const char *, PyCompilerFlags *);
+PyAPI_FUNC(int) PyRun_SimpleString(const char *s);
+PyAPI_FUNC(int) PyRun_SimpleFile(FILE *f, const char *p);
+PyAPI_FUNC(int) PyRun_SimpleFileEx(FILE *f, const char *p, int c);
+PyAPI_FUNC(int) PyRun_InteractiveOne(FILE *f, const char *p);
+PyAPI_FUNC(int) PyRun_InteractiveLoop(FILE *f, const char *p);
+PyAPI_FUNC(PyObject *) PyRun_File(FILE *fp, const char *p, int s, PyObject *g, PyObject *l);
+PyAPI_FUNC(PyObject *) PyRun_FileEx(FILE *fp, const char *p, int s, PyObject *g, PyObject *l, int c);
+PyAPI_FUNC(PyObject *) PyRun_FileFlags(FILE *fp, const char *p, int s, PyObject *g, PyObject *l, PyCompilerFlags *flags);
+
+/* Use macros for a bunch of old variants */
+#define PyRun_String(str, s, g, l) PyRun_StringFlags((str), (s), (g), (l), NULL)
+#define PyRun_AnyFile(fp, name) PyRun_AnyFileExFlags((fp), (name), 0, NULL)
+#define PyRun_AnyFileEx(fp, name, closeit) \
+    PyRun_AnyFileExFlags((fp), (name), (closeit), NULL)
+#define PyRun_AnyFileFlags(fp, name, flags) \
+    PyRun_AnyFileExFlags((fp), (name), 0, (flags))
+#define PyRun_SimpleString(s) PyRun_SimpleStringFlags((s), NULL)
+#define PyRun_SimpleFile(f, p) PyRun_SimpleFileExFlags((f), (p), 0, NULL)
+#define PyRun_SimpleFileEx(f, p, c) PyRun_SimpleFileExFlags((f), (p), (c), NULL)
+#define PyRun_InteractiveOne(f, p) PyRun_InteractiveOneFlags((f), (p), NULL)
+#define PyRun_InteractiveLoop(f, p) PyRun_InteractiveLoopFlags((f), (p), NULL)
+#define PyRun_File(fp, p, s, g, l) \
+    PyRun_FileExFlags((fp), (p), (s), (g), (l), 0, NULL)
+#define PyRun_FileEx(fp, p, s, g, l, c) \
+    PyRun_FileExFlags((fp), (p), (s), (g), (l), (c), NULL)
+#define PyRun_FileFlags(fp, p, s, g, l, flags) \
+    PyRun_FileExFlags((fp), (p), (s), (g), (l), 0, (flags))
+
+/* Stuff with no proper home (yet) */
+PyAPI_FUNC(char *) PyOS_Readline(FILE *, FILE *, const char *);
+PyAPI_DATA(char) *(*PyOS_ReadlineFunctionPointer)(FILE *, FILE *, const char *);

Dependencies/Python/include/cpython/pythread.h

@@ -0,0 +1,43 @@
+#ifndef Py_CPYTHON_PYTHREAD_H
+#  error "this header file must not be included directly"
+#endif
+
+// PY_TIMEOUT_MAX is the highest usable value (in microseconds) of PY_TIMEOUT_T
+// type, and depends on the system threading API.
+//
+// NOTE: this isn't the same value as `_thread.TIMEOUT_MAX`. The _thread module
+// exposes a higher-level API, with timeouts expressed in seconds and
+// floating-point numbers allowed.
+PyAPI_DATA(const long long) PY_TIMEOUT_MAX;
+
+#define PYTHREAD_INVALID_THREAD_ID ((unsigned long)-1)
+
+#ifdef HAVE_PTHREAD_H
+    /* Darwin needs pthread.h to know type name the pthread_key_t. */
+#   include <pthread.h>
+#   define NATIVE_TSS_KEY_T     pthread_key_t
+#elif defined(NT_THREADS)
+    /* In Windows, native TSS key type is DWORD,
+       but hardcode the unsigned long to avoid errors for include directive.
+    */
+#   define NATIVE_TSS_KEY_T     unsigned long
+#elif defined(HAVE_PTHREAD_STUBS)
+#   include "pthread_stubs.h"
+#   define NATIVE_TSS_KEY_T     pthread_key_t
+#else
+#   error "Require native threads. See https://bugs.python.org/issue31370"
+#endif
+
+/* When Py_LIMITED_API is not defined, the type layout of Py_tss_t is
+   exposed to allow static allocation in the API clients.  Even in this case,
+   you must handle TSS keys through API functions due to compatibility.
+*/
+struct _Py_tss_t {
+    int _is_initialized;
+    NATIVE_TSS_KEY_T _key;
+};
+
+#undef NATIVE_TSS_KEY_T
+
+/* When static allocation, you must initialize with Py_tss_NEEDS_INIT. */
+#define Py_tss_NEEDS_INIT   {0}

Dependencies/Python/include/cpython/pytime.h

@@ -0,0 +1,27 @@
+// PyTime_t C API: see Doc/c-api/time.rst for the documentation.
+
+#ifndef Py_LIMITED_API
+#ifndef Py_PYTIME_H
+#define Py_PYTIME_H
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+typedef int64_t PyTime_t;
+#define PyTime_MIN INT64_MIN
+#define PyTime_MAX INT64_MAX
+
+PyAPI_FUNC(double) PyTime_AsSecondsDouble(PyTime_t t);
+PyAPI_FUNC(int) PyTime_Monotonic(PyTime_t *result);
+PyAPI_FUNC(int) PyTime_PerfCounter(PyTime_t *result);
+PyAPI_FUNC(int) PyTime_Time(PyTime_t *result);
+
+PyAPI_FUNC(int) PyTime_MonotonicRaw(PyTime_t *result);
+PyAPI_FUNC(int) PyTime_PerfCounterRaw(PyTime_t *result);
+PyAPI_FUNC(int) PyTime_TimeRaw(PyTime_t *result);
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* Py_PYTIME_H */
+#endif /* Py_LIMITED_API */

Dependencies/Python/include/cpython/setobject.h

@@ -0,0 +1,71 @@
+#ifndef Py_CPYTHON_SETOBJECT_H
+#  error "this header file must not be included directly"
+#endif
+
+/* There are three kinds of entries in the table:
+
+1. Unused:  key == NULL and hash == 0
+2. Dummy:   key == dummy and hash == -1
+3. Active:  key != NULL and key != dummy and hash != -1
+
+The hash field of Unused slots is always zero.
+
+The hash field of Dummy slots are set to -1
+meaning that dummy entries can be detected by
+either entry->key==dummy or by entry->hash==-1.
+*/
+
+#define PySet_MINSIZE 8
+
+typedef struct {
+    PyObject *key;
+    Py_hash_t hash;             /* Cached hash code of the key */
+} setentry;
+
+/* The SetObject data structure is shared by set and frozenset objects.
+
+Invariant for sets:
+ - hash is -1
+
+Invariants for frozensets:
+ - data is immutable.
+ - hash is the hash of the frozenset or -1 if not computed yet.
+
+*/
+
+typedef struct {
+    PyObject_HEAD
+
+    Py_ssize_t fill;            /* Number active and dummy entries*/
+    Py_ssize_t used;            /* Number active entries */
+
+    /* The table contains mask + 1 slots, and that's a power of 2.
+     * We store the mask instead of the size because the mask is more
+     * frequently needed.
+     */
+    Py_ssize_t mask;
+
+    /* The table points to a fixed-size smalltable for small tables
+     * or to additional malloc'ed memory for bigger tables.
+     * The table pointer is never NULL which saves us from repeated
+     * runtime null-tests.
+     */
+    setentry *table;
+    Py_hash_t hash;             /* Only used by frozenset objects */
+    Py_ssize_t finger;          /* Search finger for pop() */
+
+    setentry smalltable[PySet_MINSIZE];
+    PyObject *weakreflist;      /* List of weak references */
+} PySetObject;
+
+#define _PySet_CAST(so) \
+    (assert(PyAnySet_Check(so)), _Py_CAST(PySetObject*, so))
+
+static inline Py_ssize_t PySet_GET_SIZE(PyObject *so) {
+#ifdef Py_GIL_DISABLED
+    return _Py_atomic_load_ssize_relaxed(&(_PySet_CAST(so)->used));
+#else
+    return _PySet_CAST(so)->used;
+#endif
+}
+#define PySet_GET_SIZE(so) PySet_GET_SIZE(_PyObject_CAST(so))

Dependencies/Python/include/cpython/sysmodule.h

@@ -0,0 +1,22 @@
+#ifndef Py_CPYTHON_SYSMODULE_H
+#  error "this header file must not be included directly"
+#endif
+
+typedef int(*Py_AuditHookFunction)(const char *, PyObject *, void *);
+
+PyAPI_FUNC(int) PySys_AddAuditHook(Py_AuditHookFunction, void*);
+
+typedef struct {
+    FILE* perf_map;
+    PyThread_type_lock map_lock;
+} PerfMapState;
+
+PyAPI_FUNC(int) PyUnstable_PerfMapState_Init(void);
+PyAPI_FUNC(int) PyUnstable_WritePerfMapEntry(
+    const void *code_addr,
+    unsigned int code_size,
+    const char *entry_name);
+PyAPI_FUNC(void) PyUnstable_PerfMapState_Fini(void);
+PyAPI_FUNC(int) PyUnstable_CopyPerfMapFile(const char* parent_filename);
+PyAPI_FUNC(int) PyUnstable_PerfTrampoline_CompileCode(PyCodeObject *);
+PyAPI_FUNC(int) PyUnstable_PerfTrampoline_SetPersistAfterFork(int enable);

Dependencies/Python/include/cpython/traceback.h

@@ -0,0 +1,13 @@
+#ifndef Py_CPYTHON_TRACEBACK_H
+#  error "this header file must not be included directly"
+#endif
+
+typedef struct _traceback PyTracebackObject;
+
+struct _traceback {
+    PyObject_HEAD
+    PyTracebackObject *tb_next;
+    PyFrameObject *tb_frame;
+    int tb_lasti;
+    int tb_lineno;
+};

Dependencies/Python/include/cpython/tracemalloc.h

@@ -0,0 +1,32 @@
+#ifndef Py_LIMITED_API
+#ifndef Py_TRACEMALLOC_H
+#define Py_TRACEMALLOC_H
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/* Track an allocated memory block in the tracemalloc module.
+   Return 0 on success, return -1 on error (failed to allocate memory to store
+   the trace).
+
+   Return -2 if tracemalloc is disabled.
+
+   If memory block is already tracked, update the existing trace. */
+PyAPI_FUNC(int) PyTraceMalloc_Track(
+    unsigned int domain,
+    uintptr_t ptr,
+    size_t size);
+
+/* Untrack an allocated memory block in the tracemalloc module.
+   Do nothing if the block was not tracked.
+
+   Return -2 if tracemalloc is disabled, otherwise return 0. */
+PyAPI_FUNC(int) PyTraceMalloc_Untrack(
+    unsigned int domain,
+    uintptr_t ptr);
+
+#ifdef __cplusplus
+}
+#endif
+#endif  // !Py_TRACEMALLOC_H
+#endif  // !Py_LIMITED_API

Dependencies/Python/include/cpython/tupleobject.h

@@ -0,0 +1,38 @@
+#ifndef Py_CPYTHON_TUPLEOBJECT_H
+#  error "this header file must not be included directly"
+#endif
+
+typedef struct {
+    PyObject_VAR_HEAD
+    /* ob_item contains space for 'ob_size' elements.
+       Items must normally not be NULL, except during construction when
+       the tuple is not yet visible outside the function that builds it. */
+    PyObject *ob_item[1];
+} PyTupleObject;
+
+PyAPI_FUNC(int) _PyTuple_Resize(PyObject **, Py_ssize_t);
+
+/* Cast argument to PyTupleObject* type. */
+#define _PyTuple_CAST(op) \
+    (assert(PyTuple_Check(op)), _Py_CAST(PyTupleObject*, (op)))
+
+// Macros and static inline functions, trading safety for speed
+
+static inline Py_ssize_t PyTuple_GET_SIZE(PyObject *op) {
+    PyTupleObject *tuple = _PyTuple_CAST(op);
+    return Py_SIZE(tuple);
+}
+#define PyTuple_GET_SIZE(op) PyTuple_GET_SIZE(_PyObject_CAST(op))
+
+#define PyTuple_GET_ITEM(op, index) (_PyTuple_CAST(op)->ob_item[(index)])
+
+/* Function *only* to be used to fill in brand new tuples */
+static inline void
+PyTuple_SET_ITEM(PyObject *op, Py_ssize_t index, PyObject *value) {
+    PyTupleObject *tuple = _PyTuple_CAST(op);
+    assert(0 <= index);
+    assert(index < Py_SIZE(tuple));
+    tuple->ob_item[index] = value;
+}
+#define PyTuple_SET_ITEM(op, index, value) \
+    PyTuple_SET_ITEM(_PyObject_CAST(op), (index), _PyObject_CAST(value))

Dependencies/Python/include/cpython/unicodeobject.h

@@ -0,0 +1,703 @@
+#ifndef Py_CPYTHON_UNICODEOBJECT_H
+#  error "this header file must not be included directly"
+#endif
+
+/* Py_UNICODE was the native Unicode storage format (code unit) used by
+   Python and represents a single Unicode element in the Unicode type.
+   With PEP 393, Py_UNICODE is deprecated and replaced with a
+   typedef to wchar_t. */
+Py_DEPRECATED(3.13) typedef wchar_t PY_UNICODE_TYPE;
+Py_DEPRECATED(3.13) typedef wchar_t Py_UNICODE;
+
+
+/* --- Internal Unicode Operations ---------------------------------------- */
+
+// Static inline functions to work with surrogates
+static inline int Py_UNICODE_IS_SURROGATE(Py_UCS4 ch) {
+    return (0xD800 <= ch && ch <= 0xDFFF);
+}
+static inline int Py_UNICODE_IS_HIGH_SURROGATE(Py_UCS4 ch) {
+    return (0xD800 <= ch && ch <= 0xDBFF);
+}
+static inline int Py_UNICODE_IS_LOW_SURROGATE(Py_UCS4 ch) {
+    return (0xDC00 <= ch && ch <= 0xDFFF);
+}
+
+// Join two surrogate characters and return a single Py_UCS4 value.
+static inline Py_UCS4 Py_UNICODE_JOIN_SURROGATES(Py_UCS4 high, Py_UCS4 low)  {
+    assert(Py_UNICODE_IS_HIGH_SURROGATE(high));
+    assert(Py_UNICODE_IS_LOW_SURROGATE(low));
+    return 0x10000 + (((high & 0x03FF) << 10) | (low & 0x03FF));
+}
+
+// High surrogate = top 10 bits added to 0xD800.
+// The character must be in the range [U+10000; U+10ffff].
+static inline Py_UCS4 Py_UNICODE_HIGH_SURROGATE(Py_UCS4 ch) {
+    assert(0x10000 <= ch && ch <= 0x10ffff);
+    return (0xD800 - (0x10000 >> 10) + (ch >> 10));
+}
+
+// Low surrogate = bottom 10 bits added to 0xDC00.
+// The character must be in the range [U+10000; U+10ffff].
+static inline Py_UCS4 Py_UNICODE_LOW_SURROGATE(Py_UCS4 ch) {
+    assert(0x10000 <= ch && ch <= 0x10ffff);
+    return (0xDC00 + (ch & 0x3FF));
+}
+
+
+/* --- Unicode Type ------------------------------------------------------- */
+
+/* ASCII-only strings created through PyUnicode_New use the PyASCIIObject
+   structure. state.ascii and state.compact are set, and the data
+   immediately follow the structure. utf8_length can be found
+   in the length field; the utf8 pointer is equal to the data pointer. */
+typedef struct {
+    /* There are 4 forms of Unicode strings:
+
+       - compact ascii:
+
+         * structure = PyASCIIObject
+         * test: PyUnicode_IS_COMPACT_ASCII(op)
+         * kind = PyUnicode_1BYTE_KIND
+         * compact = 1
+         * ascii = 1
+         * (length is the length of the utf8)
+         * (data starts just after the structure)
+         * (since ASCII is decoded from UTF-8, the utf8 string are the data)
+
+       - compact:
+
+         * structure = PyCompactUnicodeObject
+         * test: PyUnicode_IS_COMPACT(op) && !PyUnicode_IS_ASCII(op)
+         * kind = PyUnicode_1BYTE_KIND, PyUnicode_2BYTE_KIND or
+           PyUnicode_4BYTE_KIND
+         * compact = 1
+         * ascii = 0
+         * utf8 is not shared with data
+         * utf8_length = 0 if utf8 is NULL
+         * (data starts just after the structure)
+
+       - legacy string:
+
+         * structure = PyUnicodeObject structure
+         * test: !PyUnicode_IS_COMPACT(op)
+         * kind = PyUnicode_1BYTE_KIND, PyUnicode_2BYTE_KIND or
+           PyUnicode_4BYTE_KIND
+         * compact = 0
+         * data.any is not NULL
+         * utf8 is shared and utf8_length = length with data.any if ascii = 1
+         * utf8_length = 0 if utf8 is NULL
+
+       Compact strings use only one memory block (structure + characters),
+       whereas legacy strings use one block for the structure and one block
+       for characters.
+
+       Legacy strings are created by subclasses of Unicode.
+
+       See also _PyUnicode_CheckConsistency().
+    */
+    PyObject_HEAD
+    Py_ssize_t length;          /* Number of code points in the string */
+    Py_hash_t hash;             /* Hash value; -1 if not set */
+    struct {
+        /* If interned is non-zero, the two references from the
+           dictionary to this object are *not* counted in ob_refcnt.
+           The possible values here are:
+               0: Not Interned
+               1: Interned
+               2: Interned and Immortal
+               3: Interned, Immortal, and Static
+           This categorization allows the runtime to determine the right
+           cleanup mechanism at runtime shutdown. */
+        unsigned int interned:2;
+        /* Character size:
+
+           - PyUnicode_1BYTE_KIND (1):
+
+             * character type = Py_UCS1 (8 bits, unsigned)
+             * all characters are in the range U+0000-U+00FF (latin1)
+             * if ascii is set, all characters are in the range U+0000-U+007F
+               (ASCII), otherwise at least one character is in the range
+               U+0080-U+00FF
+
+           - PyUnicode_2BYTE_KIND (2):
+
+             * character type = Py_UCS2 (16 bits, unsigned)
+             * all characters are in the range U+0000-U+FFFF (BMP)
+             * at least one character is in the range U+0100-U+FFFF
+
+           - PyUnicode_4BYTE_KIND (4):
+
+             * character type = Py_UCS4 (32 bits, unsigned)
+             * all characters are in the range U+0000-U+10FFFF
+             * at least one character is in the range U+10000-U+10FFFF
+         */
+        unsigned int kind:3;
+        /* Compact is with respect to the allocation scheme. Compact unicode
+           objects only require one memory block while non-compact objects use
+           one block for the PyUnicodeObject struct and another for its data
+           buffer. */
+        unsigned int compact:1;
+        /* The string only contains characters in the range U+0000-U+007F (ASCII)
+           and the kind is PyUnicode_1BYTE_KIND. If ascii is set and compact is
+           set, use the PyASCIIObject structure. */
+        unsigned int ascii:1;
+        /* The object is statically allocated. */
+        unsigned int statically_allocated:1;
+        /* Padding to ensure that PyUnicode_DATA() is always aligned to
+           4 bytes (see issue #19537 on m68k). */
+        unsigned int :24;
+    } state;
+} PyASCIIObject;
+
+/* Non-ASCII strings allocated through PyUnicode_New use the
+   PyCompactUnicodeObject structure. state.compact is set, and the data
+   immediately follow the structure. */
+typedef struct {
+    PyASCIIObject _base;
+    Py_ssize_t utf8_length;     /* Number of bytes in utf8, excluding the
+                                 * terminating \0. */
+    char *utf8;                 /* UTF-8 representation (null-terminated) */
+} PyCompactUnicodeObject;
+
+/* Object format for Unicode subclasses. */
+typedef struct {
+    PyCompactUnicodeObject _base;
+    union {
+        void *any;
+        Py_UCS1 *latin1;
+        Py_UCS2 *ucs2;
+        Py_UCS4 *ucs4;
+    } data;                     /* Canonical, smallest-form Unicode buffer */
+} PyUnicodeObject;
+
+
+#define _PyASCIIObject_CAST(op) \
+    (assert(PyUnicode_Check(op)), \
+     _Py_CAST(PyASCIIObject*, (op)))
+#define _PyCompactUnicodeObject_CAST(op) \
+    (assert(PyUnicode_Check(op)), \
+     _Py_CAST(PyCompactUnicodeObject*, (op)))
+#define _PyUnicodeObject_CAST(op) \
+    (assert(PyUnicode_Check(op)), \
+     _Py_CAST(PyUnicodeObject*, (op)))
+
+
+/* --- Flexible String Representation Helper Macros (PEP 393) -------------- */
+
+/* Values for PyASCIIObject.state: */
+
+/* Interning state. */
+#define SSTATE_NOT_INTERNED 0
+#define SSTATE_INTERNED_MORTAL 1
+#define SSTATE_INTERNED_IMMORTAL 2
+#define SSTATE_INTERNED_IMMORTAL_STATIC 3
+
+/* Use only if you know it's a string */
+static inline unsigned int PyUnicode_CHECK_INTERNED(PyObject *op) {
+    return _PyASCIIObject_CAST(op)->state.interned;
+}
+#define PyUnicode_CHECK_INTERNED(op) PyUnicode_CHECK_INTERNED(_PyObject_CAST(op))
+
+/* For backward compatibility */
+static inline unsigned int PyUnicode_IS_READY(PyObject* Py_UNUSED(op)) {
+    return 1;
+}
+#define PyUnicode_IS_READY(op) PyUnicode_IS_READY(_PyObject_CAST(op))
+
+/* Return true if the string contains only ASCII characters, or 0 if not. The
+   string may be compact (PyUnicode_IS_COMPACT_ASCII) or not, but must be
+   ready. */
+static inline unsigned int PyUnicode_IS_ASCII(PyObject *op) {
+    return _PyASCIIObject_CAST(op)->state.ascii;
+}
+#define PyUnicode_IS_ASCII(op) PyUnicode_IS_ASCII(_PyObject_CAST(op))
+
+/* Return true if the string is compact or 0 if not.
+   No type checks or Ready calls are performed. */
+static inline unsigned int PyUnicode_IS_COMPACT(PyObject *op) {
+    return _PyASCIIObject_CAST(op)->state.compact;
+}
+#define PyUnicode_IS_COMPACT(op) PyUnicode_IS_COMPACT(_PyObject_CAST(op))
+
+/* Return true if the string is a compact ASCII string (use PyASCIIObject
+   structure), or 0 if not.  No type checks or Ready calls are performed. */
+static inline int PyUnicode_IS_COMPACT_ASCII(PyObject *op) {
+    return (_PyASCIIObject_CAST(op)->state.ascii && PyUnicode_IS_COMPACT(op));
+}
+#define PyUnicode_IS_COMPACT_ASCII(op) PyUnicode_IS_COMPACT_ASCII(_PyObject_CAST(op))
+
+enum PyUnicode_Kind {
+/* Return values of the PyUnicode_KIND() function: */
+    PyUnicode_1BYTE_KIND = 1,
+    PyUnicode_2BYTE_KIND = 2,
+    PyUnicode_4BYTE_KIND = 4
+};
+
+// PyUnicode_KIND(): Return one of the PyUnicode_*_KIND values defined above.
+//
+// gh-89653: Converting this macro to a static inline function would introduce
+// new compiler warnings on "kind < PyUnicode_KIND(str)" (compare signed and
+// unsigned numbers) where kind type is an int or on
+// "unsigned int kind = PyUnicode_KIND(str)" (cast signed to unsigned).
+#define PyUnicode_KIND(op) _Py_RVALUE(_PyASCIIObject_CAST(op)->state.kind)
+
+/* Return a void pointer to the raw unicode buffer. */
+static inline void* _PyUnicode_COMPACT_DATA(PyObject *op) {
+    if (PyUnicode_IS_ASCII(op)) {
+        return _Py_STATIC_CAST(void*, (_PyASCIIObject_CAST(op) + 1));
+    }
+    return _Py_STATIC_CAST(void*, (_PyCompactUnicodeObject_CAST(op) + 1));
+}
+
+static inline void* _PyUnicode_NONCOMPACT_DATA(PyObject *op) {
+    void *data;
+    assert(!PyUnicode_IS_COMPACT(op));
+    data = _PyUnicodeObject_CAST(op)->data.any;
+    assert(data != NULL);
+    return data;
+}
+
+static inline void* PyUnicode_DATA(PyObject *op) {
+    if (PyUnicode_IS_COMPACT(op)) {
+        return _PyUnicode_COMPACT_DATA(op);
+    }
+    return _PyUnicode_NONCOMPACT_DATA(op);
+}
+#define PyUnicode_DATA(op) PyUnicode_DATA(_PyObject_CAST(op))
+
+/* Return pointers to the canonical representation cast to unsigned char,
+   Py_UCS2, or Py_UCS4 for direct character access.
+   No checks are performed, use PyUnicode_KIND() before to ensure
+   these will work correctly. */
+
+#define PyUnicode_1BYTE_DATA(op) _Py_STATIC_CAST(Py_UCS1*, PyUnicode_DATA(op))
+#define PyUnicode_2BYTE_DATA(op) _Py_STATIC_CAST(Py_UCS2*, PyUnicode_DATA(op))
+#define PyUnicode_4BYTE_DATA(op) _Py_STATIC_CAST(Py_UCS4*, PyUnicode_DATA(op))
+
+/* Returns the length of the unicode string. */
+static inline Py_ssize_t PyUnicode_GET_LENGTH(PyObject *op) {
+    return _PyASCIIObject_CAST(op)->length;
+}
+#define PyUnicode_GET_LENGTH(op) PyUnicode_GET_LENGTH(_PyObject_CAST(op))
+
+/* Write into the canonical representation, this function does not do any sanity
+   checks and is intended for usage in loops.  The caller should cache the
+   kind and data pointers obtained from other function calls.
+   index is the index in the string (starts at 0) and value is the new
+   code point value which should be written to that location. */
+static inline void PyUnicode_WRITE(int kind, void *data,
+                                   Py_ssize_t index, Py_UCS4 value)
+{
+    assert(index >= 0);
+    if (kind == PyUnicode_1BYTE_KIND) {
+        assert(value <= 0xffU);
+        _Py_STATIC_CAST(Py_UCS1*, data)[index] = _Py_STATIC_CAST(Py_UCS1, value);
+    }
+    else if (kind == PyUnicode_2BYTE_KIND) {
+        assert(value <= 0xffffU);
+        _Py_STATIC_CAST(Py_UCS2*, data)[index] = _Py_STATIC_CAST(Py_UCS2, value);
+    }
+    else {
+        assert(kind == PyUnicode_4BYTE_KIND);
+        assert(value <= 0x10ffffU);
+        _Py_STATIC_CAST(Py_UCS4*, data)[index] = value;
+    }
+}
+#define PyUnicode_WRITE(kind, data, index, value) \
+    PyUnicode_WRITE(_Py_STATIC_CAST(int, kind), _Py_CAST(void*, data), \
+                    (index), _Py_STATIC_CAST(Py_UCS4, value))
+
+/* Read a code point from the string's canonical representation.  No checks
+   or ready calls are performed. */
+static inline Py_UCS4 PyUnicode_READ(int kind,
+                                     const void *data, Py_ssize_t index)
+{
+    assert(index >= 0);
+    if (kind == PyUnicode_1BYTE_KIND) {
+        return _Py_STATIC_CAST(const Py_UCS1*, data)[index];
+    }
+    if (kind == PyUnicode_2BYTE_KIND) {
+        return _Py_STATIC_CAST(const Py_UCS2*, data)[index];
+    }
+    assert(kind == PyUnicode_4BYTE_KIND);
+    return _Py_STATIC_CAST(const Py_UCS4*, data)[index];
+}
+#define PyUnicode_READ(kind, data, index) \
+    PyUnicode_READ(_Py_STATIC_CAST(int, kind), \
+                   _Py_STATIC_CAST(const void*, data), \
+                   (index))
+
+/* PyUnicode_READ_CHAR() is less efficient than PyUnicode_READ() because it
+   calls PyUnicode_KIND() and might call it twice.  For single reads, use
+   PyUnicode_READ_CHAR, for multiple consecutive reads callers should
+   cache kind and use PyUnicode_READ instead. */
+static inline Py_UCS4 PyUnicode_READ_CHAR(PyObject *unicode, Py_ssize_t index)
+{
+    int kind;
+
+    assert(index >= 0);
+    // Tolerate reading the NUL character at str[len(str)]
+    assert(index <= PyUnicode_GET_LENGTH(unicode));
+
+    kind = PyUnicode_KIND(unicode);
+    if (kind == PyUnicode_1BYTE_KIND) {
+        return PyUnicode_1BYTE_DATA(unicode)[index];
+    }
+    if (kind == PyUnicode_2BYTE_KIND) {
+        return PyUnicode_2BYTE_DATA(unicode)[index];
+    }
+    assert(kind == PyUnicode_4BYTE_KIND);
+    return PyUnicode_4BYTE_DATA(unicode)[index];
+}
+#define PyUnicode_READ_CHAR(unicode, index) \
+    PyUnicode_READ_CHAR(_PyObject_CAST(unicode), (index))
+
+/* Return a maximum character value which is suitable for creating another
+   string based on op.  This is always an approximation but more efficient
+   than iterating over the string. */
+static inline Py_UCS4 PyUnicode_MAX_CHAR_VALUE(PyObject *op)
+{
+    int kind;
+
+    if (PyUnicode_IS_ASCII(op)) {
+        return 0x7fU;
+    }
+
+    kind = PyUnicode_KIND(op);
+    if (kind == PyUnicode_1BYTE_KIND) {
+       return 0xffU;
+    }
+    if (kind == PyUnicode_2BYTE_KIND) {
+        return 0xffffU;
+    }
+    assert(kind == PyUnicode_4BYTE_KIND);
+    return 0x10ffffU;
+}
+#define PyUnicode_MAX_CHAR_VALUE(op) \
+    PyUnicode_MAX_CHAR_VALUE(_PyObject_CAST(op))
+
+
+/* === Public API ========================================================= */
+
+/* With PEP 393, this is the recommended way to allocate a new unicode object.
+   This function will allocate the object and its buffer in a single memory
+   block.  Objects created using this function are not resizable. */
+PyAPI_FUNC(PyObject*) PyUnicode_New(
+    Py_ssize_t size,            /* Number of code points in the new string */
+    Py_UCS4 maxchar             /* maximum code point value in the string */
+    );
+
+/* For backward compatibility */
+static inline int PyUnicode_READY(PyObject* Py_UNUSED(op))
+{
+    return 0;
+}
+#define PyUnicode_READY(op) PyUnicode_READY(_PyObject_CAST(op))
+
+/* Copy character from one unicode object into another, this function performs
+   character conversion when necessary and falls back to memcpy() if possible.
+
+   Fail if to is too small (smaller than *how_many* or smaller than
+   len(from)-from_start), or if kind(from[from_start:from_start+how_many]) >
+   kind(to), or if *to* has more than 1 reference.
+
+   Return the number of written character, or return -1 and raise an exception
+   on error.
+
+   Pseudo-code:
+
+       how_many = min(how_many, len(from) - from_start)
+       to[to_start:to_start+how_many] = from[from_start:from_start+how_many]
+       return how_many
+
+   Note: The function doesn't write a terminating null character.
+   */
+PyAPI_FUNC(Py_ssize_t) PyUnicode_CopyCharacters(
+    PyObject *to,
+    Py_ssize_t to_start,
+    PyObject *from,
+    Py_ssize_t from_start,
+    Py_ssize_t how_many
+    );
+
+/* Fill a string with a character: write fill_char into
+   unicode[start:start+length].
+
+   Fail if fill_char is bigger than the string maximum character, or if the
+   string has more than 1 reference.
+
+   Return the number of written character, or return -1 and raise an exception
+   on error. */
+PyAPI_FUNC(Py_ssize_t) PyUnicode_Fill(
+    PyObject *unicode,
+    Py_ssize_t start,
+    Py_ssize_t length,
+    Py_UCS4 fill_char
+    );
+
+/* Create a new string from a buffer of Py_UCS1, Py_UCS2 or Py_UCS4 characters.
+   Scan the string to find the maximum character. */
+PyAPI_FUNC(PyObject*) PyUnicode_FromKindAndData(
+    int kind,
+    const void *buffer,
+    Py_ssize_t size);
+
+
+/* --- _PyUnicodeWriter API ----------------------------------------------- */
+
+typedef struct {
+    PyObject *buffer;
+    void *data;
+    int kind;
+    Py_UCS4 maxchar;
+    Py_ssize_t size;
+    Py_ssize_t pos;
+
+    /* minimum number of allocated characters (default: 0) */
+    Py_ssize_t min_length;
+
+    /* minimum character (default: 127, ASCII) */
+    Py_UCS4 min_char;
+
+    /* If non-zero, overallocate the buffer (default: 0). */
+    unsigned char overallocate;
+
+    /* If readonly is 1, buffer is a shared string (cannot be modified)
+       and size is set to 0. */
+    unsigned char readonly;
+} _PyUnicodeWriter ;
+
+// Initialize a Unicode writer.
+//
+// By default, the minimum buffer size is 0 character and overallocation is
+// disabled. Set min_length, min_char and overallocate attributes to control
+// the allocation of the buffer.
+PyAPI_FUNC(void)
+_PyUnicodeWriter_Init(_PyUnicodeWriter *writer);
+
+/* Prepare the buffer to write 'length' characters
+   with the specified maximum character.
+
+   Return 0 on success, raise an exception and return -1 on error. */
+#define _PyUnicodeWriter_Prepare(WRITER, LENGTH, MAXCHAR)             \
+    (((MAXCHAR) <= (WRITER)->maxchar                                  \
+      && (LENGTH) <= (WRITER)->size - (WRITER)->pos)                  \
+     ? 0                                                              \
+     : (((LENGTH) == 0)                                               \
+        ? 0                                                           \
+        : _PyUnicodeWriter_PrepareInternal((WRITER), (LENGTH), (MAXCHAR))))
+
+/* Don't call this function directly, use the _PyUnicodeWriter_Prepare() macro
+   instead. */
+PyAPI_FUNC(int)
+_PyUnicodeWriter_PrepareInternal(_PyUnicodeWriter *writer,
+                                 Py_ssize_t length, Py_UCS4 maxchar);
+
+/* Prepare the buffer to have at least the kind KIND.
+   For example, kind=PyUnicode_2BYTE_KIND ensures that the writer will
+   support characters in range U+000-U+FFFF.
+
+   Return 0 on success, raise an exception and return -1 on error. */
+#define _PyUnicodeWriter_PrepareKind(WRITER, KIND)                    \
+    ((KIND) <= (WRITER)->kind                                         \
+     ? 0                                                              \
+     : _PyUnicodeWriter_PrepareKindInternal((WRITER), (KIND)))
+
+/* Don't call this function directly, use the _PyUnicodeWriter_PrepareKind()
+   macro instead. */
+PyAPI_FUNC(int)
+_PyUnicodeWriter_PrepareKindInternal(_PyUnicodeWriter *writer,
+                                     int kind);
+
+/* Append a Unicode character.
+   Return 0 on success, raise an exception and return -1 on error. */
+PyAPI_FUNC(int)
+_PyUnicodeWriter_WriteChar(_PyUnicodeWriter *writer,
+    Py_UCS4 ch
+    );
+
+/* Append a Unicode string.
+   Return 0 on success, raise an exception and return -1 on error. */
+PyAPI_FUNC(int)
+_PyUnicodeWriter_WriteStr(_PyUnicodeWriter *writer,
+    PyObject *str               /* Unicode string */
+    );
+
+/* Append a substring of a Unicode string.
+   Return 0 on success, raise an exception and return -1 on error. */
+PyAPI_FUNC(int)
+_PyUnicodeWriter_WriteSubstring(_PyUnicodeWriter *writer,
+    PyObject *str,              /* Unicode string */
+    Py_ssize_t start,
+    Py_ssize_t end
+    );
+
+/* Append an ASCII-encoded byte string.
+   Return 0 on success, raise an exception and return -1 on error. */
+PyAPI_FUNC(int)
+_PyUnicodeWriter_WriteASCIIString(_PyUnicodeWriter *writer,
+    const char *str,           /* ASCII-encoded byte string */
+    Py_ssize_t len             /* number of bytes, or -1 if unknown */
+    );
+
+/* Append a latin1-encoded byte string.
+   Return 0 on success, raise an exception and return -1 on error. */
+PyAPI_FUNC(int)
+_PyUnicodeWriter_WriteLatin1String(_PyUnicodeWriter *writer,
+    const char *str,           /* latin1-encoded byte string */
+    Py_ssize_t len             /* length in bytes */
+    );
+
+/* Get the value of the writer as a Unicode string. Clear the
+   buffer of the writer. Raise an exception and return NULL
+   on error. */
+PyAPI_FUNC(PyObject *)
+_PyUnicodeWriter_Finish(_PyUnicodeWriter *writer);
+
+/* Deallocate memory of a writer (clear its internal buffer). */
+PyAPI_FUNC(void)
+_PyUnicodeWriter_Dealloc(_PyUnicodeWriter *writer);
+
+
+/* --- Manage the default encoding ---------------------------------------- */
+
+/* Returns a pointer to the default encoding (UTF-8) of the
+   Unicode object unicode.
+
+   Like PyUnicode_AsUTF8AndSize(), this also caches the UTF-8 representation
+   in the unicodeobject.
+
+   _PyUnicode_AsString is a #define for PyUnicode_AsUTF8 to
+   support the previous internal function with the same behaviour.
+
+   Use of this API is DEPRECATED since no size information can be
+   extracted from the returned data.
+*/
+
+PyAPI_FUNC(const char *) PyUnicode_AsUTF8(PyObject *unicode);
+
+// Alias kept for backward compatibility
+#define _PyUnicode_AsString PyUnicode_AsUTF8
+
+
+/* === Characters Type APIs =============================================== */
+
+/* These should not be used directly. Use the Py_UNICODE_IS* and
+   Py_UNICODE_TO* macros instead.
+
+   These APIs are implemented in Objects/unicodectype.c.
+
+*/
+
+PyAPI_FUNC(int) _PyUnicode_IsLowercase(
+    Py_UCS4 ch       /* Unicode character */
+    );
+
+PyAPI_FUNC(int) _PyUnicode_IsUppercase(
+    Py_UCS4 ch       /* Unicode character */
+    );
+
+PyAPI_FUNC(int) _PyUnicode_IsTitlecase(
+    Py_UCS4 ch       /* Unicode character */
+    );
+
+PyAPI_FUNC(int) _PyUnicode_IsWhitespace(
+    const Py_UCS4 ch         /* Unicode character */
+    );
+
+PyAPI_FUNC(int) _PyUnicode_IsLinebreak(
+    const Py_UCS4 ch         /* Unicode character */
+    );
+
+PyAPI_FUNC(Py_UCS4) _PyUnicode_ToLowercase(
+    Py_UCS4 ch       /* Unicode character */
+    );
+
+PyAPI_FUNC(Py_UCS4) _PyUnicode_ToUppercase(
+    Py_UCS4 ch       /* Unicode character */
+    );
+
+PyAPI_FUNC(Py_UCS4) _PyUnicode_ToTitlecase(
+    Py_UCS4 ch       /* Unicode character */
+    );
+
+PyAPI_FUNC(int) _PyUnicode_ToDecimalDigit(
+    Py_UCS4 ch       /* Unicode character */
+    );
+
+PyAPI_FUNC(int) _PyUnicode_ToDigit(
+    Py_UCS4 ch       /* Unicode character */
+    );
+
+PyAPI_FUNC(double) _PyUnicode_ToNumeric(
+    Py_UCS4 ch       /* Unicode character */
+    );
+
+PyAPI_FUNC(int) _PyUnicode_IsDecimalDigit(
+    Py_UCS4 ch       /* Unicode character */
+    );
+
+PyAPI_FUNC(int) _PyUnicode_IsDigit(
+    Py_UCS4 ch       /* Unicode character */
+    );
+
+PyAPI_FUNC(int) _PyUnicode_IsNumeric(
+    Py_UCS4 ch       /* Unicode character */
+    );
+
+PyAPI_FUNC(int) _PyUnicode_IsPrintable(
+    Py_UCS4 ch       /* Unicode character */
+    );
+
+PyAPI_FUNC(int) _PyUnicode_IsAlpha(
+    Py_UCS4 ch       /* Unicode character */
+    );
+
+// Helper array used by Py_UNICODE_ISSPACE().
+PyAPI_DATA(const unsigned char) _Py_ascii_whitespace[];
+
+// Since splitting on whitespace is an important use case, and
+// whitespace in most situations is solely ASCII whitespace, we
+// optimize for the common case by using a quick look-up table
+// _Py_ascii_whitespace (see below) with an inlined check.
+static inline int Py_UNICODE_ISSPACE(Py_UCS4 ch) {
+    if (ch < 128) {
+        return _Py_ascii_whitespace[ch];
+    }
+    return _PyUnicode_IsWhitespace(ch);
+}
+
+#define Py_UNICODE_ISLOWER(ch) _PyUnicode_IsLowercase(ch)
+#define Py_UNICODE_ISUPPER(ch) _PyUnicode_IsUppercase(ch)
+#define Py_UNICODE_ISTITLE(ch) _PyUnicode_IsTitlecase(ch)
+#define Py_UNICODE_ISLINEBREAK(ch) _PyUnicode_IsLinebreak(ch)
+
+#define Py_UNICODE_TOLOWER(ch) _PyUnicode_ToLowercase(ch)
+#define Py_UNICODE_TOUPPER(ch) _PyUnicode_ToUppercase(ch)
+#define Py_UNICODE_TOTITLE(ch) _PyUnicode_ToTitlecase(ch)
+
+#define Py_UNICODE_ISDECIMAL(ch) _PyUnicode_IsDecimalDigit(ch)
+#define Py_UNICODE_ISDIGIT(ch) _PyUnicode_IsDigit(ch)
+#define Py_UNICODE_ISNUMERIC(ch) _PyUnicode_IsNumeric(ch)
+#define Py_UNICODE_ISPRINTABLE(ch) _PyUnicode_IsPrintable(ch)
+
+#define Py_UNICODE_TODECIMAL(ch) _PyUnicode_ToDecimalDigit(ch)
+#define Py_UNICODE_TODIGIT(ch) _PyUnicode_ToDigit(ch)
+#define Py_UNICODE_TONUMERIC(ch) _PyUnicode_ToNumeric(ch)
+
+#define Py_UNICODE_ISALPHA(ch) _PyUnicode_IsAlpha(ch)
+
+static inline int Py_UNICODE_ISALNUM(Py_UCS4 ch) {
+   return (Py_UNICODE_ISALPHA(ch)
+           || Py_UNICODE_ISDECIMAL(ch)
+           || Py_UNICODE_ISDIGIT(ch)
+           || Py_UNICODE_ISNUMERIC(ch));
+}
+
+
+/* === Misc functions ===================================================== */
+
+// Return an interned Unicode object for an Identifier; may fail if there is no
+// memory.
+PyAPI_FUNC(PyObject*) _PyUnicode_FromId(_Py_Identifier*);

Dependencies/Python/include/cpython/warnings.h

@@ -0,0 +1,20 @@
+#ifndef Py_CPYTHON_WARNINGS_H
+#  error "this header file must not be included directly"
+#endif
+
+PyAPI_FUNC(int) PyErr_WarnExplicitObject(
+    PyObject *category,
+    PyObject *message,
+    PyObject *filename,
+    int lineno,
+    PyObject *module,
+    PyObject *registry);
+
+PyAPI_FUNC(int) PyErr_WarnExplicitFormat(
+    PyObject *category,
+    const char *filename, int lineno,
+    const char *module, PyObject *registry,
+    const char *format, ...);
+
+// DEPRECATED: Use PyErr_WarnEx() instead.
+#define PyErr_Warn(category, msg) PyErr_WarnEx((category), (msg), 1)

Dependencies/Python/include/cpython/weakrefobject.h

@@ -0,0 +1,63 @@
+#ifndef Py_CPYTHON_WEAKREFOBJECT_H
+#  error "this header file must not be included directly"
+#endif
+
+/* PyWeakReference is the base struct for the Python ReferenceType, ProxyType,
+ * and CallableProxyType.
+ */
+struct _PyWeakReference {
+    PyObject_HEAD
+
+    /* The object to which this is a weak reference, or Py_None if none.
+     * Note that this is a stealth reference:  wr_object's refcount is
+     * not incremented to reflect this pointer.
+     */
+    PyObject *wr_object;
+
+    /* A callable to invoke when wr_object dies, or NULL if none. */
+    PyObject *wr_callback;
+
+    /* A cache for wr_object's hash code.  As usual for hashes, this is -1
+     * if the hash code isn't known yet.
+     */
+    Py_hash_t hash;
+
+    /* If wr_object is weakly referenced, wr_object has a doubly-linked NULL-
+     * terminated list of weak references to it.  These are the list pointers.
+     * If wr_object goes away, wr_object is set to Py_None, and these pointers
+     * have no meaning then.
+     */
+    PyWeakReference *wr_prev;
+    PyWeakReference *wr_next;
+    vectorcallfunc vectorcall;
+
+#ifdef Py_GIL_DISABLED
+    /* Pointer to the lock used when clearing in free-threaded builds.
+     * Normally this can be derived from wr_object, but in some cases we need
+     * to lock after wr_object has been set to Py_None.
+     */
+    PyMutex *weakrefs_lock;
+#endif
+};
+
+PyAPI_FUNC(void) _PyWeakref_ClearRef(PyWeakReference *self);
+
+Py_DEPRECATED(3.13) static inline PyObject* PyWeakref_GET_OBJECT(PyObject *ref_obj)
+{
+    PyWeakReference *ref;
+    PyObject *obj;
+    assert(PyWeakref_Check(ref_obj));
+    ref = _Py_CAST(PyWeakReference*, ref_obj);
+    obj = ref->wr_object;
+    // Explanation for the Py_REFCNT() check: when a weakref's target is part
+    // of a long chain of deallocations which triggers the trashcan mechanism,
+    // clearing the weakrefs can be delayed long after the target's refcount
+    // has dropped to zero.  In the meantime, code accessing the weakref will
+    // be able to "see" the target object even though it is supposed to be
+    // unreachable.  See issue gh-60806.
+    if (Py_REFCNT(obj) > 0) {
+        return obj;
+    }
+    return Py_None;
+}
+#define PyWeakref_GET_OBJECT(ref) PyWeakref_GET_OBJECT(_PyObject_CAST(ref))

Dependencies/Python/include/critical_section.h

@@ -0,0 +1,16 @@
+#ifndef Py_CRITICAL_SECTION_H
+#define Py_CRITICAL_SECTION_H
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+#ifndef Py_LIMITED_API
+#  define Py_CPYTHON_CRITICAL_SECTION_H
+#  include "cpython/critical_section.h"
+#  undef Py_CPYTHON_CRITICAL_SECTION_H
+#endif
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* !Py_CRITICAL_SECTION_H */

Dependencies/Python/include/datetime.h

@@ -0,0 +1,267 @@
+/*  datetime.h
+ */
+#ifndef Py_LIMITED_API
+#ifndef DATETIME_H
+#define DATETIME_H
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/* Fields are packed into successive bytes, each viewed as unsigned and
+ * big-endian, unless otherwise noted:
+ *
+ * byte offset
+ *  0           year     2 bytes, 1-9999
+ *  2           month    1 byte, 1-12
+ *  3           day      1 byte, 1-31
+ *  4           hour     1 byte, 0-23
+ *  5           minute   1 byte, 0-59
+ *  6           second   1 byte, 0-59
+ *  7           usecond  3 bytes, 0-999999
+ * 10
+ */
+
+/* # of bytes for year, month, and day. */
+#define _PyDateTime_DATE_DATASIZE 4
+
+/* # of bytes for hour, minute, second, and usecond. */
+#define _PyDateTime_TIME_DATASIZE 6
+
+/* # of bytes for year, month, day, hour, minute, second, and usecond. */
+#define _PyDateTime_DATETIME_DATASIZE 10
+
+
+typedef struct
+{
+    PyObject_HEAD
+    Py_hash_t hashcode;         /* -1 when unknown */
+    int days;                   /* -MAX_DELTA_DAYS <= days <= MAX_DELTA_DAYS */
+    int seconds;                /* 0 <= seconds < 24*3600 is invariant */
+    int microseconds;           /* 0 <= microseconds < 1000000 is invariant */
+} PyDateTime_Delta;
+
+typedef struct
+{
+    PyObject_HEAD               /* a pure abstract base class */
+} PyDateTime_TZInfo;
+
+
+/* The datetime and time types have hashcodes, and an optional tzinfo member,
+ * present if and only if hastzinfo is true.
+ */
+#define _PyTZINFO_HEAD          \
+    PyObject_HEAD               \
+    Py_hash_t hashcode;         \
+    char hastzinfo;             /* boolean flag */
+
+/* No _PyDateTime_BaseTZInfo is allocated; it's just to have something
+ * convenient to cast to, when getting at the hastzinfo member of objects
+ * starting with _PyTZINFO_HEAD.
+ */
+typedef struct
+{
+    _PyTZINFO_HEAD
+} _PyDateTime_BaseTZInfo;
+
+/* All time objects are of PyDateTime_TimeType, but that can be allocated
+ * in two ways, with or without a tzinfo member.  Without is the same as
+ * tzinfo == None, but consumes less memory.  _PyDateTime_BaseTime is an
+ * internal struct used to allocate the right amount of space for the
+ * "without" case.
+ */
+#define _PyDateTime_TIMEHEAD    \
+    _PyTZINFO_HEAD              \
+    unsigned char data[_PyDateTime_TIME_DATASIZE];
+
+typedef struct
+{
+    _PyDateTime_TIMEHEAD
+} _PyDateTime_BaseTime;         /* hastzinfo false */
+
+typedef struct
+{
+    _PyDateTime_TIMEHEAD
+    unsigned char fold;
+    PyObject *tzinfo;
+} PyDateTime_Time;              /* hastzinfo true */
+
+
+/* All datetime objects are of PyDateTime_DateTimeType, but that can be
+ * allocated in two ways too, just like for time objects above.  In addition,
+ * the plain date type is a base class for datetime, so it must also have
+ * a hastzinfo member (although it's unused there).
+ */
+typedef struct
+{
+    _PyTZINFO_HEAD
+    unsigned char data[_PyDateTime_DATE_DATASIZE];
+} PyDateTime_Date;
+
+#define _PyDateTime_DATETIMEHEAD        \
+    _PyTZINFO_HEAD                      \
+    unsigned char data[_PyDateTime_DATETIME_DATASIZE];
+
+typedef struct
+{
+    _PyDateTime_DATETIMEHEAD
+} _PyDateTime_BaseDateTime;     /* hastzinfo false */
+
+typedef struct
+{
+    _PyDateTime_DATETIMEHEAD
+    unsigned char fold;
+    PyObject *tzinfo;
+} PyDateTime_DateTime;          /* hastzinfo true */
+
+
+/* Apply for date and datetime instances. */
+
+// o is a pointer to a time or a datetime object.
+#define _PyDateTime_HAS_TZINFO(o)  (((_PyDateTime_BaseTZInfo *)(o))->hastzinfo)
+
+#define PyDateTime_GET_YEAR(o)     ((((PyDateTime_Date*)(o))->data[0] << 8) | \
+                     ((PyDateTime_Date*)(o))->data[1])
+#define PyDateTime_GET_MONTH(o)    (((PyDateTime_Date*)(o))->data[2])
+#define PyDateTime_GET_DAY(o)      (((PyDateTime_Date*)(o))->data[3])
+
+#define PyDateTime_DATE_GET_HOUR(o)        (((PyDateTime_DateTime*)(o))->data[4])
+#define PyDateTime_DATE_GET_MINUTE(o)      (((PyDateTime_DateTime*)(o))->data[5])
+#define PyDateTime_DATE_GET_SECOND(o)      (((PyDateTime_DateTime*)(o))->data[6])
+#define PyDateTime_DATE_GET_MICROSECOND(o)              \
+    ((((PyDateTime_DateTime*)(o))->data[7] << 16) |       \
+     (((PyDateTime_DateTime*)(o))->data[8] << 8)  |       \
+      ((PyDateTime_DateTime*)(o))->data[9])
+#define PyDateTime_DATE_GET_FOLD(o)        (((PyDateTime_DateTime*)(o))->fold)
+#define PyDateTime_DATE_GET_TZINFO(o)      (_PyDateTime_HAS_TZINFO((o)) ? \
+    ((PyDateTime_DateTime *)(o))->tzinfo : Py_None)
+
+/* Apply for time instances. */
+#define PyDateTime_TIME_GET_HOUR(o)        (((PyDateTime_Time*)(o))->data[0])
+#define PyDateTime_TIME_GET_MINUTE(o)      (((PyDateTime_Time*)(o))->data[1])
+#define PyDateTime_TIME_GET_SECOND(o)      (((PyDateTime_Time*)(o))->data[2])
+#define PyDateTime_TIME_GET_MICROSECOND(o)              \
+    ((((PyDateTime_Time*)(o))->data[3] << 16) |           \
+     (((PyDateTime_Time*)(o))->data[4] << 8)  |           \
+      ((PyDateTime_Time*)(o))->data[5])
+#define PyDateTime_TIME_GET_FOLD(o)        (((PyDateTime_Time*)(o))->fold)
+#define PyDateTime_TIME_GET_TZINFO(o)      (_PyDateTime_HAS_TZINFO(o) ? \
+    ((PyDateTime_Time *)(o))->tzinfo : Py_None)
+
+/* Apply for time delta instances */
+#define PyDateTime_DELTA_GET_DAYS(o)         (((PyDateTime_Delta*)(o))->days)
+#define PyDateTime_DELTA_GET_SECONDS(o)      (((PyDateTime_Delta*)(o))->seconds)
+#define PyDateTime_DELTA_GET_MICROSECONDS(o)            \
+    (((PyDateTime_Delta*)(o))->microseconds)
+
+
+/* Define structure for C API. */
+typedef struct {
+    /* type objects */
+    PyTypeObject *DateType;
+    PyTypeObject *DateTimeType;
+    PyTypeObject *TimeType;
+    PyTypeObject *DeltaType;
+    PyTypeObject *TZInfoType;
+
+    /* singletons */
+    PyObject *TimeZone_UTC;
+
+    /* constructors */
+    PyObject *(*Date_FromDate)(int, int, int, PyTypeObject*);
+    PyObject *(*DateTime_FromDateAndTime)(int, int, int, int, int, int, int,
+        PyObject*, PyTypeObject*);
+    PyObject *(*Time_FromTime)(int, int, int, int, PyObject*, PyTypeObject*);
+    PyObject *(*Delta_FromDelta)(int, int, int, int, PyTypeObject*);
+    PyObject *(*TimeZone_FromTimeZone)(PyObject *offset, PyObject *name);
+
+    /* constructors for the DB API */
+    PyObject *(*DateTime_FromTimestamp)(PyObject*, PyObject*, PyObject*);
+    PyObject *(*Date_FromTimestamp)(PyObject*, PyObject*);
+
+    /* PEP 495 constructors */
+    PyObject *(*DateTime_FromDateAndTimeAndFold)(int, int, int, int, int, int, int,
+        PyObject*, int, PyTypeObject*);
+    PyObject *(*Time_FromTimeAndFold)(int, int, int, int, PyObject*, int, PyTypeObject*);
+
+} PyDateTime_CAPI;
+
+#define PyDateTime_CAPSULE_NAME "datetime.datetime_CAPI"
+
+
+/* This block is only used as part of the public API and should not be
+ * included in _datetimemodule.c, which does not use the C API capsule.
+ * See bpo-35081 for more details.
+ * */
+#ifndef _PY_DATETIME_IMPL
+/* Define global variable for the C API and a macro for setting it. */
+static PyDateTime_CAPI *PyDateTimeAPI = NULL;
+
+#define PyDateTime_IMPORT \
+    PyDateTimeAPI = (PyDateTime_CAPI *)PyCapsule_Import(PyDateTime_CAPSULE_NAME, 0)
+
+/* Macro for access to the UTC singleton */
+#define PyDateTime_TimeZone_UTC PyDateTimeAPI->TimeZone_UTC
+
+/* Macros for type checking when not building the Python core. */
+#define PyDate_Check(op) PyObject_TypeCheck((op), PyDateTimeAPI->DateType)
+#define PyDate_CheckExact(op) Py_IS_TYPE((op), PyDateTimeAPI->DateType)
+
+#define PyDateTime_Check(op) PyObject_TypeCheck((op), PyDateTimeAPI->DateTimeType)
+#define PyDateTime_CheckExact(op) Py_IS_TYPE((op), PyDateTimeAPI->DateTimeType)
+
+#define PyTime_Check(op) PyObject_TypeCheck((op), PyDateTimeAPI->TimeType)
+#define PyTime_CheckExact(op) Py_IS_TYPE((op), PyDateTimeAPI->TimeType)
+
+#define PyDelta_Check(op) PyObject_TypeCheck((op), PyDateTimeAPI->DeltaType)
+#define PyDelta_CheckExact(op) Py_IS_TYPE((op), PyDateTimeAPI->DeltaType)
+
+#define PyTZInfo_Check(op) PyObject_TypeCheck((op), PyDateTimeAPI->TZInfoType)
+#define PyTZInfo_CheckExact(op) Py_IS_TYPE((op), PyDateTimeAPI->TZInfoType)
+
+
+/* Macros for accessing constructors in a simplified fashion. */
+#define PyDate_FromDate(year, month, day) \
+    PyDateTimeAPI->Date_FromDate((year), (month), (day), PyDateTimeAPI->DateType)
+
+#define PyDateTime_FromDateAndTime(year, month, day, hour, min, sec, usec) \
+    PyDateTimeAPI->DateTime_FromDateAndTime((year), (month), (day), (hour), \
+        (min), (sec), (usec), Py_None, PyDateTimeAPI->DateTimeType)
+
+#define PyDateTime_FromDateAndTimeAndFold(year, month, day, hour, min, sec, usec, fold) \
+    PyDateTimeAPI->DateTime_FromDateAndTimeAndFold((year), (month), (day), (hour), \
+        (min), (sec), (usec), Py_None, (fold), PyDateTimeAPI->DateTimeType)
+
+#define PyTime_FromTime(hour, minute, second, usecond) \
+    PyDateTimeAPI->Time_FromTime((hour), (minute), (second), (usecond), \
+        Py_None, PyDateTimeAPI->TimeType)
+
+#define PyTime_FromTimeAndFold(hour, minute, second, usecond, fold) \
+    PyDateTimeAPI->Time_FromTimeAndFold((hour), (minute), (second), (usecond), \
+        Py_None, (fold), PyDateTimeAPI->TimeType)
+
+#define PyDelta_FromDSU(days, seconds, useconds) \
+    PyDateTimeAPI->Delta_FromDelta((days), (seconds), (useconds), 1, \
+        PyDateTimeAPI->DeltaType)
+
+#define PyTimeZone_FromOffset(offset) \
+    PyDateTimeAPI->TimeZone_FromTimeZone((offset), NULL)
+
+#define PyTimeZone_FromOffsetAndName(offset, name) \
+    PyDateTimeAPI->TimeZone_FromTimeZone((offset), (name))
+
+/* Macros supporting the DB API. */
+#define PyDateTime_FromTimestamp(args) \
+    PyDateTimeAPI->DateTime_FromTimestamp( \
+        (PyObject*) (PyDateTimeAPI->DateTimeType), (args), NULL)
+
+#define PyDate_FromTimestamp(args) \
+    PyDateTimeAPI->Date_FromTimestamp( \
+        (PyObject*) (PyDateTimeAPI->DateType), (args))
+
+#endif   /* !defined(_PY_DATETIME_IMPL) */
+
+#ifdef __cplusplus
+}
+#endif
+#endif
+#endif /* !Py_LIMITED_API */

Dependencies/Python/include/descrobject.h

@@ -0,0 +1,100 @@
+/* Descriptors */
+#ifndef Py_DESCROBJECT_H
+#define Py_DESCROBJECT_H
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+typedef PyObject *(*getter)(PyObject *, void *);
+typedef int (*setter)(PyObject *, PyObject *, void *);
+
+struct PyGetSetDef {
+    const char *name;
+    getter get;
+    setter set;
+    const char *doc;
+    void *closure;
+};
+
+PyAPI_DATA(PyTypeObject) PyClassMethodDescr_Type;
+PyAPI_DATA(PyTypeObject) PyGetSetDescr_Type;
+PyAPI_DATA(PyTypeObject) PyMemberDescr_Type;
+PyAPI_DATA(PyTypeObject) PyMethodDescr_Type;
+PyAPI_DATA(PyTypeObject) PyWrapperDescr_Type;
+PyAPI_DATA(PyTypeObject) PyDictProxy_Type;
+PyAPI_DATA(PyTypeObject) PyProperty_Type;
+
+PyAPI_FUNC(PyObject *) PyDescr_NewMethod(PyTypeObject *, PyMethodDef *);
+PyAPI_FUNC(PyObject *) PyDescr_NewClassMethod(PyTypeObject *, PyMethodDef *);
+PyAPI_FUNC(PyObject *) PyDescr_NewMember(PyTypeObject *, PyMemberDef *);
+PyAPI_FUNC(PyObject *) PyDescr_NewGetSet(PyTypeObject *, PyGetSetDef *);
+
+PyAPI_FUNC(PyObject *) PyDictProxy_New(PyObject *);
+PyAPI_FUNC(PyObject *) PyWrapper_New(PyObject *, PyObject *);
+
+
+/* An array of PyMemberDef structures defines the name, type and offset
+   of selected members of a C structure.  These can be read by
+   PyMember_GetOne() and set by PyMember_SetOne() (except if their READONLY
+   flag is set).  The array must be terminated with an entry whose name
+   pointer is NULL. */
+struct PyMemberDef {
+    const char *name;
+    int type;
+    Py_ssize_t offset;
+    int flags;
+    const char *doc;
+};
+
+// These constants used to be in structmember.h, not prefixed by Py_.
+// (structmember.h now has aliases to the new names.)
+
+/* Types */
+#define Py_T_SHORT     0
+#define Py_T_INT       1
+#define Py_T_LONG      2
+#define Py_T_FLOAT     3
+#define Py_T_DOUBLE    4
+#define Py_T_STRING    5
+#define _Py_T_OBJECT   6  // Deprecated, use Py_T_OBJECT_EX instead
+/* the ordering here is weird for binary compatibility */
+#define Py_T_CHAR      7   /* 1-character string */
+#define Py_T_BYTE      8   /* 8-bit signed int */
+/* unsigned variants: */
+#define Py_T_UBYTE     9
+#define Py_T_USHORT    10
+#define Py_T_UINT      11
+#define Py_T_ULONG     12
+
+/* Added by Jack: strings contained in the structure */
+#define Py_T_STRING_INPLACE    13
+
+/* Added by Lillo: bools contained in the structure (assumed char) */
+#define Py_T_BOOL      14
+
+#define Py_T_OBJECT_EX 16
+#define Py_T_LONGLONG  17
+#define Py_T_ULONGLONG 18
+
+#define Py_T_PYSSIZET  19      /* Py_ssize_t */
+#define _Py_T_NONE     20 // Deprecated. Value is always None.
+
+/* Flags */
+#define Py_READONLY            1
+#define Py_AUDIT_READ          2 // Added in 3.10, harmless no-op before that
+#define _Py_WRITE_RESTRICTED   4 // Deprecated, no-op. Do not reuse the value.
+#define Py_RELATIVE_OFFSET     8
+
+PyAPI_FUNC(PyObject *) PyMember_GetOne(const char *, PyMemberDef *);
+PyAPI_FUNC(int) PyMember_SetOne(char *, PyMemberDef *, PyObject *);
+
+#ifndef Py_LIMITED_API
+#  define Py_CPYTHON_DESCROBJECT_H
+#  include "cpython/descrobject.h"
+#  undef Py_CPYTHON_DESCROBJECT_H
+#endif
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* !Py_DESCROBJECT_H */

Dependencies/Python/include/dictobject.h

@@ -0,0 +1,108 @@
+#ifndef Py_DICTOBJECT_H
+#define Py_DICTOBJECT_H
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/* Dictionary object type -- mapping from hashable object to object */
+
+/* The distribution includes a separate file, Objects/dictnotes.txt,
+   describing explorations into dictionary design and optimization.
+   It covers typical dictionary use patterns, the parameters for
+   tuning dictionaries, and several ideas for possible optimizations.
+*/
+
+PyAPI_DATA(PyTypeObject) PyDict_Type;
+
+#define PyDict_Check(op) \
+                 PyType_FastSubclass(Py_TYPE(op), Py_TPFLAGS_DICT_SUBCLASS)
+#define PyDict_CheckExact(op) Py_IS_TYPE((op), &PyDict_Type)
+
+PyAPI_FUNC(PyObject *) PyDict_New(void);
+PyAPI_FUNC(PyObject *) PyDict_GetItem(PyObject *mp, PyObject *key);
+PyAPI_FUNC(PyObject *) PyDict_GetItemWithError(PyObject *mp, PyObject *key);
+PyAPI_FUNC(int) PyDict_SetItem(PyObject *mp, PyObject *key, PyObject *item);
+PyAPI_FUNC(int) PyDict_DelItem(PyObject *mp, PyObject *key);
+PyAPI_FUNC(void) PyDict_Clear(PyObject *mp);
+PyAPI_FUNC(int) PyDict_Next(
+    PyObject *mp, Py_ssize_t *pos, PyObject **key, PyObject **value);
+PyAPI_FUNC(PyObject *) PyDict_Keys(PyObject *mp);
+PyAPI_FUNC(PyObject *) PyDict_Values(PyObject *mp);
+PyAPI_FUNC(PyObject *) PyDict_Items(PyObject *mp);
+PyAPI_FUNC(Py_ssize_t) PyDict_Size(PyObject *mp);
+PyAPI_FUNC(PyObject *) PyDict_Copy(PyObject *mp);
+PyAPI_FUNC(int) PyDict_Contains(PyObject *mp, PyObject *key);
+
+/* PyDict_Update(mp, other) is equivalent to PyDict_Merge(mp, other, 1). */
+PyAPI_FUNC(int) PyDict_Update(PyObject *mp, PyObject *other);
+
+/* PyDict_Merge updates/merges from a mapping object (an object that
+   supports PyMapping_Keys() and PyObject_GetItem()).  If override is true,
+   the last occurrence of a key wins, else the first.  The Python
+   dict.update(other) is equivalent to PyDict_Merge(dict, other, 1).
+*/
+PyAPI_FUNC(int) PyDict_Merge(PyObject *mp,
+                             PyObject *other,
+                             int override);
+
+/* PyDict_MergeFromSeq2 updates/merges from an iterable object producing
+   iterable objects of length 2.  If override is true, the last occurrence
+   of a key wins, else the first.  The Python dict constructor dict(seq2)
+   is equivalent to dict={}; PyDict_MergeFromSeq(dict, seq2, 1).
+*/
+PyAPI_FUNC(int) PyDict_MergeFromSeq2(PyObject *d,
+                                     PyObject *seq2,
+                                     int override);
+
+PyAPI_FUNC(PyObject *) PyDict_GetItemString(PyObject *dp, const char *key);
+PyAPI_FUNC(int) PyDict_SetItemString(PyObject *dp, const char *key, PyObject *item);
+PyAPI_FUNC(int) PyDict_DelItemString(PyObject *dp, const char *key);
+
+#if !defined(Py_LIMITED_API) || Py_LIMITED_API+0 >= 0x030D0000
+// Return the object from dictionary *op* which has a key *key*.
+// - If the key is present, set *result to a new strong reference to the value
+//   and return 1.
+// - If the key is missing, set *result to NULL and return 0 .
+// - On error, raise an exception and return -1.
+PyAPI_FUNC(int) PyDict_GetItemRef(PyObject *mp, PyObject *key, PyObject **result);
+PyAPI_FUNC(int) PyDict_GetItemStringRef(PyObject *mp, const char *key, PyObject **result);
+#endif
+
+#if !defined(Py_LIMITED_API) || Py_LIMITED_API+0 >= 0x030A0000
+PyAPI_FUNC(PyObject *) PyObject_GenericGetDict(PyObject *, void *);
+#endif
+
+/* Dictionary (keys, values, items) views */
+
+PyAPI_DATA(PyTypeObject) PyDictKeys_Type;
+PyAPI_DATA(PyTypeObject) PyDictValues_Type;
+PyAPI_DATA(PyTypeObject) PyDictItems_Type;
+
+#define PyDictKeys_Check(op) PyObject_TypeCheck((op), &PyDictKeys_Type)
+#define PyDictValues_Check(op) PyObject_TypeCheck((op), &PyDictValues_Type)
+#define PyDictItems_Check(op) PyObject_TypeCheck((op), &PyDictItems_Type)
+/* This excludes Values, since they are not sets. */
+# define PyDictViewSet_Check(op) \
+    (PyDictKeys_Check(op) || PyDictItems_Check(op))
+
+/* Dictionary (key, value, items) iterators */
+
+PyAPI_DATA(PyTypeObject) PyDictIterKey_Type;
+PyAPI_DATA(PyTypeObject) PyDictIterValue_Type;
+PyAPI_DATA(PyTypeObject) PyDictIterItem_Type;
+
+PyAPI_DATA(PyTypeObject) PyDictRevIterKey_Type;
+PyAPI_DATA(PyTypeObject) PyDictRevIterItem_Type;
+PyAPI_DATA(PyTypeObject) PyDictRevIterValue_Type;
+
+
+#ifndef Py_LIMITED_API
+#  define Py_CPYTHON_DICTOBJECT_H
+#  include "cpython/dictobject.h"
+#  undef Py_CPYTHON_DICTOBJECT_H
+#endif
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* !Py_DICTOBJECT_H */

Dependencies/Python/include/dynamic_annotations.h

@@ -0,0 +1,499 @@
+/* Copyright (c) 2008-2009, Google Inc.
+ * All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions are
+ * met:
+ *
+ *     * Redistributions of source code must retain the above copyright
+ * notice, this list of conditions and the following disclaimer.
+ *     * Neither the name of Google Inc. nor the names of its
+ * contributors may be used to endorse or promote products derived from
+ * this software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+ * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ * ---
+ * Author: Kostya Serebryany
+ * Copied to CPython by Jeffrey Yasskin, with all macros renamed to
+ * start with _Py_ to avoid colliding with users embedding Python, and
+ * with deprecated macros removed.
+ */
+
+/* This file defines dynamic annotations for use with dynamic analysis
+   tool such as valgrind, PIN, etc.
+
+   Dynamic annotation is a source code annotation that affects
+   the generated code (that is, the annotation is not a comment).
+   Each such annotation is attached to a particular
+   instruction and/or to a particular object (address) in the program.
+
+   The annotations that should be used by users are macros in all upper-case
+   (e.g., _Py_ANNOTATE_NEW_MEMORY).
+
+   Actual implementation of these macros may differ depending on the
+   dynamic analysis tool being used.
+
+   See https://code.google.com/p/data-race-test/  for more information.
+
+   This file supports the following dynamic analysis tools:
+   - None (DYNAMIC_ANNOTATIONS_ENABLED is not defined or zero).
+      Macros are defined empty.
+   - ThreadSanitizer, Helgrind, DRD (DYNAMIC_ANNOTATIONS_ENABLED is 1).
+      Macros are defined as calls to non-inlinable empty functions
+      that are intercepted by Valgrind. */
+
+#ifndef __DYNAMIC_ANNOTATIONS_H__
+#define __DYNAMIC_ANNOTATIONS_H__
+
+#ifndef DYNAMIC_ANNOTATIONS_ENABLED
+# define DYNAMIC_ANNOTATIONS_ENABLED 0
+#endif
+
+#if DYNAMIC_ANNOTATIONS_ENABLED != 0
+
+  /* -------------------------------------------------------------
+     Annotations useful when implementing condition variables such as CondVar,
+     using conditional critical sections (Await/LockWhen) and when constructing
+     user-defined synchronization mechanisms.
+
+     The annotations _Py_ANNOTATE_HAPPENS_BEFORE() and
+     _Py_ANNOTATE_HAPPENS_AFTER() can be used to define happens-before arcs in
+     user-defined synchronization mechanisms: the race detector will infer an
+     arc from the former to the latter when they share the same argument
+     pointer.
+
+     Example 1 (reference counting):
+
+     void Unref() {
+       _Py_ANNOTATE_HAPPENS_BEFORE(&refcount_);
+       if (AtomicDecrementByOne(&refcount_) == 0) {
+         _Py_ANNOTATE_HAPPENS_AFTER(&refcount_);
+         delete this;
+       }
+     }
+
+     Example 2 (message queue):
+
+     void MyQueue::Put(Type *e) {
+       MutexLock lock(&mu_);
+       _Py_ANNOTATE_HAPPENS_BEFORE(e);
+       PutElementIntoMyQueue(e);
+     }
+
+     Type *MyQueue::Get() {
+       MutexLock lock(&mu_);
+       Type *e = GetElementFromMyQueue();
+       _Py_ANNOTATE_HAPPENS_AFTER(e);
+       return e;
+     }
+
+     Note: when possible, please use the existing reference counting and message
+     queue implementations instead of inventing new ones. */
+
+  /* Report that wait on the condition variable at address "cv" has succeeded
+     and the lock at address "lock" is held. */
+#define _Py_ANNOTATE_CONDVAR_LOCK_WAIT(cv, lock) \
+    AnnotateCondVarWait(__FILE__, __LINE__, cv, lock)
+
+  /* Report that wait on the condition variable at "cv" has succeeded.  Variant
+     w/o lock. */
+#define _Py_ANNOTATE_CONDVAR_WAIT(cv) \
+    AnnotateCondVarWait(__FILE__, __LINE__, cv, NULL)
+
+  /* Report that we are about to signal on the condition variable at address
+     "cv". */
+#define _Py_ANNOTATE_CONDVAR_SIGNAL(cv) \
+    AnnotateCondVarSignal(__FILE__, __LINE__, cv)
+
+  /* Report that we are about to signal_all on the condition variable at "cv". */
+#define _Py_ANNOTATE_CONDVAR_SIGNAL_ALL(cv) \
+    AnnotateCondVarSignalAll(__FILE__, __LINE__, cv)
+
+  /* Annotations for user-defined synchronization mechanisms. */
+#define _Py_ANNOTATE_HAPPENS_BEFORE(obj) _Py_ANNOTATE_CONDVAR_SIGNAL(obj)
+#define _Py_ANNOTATE_HAPPENS_AFTER(obj)  _Py_ANNOTATE_CONDVAR_WAIT(obj)
+
+  /* Report that the bytes in the range [pointer, pointer+size) are about
+     to be published safely. The race checker will create a happens-before
+     arc from the call _Py_ANNOTATE_PUBLISH_MEMORY_RANGE(pointer, size) to
+     subsequent accesses to this memory.
+     Note: this annotation may not work properly if the race detector uses
+     sampling, i.e. does not observe all memory accesses.
+     */
+#define _Py_ANNOTATE_PUBLISH_MEMORY_RANGE(pointer, size) \
+    AnnotatePublishMemoryRange(__FILE__, __LINE__, pointer, size)
+
+  /* Instruct the tool to create a happens-before arc between mu->Unlock() and
+     mu->Lock(). This annotation may slow down the race detector and hide real
+     races. Normally it is used only when it would be difficult to annotate each
+     of the mutex's critical sections individually using the annotations above.
+     This annotation makes sense only for hybrid race detectors. For pure
+     happens-before detectors this is a no-op. For more details see
+     https://code.google.com/p/data-race-test/wiki/PureHappensBeforeVsHybrid . */
+#define _Py_ANNOTATE_PURE_HAPPENS_BEFORE_MUTEX(mu) \
+    AnnotateMutexIsUsedAsCondVar(__FILE__, __LINE__, mu)
+
+  /* -------------------------------------------------------------
+     Annotations useful when defining memory allocators, or when memory that
+     was protected in one way starts to be protected in another. */
+
+  /* Report that a new memory at "address" of size "size" has been allocated.
+     This might be used when the memory has been retrieved from a free list and
+     is about to be reused, or when the locking discipline for a variable
+     changes. */
+#define _Py_ANNOTATE_NEW_MEMORY(address, size) \
+    AnnotateNewMemory(__FILE__, __LINE__, address, size)
+
+  /* -------------------------------------------------------------
+     Annotations useful when defining FIFO queues that transfer data between
+     threads. */
+
+  /* Report that the producer-consumer queue (such as ProducerConsumerQueue) at
+     address "pcq" has been created.  The _Py_ANNOTATE_PCQ_* annotations should
+     be used only for FIFO queues.  For non-FIFO queues use
+     _Py_ANNOTATE_HAPPENS_BEFORE (for put) and _Py_ANNOTATE_HAPPENS_AFTER (for
+     get). */
+#define _Py_ANNOTATE_PCQ_CREATE(pcq) \
+    AnnotatePCQCreate(__FILE__, __LINE__, pcq)
+
+  /* Report that the queue at address "pcq" is about to be destroyed. */
+#define _Py_ANNOTATE_PCQ_DESTROY(pcq) \
+    AnnotatePCQDestroy(__FILE__, __LINE__, pcq)
+
+  /* Report that we are about to put an element into a FIFO queue at address
+     "pcq". */
+#define _Py_ANNOTATE_PCQ_PUT(pcq) \
+    AnnotatePCQPut(__FILE__, __LINE__, pcq)
+
+  /* Report that we've just got an element from a FIFO queue at address "pcq". */
+#define _Py_ANNOTATE_PCQ_GET(pcq) \
+    AnnotatePCQGet(__FILE__, __LINE__, pcq)
+
+  /* -------------------------------------------------------------
+     Annotations that suppress errors.  It is usually better to express the
+     program's synchronization using the other annotations, but these can
+     be used when all else fails. */
+
+  /* Report that we may have a benign race at "pointer", with size
+     "sizeof(*(pointer))". "pointer" must be a non-void* pointer.  Insert at the
+     point where "pointer" has been allocated, preferably close to the point
+     where the race happens.  See also _Py_ANNOTATE_BENIGN_RACE_STATIC. */
+#define _Py_ANNOTATE_BENIGN_RACE(pointer, description) \
+    AnnotateBenignRaceSized(__FILE__, __LINE__, pointer, \
+                            sizeof(*(pointer)), description)
+
+  /* Same as _Py_ANNOTATE_BENIGN_RACE(address, description), but applies to
+     the memory range [address, address+size). */
+#define _Py_ANNOTATE_BENIGN_RACE_SIZED(address, size, description) \
+    AnnotateBenignRaceSized(__FILE__, __LINE__, address, size, description)
+
+  /* Request the analysis tool to ignore all reads in the current thread
+     until _Py_ANNOTATE_IGNORE_READS_END is called.
+     Useful to ignore intentional racey reads, while still checking
+     other reads and all writes.
+     See also _Py_ANNOTATE_UNPROTECTED_READ. */
+#define _Py_ANNOTATE_IGNORE_READS_BEGIN() \
+    AnnotateIgnoreReadsBegin(__FILE__, __LINE__)
+
+  /* Stop ignoring reads. */
+#define _Py_ANNOTATE_IGNORE_READS_END() \
+    AnnotateIgnoreReadsEnd(__FILE__, __LINE__)
+
+  /* Similar to _Py_ANNOTATE_IGNORE_READS_BEGIN, but ignore writes. */
+#define _Py_ANNOTATE_IGNORE_WRITES_BEGIN() \
+    AnnotateIgnoreWritesBegin(__FILE__, __LINE__)
+
+  /* Stop ignoring writes. */
+#define _Py_ANNOTATE_IGNORE_WRITES_END() \
+    AnnotateIgnoreWritesEnd(__FILE__, __LINE__)
+
+  /* Start ignoring all memory accesses (reads and writes). */
+#define _Py_ANNOTATE_IGNORE_READS_AND_WRITES_BEGIN() \
+    do {\
+      _Py_ANNOTATE_IGNORE_READS_BEGIN();\
+      _Py_ANNOTATE_IGNORE_WRITES_BEGIN();\
+    }while(0)\
+
+  /* Stop ignoring all memory accesses. */
+#define _Py_ANNOTATE_IGNORE_READS_AND_WRITES_END() \
+    do {\
+      _Py_ANNOTATE_IGNORE_WRITES_END();\
+      _Py_ANNOTATE_IGNORE_READS_END();\
+    }while(0)\
+
+  /* Similar to _Py_ANNOTATE_IGNORE_READS_BEGIN, but ignore synchronization events:
+     RWLOCK* and CONDVAR*. */
+#define _Py_ANNOTATE_IGNORE_SYNC_BEGIN() \
+    AnnotateIgnoreSyncBegin(__FILE__, __LINE__)
+
+  /* Stop ignoring sync events. */
+#define _Py_ANNOTATE_IGNORE_SYNC_END() \
+    AnnotateIgnoreSyncEnd(__FILE__, __LINE__)
+
+
+  /* Enable (enable!=0) or disable (enable==0) race detection for all threads.
+     This annotation could be useful if you want to skip expensive race analysis
+     during some period of program execution, e.g. during initialization. */
+#define _Py_ANNOTATE_ENABLE_RACE_DETECTION(enable) \
+    AnnotateEnableRaceDetection(__FILE__, __LINE__, enable)
+
+  /* -------------------------------------------------------------
+     Annotations useful for debugging. */
+
+  /* Request to trace every access to "address". */
+#define _Py_ANNOTATE_TRACE_MEMORY(address) \
+    AnnotateTraceMemory(__FILE__, __LINE__, address)
+
+  /* Report the current thread name to a race detector. */
+#define _Py_ANNOTATE_THREAD_NAME(name) \
+    AnnotateThreadName(__FILE__, __LINE__, name)
+
+  /* -------------------------------------------------------------
+     Annotations useful when implementing locks.  They are not
+     normally needed by modules that merely use locks.
+     The "lock" argument is a pointer to the lock object. */
+
+  /* Report that a lock has been created at address "lock". */
+#define _Py_ANNOTATE_RWLOCK_CREATE(lock) \
+    AnnotateRWLockCreate(__FILE__, __LINE__, lock)
+
+  /* Report that the lock at address "lock" is about to be destroyed. */
+#define _Py_ANNOTATE_RWLOCK_DESTROY(lock) \
+    AnnotateRWLockDestroy(__FILE__, __LINE__, lock)
+
+  /* Report that the lock at address "lock" has been acquired.
+     is_w=1 for writer lock, is_w=0 for reader lock. */
+#define _Py_ANNOTATE_RWLOCK_ACQUIRED(lock, is_w) \
+    AnnotateRWLockAcquired(__FILE__, __LINE__, lock, is_w)
+
+  /* Report that the lock at address "lock" is about to be released. */
+#define _Py_ANNOTATE_RWLOCK_RELEASED(lock, is_w) \
+    AnnotateRWLockReleased(__FILE__, __LINE__, lock, is_w)
+
+  /* -------------------------------------------------------------
+     Annotations useful when implementing barriers.  They are not
+     normally needed by modules that merely use barriers.
+     The "barrier" argument is a pointer to the barrier object. */
+
+  /* Report that the "barrier" has been initialized with initial "count".
+   If 'reinitialization_allowed' is true, initialization is allowed to happen
+   multiple times w/o calling barrier_destroy() */
+#define _Py_ANNOTATE_BARRIER_INIT(barrier, count, reinitialization_allowed) \
+    AnnotateBarrierInit(__FILE__, __LINE__, barrier, count, \
+                        reinitialization_allowed)
+
+  /* Report that we are about to enter barrier_wait("barrier"). */
+#define _Py_ANNOTATE_BARRIER_WAIT_BEFORE(barrier) \
+    AnnotateBarrierWaitBefore(__FILE__, __LINE__, barrier)
+
+  /* Report that we just exited barrier_wait("barrier"). */
+#define _Py_ANNOTATE_BARRIER_WAIT_AFTER(barrier) \
+    AnnotateBarrierWaitAfter(__FILE__, __LINE__, barrier)
+
+  /* Report that the "barrier" has been destroyed. */
+#define _Py_ANNOTATE_BARRIER_DESTROY(barrier) \
+    AnnotateBarrierDestroy(__FILE__, __LINE__, barrier)
+
+  /* -------------------------------------------------------------
+     Annotations useful for testing race detectors. */
+
+  /* Report that we expect a race on the variable at "address".
+     Use only in unit tests for a race detector. */
+#define _Py_ANNOTATE_EXPECT_RACE(address, description) \
+    AnnotateExpectRace(__FILE__, __LINE__, address, description)
+
+  /* A no-op. Insert where you like to test the interceptors. */
+#define _Py_ANNOTATE_NO_OP(arg) \
+    AnnotateNoOp(__FILE__, __LINE__, arg)
+
+  /* Force the race detector to flush its state. The actual effect depends on
+   * the implementation of the detector. */
+#define _Py_ANNOTATE_FLUSH_STATE() \
+    AnnotateFlushState(__FILE__, __LINE__)
+
+
+#else  /* DYNAMIC_ANNOTATIONS_ENABLED == 0 */
+
+#define _Py_ANNOTATE_RWLOCK_CREATE(lock) /* empty */
+#define _Py_ANNOTATE_RWLOCK_DESTROY(lock) /* empty */
+#define _Py_ANNOTATE_RWLOCK_ACQUIRED(lock, is_w) /* empty */
+#define _Py_ANNOTATE_RWLOCK_RELEASED(lock, is_w) /* empty */
+#define _Py_ANNOTATE_BARRIER_INIT(barrier, count, reinitialization_allowed) /* */
+#define _Py_ANNOTATE_BARRIER_WAIT_BEFORE(barrier) /* empty */
+#define _Py_ANNOTATE_BARRIER_WAIT_AFTER(barrier) /* empty */
+#define _Py_ANNOTATE_BARRIER_DESTROY(barrier) /* empty */
+#define _Py_ANNOTATE_CONDVAR_LOCK_WAIT(cv, lock) /* empty */
+#define _Py_ANNOTATE_CONDVAR_WAIT(cv) /* empty */
+#define _Py_ANNOTATE_CONDVAR_SIGNAL(cv) /* empty */
+#define _Py_ANNOTATE_CONDVAR_SIGNAL_ALL(cv) /* empty */
+#define _Py_ANNOTATE_HAPPENS_BEFORE(obj) /* empty */
+#define _Py_ANNOTATE_HAPPENS_AFTER(obj) /* empty */
+#define _Py_ANNOTATE_PUBLISH_MEMORY_RANGE(address, size) /* empty */
+#define _Py_ANNOTATE_UNPUBLISH_MEMORY_RANGE(address, size)  /* empty */
+#define _Py_ANNOTATE_SWAP_MEMORY_RANGE(address, size)  /* empty */
+#define _Py_ANNOTATE_PCQ_CREATE(pcq) /* empty */
+#define _Py_ANNOTATE_PCQ_DESTROY(pcq) /* empty */
+#define _Py_ANNOTATE_PCQ_PUT(pcq) /* empty */
+#define _Py_ANNOTATE_PCQ_GET(pcq) /* empty */
+#define _Py_ANNOTATE_NEW_MEMORY(address, size) /* empty */
+#define _Py_ANNOTATE_EXPECT_RACE(address, description) /* empty */
+#define _Py_ANNOTATE_BENIGN_RACE(address, description) /* empty */
+#define _Py_ANNOTATE_BENIGN_RACE_SIZED(address, size, description) /* empty */
+#define _Py_ANNOTATE_PURE_HAPPENS_BEFORE_MUTEX(mu) /* empty */
+#define _Py_ANNOTATE_MUTEX_IS_USED_AS_CONDVAR(mu) /* empty */
+#define _Py_ANNOTATE_TRACE_MEMORY(arg) /* empty */
+#define _Py_ANNOTATE_THREAD_NAME(name) /* empty */
+#define _Py_ANNOTATE_IGNORE_READS_BEGIN() /* empty */
+#define _Py_ANNOTATE_IGNORE_READS_END() /* empty */
+#define _Py_ANNOTATE_IGNORE_WRITES_BEGIN() /* empty */
+#define _Py_ANNOTATE_IGNORE_WRITES_END() /* empty */
+#define _Py_ANNOTATE_IGNORE_READS_AND_WRITES_BEGIN() /* empty */
+#define _Py_ANNOTATE_IGNORE_READS_AND_WRITES_END() /* empty */
+#define _Py_ANNOTATE_IGNORE_SYNC_BEGIN() /* empty */
+#define _Py_ANNOTATE_IGNORE_SYNC_END() /* empty */
+#define _Py_ANNOTATE_ENABLE_RACE_DETECTION(enable) /* empty */
+#define _Py_ANNOTATE_NO_OP(arg) /* empty */
+#define _Py_ANNOTATE_FLUSH_STATE() /* empty */
+
+#endif  /* DYNAMIC_ANNOTATIONS_ENABLED */
+
+/* Use the macros above rather than using these functions directly. */
+#ifdef __cplusplus
+extern "C" {
+#endif
+void AnnotateRWLockCreate(const char *file, int line,
+                          const volatile void *lock);
+void AnnotateRWLockDestroy(const char *file, int line,
+                           const volatile void *lock);
+void AnnotateRWLockAcquired(const char *file, int line,
+                            const volatile void *lock, long is_w);
+void AnnotateRWLockReleased(const char *file, int line,
+                            const volatile void *lock, long is_w);
+void AnnotateBarrierInit(const char *file, int line,
+                         const volatile void *barrier, long count,
+                         long reinitialization_allowed);
+void AnnotateBarrierWaitBefore(const char *file, int line,
+                               const volatile void *barrier);
+void AnnotateBarrierWaitAfter(const char *file, int line,
+                              const volatile void *barrier);
+void AnnotateBarrierDestroy(const char *file, int line,
+                            const volatile void *barrier);
+void AnnotateCondVarWait(const char *file, int line,
+                         const volatile void *cv,
+                         const volatile void *lock);
+void AnnotateCondVarSignal(const char *file, int line,
+                           const volatile void *cv);
+void AnnotateCondVarSignalAll(const char *file, int line,
+                              const volatile void *cv);
+void AnnotatePublishMemoryRange(const char *file, int line,
+                                const volatile void *address,
+                                long size);
+void AnnotateUnpublishMemoryRange(const char *file, int line,
+                                  const volatile void *address,
+                                  long size);
+void AnnotatePCQCreate(const char *file, int line,
+                       const volatile void *pcq);
+void AnnotatePCQDestroy(const char *file, int line,
+                        const volatile void *pcq);
+void AnnotatePCQPut(const char *file, int line,
+                    const volatile void *pcq);
+void AnnotatePCQGet(const char *file, int line,
+                    const volatile void *pcq);
+void AnnotateNewMemory(const char *file, int line,
+                       const volatile void *address,
+                       long size);
+void AnnotateExpectRace(const char *file, int line,
+                        const volatile void *address,
+                        const char *description);
+void AnnotateBenignRace(const char *file, int line,
+                        const volatile void *address,
+                        const char *description);
+void AnnotateBenignRaceSized(const char *file, int line,
+                        const volatile void *address,
+                        long size,
+                        const char *description);
+void AnnotateMutexIsUsedAsCondVar(const char *file, int line,
+                                  const volatile void *mu);
+void AnnotateTraceMemory(const char *file, int line,
+                         const volatile void *arg);
+void AnnotateThreadName(const char *file, int line,
+                        const char *name);
+void AnnotateIgnoreReadsBegin(const char *file, int line);
+void AnnotateIgnoreReadsEnd(const char *file, int line);
+void AnnotateIgnoreWritesBegin(const char *file, int line);
+void AnnotateIgnoreWritesEnd(const char *file, int line);
+void AnnotateEnableRaceDetection(const char *file, int line, int enable);
+void AnnotateNoOp(const char *file, int line,
+                  const volatile void *arg);
+void AnnotateFlushState(const char *file, int line);
+
+/* Return non-zero value if running under valgrind.
+
+  If "valgrind.h" is included into dynamic_annotations.c,
+  the regular valgrind mechanism will be used.
+  See http://valgrind.org/docs/manual/manual-core-adv.html about
+  RUNNING_ON_VALGRIND and other valgrind "client requests".
+  The file "valgrind.h" may be obtained by doing
+     svn co svn://svn.valgrind.org/valgrind/trunk/include
+
+  If for some reason you can't use "valgrind.h" or want to fake valgrind,
+  there are two ways to make this function return non-zero:
+    - Use environment variable: export RUNNING_ON_VALGRIND=1
+    - Make your tool intercept the function RunningOnValgrind() and
+      change its return value.
+ */
+int RunningOnValgrind(void);
+
+#ifdef __cplusplus
+}
+#endif
+
+#if DYNAMIC_ANNOTATIONS_ENABLED != 0 && defined(__cplusplus)
+
+  /* _Py_ANNOTATE_UNPROTECTED_READ is the preferred way to annotate racey reads.
+
+     Instead of doing
+        _Py_ANNOTATE_IGNORE_READS_BEGIN();
+        ... = x;
+        _Py_ANNOTATE_IGNORE_READS_END();
+     one can use
+        ... = _Py_ANNOTATE_UNPROTECTED_READ(x); */
+  template <class T>
+  inline T _Py_ANNOTATE_UNPROTECTED_READ(const volatile T &x) {
+    _Py_ANNOTATE_IGNORE_READS_BEGIN();
+    T res = x;
+    _Py_ANNOTATE_IGNORE_READS_END();
+    return res;
+  }
+  /* Apply _Py_ANNOTATE_BENIGN_RACE_SIZED to a static variable. */
+#define _Py_ANNOTATE_BENIGN_RACE_STATIC(static_var, description)        \
+    namespace {                                                       \
+      class static_var ## _annotator {                                \
+       public:                                                        \
+        static_var ## _annotator() {                                  \
+          _Py_ANNOTATE_BENIGN_RACE_SIZED(&static_var,                     \
+                                      sizeof(static_var),             \
+            # static_var ": " description);                           \
+        }                                                             \
+      };                                                              \
+      static static_var ## _annotator the ## static_var ## _annotator;\
+    }
+#else /* DYNAMIC_ANNOTATIONS_ENABLED == 0 */
+
+#define _Py_ANNOTATE_UNPROTECTED_READ(x) (x)
+#define _Py_ANNOTATE_BENIGN_RACE_STATIC(static_var, description)  /* empty */
+
+#endif /* DYNAMIC_ANNOTATIONS_ENABLED */
+
+#endif  /* __DYNAMIC_ANNOTATIONS_H__ */

Dependencies/Python/include/enumobject.h

@@ -0,0 +1,17 @@
+#ifndef Py_ENUMOBJECT_H
+#define Py_ENUMOBJECT_H
+
+/* Enumerate Object */
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+PyAPI_DATA(PyTypeObject) PyEnum_Type;
+PyAPI_DATA(PyTypeObject) PyReversed_Type;
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif /* !Py_ENUMOBJECT_H */

Dependencies/Python/include/errcode.h

@@ -0,0 +1,45 @@
+// Error codes passed around between file input, tokenizer, parser and
+// interpreter.  This is necessary so we can turn them into Python
+// exceptions at a higher level.  Note that some errors have a
+// slightly different meaning when passed from the tokenizer to the
+// parser than when passed from the parser to the interpreter; e.g.
+// the parser only returns E_EOF when it hits EOF immediately, and it
+// never returns E_OK.
+//
+// The public PyRun_InteractiveOneObjectEx() function can return E_EOF,
+// same as its variants:
+//
+// * PyRun_InteractiveOneObject()
+// * PyRun_InteractiveOneFlags()
+// * PyRun_InteractiveOne()
+
+#ifndef Py_ERRCODE_H
+#define Py_ERRCODE_H
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+#define E_OK             10      /* No error */
+#define E_EOF            11      /* End Of File */
+#define E_INTR           12      /* Interrupted */
+#define E_TOKEN          13      /* Bad token */
+#define E_SYNTAX         14      /* Syntax error */
+#define E_NOMEM          15      /* Ran out of memory */
+#define E_DONE           16      /* Parsing complete */
+#define E_ERROR          17      /* Execution error */
+#define E_TABSPACE       18      /* Inconsistent mixing of tabs and spaces */
+#define E_OVERFLOW       19      /* Node had too many children */
+#define E_TOODEEP        20      /* Too many indentation levels */
+#define E_DEDENT         21      /* No matching outer block for dedent */
+#define E_DECODE         22      /* Error in decoding into Unicode */
+#define E_EOFS           23      /* EOF in triple-quoted string */
+#define E_EOLS           24      /* EOL in single-quoted string */
+#define E_LINECONT       25      /* Unexpected characters after a line continuation */
+#define E_BADSINGLE      27      /* Ill-formed single statement input */
+#define E_INTERACT_STOP  28      /* Interactive mode stopped tokenization */
+#define E_COLUMNOVERFLOW 29      /* Column offset overflow */
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* !Py_ERRCODE_H */

Dependencies/Python/include/exports.h

@@ -0,0 +1,108 @@
+#ifndef Py_EXPORTS_H
+#define Py_EXPORTS_H
+
+/* Declarations for symbol visibility.
+
+  PyAPI_FUNC(type): Declares a public Python API function and return type
+  PyAPI_DATA(type): Declares public Python data and its type
+  PyMODINIT_FUNC:   A Python module init function.  If these functions are
+                    inside the Python core, they are private to the core.
+                    If in an extension module, it may be declared with
+                    external linkage depending on the platform.
+
+  As a number of platforms support/require "__declspec(dllimport/dllexport)",
+  we support a HAVE_DECLSPEC_DLL macro to save duplication.
+*/
+
+/*
+  All windows ports, except cygwin, are handled in PC/pyconfig.h.
+
+  Cygwin is the only other autoconf platform requiring special
+  linkage handling and it uses __declspec().
+*/
+#if defined(__CYGWIN__)
+#       define HAVE_DECLSPEC_DLL
+#endif
+
+#if defined(_WIN32) || defined(__CYGWIN__)
+    #if defined(Py_ENABLE_SHARED)
+        #define Py_IMPORTED_SYMBOL __declspec(dllimport)
+        #define Py_EXPORTED_SYMBOL __declspec(dllexport)
+        #define Py_LOCAL_SYMBOL
+    #else
+        #define Py_IMPORTED_SYMBOL
+        #define Py_EXPORTED_SYMBOL
+        #define Py_LOCAL_SYMBOL
+    #endif
+#else
+/*
+ * If we only ever used gcc >= 5, we could use __has_attribute(visibility)
+ * as a cross-platform way to determine if visibility is supported. However,
+ * we may still need to support gcc >= 4, as some Ubuntu LTS and Centos versions
+ * have 4 < gcc < 5.
+ */
+    #ifndef __has_attribute
+      #define __has_attribute(x) 0  // Compatibility with non-clang compilers.
+    #endif
+    #if (defined(__GNUC__) && (__GNUC__ >= 4)) ||\
+        (defined(__clang__) && __has_attribute(visibility))
+        #define Py_IMPORTED_SYMBOL __attribute__ ((visibility ("default")))
+        #define Py_EXPORTED_SYMBOL __attribute__ ((visibility ("default")))
+        #define Py_LOCAL_SYMBOL  __attribute__ ((visibility ("hidden")))
+    #else
+        #define Py_IMPORTED_SYMBOL
+        #define Py_EXPORTED_SYMBOL
+        #define Py_LOCAL_SYMBOL
+    #endif
+#endif
+
+/* only get special linkage if built as shared or platform is Cygwin */
+#if defined(Py_ENABLE_SHARED) || defined(__CYGWIN__)
+#       if defined(HAVE_DECLSPEC_DLL)
+#               if defined(Py_BUILD_CORE) && !defined(Py_BUILD_CORE_MODULE)
+#                       define PyAPI_FUNC(RTYPE) Py_EXPORTED_SYMBOL RTYPE
+#                       define PyAPI_DATA(RTYPE) extern Py_EXPORTED_SYMBOL RTYPE
+        /* module init functions inside the core need no external linkage */
+        /* except for Cygwin to handle embedding */
+#                       if defined(__CYGWIN__)
+#                               define PyMODINIT_FUNC Py_EXPORTED_SYMBOL PyObject*
+#                       else /* __CYGWIN__ */
+#                               define PyMODINIT_FUNC PyObject*
+#                       endif /* __CYGWIN__ */
+#               else /* Py_BUILD_CORE */
+        /* Building an extension module, or an embedded situation */
+        /* public Python functions and data are imported */
+        /* Under Cygwin, auto-import functions to prevent compilation */
+        /* failures similar to those described at the bottom of 4.1: */
+        /* http://docs.python.org/extending/windows.html#a-cookbook-approach */
+#                       if !defined(__CYGWIN__)
+#                               define PyAPI_FUNC(RTYPE) Py_IMPORTED_SYMBOL RTYPE
+#                       endif /* !__CYGWIN__ */
+#                       define PyAPI_DATA(RTYPE) extern Py_IMPORTED_SYMBOL RTYPE
+        /* module init functions outside the core must be exported */
+#                       if defined(__cplusplus)
+#                               define PyMODINIT_FUNC extern "C" Py_EXPORTED_SYMBOL PyObject*
+#                       else /* __cplusplus */
+#                               define PyMODINIT_FUNC Py_EXPORTED_SYMBOL PyObject*
+#                       endif /* __cplusplus */
+#               endif /* Py_BUILD_CORE */
+#       endif /* HAVE_DECLSPEC_DLL */
+#endif /* Py_ENABLE_SHARED */
+
+/* If no external linkage macros defined by now, create defaults */
+#ifndef PyAPI_FUNC
+#       define PyAPI_FUNC(RTYPE) Py_EXPORTED_SYMBOL RTYPE
+#endif
+#ifndef PyAPI_DATA
+#       define PyAPI_DATA(RTYPE) extern Py_EXPORTED_SYMBOL RTYPE
+#endif
+#ifndef PyMODINIT_FUNC
+#       if defined(__cplusplus)
+#               define PyMODINIT_FUNC extern "C" Py_EXPORTED_SYMBOL PyObject*
+#       else /* __cplusplus */
+#               define PyMODINIT_FUNC Py_EXPORTED_SYMBOL PyObject*
+#       endif /* __cplusplus */
+#endif
+
+
+#endif /* Py_EXPORTS_H */

Dependencies/Python/include/fileobject.h

@@ -0,0 +1,41 @@
+/* File object interface (what's left of it -- see io.py) */
+
+#ifndef Py_FILEOBJECT_H
+#define Py_FILEOBJECT_H
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+#define PY_STDIOTEXTMODE "b"
+
+PyAPI_FUNC(PyObject *) PyFile_FromFd(int, const char *, const char *, int,
+                                     const char *, const char *,
+                                     const char *, int);
+PyAPI_FUNC(PyObject *) PyFile_GetLine(PyObject *, int);
+PyAPI_FUNC(int) PyFile_WriteObject(PyObject *, PyObject *, int);
+PyAPI_FUNC(int) PyFile_WriteString(const char *, PyObject *);
+PyAPI_FUNC(int) PyObject_AsFileDescriptor(PyObject *);
+
+/* The default encoding used by the platform file system APIs
+   If non-NULL, this is different than the default encoding for strings
+*/
+Py_DEPRECATED(3.12) PyAPI_DATA(const char *) Py_FileSystemDefaultEncoding;
+#if !defined(Py_LIMITED_API) || Py_LIMITED_API+0 >= 0x03060000
+Py_DEPRECATED(3.12) PyAPI_DATA(const char *) Py_FileSystemDefaultEncodeErrors;
+#endif
+Py_DEPRECATED(3.12) PyAPI_DATA(int) Py_HasFileSystemDefaultEncoding;
+
+#if !defined(Py_LIMITED_API) || Py_LIMITED_API+0 >= 0x03070000
+Py_DEPRECATED(3.12) PyAPI_DATA(int) Py_UTF8Mode;
+#endif
+
+#ifndef Py_LIMITED_API
+#  define Py_CPYTHON_FILEOBJECT_H
+#  include "cpython/fileobject.h"
+#  undef Py_CPYTHON_FILEOBJECT_H
+#endif
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* !Py_FILEOBJECT_H */

Dependencies/Python/include/fileutils.h

@@ -0,0 +1,62 @@
+#ifndef Py_FILEUTILS_H
+#define Py_FILEUTILS_H
+
+/*******************************
+ * stat() and fstat() fiddling *
+ *******************************/
+
+#ifdef HAVE_SYS_STAT_H
+#  include <sys/stat.h>           // S_ISREG()
+#elif defined(HAVE_STAT_H)
+#  include <stat.h>               // S_ISREG()
+#endif
+
+#ifndef S_IFMT
+   // VisualAge C/C++ Failed to Define MountType Field in sys/stat.h.
+#  define S_IFMT 0170000
+#endif
+#ifndef S_IFLNK
+   // Windows doesn't define S_IFLNK, but posixmodule.c maps
+   // IO_REPARSE_TAG_SYMLINK to S_IFLNK.
+#  define S_IFLNK 0120000
+#endif
+#ifndef S_ISREG
+#  define S_ISREG(x) (((x) & S_IFMT) == S_IFREG)
+#endif
+#ifndef S_ISDIR
+#  define S_ISDIR(x) (((x) & S_IFMT) == S_IFDIR)
+#endif
+#ifndef S_ISCHR
+#  define S_ISCHR(x) (((x) & S_IFMT) == S_IFCHR)
+#endif
+#ifndef S_ISLNK
+#  define S_ISLNK(x) (((x) & S_IFMT) == S_IFLNK)
+#endif
+
+
+// Move this down here since some C++ #include's don't like to be included
+// inside an extern "C".
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+#if !defined(Py_LIMITED_API) || Py_LIMITED_API+0 >= 0x03050000
+PyAPI_FUNC(wchar_t *) Py_DecodeLocale(
+    const char *arg,
+    size_t *size);
+
+PyAPI_FUNC(char*) Py_EncodeLocale(
+    const wchar_t *text,
+    size_t *error_pos);
+#endif
+
+#ifndef Py_LIMITED_API
+#  define Py_CPYTHON_FILEUTILS_H
+#  include "cpython/fileutils.h"
+#  undef Py_CPYTHON_FILEUTILS_H
+#endif
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* !Py_FILEUTILS_H */

Dependencies/Python/include/floatobject.h

@@ -0,0 +1,54 @@
+
+/* Float object interface */
+
+/*
+PyFloatObject represents a (double precision) floating-point number.
+*/
+
+#ifndef Py_FLOATOBJECT_H
+#define Py_FLOATOBJECT_H
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+PyAPI_DATA(PyTypeObject) PyFloat_Type;
+
+#define PyFloat_Check(op) PyObject_TypeCheck(op, &PyFloat_Type)
+#define PyFloat_CheckExact(op) Py_IS_TYPE((op), &PyFloat_Type)
+
+#define Py_RETURN_NAN return PyFloat_FromDouble(Py_NAN)
+
+#define Py_RETURN_INF(sign)                          \
+    do {                                             \
+        if (copysign(1., sign) == 1.) {              \
+            return PyFloat_FromDouble(Py_HUGE_VAL);  \
+        }                                            \
+        else {                                       \
+            return PyFloat_FromDouble(-Py_HUGE_VAL); \
+        }                                            \
+    } while(0)
+
+PyAPI_FUNC(double) PyFloat_GetMax(void);
+PyAPI_FUNC(double) PyFloat_GetMin(void);
+PyAPI_FUNC(PyObject*) PyFloat_GetInfo(void);
+
+/* Return Python float from string PyObject. */
+PyAPI_FUNC(PyObject*) PyFloat_FromString(PyObject*);
+
+/* Return Python float from C double. */
+PyAPI_FUNC(PyObject*) PyFloat_FromDouble(double);
+
+/* Extract C double from Python float.  The macro version trades safety for
+   speed. */
+PyAPI_FUNC(double) PyFloat_AsDouble(PyObject*);
+
+#ifndef Py_LIMITED_API
+#  define Py_CPYTHON_FLOATOBJECT_H
+#  include "cpython/floatobject.h"
+#  undef Py_CPYTHON_FLOATOBJECT_H
+#endif
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* !Py_FLOATOBJECT_H */

Dependencies/Python/include/frameobject.h

@@ -0,0 +1,20 @@
+/* Frame object interface */
+
+#ifndef Py_FRAMEOBJECT_H
+#define Py_FRAMEOBJECT_H
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+#include "pyframe.h"
+
+#ifndef Py_LIMITED_API
+#  define Py_CPYTHON_FRAMEOBJECT_H
+#  include "cpython/frameobject.h"
+#  undef Py_CPYTHON_FRAMEOBJECT_H
+#endif
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* !Py_FRAMEOBJECT_H */

Dependencies/Python/include/genericaliasobject.h

@@ -0,0 +1,14 @@
+// Implementation of PEP 585: support list[int] etc.
+#ifndef Py_GENERICALIASOBJECT_H
+#define Py_GENERICALIASOBJECT_H
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+PyAPI_FUNC(PyObject *) Py_GenericAlias(PyObject *, PyObject *);
+PyAPI_DATA(PyTypeObject) Py_GenericAliasType;
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* !Py_GENERICALIASOBJECT_H */

Dependencies/Python/include/import.h

@@ -0,0 +1,103 @@
+/* Module definition and import interface */
+
+#ifndef Py_IMPORT_H
+#define Py_IMPORT_H
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+PyAPI_FUNC(long) PyImport_GetMagicNumber(void);
+PyAPI_FUNC(const char *) PyImport_GetMagicTag(void);
+PyAPI_FUNC(PyObject *) PyImport_ExecCodeModule(
+    const char *name,           /* UTF-8 encoded string */
+    PyObject *co
+    );
+PyAPI_FUNC(PyObject *) PyImport_ExecCodeModuleEx(
+    const char *name,           /* UTF-8 encoded string */
+    PyObject *co,
+    const char *pathname        /* decoded from the filesystem encoding */
+    );
+PyAPI_FUNC(PyObject *) PyImport_ExecCodeModuleWithPathnames(
+    const char *name,           /* UTF-8 encoded string */
+    PyObject *co,
+    const char *pathname,       /* decoded from the filesystem encoding */
+    const char *cpathname       /* decoded from the filesystem encoding */
+    );
+#if !defined(Py_LIMITED_API) || Py_LIMITED_API+0 >= 0x03030000
+PyAPI_FUNC(PyObject *) PyImport_ExecCodeModuleObject(
+    PyObject *name,
+    PyObject *co,
+    PyObject *pathname,
+    PyObject *cpathname
+    );
+#endif
+PyAPI_FUNC(PyObject *) PyImport_GetModuleDict(void);
+#if !defined(Py_LIMITED_API) || Py_LIMITED_API+0 >= 0x03070000
+PyAPI_FUNC(PyObject *) PyImport_GetModule(PyObject *name);
+#endif
+#if !defined(Py_LIMITED_API) || Py_LIMITED_API+0 >= 0x03030000
+PyAPI_FUNC(PyObject *) PyImport_AddModuleObject(
+    PyObject *name
+    );
+#endif
+PyAPI_FUNC(PyObject *) PyImport_AddModule(
+    const char *name            /* UTF-8 encoded string */
+    );
+#if !defined(Py_LIMITED_API) || Py_LIMITED_API+0 >= 0x030d0000
+PyAPI_FUNC(PyObject *) PyImport_AddModuleRef(
+    const char *name            /* UTF-8 encoded string */
+    );
+#endif
+PyAPI_FUNC(PyObject *) PyImport_ImportModule(
+    const char *name            /* UTF-8 encoded string */
+    );
+Py_DEPRECATED(3.13) PyAPI_FUNC(PyObject *) PyImport_ImportModuleNoBlock(
+    const char *name            /* UTF-8 encoded string */
+    );
+PyAPI_FUNC(PyObject *) PyImport_ImportModuleLevel(
+    const char *name,           /* UTF-8 encoded string */
+    PyObject *globals,
+    PyObject *locals,
+    PyObject *fromlist,
+    int level
+    );
+#if !defined(Py_LIMITED_API) || Py_LIMITED_API+0 >= 0x03050000
+PyAPI_FUNC(PyObject *) PyImport_ImportModuleLevelObject(
+    PyObject *name,
+    PyObject *globals,
+    PyObject *locals,
+    PyObject *fromlist,
+    int level
+    );
+#endif
+
+#define PyImport_ImportModuleEx(n, g, l, f) \
+    PyImport_ImportModuleLevel((n), (g), (l), (f), 0)
+
+PyAPI_FUNC(PyObject *) PyImport_GetImporter(PyObject *path);
+PyAPI_FUNC(PyObject *) PyImport_Import(PyObject *name);
+PyAPI_FUNC(PyObject *) PyImport_ReloadModule(PyObject *m);
+#if !defined(Py_LIMITED_API) || Py_LIMITED_API+0 >= 0x03030000
+PyAPI_FUNC(int) PyImport_ImportFrozenModuleObject(
+    PyObject *name
+    );
+#endif
+PyAPI_FUNC(int) PyImport_ImportFrozenModule(
+    const char *name            /* UTF-8 encoded string */
+    );
+
+PyAPI_FUNC(int) PyImport_AppendInittab(
+    const char *name,           /* ASCII encoded string */
+    PyObject* (*initfunc)(void)
+    );
+
+#ifndef Py_LIMITED_API
+#  define Py_CPYTHON_IMPORT_H
+#  include "cpython/import.h"
+#  undef Py_CPYTHON_IMPORT_H
+#endif
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* !Py_IMPORT_H */

Dependencies/Python/include/internal/mimalloc/mimalloc.h

@@ -0,0 +1,565 @@
+/* ----------------------------------------------------------------------------
+Copyright (c) 2018-2023, Microsoft Research, Daan Leijen
+This is free software; you can redistribute it and/or modify it under the
+terms of the MIT license. A copy of the license can be found in the file
+"LICENSE" at the root of this distribution.
+-----------------------------------------------------------------------------*/
+#pragma once
+#ifndef MIMALLOC_H
+#define MIMALLOC_H
+
+#define MI_MALLOC_VERSION 212   // major + 2 digits minor
+
+// ------------------------------------------------------
+// Compiler specific attributes
+// ------------------------------------------------------
+
+#ifdef __cplusplus
+  #if (__cplusplus >= 201103L) || (_MSC_VER > 1900)  // C++11
+    #define mi_attr_noexcept   noexcept
+  #else
+    #define mi_attr_noexcept   throw()
+  #endif
+#else
+  #define mi_attr_noexcept
+#endif
+
+#if defined(__cplusplus) && (__cplusplus >= 201703)
+  #define mi_decl_nodiscard    [[nodiscard]]
+#elif (defined(__GNUC__) && (__GNUC__ >= 4)) || defined(__clang__)  // includes clang, icc, and clang-cl
+  #define mi_decl_nodiscard    __attribute__((warn_unused_result))
+#elif defined(_HAS_NODISCARD)
+  #define mi_decl_nodiscard    _NODISCARD
+#elif (_MSC_VER >= 1700)
+  #define mi_decl_nodiscard    _Check_return_
+#else
+  #define mi_decl_nodiscard
+#endif
+
+#if defined(_MSC_VER) || defined(__MINGW32__)
+  #if !defined(MI_SHARED_LIB)
+    #define mi_decl_export
+  #elif defined(MI_SHARED_LIB_EXPORT)
+    #define mi_decl_export              __declspec(dllexport)
+  #else
+    #define mi_decl_export              __declspec(dllimport)
+  #endif
+  #if defined(__MINGW32__)
+    #define mi_decl_restrict
+    #define mi_attr_malloc              __attribute__((malloc))
+  #else
+    #if (_MSC_VER >= 1900) && !defined(__EDG__)
+      #define mi_decl_restrict          __declspec(allocator) __declspec(restrict)
+    #else
+      #define mi_decl_restrict          __declspec(restrict)
+    #endif
+    #define mi_attr_malloc
+  #endif
+  #define mi_cdecl                      __cdecl
+  #define mi_attr_alloc_size(s)
+  #define mi_attr_alloc_size2(s1,s2)
+  #define mi_attr_alloc_align(p)
+#elif defined(__GNUC__)                 // includes clang and icc
+  #if defined(MI_SHARED_LIB) && defined(MI_SHARED_LIB_EXPORT)
+    #define mi_decl_export              __attribute__((visibility("default")))
+  #else
+    #define mi_decl_export
+  #endif
+  #define mi_cdecl                      // leads to warnings... __attribute__((cdecl))
+  #define mi_decl_restrict
+  #define mi_attr_malloc                __attribute__((malloc))
+  #if (defined(__clang_major__) && (__clang_major__ < 4)) || (__GNUC__ < 5)
+    #define mi_attr_alloc_size(s)
+    #define mi_attr_alloc_size2(s1,s2)
+    #define mi_attr_alloc_align(p)
+  #elif defined(__INTEL_COMPILER)
+    #define mi_attr_alloc_size(s)       __attribute__((alloc_size(s)))
+    #define mi_attr_alloc_size2(s1,s2)  __attribute__((alloc_size(s1,s2)))
+    #define mi_attr_alloc_align(p)
+  #else
+    #define mi_attr_alloc_size(s)       __attribute__((alloc_size(s)))
+    #define mi_attr_alloc_size2(s1,s2)  __attribute__((alloc_size(s1,s2)))
+    #define mi_attr_alloc_align(p)      __attribute__((alloc_align(p)))
+  #endif
+#else
+  #define mi_cdecl
+  #define mi_decl_export
+  #define mi_decl_restrict
+  #define mi_attr_malloc
+  #define mi_attr_alloc_size(s)
+  #define mi_attr_alloc_size2(s1,s2)
+  #define mi_attr_alloc_align(p)
+#endif
+
+// ------------------------------------------------------
+// Includes
+// ------------------------------------------------------
+
+#include <stddef.h>     // size_t
+#include <stdbool.h>    // bool
+#include <stdint.h>     // INTPTR_MAX
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+// ------------------------------------------------------
+// Standard malloc interface
+// ------------------------------------------------------
+
+mi_decl_nodiscard mi_decl_export mi_decl_restrict void* mi_malloc(size_t size)  mi_attr_noexcept mi_attr_malloc mi_attr_alloc_size(1);
+mi_decl_nodiscard mi_decl_export mi_decl_restrict void* mi_calloc(size_t count, size_t size)  mi_attr_noexcept mi_attr_malloc mi_attr_alloc_size2(1,2);
+mi_decl_nodiscard mi_decl_export void* mi_realloc(void* p, size_t newsize)      mi_attr_noexcept mi_attr_alloc_size(2);
+mi_decl_export void* mi_expand(void* p, size_t newsize)                         mi_attr_noexcept mi_attr_alloc_size(2);
+
+mi_decl_export void mi_free(void* p) mi_attr_noexcept;
+mi_decl_nodiscard mi_decl_export mi_decl_restrict char* mi_strdup(const char* s) mi_attr_noexcept mi_attr_malloc;
+mi_decl_nodiscard mi_decl_export mi_decl_restrict char* mi_strndup(const char* s, size_t n) mi_attr_noexcept mi_attr_malloc;
+mi_decl_nodiscard mi_decl_export mi_decl_restrict char* mi_realpath(const char* fname, char* resolved_name) mi_attr_noexcept mi_attr_malloc;
+
+// ------------------------------------------------------
+// Extended functionality
+// ------------------------------------------------------
+#define MI_SMALL_WSIZE_MAX  (128)
+#define MI_SMALL_SIZE_MAX   (MI_SMALL_WSIZE_MAX*sizeof(void*))
+
+mi_decl_nodiscard mi_decl_export mi_decl_restrict void* mi_malloc_small(size_t size) mi_attr_noexcept mi_attr_malloc mi_attr_alloc_size(1);
+mi_decl_nodiscard mi_decl_export mi_decl_restrict void* mi_zalloc_small(size_t size) mi_attr_noexcept mi_attr_malloc mi_attr_alloc_size(1);
+mi_decl_nodiscard mi_decl_export mi_decl_restrict void* mi_zalloc(size_t size)       mi_attr_noexcept mi_attr_malloc mi_attr_alloc_size(1);
+
+mi_decl_nodiscard mi_decl_export mi_decl_restrict void* mi_mallocn(size_t count, size_t size) mi_attr_noexcept mi_attr_malloc mi_attr_alloc_size2(1,2);
+mi_decl_nodiscard mi_decl_export void* mi_reallocn(void* p, size_t count, size_t size)        mi_attr_noexcept mi_attr_alloc_size2(2,3);
+mi_decl_nodiscard mi_decl_export void* mi_reallocf(void* p, size_t newsize)                   mi_attr_noexcept mi_attr_alloc_size(2);
+
+mi_decl_nodiscard mi_decl_export size_t mi_usable_size(const void* p) mi_attr_noexcept;
+mi_decl_nodiscard mi_decl_export size_t mi_good_size(size_t size)     mi_attr_noexcept;
+
+
+// ------------------------------------------------------
+// Internals
+// ------------------------------------------------------
+
+typedef void (mi_cdecl mi_deferred_free_fun)(bool force, unsigned long long heartbeat, void* arg);
+mi_decl_export void mi_register_deferred_free(mi_deferred_free_fun* deferred_free, void* arg) mi_attr_noexcept;
+
+typedef void (mi_cdecl mi_output_fun)(const char* msg, void* arg);
+mi_decl_export void mi_register_output(mi_output_fun* out, void* arg) mi_attr_noexcept;
+
+typedef void (mi_cdecl mi_error_fun)(int err, void* arg);
+mi_decl_export void mi_register_error(mi_error_fun* fun, void* arg);
+
+mi_decl_export void mi_collect(bool force)    mi_attr_noexcept;
+mi_decl_export int  mi_version(void)          mi_attr_noexcept;
+mi_decl_export void mi_stats_reset(void)      mi_attr_noexcept;
+mi_decl_export void mi_stats_merge(void)      mi_attr_noexcept;
+mi_decl_export void mi_stats_print(void* out) mi_attr_noexcept;  // backward compatibility: `out` is ignored and should be NULL
+mi_decl_export void mi_stats_print_out(mi_output_fun* out, void* arg) mi_attr_noexcept;
+
+mi_decl_export void mi_process_init(void)     mi_attr_noexcept;
+mi_decl_export void mi_thread_init(void)      mi_attr_noexcept;
+mi_decl_export void mi_thread_done(void)      mi_attr_noexcept;
+mi_decl_export void mi_thread_stats_print_out(mi_output_fun* out, void* arg) mi_attr_noexcept;
+
+mi_decl_export void mi_process_info(size_t* elapsed_msecs, size_t* user_msecs, size_t* system_msecs,
+                                    size_t* current_rss, size_t* peak_rss,
+                                    size_t* current_commit, size_t* peak_commit, size_t* page_faults) mi_attr_noexcept;
+
+// -------------------------------------------------------------------------------------
+// Aligned allocation
+// Note that `alignment` always follows `size` for consistency with unaligned
+// allocation, but unfortunately this differs from `posix_memalign` and `aligned_alloc`.
+// -------------------------------------------------------------------------------------
+
+mi_decl_nodiscard mi_decl_export mi_decl_restrict void* mi_malloc_aligned(size_t size, size_t alignment) mi_attr_noexcept mi_attr_malloc mi_attr_alloc_size(1) mi_attr_alloc_align(2);
+mi_decl_nodiscard mi_decl_export mi_decl_restrict void* mi_malloc_aligned_at(size_t size, size_t alignment, size_t offset) mi_attr_noexcept mi_attr_malloc mi_attr_alloc_size(1);
+mi_decl_nodiscard mi_decl_export mi_decl_restrict void* mi_zalloc_aligned(size_t size, size_t alignment) mi_attr_noexcept mi_attr_malloc mi_attr_alloc_size(1) mi_attr_alloc_align(2);
+mi_decl_nodiscard mi_decl_export mi_decl_restrict void* mi_zalloc_aligned_at(size_t size, size_t alignment, size_t offset) mi_attr_noexcept mi_attr_malloc mi_attr_alloc_size(1);
+mi_decl_nodiscard mi_decl_export mi_decl_restrict void* mi_calloc_aligned(size_t count, size_t size, size_t alignment) mi_attr_noexcept mi_attr_malloc mi_attr_alloc_size2(1,2) mi_attr_alloc_align(3);
+mi_decl_nodiscard mi_decl_export mi_decl_restrict void* mi_calloc_aligned_at(size_t count, size_t size, size_t alignment, size_t offset) mi_attr_noexcept mi_attr_malloc mi_attr_alloc_size2(1,2);
+mi_decl_nodiscard mi_decl_export void* mi_realloc_aligned(void* p, size_t newsize, size_t alignment) mi_attr_noexcept mi_attr_alloc_size(2) mi_attr_alloc_align(3);
+mi_decl_nodiscard mi_decl_export void* mi_realloc_aligned_at(void* p, size_t newsize, size_t alignment, size_t offset) mi_attr_noexcept mi_attr_alloc_size(2);
+
+
+// -------------------------------------------------------------------------------------
+// Heaps: first-class, but can only allocate from the same thread that created it.
+// -------------------------------------------------------------------------------------
+
+struct mi_heap_s;
+typedef struct mi_heap_s mi_heap_t;
+
+mi_decl_nodiscard mi_decl_export mi_heap_t* mi_heap_new(void);
+mi_decl_export void       mi_heap_delete(mi_heap_t* heap);
+mi_decl_export void       mi_heap_destroy(mi_heap_t* heap);
+mi_decl_export mi_heap_t* mi_heap_set_default(mi_heap_t* heap);
+mi_decl_export mi_heap_t* mi_heap_get_default(void);
+mi_decl_export mi_heap_t* mi_heap_get_backing(void);
+mi_decl_export void       mi_heap_collect(mi_heap_t* heap, bool force) mi_attr_noexcept;
+
+mi_decl_nodiscard mi_decl_export mi_decl_restrict void* mi_heap_malloc(mi_heap_t* heap, size_t size) mi_attr_noexcept mi_attr_malloc mi_attr_alloc_size(2);
+mi_decl_nodiscard mi_decl_export mi_decl_restrict void* mi_heap_zalloc(mi_heap_t* heap, size_t size) mi_attr_noexcept mi_attr_malloc mi_attr_alloc_size(2);
+mi_decl_nodiscard mi_decl_export mi_decl_restrict void* mi_heap_calloc(mi_heap_t* heap, size_t count, size_t size) mi_attr_noexcept mi_attr_malloc mi_attr_alloc_size2(2, 3);
+mi_decl_nodiscard mi_decl_export mi_decl_restrict void* mi_heap_mallocn(mi_heap_t* heap, size_t count, size_t size) mi_attr_noexcept mi_attr_malloc mi_attr_alloc_size2(2, 3);
+mi_decl_nodiscard mi_decl_export mi_decl_restrict void* mi_heap_malloc_small(mi_heap_t* heap, size_t size) mi_attr_noexcept mi_attr_malloc mi_attr_alloc_size(2);
+
+mi_decl_nodiscard mi_decl_export void* mi_heap_realloc(mi_heap_t* heap, void* p, size_t newsize)              mi_attr_noexcept mi_attr_alloc_size(3);
+mi_decl_nodiscard mi_decl_export void* mi_heap_reallocn(mi_heap_t* heap, void* p, size_t count, size_t size)  mi_attr_noexcept mi_attr_alloc_size2(3,4);
+mi_decl_nodiscard mi_decl_export void* mi_heap_reallocf(mi_heap_t* heap, void* p, size_t newsize)             mi_attr_noexcept mi_attr_alloc_size(3);
+
+mi_decl_nodiscard mi_decl_export mi_decl_restrict char* mi_heap_strdup(mi_heap_t* heap, const char* s)            mi_attr_noexcept mi_attr_malloc;
+mi_decl_nodiscard mi_decl_export mi_decl_restrict char* mi_heap_strndup(mi_heap_t* heap, const char* s, size_t n) mi_attr_noexcept mi_attr_malloc;
+mi_decl_nodiscard mi_decl_export mi_decl_restrict char* mi_heap_realpath(mi_heap_t* heap, const char* fname, char* resolved_name) mi_attr_noexcept mi_attr_malloc;
+
+mi_decl_nodiscard mi_decl_export mi_decl_restrict void* mi_heap_malloc_aligned(mi_heap_t* heap, size_t size, size_t alignment) mi_attr_noexcept mi_attr_malloc mi_attr_alloc_size(2) mi_attr_alloc_align(3);
+mi_decl_nodiscard mi_decl_export mi_decl_restrict void* mi_heap_malloc_aligned_at(mi_heap_t* heap, size_t size, size_t alignment, size_t offset) mi_attr_noexcept mi_attr_malloc mi_attr_alloc_size(2);
+mi_decl_nodiscard mi_decl_export mi_decl_restrict void* mi_heap_zalloc_aligned(mi_heap_t* heap, size_t size, size_t alignment) mi_attr_noexcept mi_attr_malloc mi_attr_alloc_size(2) mi_attr_alloc_align(3);
+mi_decl_nodiscard mi_decl_export mi_decl_restrict void* mi_heap_zalloc_aligned_at(mi_heap_t* heap, size_t size, size_t alignment, size_t offset) mi_attr_noexcept mi_attr_malloc mi_attr_alloc_size(2);
+mi_decl_nodiscard mi_decl_export mi_decl_restrict void* mi_heap_calloc_aligned(mi_heap_t* heap, size_t count, size_t size, size_t alignment) mi_attr_noexcept mi_attr_malloc mi_attr_alloc_size2(2, 3) mi_attr_alloc_align(4);
+mi_decl_nodiscard mi_decl_export mi_decl_restrict void* mi_heap_calloc_aligned_at(mi_heap_t* heap, size_t count, size_t size, size_t alignment, size_t offset) mi_attr_noexcept mi_attr_malloc mi_attr_alloc_size2(2, 3);
+mi_decl_nodiscard mi_decl_export void* mi_heap_realloc_aligned(mi_heap_t* heap, void* p, size_t newsize, size_t alignment) mi_attr_noexcept mi_attr_alloc_size(3) mi_attr_alloc_align(4);
+mi_decl_nodiscard mi_decl_export void* mi_heap_realloc_aligned_at(mi_heap_t* heap, void* p, size_t newsize, size_t alignment, size_t offset) mi_attr_noexcept mi_attr_alloc_size(3);
+
+
+// --------------------------------------------------------------------------------
+// Zero initialized re-allocation.
+// Only valid on memory that was originally allocated with zero initialization too.
+// e.g. `mi_calloc`, `mi_zalloc`, `mi_zalloc_aligned` etc.
+// see <https://github.com/microsoft/mimalloc/issues/63#issuecomment-508272992>
+// --------------------------------------------------------------------------------
+
+mi_decl_nodiscard mi_decl_export void* mi_rezalloc(void* p, size_t newsize)                mi_attr_noexcept mi_attr_alloc_size(2);
+mi_decl_nodiscard mi_decl_export void* mi_recalloc(void* p, size_t newcount, size_t size)  mi_attr_noexcept mi_attr_alloc_size2(2,3);
+
+mi_decl_nodiscard mi_decl_export void* mi_rezalloc_aligned(void* p, size_t newsize, size_t alignment) mi_attr_noexcept mi_attr_alloc_size(2) mi_attr_alloc_align(3);
+mi_decl_nodiscard mi_decl_export void* mi_rezalloc_aligned_at(void* p, size_t newsize, size_t alignment, size_t offset) mi_attr_noexcept mi_attr_alloc_size(2);
+mi_decl_nodiscard mi_decl_export void* mi_recalloc_aligned(void* p, size_t newcount, size_t size, size_t alignment) mi_attr_noexcept mi_attr_alloc_size2(2,3) mi_attr_alloc_align(4);
+mi_decl_nodiscard mi_decl_export void* mi_recalloc_aligned_at(void* p, size_t newcount, size_t size, size_t alignment, size_t offset) mi_attr_noexcept mi_attr_alloc_size2(2,3);
+
+mi_decl_nodiscard mi_decl_export void* mi_heap_rezalloc(mi_heap_t* heap, void* p, size_t newsize)                mi_attr_noexcept mi_attr_alloc_size(3);
+mi_decl_nodiscard mi_decl_export void* mi_heap_recalloc(mi_heap_t* heap, void* p, size_t newcount, size_t size)  mi_attr_noexcept mi_attr_alloc_size2(3,4);
+
+mi_decl_nodiscard mi_decl_export void* mi_heap_rezalloc_aligned(mi_heap_t* heap, void* p, size_t newsize, size_t alignment) mi_attr_noexcept mi_attr_alloc_size(3) mi_attr_alloc_align(4);
+mi_decl_nodiscard mi_decl_export void* mi_heap_rezalloc_aligned_at(mi_heap_t* heap, void* p, size_t newsize, size_t alignment, size_t offset) mi_attr_noexcept mi_attr_alloc_size(3);
+mi_decl_nodiscard mi_decl_export void* mi_heap_recalloc_aligned(mi_heap_t* heap, void* p, size_t newcount, size_t size, size_t alignment) mi_attr_noexcept mi_attr_alloc_size2(3,4) mi_attr_alloc_align(5);
+mi_decl_nodiscard mi_decl_export void* mi_heap_recalloc_aligned_at(mi_heap_t* heap, void* p, size_t newcount, size_t size, size_t alignment, size_t offset) mi_attr_noexcept mi_attr_alloc_size2(3,4);
+
+
+// ------------------------------------------------------
+// Analysis
+// ------------------------------------------------------
+
+mi_decl_export bool mi_heap_contains_block(mi_heap_t* heap, const void* p);
+mi_decl_export bool mi_heap_check_owned(mi_heap_t* heap, const void* p);
+mi_decl_export bool mi_check_owned(const void* p);
+
+// An area of heap space contains blocks of a single size.
+typedef struct mi_heap_area_s {
+  void*  blocks;      // start of the area containing heap blocks
+  size_t reserved;    // bytes reserved for this area (virtual)
+  size_t committed;   // current available bytes for this area
+  size_t used;        // number of allocated blocks
+  size_t block_size;  // size in bytes of each block
+  size_t full_block_size; // size in bytes of a full block including padding and metadata.
+} mi_heap_area_t;
+
+typedef bool (mi_cdecl mi_block_visit_fun)(const mi_heap_t* heap, const mi_heap_area_t* area, void* block, size_t block_size, void* arg);
+
+mi_decl_export bool mi_heap_visit_blocks(const mi_heap_t* heap, bool visit_all_blocks, mi_block_visit_fun* visitor, void* arg);
+
+// Experimental
+mi_decl_nodiscard mi_decl_export bool mi_is_in_heap_region(const void* p) mi_attr_noexcept;
+mi_decl_nodiscard mi_decl_export bool mi_is_redirected(void) mi_attr_noexcept;
+
+mi_decl_export int mi_reserve_huge_os_pages_interleave(size_t pages, size_t numa_nodes, size_t timeout_msecs) mi_attr_noexcept;
+mi_decl_export int mi_reserve_huge_os_pages_at(size_t pages, int numa_node, size_t timeout_msecs) mi_attr_noexcept;
+
+mi_decl_export int  mi_reserve_os_memory(size_t size, bool commit, bool allow_large) mi_attr_noexcept;
+mi_decl_export bool mi_manage_os_memory(void* start, size_t size, bool is_committed, bool is_large, bool is_zero, int numa_node) mi_attr_noexcept;
+
+mi_decl_export void mi_debug_show_arenas(void) mi_attr_noexcept;
+
+// Experimental: heaps associated with specific memory arena's
+typedef int mi_arena_id_t;
+mi_decl_export void* mi_arena_area(mi_arena_id_t arena_id, size_t* size);
+mi_decl_export int   mi_reserve_huge_os_pages_at_ex(size_t pages, int numa_node, size_t timeout_msecs, bool exclusive, mi_arena_id_t* arena_id) mi_attr_noexcept;
+mi_decl_export int   mi_reserve_os_memory_ex(size_t size, bool commit, bool allow_large, bool exclusive, mi_arena_id_t* arena_id) mi_attr_noexcept;
+mi_decl_export bool  mi_manage_os_memory_ex(void* start, size_t size, bool is_committed, bool is_large, bool is_zero, int numa_node, bool exclusive, mi_arena_id_t* arena_id) mi_attr_noexcept;
+
+#if MI_MALLOC_VERSION >= 182
+// Create a heap that only allocates in the specified arena
+mi_decl_nodiscard mi_decl_export mi_heap_t* mi_heap_new_in_arena(mi_arena_id_t arena_id);
+#endif
+
+// deprecated
+mi_decl_export int  mi_reserve_huge_os_pages(size_t pages, double max_secs, size_t* pages_reserved) mi_attr_noexcept;
+
+
+// ------------------------------------------------------
+// Convenience
+// ------------------------------------------------------
+
+#define mi_malloc_tp(tp)                ((tp*)mi_malloc(sizeof(tp)))
+#define mi_zalloc_tp(tp)                ((tp*)mi_zalloc(sizeof(tp)))
+#define mi_calloc_tp(tp,n)              ((tp*)mi_calloc(n,sizeof(tp)))
+#define mi_mallocn_tp(tp,n)             ((tp*)mi_mallocn(n,sizeof(tp)))
+#define mi_reallocn_tp(p,tp,n)          ((tp*)mi_reallocn(p,n,sizeof(tp)))
+#define mi_recalloc_tp(p,tp,n)          ((tp*)mi_recalloc(p,n,sizeof(tp)))
+
+#define mi_heap_malloc_tp(hp,tp)        ((tp*)mi_heap_malloc(hp,sizeof(tp)))
+#define mi_heap_zalloc_tp(hp,tp)        ((tp*)mi_heap_zalloc(hp,sizeof(tp)))
+#define mi_heap_calloc_tp(hp,tp,n)      ((tp*)mi_heap_calloc(hp,n,sizeof(tp)))
+#define mi_heap_mallocn_tp(hp,tp,n)     ((tp*)mi_heap_mallocn(hp,n,sizeof(tp)))
+#define mi_heap_reallocn_tp(hp,p,tp,n)  ((tp*)mi_heap_reallocn(hp,p,n,sizeof(tp)))
+#define mi_heap_recalloc_tp(hp,p,tp,n)  ((tp*)mi_heap_recalloc(hp,p,n,sizeof(tp)))
+
+
+// ------------------------------------------------------
+// Options
+// ------------------------------------------------------
+
+typedef enum mi_option_e {
+  // stable options
+  mi_option_show_errors,              // print error messages
+  mi_option_show_stats,               // print statistics on termination
+  mi_option_verbose,                  // print verbose messages
+  // the following options are experimental (see src/options.h)
+  mi_option_eager_commit,             // eager commit segments? (after `eager_commit_delay` segments) (=1)
+  mi_option_arena_eager_commit,       // eager commit arenas? Use 2 to enable just on overcommit systems (=2)
+  mi_option_purge_decommits,          // should a memory purge decommit (or only reset) (=1)
+  mi_option_allow_large_os_pages,     // allow large (2MiB) OS pages, implies eager commit
+  mi_option_reserve_huge_os_pages,    // reserve N huge OS pages (1GiB/page) at startup
+  mi_option_reserve_huge_os_pages_at, // reserve huge OS pages at a specific NUMA node
+  mi_option_reserve_os_memory,        // reserve specified amount of OS memory in an arena at startup
+  mi_option_deprecated_segment_cache,
+  mi_option_deprecated_page_reset,
+  mi_option_abandoned_page_purge,     // immediately purge delayed purges on thread termination
+  mi_option_deprecated_segment_reset,
+  mi_option_eager_commit_delay,
+  mi_option_purge_delay,              // memory purging is delayed by N milli seconds; use 0 for immediate purging or -1 for no purging at all.
+  mi_option_use_numa_nodes,           // 0 = use all available numa nodes, otherwise use at most N nodes.
+  mi_option_limit_os_alloc,           // 1 = do not use OS memory for allocation (but only programmatically reserved arenas)
+  mi_option_os_tag,                   // tag used for OS logging (macOS only for now)
+  mi_option_max_errors,               // issue at most N error messages
+  mi_option_max_warnings,             // issue at most N warning messages
+  mi_option_max_segment_reclaim,
+  mi_option_destroy_on_exit,          // if set, release all memory on exit; sometimes used for dynamic unloading but can be unsafe.
+  mi_option_arena_reserve,            // initial memory size in KiB for arena reservation (1GiB on 64-bit)
+  mi_option_arena_purge_mult,
+  mi_option_purge_extend_delay,
+  _mi_option_last,
+  // legacy option names
+  mi_option_large_os_pages = mi_option_allow_large_os_pages,
+  mi_option_eager_region_commit = mi_option_arena_eager_commit,
+  mi_option_reset_decommits = mi_option_purge_decommits,
+  mi_option_reset_delay = mi_option_purge_delay,
+  mi_option_abandoned_page_reset = mi_option_abandoned_page_purge
+} mi_option_t;
+
+
+mi_decl_nodiscard mi_decl_export bool mi_option_is_enabled(mi_option_t option);
+mi_decl_export void mi_option_enable(mi_option_t option);
+mi_decl_export void mi_option_disable(mi_option_t option);
+mi_decl_export void mi_option_set_enabled(mi_option_t option, bool enable);
+mi_decl_export void mi_option_set_enabled_default(mi_option_t option, bool enable);
+
+mi_decl_nodiscard mi_decl_export long   mi_option_get(mi_option_t option);
+mi_decl_nodiscard mi_decl_export long   mi_option_get_clamp(mi_option_t option, long min, long max);
+mi_decl_nodiscard mi_decl_export size_t mi_option_get_size(mi_option_t option);
+mi_decl_export void mi_option_set(mi_option_t option, long value);
+mi_decl_export void mi_option_set_default(mi_option_t option, long value);
+
+
+// -------------------------------------------------------------------------------------------------------
+// "mi" prefixed implementations of various posix, Unix, Windows, and C++ allocation functions.
+// (This can be convenient when providing overrides of these functions as done in `mimalloc-override.h`.)
+// note: we use `mi_cfree` as "checked free" and it checks if the pointer is in our heap before free-ing.
+// -------------------------------------------------------------------------------------------------------
+
+mi_decl_export void  mi_cfree(void* p) mi_attr_noexcept;
+mi_decl_export void* mi__expand(void* p, size_t newsize) mi_attr_noexcept;
+mi_decl_nodiscard mi_decl_export size_t mi_malloc_size(const void* p)        mi_attr_noexcept;
+mi_decl_nodiscard mi_decl_export size_t mi_malloc_good_size(size_t size)     mi_attr_noexcept;
+mi_decl_nodiscard mi_decl_export size_t mi_malloc_usable_size(const void *p) mi_attr_noexcept;
+
+mi_decl_export int mi_posix_memalign(void** p, size_t alignment, size_t size)   mi_attr_noexcept;
+mi_decl_nodiscard mi_decl_export mi_decl_restrict void* mi_memalign(size_t alignment, size_t size) mi_attr_noexcept mi_attr_malloc mi_attr_alloc_size(2) mi_attr_alloc_align(1);
+mi_decl_nodiscard mi_decl_export mi_decl_restrict void* mi_valloc(size_t size)  mi_attr_noexcept mi_attr_malloc mi_attr_alloc_size(1);
+mi_decl_nodiscard mi_decl_export mi_decl_restrict void* mi_pvalloc(size_t size) mi_attr_noexcept mi_attr_malloc mi_attr_alloc_size(1);
+mi_decl_nodiscard mi_decl_export mi_decl_restrict void* mi_aligned_alloc(size_t alignment, size_t size) mi_attr_noexcept mi_attr_malloc mi_attr_alloc_size(2) mi_attr_alloc_align(1);
+
+mi_decl_nodiscard mi_decl_export void* mi_reallocarray(void* p, size_t count, size_t size) mi_attr_noexcept mi_attr_alloc_size2(2,3);
+mi_decl_nodiscard mi_decl_export int   mi_reallocarr(void* p, size_t count, size_t size) mi_attr_noexcept;
+mi_decl_nodiscard mi_decl_export void* mi_aligned_recalloc(void* p, size_t newcount, size_t size, size_t alignment) mi_attr_noexcept;
+mi_decl_nodiscard mi_decl_export void* mi_aligned_offset_recalloc(void* p, size_t newcount, size_t size, size_t alignment, size_t offset) mi_attr_noexcept;
+
+mi_decl_nodiscard mi_decl_export mi_decl_restrict unsigned short* mi_wcsdup(const unsigned short* s) mi_attr_noexcept mi_attr_malloc;
+mi_decl_nodiscard mi_decl_export mi_decl_restrict unsigned char*  mi_mbsdup(const unsigned char* s)  mi_attr_noexcept mi_attr_malloc;
+mi_decl_export int mi_dupenv_s(char** buf, size_t* size, const char* name)                      mi_attr_noexcept;
+mi_decl_export int mi_wdupenv_s(unsigned short** buf, size_t* size, const unsigned short* name) mi_attr_noexcept;
+
+mi_decl_export void mi_free_size(void* p, size_t size)                           mi_attr_noexcept;
+mi_decl_export void mi_free_size_aligned(void* p, size_t size, size_t alignment) mi_attr_noexcept;
+mi_decl_export void mi_free_aligned(void* p, size_t alignment)                   mi_attr_noexcept;
+
+// The `mi_new` wrappers implement C++ semantics on out-of-memory instead of directly returning `NULL`.
+// (and call `std::get_new_handler` and potentially raise a `std::bad_alloc` exception).
+mi_decl_nodiscard mi_decl_export mi_decl_restrict void* mi_new(size_t size)                   mi_attr_malloc mi_attr_alloc_size(1);
+mi_decl_nodiscard mi_decl_export mi_decl_restrict void* mi_new_aligned(size_t size, size_t alignment) mi_attr_malloc mi_attr_alloc_size(1) mi_attr_alloc_align(2);
+mi_decl_nodiscard mi_decl_export mi_decl_restrict void* mi_new_nothrow(size_t size)           mi_attr_noexcept mi_attr_malloc mi_attr_alloc_size(1);
+mi_decl_nodiscard mi_decl_export mi_decl_restrict void* mi_new_aligned_nothrow(size_t size, size_t alignment) mi_attr_noexcept mi_attr_malloc mi_attr_alloc_size(1) mi_attr_alloc_align(2);
+mi_decl_nodiscard mi_decl_export mi_decl_restrict void* mi_new_n(size_t count, size_t size)   mi_attr_malloc mi_attr_alloc_size2(1, 2);
+mi_decl_nodiscard mi_decl_export void* mi_new_realloc(void* p, size_t newsize)                mi_attr_alloc_size(2);
+mi_decl_nodiscard mi_decl_export void* mi_new_reallocn(void* p, size_t newcount, size_t size) mi_attr_alloc_size2(2, 3);
+
+mi_decl_nodiscard mi_decl_export mi_decl_restrict void* mi_heap_alloc_new(mi_heap_t* heap, size_t size)                mi_attr_malloc mi_attr_alloc_size(2);
+mi_decl_nodiscard mi_decl_export mi_decl_restrict void* mi_heap_alloc_new_n(mi_heap_t* heap, size_t count, size_t size) mi_attr_malloc mi_attr_alloc_size2(2, 3);
+
+#ifdef __cplusplus
+}
+#endif
+
+// ---------------------------------------------------------------------------------------------
+// Implement the C++ std::allocator interface for use in STL containers.
+// (note: see `mimalloc-new-delete.h` for overriding the new/delete operators globally)
+// ---------------------------------------------------------------------------------------------
+#ifdef __cplusplus
+
+#include <cstddef>     // std::size_t
+#include <cstdint>     // PTRDIFF_MAX
+#if (__cplusplus >= 201103L) || (_MSC_VER > 1900)  // C++11
+#include <type_traits> // std::true_type
+#include <utility>     // std::forward
+#endif
+
+template<class T> struct _mi_stl_allocator_common {
+  typedef T                 value_type;
+  typedef std::size_t       size_type;
+  typedef std::ptrdiff_t    difference_type;
+  typedef value_type&       reference;
+  typedef value_type const& const_reference;
+  typedef value_type*       pointer;
+  typedef value_type const* const_pointer;
+
+  #if ((__cplusplus >= 201103L) || (_MSC_VER > 1900))  // C++11
+  using propagate_on_container_copy_assignment = std::true_type;
+  using propagate_on_container_move_assignment = std::true_type;
+  using propagate_on_container_swap            = std::true_type;
+  template <class U, class ...Args> void construct(U* p, Args&& ...args) { ::new(p) U(std::forward<Args>(args)...); }
+  template <class U> void destroy(U* p) mi_attr_noexcept { p->~U(); }
+  #else
+  void construct(pointer p, value_type const& val) { ::new(p) value_type(val); }
+  void destroy(pointer p) { p->~value_type(); }
+  #endif
+
+  size_type     max_size() const mi_attr_noexcept { return (PTRDIFF_MAX/sizeof(value_type)); }
+  pointer       address(reference x) const        { return &x; }
+  const_pointer address(const_reference x) const  { return &x; }
+};
+
+template<class T> struct mi_stl_allocator : public _mi_stl_allocator_common<T> {
+  using typename _mi_stl_allocator_common<T>::size_type;
+  using typename _mi_stl_allocator_common<T>::value_type;
+  using typename _mi_stl_allocator_common<T>::pointer;
+  template <class U> struct rebind { typedef mi_stl_allocator<U> other; };
+
+  mi_stl_allocator()                                             mi_attr_noexcept = default;
+  mi_stl_allocator(const mi_stl_allocator&)                      mi_attr_noexcept = default;
+  template<class U> mi_stl_allocator(const mi_stl_allocator<U>&) mi_attr_noexcept { }
+  mi_stl_allocator  select_on_container_copy_construction() const { return *this; }
+  void              deallocate(T* p, size_type) { mi_free(p); }
+
+  #if (__cplusplus >= 201703L)  // C++17
+  mi_decl_nodiscard T* allocate(size_type count) { return static_cast<T*>(mi_new_n(count, sizeof(T))); }
+  mi_decl_nodiscard T* allocate(size_type count, const void*) { return allocate(count); }
+  #else
+  mi_decl_nodiscard pointer allocate(size_type count, const void* = 0) { return static_cast<pointer>(mi_new_n(count, sizeof(value_type))); }
+  #endif
+
+  #if ((__cplusplus >= 201103L) || (_MSC_VER > 1900))  // C++11
+  using is_always_equal = std::true_type;
+  #endif
+};
+
+template<class T1,class T2> bool operator==(const mi_stl_allocator<T1>& , const mi_stl_allocator<T2>& ) mi_attr_noexcept { return true; }
+template<class T1,class T2> bool operator!=(const mi_stl_allocator<T1>& , const mi_stl_allocator<T2>& ) mi_attr_noexcept { return false; }
+
+
+#if (__cplusplus >= 201103L) || (_MSC_VER >= 1900)  // C++11
+#define MI_HAS_HEAP_STL_ALLOCATOR 1
+
+#include <memory>      // std::shared_ptr
+
+// Common base class for STL allocators in a specific heap
+template<class T, bool _mi_destroy> struct _mi_heap_stl_allocator_common : public _mi_stl_allocator_common<T> {
+  using typename _mi_stl_allocator_common<T>::size_type;
+  using typename _mi_stl_allocator_common<T>::value_type;
+  using typename _mi_stl_allocator_common<T>::pointer;
+
+  _mi_heap_stl_allocator_common(mi_heap_t* hp) : heap(hp) { }    /* will not delete nor destroy the passed in heap */
+
+  #if (__cplusplus >= 201703L)  // C++17
+  mi_decl_nodiscard T* allocate(size_type count) { return static_cast<T*>(mi_heap_alloc_new_n(this->heap.get(), count, sizeof(T))); }
+  mi_decl_nodiscard T* allocate(size_type count, const void*) { return allocate(count); }
+  #else
+  mi_decl_nodiscard pointer allocate(size_type count, const void* = 0) { return static_cast<pointer>(mi_heap_alloc_new_n(this->heap.get(), count, sizeof(value_type))); }
+  #endif
+
+  #if ((__cplusplus >= 201103L) || (_MSC_VER > 1900))  // C++11
+  using is_always_equal = std::false_type;
+  #endif
+
+  void collect(bool force) { mi_heap_collect(this->heap.get(), force); }
+  template<class U> bool is_equal(const _mi_heap_stl_allocator_common<U, _mi_destroy>& x) const { return (this->heap == x.heap); }
+
+protected:
+  std::shared_ptr<mi_heap_t> heap;
+  template<class U, bool D> friend struct _mi_heap_stl_allocator_common;
+
+  _mi_heap_stl_allocator_common() {
+    mi_heap_t* hp = mi_heap_new();
+    this->heap.reset(hp, (_mi_destroy ? &heap_destroy : &heap_delete));  /* calls heap_delete/destroy when the refcount drops to zero */
+  }
+  _mi_heap_stl_allocator_common(const _mi_heap_stl_allocator_common& x) mi_attr_noexcept : heap(x.heap) { }
+  template<class U> _mi_heap_stl_allocator_common(const _mi_heap_stl_allocator_common<U, _mi_destroy>& x) mi_attr_noexcept : heap(x.heap) { }
+
+private:
+  static void heap_delete(mi_heap_t* hp)  { if (hp != NULL) { mi_heap_delete(hp); } }
+  static void heap_destroy(mi_heap_t* hp) { if (hp != NULL) { mi_heap_destroy(hp); } }
+};
+
+// STL allocator allocation in a specific heap
+template<class T> struct mi_heap_stl_allocator : public _mi_heap_stl_allocator_common<T, false> {
+  using typename _mi_heap_stl_allocator_common<T, false>::size_type;
+  mi_heap_stl_allocator() : _mi_heap_stl_allocator_common<T, false>() { } // creates fresh heap that is deleted when the destructor is called
+  mi_heap_stl_allocator(mi_heap_t* hp) : _mi_heap_stl_allocator_common<T, false>(hp) { }  // no delete nor destroy on the passed in heap
+  template<class U> mi_heap_stl_allocator(const mi_heap_stl_allocator<U>& x) mi_attr_noexcept : _mi_heap_stl_allocator_common<T, false>(x) { }
+
+  mi_heap_stl_allocator select_on_container_copy_construction() const { return *this; }
+  void deallocate(T* p, size_type) { mi_free(p); }
+  template<class U> struct rebind { typedef mi_heap_stl_allocator<U> other; };
+};
+
+template<class T1, class T2> bool operator==(const mi_heap_stl_allocator<T1>& x, const mi_heap_stl_allocator<T2>& y) mi_attr_noexcept { return (x.is_equal(y)); }
+template<class T1, class T2> bool operator!=(const mi_heap_stl_allocator<T1>& x, const mi_heap_stl_allocator<T2>& y) mi_attr_noexcept { return (!x.is_equal(y)); }
+
+
+// STL allocator allocation in a specific heap, where `free` does nothing and
+// the heap is destroyed in one go on destruction -- use with care!
+template<class T> struct mi_heap_destroy_stl_allocator : public _mi_heap_stl_allocator_common<T, true> {
+  using typename _mi_heap_stl_allocator_common<T, true>::size_type;
+  mi_heap_destroy_stl_allocator() : _mi_heap_stl_allocator_common<T, true>() { } // creates fresh heap that is destroyed when the destructor is called
+  mi_heap_destroy_stl_allocator(mi_heap_t* hp) : _mi_heap_stl_allocator_common<T, true>(hp) { }  // no delete nor destroy on the passed in heap
+  template<class U> mi_heap_destroy_stl_allocator(const mi_heap_destroy_stl_allocator<U>& x) mi_attr_noexcept : _mi_heap_stl_allocator_common<T, true>(x) { }
+
+  mi_heap_destroy_stl_allocator select_on_container_copy_construction() const { return *this; }
+  void deallocate(T*, size_type) { /* do nothing as we destroy the heap on destruct. */ }
+  template<class U> struct rebind { typedef mi_heap_destroy_stl_allocator<U> other; };
+};
+
+template<class T1, class T2> bool operator==(const mi_heap_destroy_stl_allocator<T1>& x, const mi_heap_destroy_stl_allocator<T2>& y) mi_attr_noexcept { return (x.is_equal(y)); }
+template<class T1, class T2> bool operator!=(const mi_heap_destroy_stl_allocator<T1>& x, const mi_heap_destroy_stl_allocator<T2>& y) mi_attr_noexcept { return (!x.is_equal(y)); }
+
+#endif // C++11
+
+#endif // __cplusplus
+
+#endif

Dependencies/Python/include/internal/mimalloc/mimalloc/atomic.h

@@ -0,0 +1,392 @@
+/* ----------------------------------------------------------------------------
+Copyright (c) 2018-2023 Microsoft Research, Daan Leijen
+This is free software; you can redistribute it and/or modify it under the
+terms of the MIT license. A copy of the license can be found in the file
+"LICENSE" at the root of this distribution.
+-----------------------------------------------------------------------------*/
+#pragma once
+#ifndef MIMALLOC_ATOMIC_H
+#define MIMALLOC_ATOMIC_H
+
+// --------------------------------------------------------------------------------------------
+// Atomics
+// We need to be portable between C, C++, and MSVC.
+// We base the primitives on the C/C++ atomics and create a mimimal wrapper for MSVC in C compilation mode.
+// This is why we try to use only `uintptr_t` and `<type>*` as atomic types.
+// To gain better insight in the range of used atomics, we use explicitly named memory order operations
+// instead of passing the memory order as a parameter.
+// -----------------------------------------------------------------------------------------------
+
+#if defined(__cplusplus)
+// Use C++ atomics
+#include <atomic>
+#define  _Atomic(tp)            std::atomic<tp>
+#define  mi_atomic(name)        std::atomic_##name
+#define  mi_memory_order(name)  std::memory_order_##name
+#if (__cplusplus >= 202002L)    // c++20, see issue #571
+ #define MI_ATOMIC_VAR_INIT(x)  x
+#elif !defined(ATOMIC_VAR_INIT)
+ #define MI_ATOMIC_VAR_INIT(x)  x
+#else
+ #define MI_ATOMIC_VAR_INIT(x)  ATOMIC_VAR_INIT(x)
+#endif
+#elif defined(_MSC_VER)
+// Use MSVC C wrapper for C11 atomics
+#define  _Atomic(tp)            tp
+#define  MI_ATOMIC_VAR_INIT(x)  x
+#define  mi_atomic(name)        mi_atomic_##name
+#define  mi_memory_order(name)  mi_memory_order_##name
+#else
+// Use C11 atomics
+#include <stdatomic.h>
+#define  mi_atomic(name)        atomic_##name
+#define  mi_memory_order(name)  memory_order_##name
+#if (__STDC_VERSION__ >= 201710L) // c17, see issue #735
+ #define MI_ATOMIC_VAR_INIT(x) x
+#elif !defined(ATOMIC_VAR_INIT)
+ #define MI_ATOMIC_VAR_INIT(x) x
+#else
+ #define MI_ATOMIC_VAR_INIT(x) ATOMIC_VAR_INIT(x)
+#endif
+#endif
+
+// Various defines for all used memory orders in mimalloc
+#define mi_atomic_cas_weak(p,expected,desired,mem_success,mem_fail)  \
+  mi_atomic(compare_exchange_weak_explicit)(p,expected,desired,mem_success,mem_fail)
+
+#define mi_atomic_cas_strong(p,expected,desired,mem_success,mem_fail)  \
+  mi_atomic(compare_exchange_strong_explicit)(p,expected,desired,mem_success,mem_fail)
+
+#define mi_atomic_load_acquire(p)                mi_atomic(load_explicit)(p,mi_memory_order(acquire))
+#define mi_atomic_load_relaxed(p)                mi_atomic(load_explicit)(p,mi_memory_order(relaxed))
+#define mi_atomic_store_release(p,x)             mi_atomic(store_explicit)(p,x,mi_memory_order(release))
+#define mi_atomic_store_relaxed(p,x)             mi_atomic(store_explicit)(p,x,mi_memory_order(relaxed))
+#define mi_atomic_exchange_release(p,x)          mi_atomic(exchange_explicit)(p,x,mi_memory_order(release))
+#define mi_atomic_exchange_acq_rel(p,x)          mi_atomic(exchange_explicit)(p,x,mi_memory_order(acq_rel))
+#define mi_atomic_cas_weak_release(p,exp,des)    mi_atomic_cas_weak(p,exp,des,mi_memory_order(release),mi_memory_order(relaxed))
+#define mi_atomic_cas_weak_acq_rel(p,exp,des)    mi_atomic_cas_weak(p,exp,des,mi_memory_order(acq_rel),mi_memory_order(acquire))
+#define mi_atomic_cas_strong_release(p,exp,des)  mi_atomic_cas_strong(p,exp,des,mi_memory_order(release),mi_memory_order(relaxed))
+#define mi_atomic_cas_strong_acq_rel(p,exp,des)  mi_atomic_cas_strong(p,exp,des,mi_memory_order(acq_rel),mi_memory_order(acquire))
+
+#define mi_atomic_add_relaxed(p,x)               mi_atomic(fetch_add_explicit)(p,x,mi_memory_order(relaxed))
+#define mi_atomic_sub_relaxed(p,x)               mi_atomic(fetch_sub_explicit)(p,x,mi_memory_order(relaxed))
+#define mi_atomic_add_acq_rel(p,x)               mi_atomic(fetch_add_explicit)(p,x,mi_memory_order(acq_rel))
+#define mi_atomic_sub_acq_rel(p,x)               mi_atomic(fetch_sub_explicit)(p,x,mi_memory_order(acq_rel))
+#define mi_atomic_and_acq_rel(p,x)               mi_atomic(fetch_and_explicit)(p,x,mi_memory_order(acq_rel))
+#define mi_atomic_or_acq_rel(p,x)                mi_atomic(fetch_or_explicit)(p,x,mi_memory_order(acq_rel))
+
+#define mi_atomic_increment_relaxed(p)           mi_atomic_add_relaxed(p,(uintptr_t)1)
+#define mi_atomic_decrement_relaxed(p)           mi_atomic_sub_relaxed(p,(uintptr_t)1)
+#define mi_atomic_increment_acq_rel(p)           mi_atomic_add_acq_rel(p,(uintptr_t)1)
+#define mi_atomic_decrement_acq_rel(p)           mi_atomic_sub_acq_rel(p,(uintptr_t)1)
+
+static inline void mi_atomic_yield(void);
+static inline intptr_t mi_atomic_addi(_Atomic(intptr_t)*p, intptr_t add);
+static inline intptr_t mi_atomic_subi(_Atomic(intptr_t)*p, intptr_t sub);
+
+
+#if defined(__cplusplus) || !defined(_MSC_VER)
+
+// In C++/C11 atomics we have polymorphic atomics so can use the typed `ptr` variants (where `tp` is the type of atomic value)
+// We use these macros so we can provide a typed wrapper in MSVC in C compilation mode as well
+#define mi_atomic_load_ptr_acquire(tp,p)                mi_atomic_load_acquire(p)
+#define mi_atomic_load_ptr_relaxed(tp,p)                mi_atomic_load_relaxed(p)
+
+// In C++ we need to add casts to help resolve templates if NULL is passed
+#if defined(__cplusplus)
+#define mi_atomic_store_ptr_release(tp,p,x)             mi_atomic_store_release(p,(tp*)x)
+#define mi_atomic_store_ptr_relaxed(tp,p,x)             mi_atomic_store_relaxed(p,(tp*)x)
+#define mi_atomic_cas_ptr_weak_release(tp,p,exp,des)    mi_atomic_cas_weak_release(p,exp,(tp*)des)
+#define mi_atomic_cas_ptr_weak_acq_rel(tp,p,exp,des)    mi_atomic_cas_weak_acq_rel(p,exp,(tp*)des)
+#define mi_atomic_cas_ptr_strong_release(tp,p,exp,des)  mi_atomic_cas_strong_release(p,exp,(tp*)des)
+#define mi_atomic_exchange_ptr_release(tp,p,x)          mi_atomic_exchange_release(p,(tp*)x)
+#define mi_atomic_exchange_ptr_acq_rel(tp,p,x)          mi_atomic_exchange_acq_rel(p,(tp*)x)
+#else
+#define mi_atomic_store_ptr_release(tp,p,x)             mi_atomic_store_release(p,x)
+#define mi_atomic_store_ptr_relaxed(tp,p,x)             mi_atomic_store_relaxed(p,x)
+#define mi_atomic_cas_ptr_weak_release(tp,p,exp,des)    mi_atomic_cas_weak_release(p,exp,des)
+#define mi_atomic_cas_ptr_weak_acq_rel(tp,p,exp,des)    mi_atomic_cas_weak_acq_rel(p,exp,des)
+#define mi_atomic_cas_ptr_strong_release(tp,p,exp,des)  mi_atomic_cas_strong_release(p,exp,des)
+#define mi_atomic_exchange_ptr_release(tp,p,x)          mi_atomic_exchange_release(p,x)
+#define mi_atomic_exchange_ptr_acq_rel(tp,p,x)          mi_atomic_exchange_acq_rel(p,x)
+#endif
+
+// These are used by the statistics
+static inline int64_t mi_atomic_addi64_relaxed(volatile int64_t* p, int64_t add) {
+  return mi_atomic(fetch_add_explicit)((_Atomic(int64_t)*)p, add, mi_memory_order(relaxed));
+}
+static inline void mi_atomic_maxi64_relaxed(volatile int64_t* p, int64_t x) {
+  int64_t current = mi_atomic_load_relaxed((_Atomic(int64_t)*)p);
+  while (current < x && !mi_atomic_cas_weak_release((_Atomic(int64_t)*)p, &current, x)) { /* nothing */ };
+}
+
+// Used by timers
+#define mi_atomic_loadi64_acquire(p)            mi_atomic(load_explicit)(p,mi_memory_order(acquire))
+#define mi_atomic_loadi64_relaxed(p)            mi_atomic(load_explicit)(p,mi_memory_order(relaxed))
+#define mi_atomic_storei64_release(p,x)         mi_atomic(store_explicit)(p,x,mi_memory_order(release))
+#define mi_atomic_storei64_relaxed(p,x)         mi_atomic(store_explicit)(p,x,mi_memory_order(relaxed))
+
+#define mi_atomic_casi64_strong_acq_rel(p,e,d)  mi_atomic_cas_strong_acq_rel(p,e,d)
+#define mi_atomic_addi64_acq_rel(p,i)           mi_atomic_add_acq_rel(p,i)
+
+
+#elif defined(_MSC_VER)
+
+// MSVC C compilation wrapper that uses Interlocked operations to model C11 atomics.
+#define WIN32_LEAN_AND_MEAN
+#include <windows.h>
+#include <intrin.h>
+#ifdef _WIN64
+typedef LONG64   msc_intptr_t;
+#define MI_64(f) f##64
+#else
+typedef LONG     msc_intptr_t;
+#define MI_64(f) f
+#endif
+
+typedef enum mi_memory_order_e {
+  mi_memory_order_relaxed,
+  mi_memory_order_consume,
+  mi_memory_order_acquire,
+  mi_memory_order_release,
+  mi_memory_order_acq_rel,
+  mi_memory_order_seq_cst
+} mi_memory_order;
+
+static inline uintptr_t mi_atomic_fetch_add_explicit(_Atomic(uintptr_t)*p, uintptr_t add, mi_memory_order mo) {
+  (void)(mo);
+  return (uintptr_t)MI_64(_InterlockedExchangeAdd)((volatile msc_intptr_t*)p, (msc_intptr_t)add);
+}
+static inline uintptr_t mi_atomic_fetch_sub_explicit(_Atomic(uintptr_t)*p, uintptr_t sub, mi_memory_order mo) {
+  (void)(mo);
+  return (uintptr_t)MI_64(_InterlockedExchangeAdd)((volatile msc_intptr_t*)p, -((msc_intptr_t)sub));
+}
+static inline uintptr_t mi_atomic_fetch_and_explicit(_Atomic(uintptr_t)*p, uintptr_t x, mi_memory_order mo) {
+  (void)(mo);
+  return (uintptr_t)MI_64(_InterlockedAnd)((volatile msc_intptr_t*)p, (msc_intptr_t)x);
+}
+static inline uintptr_t mi_atomic_fetch_or_explicit(_Atomic(uintptr_t)*p, uintptr_t x, mi_memory_order mo) {
+  (void)(mo);
+  return (uintptr_t)MI_64(_InterlockedOr)((volatile msc_intptr_t*)p, (msc_intptr_t)x);
+}
+static inline bool mi_atomic_compare_exchange_strong_explicit(_Atomic(uintptr_t)*p, uintptr_t* expected, uintptr_t desired, mi_memory_order mo1, mi_memory_order mo2) {
+  (void)(mo1); (void)(mo2);
+  uintptr_t read = (uintptr_t)MI_64(_InterlockedCompareExchange)((volatile msc_intptr_t*)p, (msc_intptr_t)desired, (msc_intptr_t)(*expected));
+  if (read == *expected) {
+    return true;
+  }
+  else {
+    *expected = read;
+    return false;
+  }
+}
+static inline bool mi_atomic_compare_exchange_weak_explicit(_Atomic(uintptr_t)*p, uintptr_t* expected, uintptr_t desired, mi_memory_order mo1, mi_memory_order mo2) {
+  return mi_atomic_compare_exchange_strong_explicit(p, expected, desired, mo1, mo2);
+}
+static inline uintptr_t mi_atomic_exchange_explicit(_Atomic(uintptr_t)*p, uintptr_t exchange, mi_memory_order mo) {
+  (void)(mo);
+  return (uintptr_t)MI_64(_InterlockedExchange)((volatile msc_intptr_t*)p, (msc_intptr_t)exchange);
+}
+static inline void mi_atomic_thread_fence(mi_memory_order mo) {
+  (void)(mo);
+  _Atomic(uintptr_t) x = 0;
+  mi_atomic_exchange_explicit(&x, 1, mo);
+}
+static inline uintptr_t mi_atomic_load_explicit(_Atomic(uintptr_t) const* p, mi_memory_order mo) {
+  (void)(mo);
+#if defined(_M_IX86) || defined(_M_X64)
+  return *p;
+#else
+  uintptr_t x = *p;
+  if (mo > mi_memory_order_relaxed) {
+    while (!mi_atomic_compare_exchange_weak_explicit((_Atomic(uintptr_t)*)p, &x, x, mo, mi_memory_order_relaxed)) { /* nothing */ };
+  }
+  return x;
+#endif
+}
+static inline void mi_atomic_store_explicit(_Atomic(uintptr_t)*p, uintptr_t x, mi_memory_order mo) {
+  (void)(mo);
+#if defined(_M_IX86) || defined(_M_X64)
+  *p = x;
+#else
+  mi_atomic_exchange_explicit(p, x, mo);
+#endif
+}
+static inline int64_t mi_atomic_loadi64_explicit(_Atomic(int64_t)*p, mi_memory_order mo) {
+  (void)(mo);
+#if defined(_M_X64)
+  return *p;
+#else
+  int64_t old = *p;
+  int64_t x = old;
+  while ((old = InterlockedCompareExchange64(p, x, old)) != x) {
+    x = old;
+  }
+  return x;
+#endif
+}
+static inline void mi_atomic_storei64_explicit(_Atomic(int64_t)*p, int64_t x, mi_memory_order mo) {
+  (void)(mo);
+#if defined(x_M_IX86) || defined(_M_X64)
+  *p = x;
+#else
+  InterlockedExchange64(p, x);
+#endif
+}
+
+// These are used by the statistics
+static inline int64_t mi_atomic_addi64_relaxed(volatile _Atomic(int64_t)*p, int64_t add) {
+#ifdef _WIN64
+  return (int64_t)mi_atomic_addi((int64_t*)p, add);
+#else
+  int64_t current;
+  int64_t sum;
+  do {
+    current = *p;
+    sum = current + add;
+  } while (_InterlockedCompareExchange64(p, sum, current) != current);
+  return current;
+#endif
+}
+static inline void mi_atomic_maxi64_relaxed(volatile _Atomic(int64_t)*p, int64_t x) {
+  int64_t current;
+  do {
+    current = *p;
+  } while (current < x && _InterlockedCompareExchange64(p, x, current) != current);
+}
+
+static inline void mi_atomic_addi64_acq_rel(volatile _Atomic(int64_t*)p, int64_t i) {
+  mi_atomic_addi64_relaxed(p, i);
+}
+
+static inline bool mi_atomic_casi64_strong_acq_rel(volatile _Atomic(int64_t*)p, int64_t* exp, int64_t des) {
+  int64_t read = _InterlockedCompareExchange64(p, des, *exp);
+  if (read == *exp) {
+    return true;
+  }
+  else {
+    *exp = read;
+    return false;
+  }
+}
+
+// The pointer macros cast to `uintptr_t`.
+#define mi_atomic_load_ptr_acquire(tp,p)                (tp*)mi_atomic_load_acquire((_Atomic(uintptr_t)*)(p))
+#define mi_atomic_load_ptr_relaxed(tp,p)                (tp*)mi_atomic_load_relaxed((_Atomic(uintptr_t)*)(p))
+#define mi_atomic_store_ptr_release(tp,p,x)             mi_atomic_store_release((_Atomic(uintptr_t)*)(p),(uintptr_t)(x))
+#define mi_atomic_store_ptr_relaxed(tp,p,x)             mi_atomic_store_relaxed((_Atomic(uintptr_t)*)(p),(uintptr_t)(x))
+#define mi_atomic_cas_ptr_weak_release(tp,p,exp,des)    mi_atomic_cas_weak_release((_Atomic(uintptr_t)*)(p),(uintptr_t*)exp,(uintptr_t)des)
+#define mi_atomic_cas_ptr_weak_acq_rel(tp,p,exp,des)    mi_atomic_cas_weak_acq_rel((_Atomic(uintptr_t)*)(p),(uintptr_t*)exp,(uintptr_t)des)
+#define mi_atomic_cas_ptr_strong_release(tp,p,exp,des)  mi_atomic_cas_strong_release((_Atomic(uintptr_t)*)(p),(uintptr_t*)exp,(uintptr_t)des)
+#define mi_atomic_exchange_ptr_release(tp,p,x)          (tp*)mi_atomic_exchange_release((_Atomic(uintptr_t)*)(p),(uintptr_t)x)
+#define mi_atomic_exchange_ptr_acq_rel(tp,p,x)          (tp*)mi_atomic_exchange_acq_rel((_Atomic(uintptr_t)*)(p),(uintptr_t)x)
+
+#define mi_atomic_loadi64_acquire(p)    mi_atomic(loadi64_explicit)(p,mi_memory_order(acquire))
+#define mi_atomic_loadi64_relaxed(p)    mi_atomic(loadi64_explicit)(p,mi_memory_order(relaxed))
+#define mi_atomic_storei64_release(p,x) mi_atomic(storei64_explicit)(p,x,mi_memory_order(release))
+#define mi_atomic_storei64_relaxed(p,x) mi_atomic(storei64_explicit)(p,x,mi_memory_order(relaxed))
+
+
+#endif
+
+
+// Atomically add a signed value; returns the previous value.
+static inline intptr_t mi_atomic_addi(_Atomic(intptr_t)*p, intptr_t add) {
+  return (intptr_t)mi_atomic_add_acq_rel((_Atomic(uintptr_t)*)p, (uintptr_t)add);
+}
+
+// Atomically subtract a signed value; returns the previous value.
+static inline intptr_t mi_atomic_subi(_Atomic(intptr_t)*p, intptr_t sub) {
+  return (intptr_t)mi_atomic_addi(p, -sub);
+}
+
+typedef _Atomic(uintptr_t) mi_atomic_once_t;
+
+// Returns true only on the first invocation
+static inline bool mi_atomic_once( mi_atomic_once_t* once ) {
+  if (mi_atomic_load_relaxed(once) != 0) return false;     // quick test
+  uintptr_t expected = 0;
+  return mi_atomic_cas_strong_acq_rel(once, &expected, (uintptr_t)1); // try to set to 1
+}
+
+typedef _Atomic(uintptr_t) mi_atomic_guard_t;
+
+// Allows only one thread to execute at a time
+#define mi_atomic_guard(guard) \
+  uintptr_t _mi_guard_expected = 0; \
+  for(bool _mi_guard_once = true; \
+      _mi_guard_once && mi_atomic_cas_strong_acq_rel(guard,&_mi_guard_expected,(uintptr_t)1); \
+      (mi_atomic_store_release(guard,(uintptr_t)0), _mi_guard_once = false) )
+
+
+
+// Yield
+#if defined(__cplusplus)
+#include <thread>
+static inline void mi_atomic_yield(void) {
+  std::this_thread::yield();
+}
+#elif defined(_WIN32)
+#define WIN32_LEAN_AND_MEAN
+#include <windows.h>
+static inline void mi_atomic_yield(void) {
+  YieldProcessor();
+}
+#elif defined(__SSE2__)
+#include <emmintrin.h>
+static inline void mi_atomic_yield(void) {
+  _mm_pause();
+}
+#elif (defined(__GNUC__) || defined(__clang__)) && \
+      (defined(__x86_64__) || defined(__i386__) || \
+       defined(__aarch64__) || defined(__arm__) || \
+       defined(__powerpc__) || defined(__ppc__) || defined(__PPC__) || defined(__POWERPC__))
+#if defined(__x86_64__) || defined(__i386__)
+static inline void mi_atomic_yield(void) {
+  __asm__ volatile ("pause" ::: "memory");
+}
+#elif defined(__aarch64__)
+static inline void mi_atomic_yield(void) {
+  __asm__ volatile("wfe");
+}
+#elif defined(__arm__)
+#if __ARM_ARCH >= 7
+static inline void mi_atomic_yield(void) {
+  __asm__ volatile("yield" ::: "memory");
+}
+#else
+static inline void mi_atomic_yield(void) {
+  __asm__ volatile ("nop" ::: "memory");
+}
+#endif
+#elif defined(__powerpc__) || defined(__ppc__) || defined(__PPC__) || defined(__POWERPC__)
+#ifdef __APPLE__
+static inline void mi_atomic_yield(void) {
+  __asm__ volatile ("or r27,r27,r27" ::: "memory");
+}
+#else
+static inline void mi_atomic_yield(void) {
+  __asm__ __volatile__ ("or 27,27,27" ::: "memory");
+}
+#endif
+#endif
+#elif defined(__sun)
+// Fallback for other archs
+#include <synch.h>
+static inline void mi_atomic_yield(void) {
+  smt_pause();
+}
+#elif defined(__wasi__)
+#include <sched.h>
+static inline void mi_atomic_yield(void) {
+  sched_yield();
+}
+#else
+#include <unistd.h>
+static inline void mi_atomic_yield(void) {
+  sleep(0);
+}
+#endif
+
+
+#endif // __MIMALLOC_ATOMIC_H

Dependencies/Python/include/internal/mimalloc/mimalloc/internal.h

@@ -0,0 +1,969 @@
+/* ----------------------------------------------------------------------------
+Copyright (c) 2018-2023, Microsoft Research, Daan Leijen
+This is free software; you can redistribute it and/or modify it under the
+terms of the MIT license. A copy of the license can be found in the file
+"LICENSE" at the root of this distribution.
+-----------------------------------------------------------------------------*/
+#pragma once
+#ifndef MIMALLOC_INTERNAL_H
+#define MIMALLOC_INTERNAL_H
+
+
+// --------------------------------------------------------------------------
+// This file contains the interal API's of mimalloc and various utility
+// functions and macros.
+// --------------------------------------------------------------------------
+
+#include "types.h"
+#include "track.h"
+
+#if (MI_DEBUG>0)
+#define mi_trace_message(...)  _mi_trace_message(__VA_ARGS__)
+#else
+#define mi_trace_message(...)
+#endif
+
+#if defined(__EMSCRIPTEN__) && !defined(__wasi__)
+#define __wasi__
+#endif
+
+#if defined(__cplusplus)
+#define mi_decl_externc       extern "C"
+#else
+#define mi_decl_externc
+#endif
+
+// pthreads
+#if !defined(_WIN32) && !defined(__wasi__)
+#define  MI_USE_PTHREADS
+#include <pthread.h>
+#endif
+
+// "options.c"
+void       _mi_fputs(mi_output_fun* out, void* arg, const char* prefix, const char* message);
+void       _mi_fprintf(mi_output_fun* out, void* arg, const char* fmt, ...);
+void       _mi_warning_message(const char* fmt, ...);
+void       _mi_verbose_message(const char* fmt, ...);
+void       _mi_trace_message(const char* fmt, ...);
+void       _mi_options_init(void);
+void       _mi_error_message(int err, const char* fmt, ...);
+
+// random.c
+void       _mi_random_init(mi_random_ctx_t* ctx);
+void       _mi_random_init_weak(mi_random_ctx_t* ctx);
+void       _mi_random_reinit_if_weak(mi_random_ctx_t * ctx);
+void       _mi_random_split(mi_random_ctx_t* ctx, mi_random_ctx_t* new_ctx);
+uintptr_t  _mi_random_next(mi_random_ctx_t* ctx);
+uintptr_t  _mi_heap_random_next(mi_heap_t* heap);
+uintptr_t  _mi_os_random_weak(uintptr_t extra_seed);
+static inline uintptr_t _mi_random_shuffle(uintptr_t x);
+
+// init.c
+extern mi_decl_cache_align mi_stats_t       _mi_stats_main;
+extern mi_decl_cache_align const mi_page_t  _mi_page_empty;
+bool       _mi_is_main_thread(void);
+size_t     _mi_current_thread_count(void);
+bool       _mi_preloading(void);           // true while the C runtime is not initialized yet
+mi_threadid_t _mi_thread_id(void) mi_attr_noexcept;
+mi_heap_t*    _mi_heap_main_get(void);     // statically allocated main backing heap
+void       _mi_thread_done(mi_heap_t* heap);
+void       _mi_thread_data_collect(void);
+void       _mi_tld_init(mi_tld_t* tld, mi_heap_t* bheap);
+
+// os.c
+void       _mi_os_init(void);                                            // called from process init
+void*      _mi_os_alloc(size_t size, mi_memid_t* memid, mi_stats_t* stats);
+void       _mi_os_free(void* p, size_t size, mi_memid_t memid, mi_stats_t* stats);
+void       _mi_os_free_ex(void* p, size_t size, bool still_committed, mi_memid_t memid, mi_stats_t* stats);
+
+size_t     _mi_os_page_size(void);
+size_t     _mi_os_good_alloc_size(size_t size);
+bool       _mi_os_has_overcommit(void);
+bool       _mi_os_has_virtual_reserve(void);
+
+bool       _mi_os_purge(void* p, size_t size, mi_stats_t* stats);
+bool       _mi_os_reset(void* addr, size_t size, mi_stats_t* tld_stats);
+bool       _mi_os_commit(void* p, size_t size, bool* is_zero, mi_stats_t* stats);
+bool       _mi_os_decommit(void* addr, size_t size, mi_stats_t* stats);
+bool       _mi_os_protect(void* addr, size_t size);
+bool       _mi_os_unprotect(void* addr, size_t size);
+bool       _mi_os_purge(void* p, size_t size, mi_stats_t* stats);
+bool       _mi_os_purge_ex(void* p, size_t size, bool allow_reset, mi_stats_t* stats);
+
+void*      _mi_os_alloc_aligned(size_t size, size_t alignment, bool commit, bool allow_large, mi_memid_t* memid, mi_stats_t* stats);
+void*      _mi_os_alloc_aligned_at_offset(size_t size, size_t alignment, size_t align_offset, bool commit, bool allow_large, mi_memid_t* memid, mi_stats_t* tld_stats);
+
+void*      _mi_os_get_aligned_hint(size_t try_alignment, size_t size);
+bool       _mi_os_use_large_page(size_t size, size_t alignment);
+size_t     _mi_os_large_page_size(void);
+
+void*      _mi_os_alloc_huge_os_pages(size_t pages, int numa_node, mi_msecs_t max_secs, size_t* pages_reserved, size_t* psize, mi_memid_t* memid);
+
+// arena.c
+mi_arena_id_t _mi_arena_id_none(void);
+void       _mi_arena_free(void* p, size_t size, size_t still_committed_size, mi_memid_t memid, mi_stats_t* stats);
+void*      _mi_arena_alloc(size_t size, bool commit, bool allow_large, mi_arena_id_t req_arena_id, mi_memid_t* memid, mi_os_tld_t* tld);
+void*      _mi_arena_alloc_aligned(size_t size, size_t alignment, size_t align_offset, bool commit, bool allow_large, mi_arena_id_t req_arena_id, mi_memid_t* memid, mi_os_tld_t* tld);
+bool       _mi_arena_memid_is_suitable(mi_memid_t memid, mi_arena_id_t request_arena_id);
+bool       _mi_arena_contains(const void* p);
+void       _mi_arena_collect(bool force_purge, mi_stats_t* stats);
+void       _mi_arena_unsafe_destroy_all(mi_stats_t* stats);
+
+// "segment-map.c"
+void       _mi_segment_map_allocated_at(const mi_segment_t* segment);
+void       _mi_segment_map_freed_at(const mi_segment_t* segment);
+
+// "segment.c"
+extern mi_abandoned_pool_t _mi_abandoned_default;  // global abandoned pool
+mi_page_t* _mi_segment_page_alloc(mi_heap_t* heap, size_t block_size, size_t page_alignment, mi_segments_tld_t* tld, mi_os_tld_t* os_tld);
+void       _mi_segment_page_free(mi_page_t* page, bool force, mi_segments_tld_t* tld);
+void       _mi_segment_page_abandon(mi_page_t* page, mi_segments_tld_t* tld);
+bool       _mi_segment_try_reclaim_abandoned( mi_heap_t* heap, bool try_all, mi_segments_tld_t* tld);
+void       _mi_segment_thread_collect(mi_segments_tld_t* tld);
+bool       _mi_abandoned_pool_visit_blocks(mi_abandoned_pool_t* pool, uint8_t page_tag, bool visit_blocks, mi_block_visit_fun* visitor, void* arg);
+
+
+#if MI_HUGE_PAGE_ABANDON
+void       _mi_segment_huge_page_free(mi_segment_t* segment, mi_page_t* page, mi_block_t* block);
+#else
+void       _mi_segment_huge_page_reset(mi_segment_t* segment, mi_page_t* page, mi_block_t* block);
+#endif
+
+uint8_t*   _mi_segment_page_start(const mi_segment_t* segment, const mi_page_t* page, size_t* page_size); // page start for any page
+void       _mi_abandoned_reclaim_all(mi_heap_t* heap, mi_segments_tld_t* tld);
+void       _mi_abandoned_await_readers(mi_abandoned_pool_t *pool);
+void       _mi_abandoned_collect(mi_heap_t* heap, bool force, mi_segments_tld_t* tld);
+
+// "page.c"
+void*      _mi_malloc_generic(mi_heap_t* heap, size_t size, bool zero, size_t huge_alignment)  mi_attr_noexcept mi_attr_malloc;
+
+void       _mi_page_retire(mi_page_t* page) mi_attr_noexcept;                  // free the page if there are no other pages with many free blocks
+void       _mi_page_unfull(mi_page_t* page);
+void       _mi_page_free(mi_page_t* page, mi_page_queue_t* pq, bool force);   // free the page
+void       _mi_page_abandon(mi_page_t* page, mi_page_queue_t* pq);            // abandon the page, to be picked up by another thread...
+void       _mi_heap_delayed_free_all(mi_heap_t* heap);
+bool       _mi_heap_delayed_free_partial(mi_heap_t* heap);
+void       _mi_heap_collect_retired(mi_heap_t* heap, bool force);
+
+void       _mi_page_use_delayed_free(mi_page_t* page, mi_delayed_t delay, bool override_never);
+bool       _mi_page_try_use_delayed_free(mi_page_t* page, mi_delayed_t delay, bool override_never);
+size_t     _mi_page_queue_append(mi_heap_t* heap, mi_page_queue_t* pq, mi_page_queue_t* append);
+void       _mi_deferred_free(mi_heap_t* heap, bool force);
+
+void       _mi_page_free_collect(mi_page_t* page,bool force);
+void       _mi_page_reclaim(mi_heap_t* heap, mi_page_t* page);   // callback from segments
+
+size_t     _mi_bin_size(uint8_t bin);           // for stats
+uint8_t    _mi_bin(size_t size);                // for stats
+
+// "heap.c"
+void       _mi_heap_init_ex(mi_heap_t* heap, mi_tld_t* tld, mi_arena_id_t arena_id, bool no_reclaim, uint8_t tag);
+void       _mi_heap_destroy_pages(mi_heap_t* heap);
+void       _mi_heap_collect_abandon(mi_heap_t* heap);
+void       _mi_heap_set_default_direct(mi_heap_t* heap);
+bool       _mi_heap_memid_is_suitable(mi_heap_t* heap, mi_memid_t memid);
+void       _mi_heap_unsafe_destroy_all(void);
+void       _mi_heap_area_init(mi_heap_area_t* area, mi_page_t* page);
+bool       _mi_heap_area_visit_blocks(const mi_heap_area_t* area, mi_page_t *page, mi_block_visit_fun* visitor, void* arg);
+
+// "stats.c"
+void       _mi_stats_done(mi_stats_t* stats);
+mi_msecs_t  _mi_clock_now(void);
+mi_msecs_t  _mi_clock_end(mi_msecs_t start);
+mi_msecs_t  _mi_clock_start(void);
+
+// "alloc.c"
+void*       _mi_page_malloc(mi_heap_t* heap, mi_page_t* page, size_t size, bool zero) mi_attr_noexcept;  // called from `_mi_malloc_generic`
+void*       _mi_heap_malloc_zero(mi_heap_t* heap, size_t size, bool zero) mi_attr_noexcept;
+void*       _mi_heap_malloc_zero_ex(mi_heap_t* heap, size_t size, bool zero, size_t huge_alignment) mi_attr_noexcept;     // called from `_mi_heap_malloc_aligned`
+void*       _mi_heap_realloc_zero(mi_heap_t* heap, void* p, size_t newsize, bool zero) mi_attr_noexcept;
+mi_block_t* _mi_page_ptr_unalign(const mi_segment_t* segment, const mi_page_t* page, const void* p);
+bool        _mi_free_delayed_block(mi_block_t* block);
+void        _mi_free_generic(const mi_segment_t* segment, mi_page_t* page, bool is_local, void* p) mi_attr_noexcept;  // for runtime integration
+void        _mi_padding_shrink(const mi_page_t* page, const mi_block_t* block, const size_t min_size);
+
+// option.c, c primitives
+char        _mi_toupper(char c);
+int         _mi_strnicmp(const char* s, const char* t, size_t n);
+void        _mi_strlcpy(char* dest, const char* src, size_t dest_size);
+void        _mi_strlcat(char* dest, const char* src, size_t dest_size);
+size_t      _mi_strlen(const char* s);
+size_t      _mi_strnlen(const char* s, size_t max_len);
+
+
+#if MI_DEBUG>1
+bool        _mi_page_is_valid(mi_page_t* page);
+#endif
+
+
+// ------------------------------------------------------
+// Branches
+// ------------------------------------------------------
+
+#if defined(__GNUC__) || defined(__clang__)
+#define mi_unlikely(x)     (__builtin_expect(!!(x),false))
+#define mi_likely(x)       (__builtin_expect(!!(x),true))
+#elif (defined(__cplusplus) && (__cplusplus >= 202002L)) || (defined(_MSVC_LANG) && _MSVC_LANG >= 202002L)
+#define mi_unlikely(x)     (x) [[unlikely]]
+#define mi_likely(x)       (x) [[likely]]
+#else
+#define mi_unlikely(x)     (x)
+#define mi_likely(x)       (x)
+#endif
+
+#ifndef __has_builtin
+#define __has_builtin(x)  0
+#endif
+
+
+/* -----------------------------------------------------------
+  Error codes passed to `_mi_fatal_error`
+  All are recoverable but EFAULT is a serious error and aborts by default in secure mode.
+  For portability define undefined error codes using common Unix codes:
+  <https://www-numi.fnal.gov/offline_software/srt_public_context/WebDocs/Errors/unix_system_errors.html>
+----------------------------------------------------------- */
+#include <errno.h>
+#ifndef EAGAIN         // double free
+#define EAGAIN (11)
+#endif
+#ifndef ENOMEM         // out of memory
+#define ENOMEM (12)
+#endif
+#ifndef EFAULT         // corrupted free-list or meta-data
+#define EFAULT (14)
+#endif
+#ifndef EINVAL         // trying to free an invalid pointer
+#define EINVAL (22)
+#endif
+#ifndef EOVERFLOW      // count*size overflow
+#define EOVERFLOW (75)
+#endif
+
+
+/* -----------------------------------------------------------
+  Inlined definitions
+----------------------------------------------------------- */
+#define MI_UNUSED(x)     (void)(x)
+#if (MI_DEBUG>0)
+#define MI_UNUSED_RELEASE(x)
+#else
+#define MI_UNUSED_RELEASE(x)  MI_UNUSED(x)
+#endif
+
+#define MI_INIT4(x)   x(),x(),x(),x()
+#define MI_INIT8(x)   MI_INIT4(x),MI_INIT4(x)
+#define MI_INIT16(x)  MI_INIT8(x),MI_INIT8(x)
+#define MI_INIT32(x)  MI_INIT16(x),MI_INIT16(x)
+#define MI_INIT64(x)  MI_INIT32(x),MI_INIT32(x)
+#define MI_INIT128(x) MI_INIT64(x),MI_INIT64(x)
+#define MI_INIT256(x) MI_INIT128(x),MI_INIT128(x)
+
+
+#include <string.h>
+// initialize a local variable to zero; use memset as compilers optimize constant sized memset's
+#define _mi_memzero_var(x)  memset(&x,0,sizeof(x))
+
+// Is `x` a power of two? (0 is considered a power of two)
+static inline bool _mi_is_power_of_two(uintptr_t x) {
+  return ((x & (x - 1)) == 0);
+}
+
+// Is a pointer aligned?
+static inline bool _mi_is_aligned(void* p, size_t alignment) {
+  mi_assert_internal(alignment != 0);
+  return (((uintptr_t)p % alignment) == 0);
+}
+
+// Align upwards
+static inline uintptr_t _mi_align_up(uintptr_t sz, size_t alignment) {
+  mi_assert_internal(alignment != 0);
+  uintptr_t mask = alignment - 1;
+  if ((alignment & mask) == 0) {  // power of two?
+    return ((sz + mask) & ~mask);
+  }
+  else {
+    return (((sz + mask)/alignment)*alignment);
+  }
+}
+
+// Align downwards
+static inline uintptr_t _mi_align_down(uintptr_t sz, size_t alignment) {
+  mi_assert_internal(alignment != 0);
+  uintptr_t mask = alignment - 1;
+  if ((alignment & mask) == 0) { // power of two?
+    return (sz & ~mask);
+  }
+  else {
+    return ((sz / alignment) * alignment);
+  }
+}
+
+// Divide upwards: `s <= _mi_divide_up(s,d)*d < s+d`.
+static inline uintptr_t _mi_divide_up(uintptr_t size, size_t divider) {
+  mi_assert_internal(divider != 0);
+  return (divider == 0 ? size : ((size + divider - 1) / divider));
+}
+
+// Is memory zero initialized?
+static inline bool mi_mem_is_zero(const void* p, size_t size) {
+  for (size_t i = 0; i < size; i++) {
+    if (((uint8_t*)p)[i] != 0) return false;
+  }
+  return true;
+}
+
+
+// Align a byte size to a size in _machine words_,
+// i.e. byte size == `wsize*sizeof(void*)`.
+static inline size_t _mi_wsize_from_size(size_t size) {
+  mi_assert_internal(size <= SIZE_MAX - sizeof(uintptr_t));
+  return (size + sizeof(uintptr_t) - 1) / sizeof(uintptr_t);
+}
+
+// Overflow detecting multiply
+#if __has_builtin(__builtin_umul_overflow) || (defined(__GNUC__) && (__GNUC__ >= 5))
+#include <limits.h>      // UINT_MAX, ULONG_MAX
+#if defined(_CLOCK_T)    // for Illumos
+#undef _CLOCK_T
+#endif
+static inline bool mi_mul_overflow(size_t count, size_t size, size_t* total) {
+  #if (SIZE_MAX == ULONG_MAX)
+    return __builtin_umull_overflow(count, size, (unsigned long *)total);
+  #elif (SIZE_MAX == UINT_MAX)
+    return __builtin_umul_overflow(count, size, (unsigned int *)total);
+  #else
+    return __builtin_umulll_overflow(count, size, (unsigned long long *)total);
+  #endif
+}
+#else /* __builtin_umul_overflow is unavailable */
+static inline bool mi_mul_overflow(size_t count, size_t size, size_t* total) {
+  #define MI_MUL_NO_OVERFLOW ((size_t)1 << (4*sizeof(size_t)))  // sqrt(SIZE_MAX)
+  *total = count * size;
+  // note: gcc/clang optimize this to directly check the overflow flag
+  return ((size >= MI_MUL_NO_OVERFLOW || count >= MI_MUL_NO_OVERFLOW) && size > 0 && (SIZE_MAX / size) < count);
+}
+#endif
+
+// Safe multiply `count*size` into `total`; return `true` on overflow.
+static inline bool mi_count_size_overflow(size_t count, size_t size, size_t* total) {
+  if (count==1) {  // quick check for the case where count is one (common for C++ allocators)
+    *total = size;
+    return false;
+  }
+  else if mi_unlikely(mi_mul_overflow(count, size, total)) {
+    #if MI_DEBUG > 0
+    _mi_error_message(EOVERFLOW, "allocation request is too large (%zu * %zu bytes)\n", count, size);
+    #endif
+    *total = SIZE_MAX;
+    return true;
+  }
+  else return false;
+}
+
+
+/*----------------------------------------------------------------------------------------
+  Heap functions
+------------------------------------------------------------------------------------------- */
+
+extern const mi_heap_t _mi_heap_empty;  // read-only empty heap, initial value of the thread local default heap
+
+static inline bool mi_heap_is_backing(const mi_heap_t* heap) {
+  return (heap->tld->heap_backing == heap);
+}
+
+static inline bool mi_heap_is_initialized(mi_heap_t* heap) {
+  mi_assert_internal(heap != NULL);
+  return (heap != &_mi_heap_empty);
+}
+
+static inline uintptr_t _mi_ptr_cookie(const void* p) {
+  extern mi_heap_t _mi_heap_main;
+  mi_assert_internal(_mi_heap_main.cookie != 0);
+  return ((uintptr_t)p ^ _mi_heap_main.cookie);
+}
+
+/* -----------------------------------------------------------
+  Pages
+----------------------------------------------------------- */
+
+static inline mi_page_t* _mi_heap_get_free_small_page(mi_heap_t* heap, size_t size) {
+  mi_assert_internal(size <= (MI_SMALL_SIZE_MAX + MI_PADDING_SIZE));
+  const size_t idx = _mi_wsize_from_size(size);
+  mi_assert_internal(idx < MI_PAGES_DIRECT);
+  return heap->pages_free_direct[idx];
+}
+
+// Segment that contains the pointer
+// Large aligned blocks may be aligned at N*MI_SEGMENT_SIZE (inside a huge segment > MI_SEGMENT_SIZE),
+// and we need align "down" to the segment info which is `MI_SEGMENT_SIZE` bytes before it;
+// therefore we align one byte before `p`.
+static inline mi_segment_t* _mi_ptr_segment(const void* p) {
+  mi_assert_internal(p != NULL);
+  return (mi_segment_t*)(((uintptr_t)p - 1) & ~MI_SEGMENT_MASK);
+}
+
+static inline mi_page_t* mi_slice_to_page(mi_slice_t* s) {
+  mi_assert_internal(s->slice_offset== 0 && s->slice_count > 0);
+  return (mi_page_t*)(s);
+}
+
+static inline mi_slice_t* mi_page_to_slice(mi_page_t* p) {
+  mi_assert_internal(p->slice_offset== 0 && p->slice_count > 0);
+  return (mi_slice_t*)(p);
+}
+
+// Segment belonging to a page
+static inline mi_segment_t* _mi_page_segment(const mi_page_t* page) {
+  mi_segment_t* segment = _mi_ptr_segment(page);
+  mi_assert_internal(segment == NULL || ((mi_slice_t*)page >= segment->slices && (mi_slice_t*)page < segment->slices + segment->slice_entries));
+  return segment;
+}
+
+static inline mi_slice_t* mi_slice_first(const mi_slice_t* slice) {
+  mi_slice_t* start = (mi_slice_t*)((uint8_t*)slice - slice->slice_offset);
+  mi_assert_internal(start >= _mi_ptr_segment(slice)->slices);
+  mi_assert_internal(start->slice_offset == 0);
+  mi_assert_internal(start + start->slice_count > slice);
+  return start;
+}
+
+// Get the page containing the pointer (performance critical as it is called in mi_free)
+static inline mi_page_t* _mi_segment_page_of(const mi_segment_t* segment, const void* p) {
+  mi_assert_internal(p > (void*)segment);
+  ptrdiff_t diff = (uint8_t*)p - (uint8_t*)segment;
+  mi_assert_internal(diff > 0 && diff <= (ptrdiff_t)MI_SEGMENT_SIZE);
+  size_t idx = (size_t)diff >> MI_SEGMENT_SLICE_SHIFT;
+  mi_assert_internal(idx <= segment->slice_entries);
+  mi_slice_t* slice0 = (mi_slice_t*)&segment->slices[idx];
+  mi_slice_t* slice = mi_slice_first(slice0);  // adjust to the block that holds the page data
+  mi_assert_internal(slice->slice_offset == 0);
+  mi_assert_internal(slice >= segment->slices && slice < segment->slices + segment->slice_entries);
+  return mi_slice_to_page(slice);
+}
+
+// Quick page start for initialized pages
+static inline uint8_t* _mi_page_start(const mi_segment_t* segment, const mi_page_t* page, size_t* page_size) {
+  return _mi_segment_page_start(segment, page, page_size);
+}
+
+// Get the page containing the pointer
+static inline mi_page_t* _mi_ptr_page(void* p) {
+  return _mi_segment_page_of(_mi_ptr_segment(p), p);
+}
+
+// Get the block size of a page (special case for huge objects)
+static inline size_t mi_page_block_size(const mi_page_t* page) {
+  const size_t bsize = page->xblock_size;
+  mi_assert_internal(bsize > 0);
+  if mi_likely(bsize < MI_HUGE_BLOCK_SIZE) {
+    return bsize;
+  }
+  else {
+    size_t psize;
+    _mi_segment_page_start(_mi_page_segment(page), page, &psize);
+    return psize;
+  }
+}
+
+static inline bool mi_page_is_huge(const mi_page_t* page) {
+  return (_mi_page_segment(page)->kind == MI_SEGMENT_HUGE);
+}
+
+// Get the usable block size of a page without fixed padding.
+// This may still include internal padding due to alignment and rounding up size classes.
+static inline size_t mi_page_usable_block_size(const mi_page_t* page) {
+  return mi_page_block_size(page) - MI_PADDING_SIZE;
+}
+
+// size of a segment
+static inline size_t mi_segment_size(mi_segment_t* segment) {
+  return segment->segment_slices * MI_SEGMENT_SLICE_SIZE;
+}
+
+static inline uint8_t* mi_segment_end(mi_segment_t* segment) {
+  return (uint8_t*)segment + mi_segment_size(segment);
+}
+
+// Thread free access
+static inline mi_block_t* mi_page_thread_free(const mi_page_t* page) {
+  return (mi_block_t*)(mi_atomic_load_relaxed(&((mi_page_t*)page)->xthread_free) & ~3);
+}
+
+static inline mi_delayed_t mi_page_thread_free_flag(const mi_page_t* page) {
+  return (mi_delayed_t)(mi_atomic_load_relaxed(&((mi_page_t*)page)->xthread_free) & 3);
+}
+
+// Heap access
+static inline mi_heap_t* mi_page_heap(const mi_page_t* page) {
+  return (mi_heap_t*)(mi_atomic_load_relaxed(&((mi_page_t*)page)->xheap));
+}
+
+static inline void mi_page_set_heap(mi_page_t* page, mi_heap_t* heap) {
+  mi_assert_internal(mi_page_thread_free_flag(page) != MI_DELAYED_FREEING);
+  mi_atomic_store_release(&page->xheap,(uintptr_t)heap);
+}
+
+// Thread free flag helpers
+static inline mi_block_t* mi_tf_block(mi_thread_free_t tf) {
+  return (mi_block_t*)(tf & ~0x03);
+}
+static inline mi_delayed_t mi_tf_delayed(mi_thread_free_t tf) {
+  return (mi_delayed_t)(tf & 0x03);
+}
+static inline mi_thread_free_t mi_tf_make(mi_block_t* block, mi_delayed_t delayed) {
+  return (mi_thread_free_t)((uintptr_t)block | (uintptr_t)delayed);
+}
+static inline mi_thread_free_t mi_tf_set_delayed(mi_thread_free_t tf, mi_delayed_t delayed) {
+  return mi_tf_make(mi_tf_block(tf),delayed);
+}
+static inline mi_thread_free_t mi_tf_set_block(mi_thread_free_t tf, mi_block_t* block) {
+  return mi_tf_make(block, mi_tf_delayed(tf));
+}
+
+// are all blocks in a page freed?
+// note: needs up-to-date used count, (as the `xthread_free` list may not be empty). see `_mi_page_collect_free`.
+static inline bool mi_page_all_free(const mi_page_t* page) {
+  mi_assert_internal(page != NULL);
+  return (page->used == 0);
+}
+
+// are there any available blocks?
+static inline bool mi_page_has_any_available(const mi_page_t* page) {
+  mi_assert_internal(page != NULL && page->reserved > 0);
+  return (page->used < page->reserved || (mi_page_thread_free(page) != NULL));
+}
+
+// are there immediately available blocks, i.e. blocks available on the free list.
+static inline bool mi_page_immediate_available(const mi_page_t* page) {
+  mi_assert_internal(page != NULL);
+  return (page->free != NULL);
+}
+
+// is more than 7/8th of a page in use?
+static inline bool mi_page_mostly_used(const mi_page_t* page) {
+  if (page==NULL) return true;
+  uint16_t frac = page->reserved / 8U;
+  return (page->reserved - page->used <= frac);
+}
+
+static inline mi_page_queue_t* mi_page_queue(const mi_heap_t* heap, size_t size) {
+  return &((mi_heap_t*)heap)->pages[_mi_bin(size)];
+}
+
+
+
+//-----------------------------------------------------------
+// Page flags
+//-----------------------------------------------------------
+static inline bool mi_page_is_in_full(const mi_page_t* page) {
+  return page->flags.x.in_full;
+}
+
+static inline void mi_page_set_in_full(mi_page_t* page, bool in_full) {
+  page->flags.x.in_full = in_full;
+}
+
+static inline bool mi_page_has_aligned(const mi_page_t* page) {
+  return page->flags.x.has_aligned;
+}
+
+static inline void mi_page_set_has_aligned(mi_page_t* page, bool has_aligned) {
+  page->flags.x.has_aligned = has_aligned;
+}
+
+
+/* -------------------------------------------------------------------
+Encoding/Decoding the free list next pointers
+
+This is to protect against buffer overflow exploits where the
+free list is mutated. Many hardened allocators xor the next pointer `p`
+with a secret key `k1`, as `p^k1`. This prevents overwriting with known
+values but might be still too weak: if the attacker can guess
+the pointer `p` this  can reveal `k1` (since `p^k1^p == k1`).
+Moreover, if multiple blocks can be read as well, the attacker can
+xor both as `(p1^k1) ^ (p2^k1) == p1^p2` which may reveal a lot
+about the pointers (and subsequently `k1`).
+
+Instead mimalloc uses an extra key `k2` and encodes as `((p^k2)<<<k1)+k1`.
+Since these operations are not associative, the above approaches do not
+work so well any more even if the `p` can be guesstimated. For example,
+for the read case we can subtract two entries to discard the `+k1` term,
+but that leads to `((p1^k2)<<<k1) - ((p2^k2)<<<k1)` at best.
+We include the left-rotation since xor and addition are otherwise linear
+in the lowest bit. Finally, both keys are unique per page which reduces
+the re-use of keys by a large factor.
+
+We also pass a separate `null` value to be used as `NULL` or otherwise
+`(k2<<<k1)+k1` would appear (too) often as a sentinel value.
+------------------------------------------------------------------- */
+
+static inline bool mi_is_in_same_segment(const void* p, const void* q) {
+  return (_mi_ptr_segment(p) == _mi_ptr_segment(q));
+}
+
+static inline bool mi_is_in_same_page(const void* p, const void* q) {
+  mi_segment_t* segment = _mi_ptr_segment(p);
+  if (_mi_ptr_segment(q) != segment) return false;
+  // assume q may be invalid // return (_mi_segment_page_of(segment, p) == _mi_segment_page_of(segment, q));
+  mi_page_t* page = _mi_segment_page_of(segment, p);
+  size_t psize;
+  uint8_t* start = _mi_segment_page_start(segment, page, &psize);
+  return (start <= (uint8_t*)q && (uint8_t*)q < start + psize);
+}
+
+static inline uintptr_t mi_rotl(uintptr_t x, uintptr_t shift) {
+  shift %= MI_INTPTR_BITS;
+  return (shift==0 ? x : ((x << shift) | (x >> (MI_INTPTR_BITS - shift))));
+}
+static inline uintptr_t mi_rotr(uintptr_t x, uintptr_t shift) {
+  shift %= MI_INTPTR_BITS;
+  return (shift==0 ? x : ((x >> shift) | (x << (MI_INTPTR_BITS - shift))));
+}
+
+static inline void* mi_ptr_decode(const void* null, const mi_encoded_t x, const uintptr_t* keys) {
+  void* p = (void*)(mi_rotr(x - keys[0], keys[0]) ^ keys[1]);
+  return (p==null ? NULL : p);
+}
+
+static inline mi_encoded_t mi_ptr_encode(const void* null, const void* p, const uintptr_t* keys) {
+  uintptr_t x = (uintptr_t)(p==NULL ? null : p);
+  return mi_rotl(x ^ keys[1], keys[0]) + keys[0];
+}
+
+static inline mi_block_t* mi_block_nextx( const void* null, const mi_block_t* block, const uintptr_t* keys ) {
+  mi_track_mem_defined(block,sizeof(mi_block_t));
+  mi_block_t* next;
+  #ifdef MI_ENCODE_FREELIST
+  next = (mi_block_t*)mi_ptr_decode(null, block->next, keys);
+  #else
+  MI_UNUSED(keys); MI_UNUSED(null);
+  next = (mi_block_t*)block->next;
+  #endif
+  mi_track_mem_noaccess(block,sizeof(mi_block_t));
+  return next;
+}
+
+static inline void mi_block_set_nextx(const void* null, mi_block_t* block, const mi_block_t* next, const uintptr_t* keys) {
+  mi_track_mem_undefined(block,sizeof(mi_block_t));
+  #ifdef MI_ENCODE_FREELIST
+  block->next = mi_ptr_encode(null, next, keys);
+  #else
+  MI_UNUSED(keys); MI_UNUSED(null);
+  block->next = (mi_encoded_t)next;
+  #endif
+  mi_track_mem_noaccess(block,sizeof(mi_block_t));
+}
+
+static inline mi_block_t* mi_block_next(const mi_page_t* page, const mi_block_t* block) {
+  #ifdef MI_ENCODE_FREELIST
+  mi_block_t* next = mi_block_nextx(page,block,page->keys);
+  // check for free list corruption: is `next` at least in the same page?
+  // TODO: check if `next` is `page->block_size` aligned?
+  if mi_unlikely(next!=NULL && !mi_is_in_same_page(block, next)) {
+    _mi_error_message(EFAULT, "corrupted free list entry of size %zub at %p: value 0x%zx\n", mi_page_block_size(page), block, (uintptr_t)next);
+    next = NULL;
+  }
+  return next;
+  #else
+  MI_UNUSED(page);
+  return mi_block_nextx(page,block,NULL);
+  #endif
+}
+
+static inline void mi_block_set_next(const mi_page_t* page, mi_block_t* block, const mi_block_t* next) {
+  #ifdef MI_ENCODE_FREELIST
+  mi_block_set_nextx(page,block,next, page->keys);
+  #else
+  MI_UNUSED(page);
+  mi_block_set_nextx(page,block,next,NULL);
+  #endif
+}
+
+
+// -------------------------------------------------------------------
+// commit mask
+// -------------------------------------------------------------------
+
+static inline void mi_commit_mask_create_empty(mi_commit_mask_t* cm) {
+  for (size_t i = 0; i < MI_COMMIT_MASK_FIELD_COUNT; i++) {
+    cm->mask[i] = 0;
+  }
+}
+
+static inline void mi_commit_mask_create_full(mi_commit_mask_t* cm) {
+  for (size_t i = 0; i < MI_COMMIT_MASK_FIELD_COUNT; i++) {
+    cm->mask[i] = ~((size_t)0);
+  }
+}
+
+static inline bool mi_commit_mask_is_empty(const mi_commit_mask_t* cm) {
+  for (size_t i = 0; i < MI_COMMIT_MASK_FIELD_COUNT; i++) {
+    if (cm->mask[i] != 0) return false;
+  }
+  return true;
+}
+
+static inline bool mi_commit_mask_is_full(const mi_commit_mask_t* cm) {
+  for (size_t i = 0; i < MI_COMMIT_MASK_FIELD_COUNT; i++) {
+    if (cm->mask[i] != ~((size_t)0)) return false;
+  }
+  return true;
+}
+
+// defined in `segment.c`:
+size_t _mi_commit_mask_committed_size(const mi_commit_mask_t* cm, size_t total);
+size_t _mi_commit_mask_next_run(const mi_commit_mask_t* cm, size_t* idx);
+
+#define mi_commit_mask_foreach(cm,idx,count) \
+  idx = 0; \
+  while ((count = _mi_commit_mask_next_run(cm,&idx)) > 0) {
+
+#define mi_commit_mask_foreach_end() \
+    idx += count; \
+  }
+
+
+
+/* -----------------------------------------------------------
+  memory id's
+----------------------------------------------------------- */
+
+static inline mi_memid_t _mi_memid_create(mi_memkind_t memkind) {
+  mi_memid_t memid;
+  _mi_memzero_var(memid);
+  memid.memkind = memkind;
+  return memid;
+}
+
+static inline mi_memid_t _mi_memid_none(void) {
+  return _mi_memid_create(MI_MEM_NONE);
+}
+
+static inline mi_memid_t _mi_memid_create_os(bool committed, bool is_zero, bool is_large) {
+  mi_memid_t memid = _mi_memid_create(MI_MEM_OS);
+  memid.initially_committed = committed;
+  memid.initially_zero = is_zero;
+  memid.is_pinned = is_large;
+  return memid;
+}
+
+
+// -------------------------------------------------------------------
+// Fast "random" shuffle
+// -------------------------------------------------------------------
+
+static inline uintptr_t _mi_random_shuffle(uintptr_t x) {
+  if (x==0) { x = 17; }   // ensure we don't get stuck in generating zeros
+#if (MI_INTPTR_SIZE==8)
+  // by Sebastiano Vigna, see: <http://xoshiro.di.unimi.it/splitmix64.c>
+  x ^= x >> 30;
+  x *= 0xbf58476d1ce4e5b9UL;
+  x ^= x >> 27;
+  x *= 0x94d049bb133111ebUL;
+  x ^= x >> 31;
+#elif (MI_INTPTR_SIZE==4)
+  // by Chris Wellons, see: <https://nullprogram.com/blog/2018/07/31/>
+  x ^= x >> 16;
+  x *= 0x7feb352dUL;
+  x ^= x >> 15;
+  x *= 0x846ca68bUL;
+  x ^= x >> 16;
+#endif
+  return x;
+}
+
+// -------------------------------------------------------------------
+// Optimize numa node access for the common case (= one node)
+// -------------------------------------------------------------------
+
+int    _mi_os_numa_node_get(mi_os_tld_t* tld);
+size_t _mi_os_numa_node_count_get(void);
+
+extern _Atomic(size_t) _mi_numa_node_count;
+static inline int _mi_os_numa_node(mi_os_tld_t* tld) {
+  if mi_likely(mi_atomic_load_relaxed(&_mi_numa_node_count) == 1) { return 0; }
+  else return _mi_os_numa_node_get(tld);
+}
+static inline size_t _mi_os_numa_node_count(void) {
+  const size_t count = mi_atomic_load_relaxed(&_mi_numa_node_count);
+  if mi_likely(count > 0) { return count; }
+  else return _mi_os_numa_node_count_get();
+}
+
+
+
+// -----------------------------------------------------------------------
+// Count bits: trailing or leading zeros (with MI_INTPTR_BITS on all zero)
+// -----------------------------------------------------------------------
+
+#if defined(__GNUC__)
+
+#include <limits.h>       // LONG_MAX
+#define MI_HAVE_FAST_BITSCAN
+static inline size_t mi_clz(uintptr_t x) {
+  if (x==0) return MI_INTPTR_BITS;
+#if (INTPTR_MAX == LONG_MAX)
+  return __builtin_clzl(x);
+#else
+  return __builtin_clzll(x);
+#endif
+}
+static inline size_t mi_ctz(uintptr_t x) {
+  if (x==0) return MI_INTPTR_BITS;
+#if (INTPTR_MAX == LONG_MAX)
+  return __builtin_ctzl(x);
+#else
+  return __builtin_ctzll(x);
+#endif
+}
+
+#elif defined(_MSC_VER)
+
+#include <limits.h>       // LONG_MAX
+#include <intrin.h>       // BitScanReverse64
+#define MI_HAVE_FAST_BITSCAN
+static inline size_t mi_clz(uintptr_t x) {
+  if (x==0) return MI_INTPTR_BITS;
+  unsigned long idx;
+#if (INTPTR_MAX == LONG_MAX)
+  _BitScanReverse(&idx, x);
+#else
+  _BitScanReverse64(&idx, x);
+#endif
+  return ((MI_INTPTR_BITS - 1) - idx);
+}
+static inline size_t mi_ctz(uintptr_t x) {
+  if (x==0) return MI_INTPTR_BITS;
+  unsigned long idx;
+#if (INTPTR_MAX == LONG_MAX)
+  _BitScanForward(&idx, x);
+#else
+  _BitScanForward64(&idx, x);
+#endif
+  return idx;
+}
+
+#else
+static inline size_t mi_ctz32(uint32_t x) {
+  // de Bruijn multiplication, see <http://supertech.csail.mit.edu/papers/debruijn.pdf>
+  static const unsigned char debruijn[32] = {
+    0, 1, 28, 2, 29, 14, 24, 3, 30, 22, 20, 15, 25, 17, 4, 8,
+    31, 27, 13, 23, 21, 19, 16, 7, 26, 12, 18, 6, 11, 5, 10, 9
+  };
+  if (x==0) return 32;
+  return debruijn[((x & -(int32_t)x) * 0x077CB531UL) >> 27];
+}
+static inline size_t mi_clz32(uint32_t x) {
+  // de Bruijn multiplication, see <http://supertech.csail.mit.edu/papers/debruijn.pdf>
+  static const uint8_t debruijn[32] = {
+    31, 22, 30, 21, 18, 10, 29, 2, 20, 17, 15, 13, 9, 6, 28, 1,
+    23, 19, 11, 3, 16, 14, 7, 24, 12, 4, 8, 25, 5, 26, 27, 0
+  };
+  if (x==0) return 32;
+  x |= x >> 1;
+  x |= x >> 2;
+  x |= x >> 4;
+  x |= x >> 8;
+  x |= x >> 16;
+  return debruijn[(uint32_t)(x * 0x07C4ACDDUL) >> 27];
+}
+
+static inline size_t mi_clz(uintptr_t x) {
+  if (x==0) return MI_INTPTR_BITS;
+#if (MI_INTPTR_BITS <= 32)
+  return mi_clz32((uint32_t)x);
+#else
+  size_t count = mi_clz32((uint32_t)(x >> 32));
+  if (count < 32) return count;
+  return (32 + mi_clz32((uint32_t)x));
+#endif
+}
+static inline size_t mi_ctz(uintptr_t x) {
+  if (x==0) return MI_INTPTR_BITS;
+#if (MI_INTPTR_BITS <= 32)
+  return mi_ctz32((uint32_t)x);
+#else
+  size_t count = mi_ctz32((uint32_t)x);
+  if (count < 32) return count;
+  return (32 + mi_ctz32((uint32_t)(x>>32)));
+#endif
+}
+
+#endif
+
+// "bit scan reverse": Return index of the highest bit (or MI_INTPTR_BITS if `x` is zero)
+static inline size_t mi_bsr(uintptr_t x) {
+  return (x==0 ? MI_INTPTR_BITS : MI_INTPTR_BITS - 1 - mi_clz(x));
+}
+
+
+// ---------------------------------------------------------------------------------
+// Provide our own `_mi_memcpy` for potential performance optimizations.
+//
+// For now, only on Windows with msvc/clang-cl we optimize to `rep movsb` if
+// we happen to run on x86/x64 cpu's that have "fast short rep movsb" (FSRM) support
+// (AMD Zen3+ (~2020) or Intel Ice Lake+ (~2017). See also issue #201 and pr #253.
+// ---------------------------------------------------------------------------------
+
+#if !MI_TRACK_ENABLED && defined(_WIN32) && (defined(_M_IX86) || defined(_M_X64))
+#include <intrin.h>
+extern bool _mi_cpu_has_fsrm;
+static inline void _mi_memcpy(void* dst, const void* src, size_t n) {
+  if (_mi_cpu_has_fsrm) {
+    __movsb((unsigned char*)dst, (const unsigned char*)src, n);
+  }
+  else {
+    memcpy(dst, src, n);
+  }
+}
+static inline void _mi_memzero(void* dst, size_t n) {
+  if (_mi_cpu_has_fsrm) {
+    __stosb((unsigned char*)dst, 0, n);
+  }
+  else {
+    memset(dst, 0, n);
+  }
+}
+#else
+static inline void _mi_memcpy(void* dst, const void* src, size_t n) {
+  memcpy(dst, src, n);
+}
+static inline void _mi_memzero(void* dst, size_t n) {
+  memset(dst, 0, n);
+}
+#endif
+
+// -------------------------------------------------------------------------------
+// The `_mi_memcpy_aligned` can be used if the pointers are machine-word aligned
+// This is used for example in `mi_realloc`.
+// -------------------------------------------------------------------------------
+
+#if (defined(__GNUC__) && (__GNUC__ >= 4)) || defined(__clang__)
+// On GCC/CLang we provide a hint that the pointers are word aligned.
+static inline void _mi_memcpy_aligned(void* dst, const void* src, size_t n) {
+  mi_assert_internal(((uintptr_t)dst % MI_INTPTR_SIZE == 0) && ((uintptr_t)src % MI_INTPTR_SIZE == 0));
+  void* adst = __builtin_assume_aligned(dst, MI_INTPTR_SIZE);
+  const void* asrc = __builtin_assume_aligned(src, MI_INTPTR_SIZE);
+  _mi_memcpy(adst, asrc, n);
+}
+
+static inline void _mi_memzero_aligned(void* dst, size_t n) {
+  mi_assert_internal((uintptr_t)dst % MI_INTPTR_SIZE == 0);
+  void* adst = __builtin_assume_aligned(dst, MI_INTPTR_SIZE);
+  _mi_memzero(adst, n);
+}
+#else
+// Default fallback on `_mi_memcpy`
+static inline void _mi_memcpy_aligned(void* dst, const void* src, size_t n) {
+  mi_assert_internal(((uintptr_t)dst % MI_INTPTR_SIZE == 0) && ((uintptr_t)src % MI_INTPTR_SIZE == 0));
+  _mi_memcpy(dst, src, n);
+}
+
+static inline void _mi_memzero_aligned(void* dst, size_t n) {
+  mi_assert_internal((uintptr_t)dst % MI_INTPTR_SIZE == 0);
+  _mi_memzero(dst, n);
+}
+#endif
+
+
+#endif

Dependencies/Python/include/internal/mimalloc/mimalloc/prim.h

@@ -0,0 +1,329 @@
+/* ----------------------------------------------------------------------------
+Copyright (c) 2018-2023, Microsoft Research, Daan Leijen
+This is free software; you can redistribute it and/or modify it under the
+terms of the MIT license. A copy of the license can be found in the file
+"LICENSE" at the root of this distribution.
+-----------------------------------------------------------------------------*/
+#pragma once
+#ifndef MIMALLOC_PRIM_H
+#define MIMALLOC_PRIM_H
+
+
+// --------------------------------------------------------------------------
+// This file specifies the primitive portability API.
+// Each OS/host needs to implement these primitives, see `src/prim`
+// for implementations on Window, macOS, WASI, and Linux/Unix.
+//
+// note: on all primitive functions, we always have result parameters != NUL, and:
+//  addr != NULL and page aligned
+//  size > 0     and page aligned
+//  return value is an error code an int where 0 is success.
+// --------------------------------------------------------------------------
+
+// OS memory configuration
+typedef struct mi_os_mem_config_s {
+  size_t  page_size;            // 4KiB
+  size_t  large_page_size;      // 2MiB
+  size_t  alloc_granularity;    // smallest allocation size (on Windows 64KiB)
+  bool    has_overcommit;       // can we reserve more memory than can be actually committed?
+  bool    must_free_whole;      // must allocated blocks be freed as a whole (false for mmap, true for VirtualAlloc)
+  bool    has_virtual_reserve;  // supports virtual address space reservation? (if true we can reserve virtual address space without using commit or physical memory)
+} mi_os_mem_config_t;
+
+// Initialize
+void _mi_prim_mem_init( mi_os_mem_config_t* config );
+
+// Free OS memory
+int _mi_prim_free(void* addr, size_t size );
+
+// Allocate OS memory. Return NULL on error.
+// The `try_alignment` is just a hint and the returned pointer does not have to be aligned.
+// If `commit` is false, the virtual memory range only needs to be reserved (with no access)
+// which will later be committed explicitly using `_mi_prim_commit`.
+// `is_zero` is set to true if the memory was zero initialized (as on most OS's)
+// pre: !commit => !allow_large
+//      try_alignment >= _mi_os_page_size() and a power of 2
+int _mi_prim_alloc(size_t size, size_t try_alignment, bool commit, bool allow_large, bool* is_large, bool* is_zero, void** addr);
+
+// Commit memory. Returns error code or 0 on success.
+// For example, on Linux this would make the memory PROT_READ|PROT_WRITE.
+// `is_zero` is set to true if the memory was zero initialized (e.g. on Windows)
+int _mi_prim_commit(void* addr, size_t size, bool* is_zero);
+
+// Decommit memory. Returns error code or 0 on success. The `needs_recommit` result is true
+// if the memory would need to be re-committed. For example, on Windows this is always true,
+// but on Linux we could use MADV_DONTNEED to decommit which does not need a recommit.
+// pre: needs_recommit != NULL
+int _mi_prim_decommit(void* addr, size_t size, bool* needs_recommit);
+
+// Reset memory. The range keeps being accessible but the content might be reset.
+// Returns error code or 0 on success.
+int _mi_prim_reset(void* addr, size_t size);
+
+// Protect memory. Returns error code or 0 on success.
+int _mi_prim_protect(void* addr, size_t size, bool protect);
+
+// Allocate huge (1GiB) pages possibly associated with a NUMA node.
+// `is_zero` is set to true if the memory was zero initialized (as on most OS's)
+// pre: size > 0  and a multiple of 1GiB.
+//      numa_node is either negative (don't care), or a numa node number.
+int _mi_prim_alloc_huge_os_pages(void* hint_addr, size_t size, int numa_node, bool* is_zero, void** addr);
+
+// Return the current NUMA node
+size_t _mi_prim_numa_node(void);
+
+// Return the number of logical NUMA nodes
+size_t _mi_prim_numa_node_count(void);
+
+// Clock ticks
+mi_msecs_t _mi_prim_clock_now(void);
+
+// Return process information (only for statistics)
+typedef struct mi_process_info_s {
+  mi_msecs_t  elapsed;
+  mi_msecs_t  utime;
+  mi_msecs_t  stime;
+  size_t      current_rss;
+  size_t      peak_rss;
+  size_t      current_commit;
+  size_t      peak_commit;
+  size_t      page_faults;
+} mi_process_info_t;
+
+void _mi_prim_process_info(mi_process_info_t* pinfo);
+
+// Default stderr output. (only for warnings etc. with verbose enabled)
+// msg != NULL && _mi_strlen(msg) > 0
+void _mi_prim_out_stderr( const char* msg );
+
+// Get an environment variable. (only for options)
+// name != NULL, result != NULL, result_size >= 64
+bool _mi_prim_getenv(const char* name, char* result, size_t result_size);
+
+
+// Fill a buffer with strong randomness; return `false` on error or if
+// there is no strong randomization available.
+bool _mi_prim_random_buf(void* buf, size_t buf_len);
+
+// Called on the first thread start, and should ensure `_mi_thread_done` is called on thread termination.
+void _mi_prim_thread_init_auto_done(void);
+
+// Called on process exit and may take action to clean up resources associated with the thread auto done.
+void _mi_prim_thread_done_auto_done(void);
+
+// Called when the default heap for a thread changes
+void _mi_prim_thread_associate_default_heap(mi_heap_t* heap);
+
+
+//-------------------------------------------------------------------
+// Thread id: `_mi_prim_thread_id()`
+//
+// Getting the thread id should be performant as it is called in the
+// fast path of `_mi_free` and we specialize for various platforms as
+// inlined definitions. Regular code should call `init.c:_mi_thread_id()`.
+// We only require _mi_prim_thread_id() to return a unique id
+// for each thread (unequal to zero).
+//-------------------------------------------------------------------
+
+// defined in `init.c`; do not use these directly
+extern mi_decl_thread mi_heap_t* _mi_heap_default;  // default heap to allocate from
+extern bool _mi_process_is_initialized;             // has mi_process_init been called?
+
+static inline mi_threadid_t _mi_prim_thread_id(void) mi_attr_noexcept;
+
+#ifdef MI_PRIM_THREAD_ID
+
+static inline mi_threadid_t _mi_prim_thread_id(void) mi_attr_noexcept {
+  return MI_PRIM_THREAD_ID();
+}
+
+#elif defined(_WIN32)
+
+#define WIN32_LEAN_AND_MEAN
+#include <windows.h>
+static inline mi_threadid_t _mi_prim_thread_id(void) mi_attr_noexcept {
+  // Windows: works on Intel and ARM in both 32- and 64-bit
+  return (uintptr_t)NtCurrentTeb();
+}
+
+// We use assembly for a fast thread id on the main platforms. The TLS layout depends on
+// both the OS and libc implementation so we use specific tests for each main platform.
+// If you test on another platform and it works please send a PR :-)
+// see also https://akkadia.org/drepper/tls.pdf for more info on the TLS register.
+#elif defined(__GNUC__) && ( \
+           (defined(__GLIBC__)   && (defined(__x86_64__) || defined(__i386__) || (defined(__arm__) && __ARM_ARCH >= 7) || defined(__aarch64__))) \
+        || (defined(__APPLE__)   && (defined(__x86_64__) || defined(__aarch64__))) \
+        || (defined(__BIONIC__)  && (defined(__x86_64__) || defined(__i386__) || (defined(__arm__) && __ARM_ARCH >= 7) || defined(__aarch64__))) \
+        || (defined(__FreeBSD__) && (defined(__x86_64__) || defined(__i386__) || defined(__aarch64__))) \
+        || (defined(__OpenBSD__) && (defined(__x86_64__) || defined(__i386__) || defined(__aarch64__))) \
+      )
+
+static inline void* mi_prim_tls_slot(size_t slot) mi_attr_noexcept {
+  void* res;
+  const size_t ofs = (slot*sizeof(void*));
+  #if defined(__i386__)
+    __asm__("movl %%gs:%1, %0" : "=r" (res) : "m" (*((void**)ofs)) : );  // x86 32-bit always uses GS
+  #elif defined(__APPLE__) && defined(__x86_64__)
+    __asm__("movq %%gs:%1, %0" : "=r" (res) : "m" (*((void**)ofs)) : );  // x86_64 macOSX uses GS
+  #elif defined(__x86_64__) && (MI_INTPTR_SIZE==4)
+    __asm__("movl %%fs:%1, %0" : "=r" (res) : "m" (*((void**)ofs)) : );  // x32 ABI
+  #elif defined(__x86_64__)
+    __asm__("movq %%fs:%1, %0" : "=r" (res) : "m" (*((void**)ofs)) : );  // x86_64 Linux, BSD uses FS
+  #elif defined(__arm__)
+    void** tcb; MI_UNUSED(ofs);
+    __asm__ volatile ("mrc p15, 0, %0, c13, c0, 3\nbic %0, %0, #3" : "=r" (tcb));
+    res = tcb[slot];
+  #elif defined(__aarch64__)
+    void** tcb; MI_UNUSED(ofs);
+    #if defined(__APPLE__) // M1, issue #343
+    __asm__ volatile ("mrs %0, tpidrro_el0\nbic %0, %0, #7" : "=r" (tcb));
+    #else
+    __asm__ volatile ("mrs %0, tpidr_el0" : "=r" (tcb));
+    #endif
+    res = tcb[slot];
+  #endif
+  return res;
+}
+
+// setting a tls slot is only used on macOS for now
+static inline void mi_prim_tls_slot_set(size_t slot, void* value) mi_attr_noexcept {
+  const size_t ofs = (slot*sizeof(void*));
+  #if defined(__i386__)
+    __asm__("movl %1,%%gs:%0" : "=m" (*((void**)ofs)) : "rn" (value) : );  // 32-bit always uses GS
+  #elif defined(__APPLE__) && defined(__x86_64__)
+    __asm__("movq %1,%%gs:%0" : "=m" (*((void**)ofs)) : "rn" (value) : );  // x86_64 macOS uses GS
+  #elif defined(__x86_64__) && (MI_INTPTR_SIZE==4)
+    __asm__("movl %1,%%fs:%0" : "=m" (*((void**)ofs)) : "rn" (value) : );  // x32 ABI
+  #elif defined(__x86_64__)
+    __asm__("movq %1,%%fs:%0" : "=m" (*((void**)ofs)) : "rn" (value) : );  // x86_64 Linux, BSD uses FS
+  #elif defined(__arm__)
+    void** tcb; MI_UNUSED(ofs);
+    __asm__ volatile ("mrc p15, 0, %0, c13, c0, 3\nbic %0, %0, #3" : "=r" (tcb));
+    tcb[slot] = value;
+  #elif defined(__aarch64__)
+    void** tcb; MI_UNUSED(ofs);
+    #if defined(__APPLE__) // M1, issue #343
+    __asm__ volatile ("mrs %0, tpidrro_el0\nbic %0, %0, #7" : "=r" (tcb));
+    #else
+    __asm__ volatile ("mrs %0, tpidr_el0" : "=r" (tcb));
+    #endif
+    tcb[slot] = value;
+  #endif
+}
+
+static inline mi_threadid_t _mi_prim_thread_id(void) mi_attr_noexcept {
+  #if defined(__BIONIC__)
+    // issue #384, #495: on the Bionic libc (Android), slot 1 is the thread id
+    // see: https://github.com/aosp-mirror/platform_bionic/blob/c44b1d0676ded732df4b3b21c5f798eacae93228/libc/platform/bionic/tls_defines.h#L86
+    return (uintptr_t)mi_prim_tls_slot(1);
+  #else
+    // in all our other targets, slot 0 is the thread id
+    // glibc: https://sourceware.org/git/?p=glibc.git;a=blob_plain;f=sysdeps/x86_64/nptl/tls.h
+    // apple: https://github.com/apple/darwin-xnu/blob/main/libsyscall/os/tsd.h#L36
+    return (uintptr_t)mi_prim_tls_slot(0);
+  #endif
+}
+
+#else
+
+// otherwise use portable C, taking the address of a thread local variable (this is still very fast on most platforms).
+static inline mi_threadid_t _mi_prim_thread_id(void) mi_attr_noexcept {
+  return (uintptr_t)&_mi_heap_default;
+}
+
+#endif
+
+
+
+/* ----------------------------------------------------------------------------------------
+The thread local default heap: `_mi_prim_get_default_heap()`
+This is inlined here as it is on the fast path for allocation functions.
+
+On most platforms (Windows, Linux, FreeBSD, NetBSD, etc), this just returns a
+__thread local variable (`_mi_heap_default`). With the initial-exec TLS model this ensures
+that the storage will always be available (allocated on the thread stacks).
+
+On some platforms though we cannot use that when overriding `malloc` since the underlying
+TLS implementation (or the loader) will call itself `malloc` on a first access and recurse.
+We try to circumvent this in an efficient way:
+- macOSX : we use an unused TLS slot from the OS allocated slots (MI_TLS_SLOT). On OSX, the
+           loader itself calls `malloc` even before the modules are initialized.
+- OpenBSD: we use an unused slot from the pthread block (MI_TLS_PTHREAD_SLOT_OFS).
+- DragonFly: defaults are working but seem slow compared to freeBSD (see PR #323)
+------------------------------------------------------------------------------------------- */
+
+static inline mi_heap_t* mi_prim_get_default_heap(void);
+
+#if defined(MI_MALLOC_OVERRIDE)
+#if defined(__APPLE__) // macOS
+  #define MI_TLS_SLOT               89  // seems unused?
+  // #define MI_TLS_RECURSE_GUARD 1
+  // other possible unused ones are 9, 29, __PTK_FRAMEWORK_JAVASCRIPTCORE_KEY4 (94), __PTK_FRAMEWORK_GC_KEY9 (112) and __PTK_FRAMEWORK_OLDGC_KEY9 (89)
+  // see <https://github.com/rweichler/substrate/blob/master/include/pthread_machdep.h>
+#elif defined(__OpenBSD__)
+  // use end bytes of a name; goes wrong if anyone uses names > 23 characters (ptrhread specifies 16)
+  // see <https://github.com/openbsd/src/blob/master/lib/libc/include/thread_private.h#L371>
+  #define MI_TLS_PTHREAD_SLOT_OFS   (6*sizeof(int) + 4*sizeof(void*) + 24)
+  // #elif defined(__DragonFly__)
+  // #warning "mimalloc is not working correctly on DragonFly yet."
+  // #define MI_TLS_PTHREAD_SLOT_OFS   (4 + 1*sizeof(void*))  // offset `uniqueid` (also used by gdb?) <https://github.com/DragonFlyBSD/DragonFlyBSD/blob/master/lib/libthread_xu/thread/thr_private.h#L458>
+#elif defined(__ANDROID__)
+  // See issue #381
+  #define MI_TLS_PTHREAD
+#endif
+#endif
+
+
+#if defined(MI_TLS_SLOT)
+
+static inline mi_heap_t* mi_prim_get_default_heap(void) {
+  mi_heap_t* heap = (mi_heap_t*)mi_prim_tls_slot(MI_TLS_SLOT);
+  if mi_unlikely(heap == NULL) {
+    #ifdef __GNUC__
+    __asm(""); // prevent conditional load of the address of _mi_heap_empty
+    #endif
+    heap = (mi_heap_t*)&_mi_heap_empty;
+  }
+  return heap;
+}
+
+#elif defined(MI_TLS_PTHREAD_SLOT_OFS)
+
+static inline mi_heap_t** mi_prim_tls_pthread_heap_slot(void) {
+  pthread_t self = pthread_self();
+  #if defined(__DragonFly__)
+  if (self==NULL) return NULL;
+  #endif
+  return (mi_heap_t**)((uint8_t*)self + MI_TLS_PTHREAD_SLOT_OFS);
+}
+
+static inline mi_heap_t* mi_prim_get_default_heap(void) {
+  mi_heap_t** pheap = mi_prim_tls_pthread_heap_slot();
+  if mi_unlikely(pheap == NULL) return _mi_heap_main_get();
+  mi_heap_t* heap = *pheap;
+  if mi_unlikely(heap == NULL) return (mi_heap_t*)&_mi_heap_empty;
+  return heap;
+}
+
+#elif defined(MI_TLS_PTHREAD)
+
+extern pthread_key_t _mi_heap_default_key;
+static inline mi_heap_t* mi_prim_get_default_heap(void) {
+  mi_heap_t* heap = (mi_unlikely(_mi_heap_default_key == (pthread_key_t)(-1)) ? _mi_heap_main_get() : (mi_heap_t*)pthread_getspecific(_mi_heap_default_key));
+  return (mi_unlikely(heap == NULL) ? (mi_heap_t*)&_mi_heap_empty : heap);
+}
+
+#else // default using a thread local variable; used on most platforms.
+
+static inline mi_heap_t* mi_prim_get_default_heap(void) {
+  #if defined(MI_TLS_RECURSE_GUARD)
+  if (mi_unlikely(!_mi_process_is_initialized)) return _mi_heap_main_get();
+  #endif
+  return _mi_heap_default;
+}
+
+#endif  // mi_prim_get_default_heap()
+
+
+
+#endif  // MIMALLOC_PRIM_H

Dependencies/Python/include/internal/mimalloc/mimalloc/track.h

@@ -0,0 +1,147 @@
+/* ----------------------------------------------------------------------------
+Copyright (c) 2018-2023, Microsoft Research, Daan Leijen
+This is free software; you can redistribute it and/or modify it under the
+terms of the MIT license. A copy of the license can be found in the file
+"LICENSE" at the root of this distribution.
+-----------------------------------------------------------------------------*/
+#pragma once
+#ifndef MIMALLOC_TRACK_H
+#define MIMALLOC_TRACK_H
+
+/* ------------------------------------------------------------------------------------------------------
+Track memory ranges with macros for tools like Valgrind address sanitizer, or other memory checkers.
+These can be defined for tracking allocation:
+
+  #define mi_track_malloc_size(p,reqsize,size,zero)
+  #define mi_track_free_size(p,_size)
+
+The macros are set up such that the size passed to `mi_track_free_size`
+always matches the size of `mi_track_malloc_size`. (currently, `size == mi_usable_size(p)`).
+The `reqsize` is what the user requested, and `size >= reqsize`.
+The `size` is either byte precise (and `size==reqsize`) if `MI_PADDING` is enabled,
+or otherwise it is the usable block size which may be larger than the original request.
+Use `_mi_block_size_of(void* p)` to get the full block size that was allocated (including padding etc).
+The `zero` parameter is `true` if the allocated block is zero initialized.
+
+Optional:
+
+  #define mi_track_align(p,alignedp,offset,size)
+  #define mi_track_resize(p,oldsize,newsize)
+  #define mi_track_init()
+
+The `mi_track_align` is called right after a `mi_track_malloc` for aligned pointers in a block.
+The corresponding `mi_track_free` still uses the block start pointer and original size (corresponding to the `mi_track_malloc`).
+The `mi_track_resize` is currently unused but could be called on reallocations within a block.
+`mi_track_init` is called at program start.
+
+The following macros are for tools like asan and valgrind to track whether memory is
+defined, undefined, or not accessible at all:
+
+  #define mi_track_mem_defined(p,size)
+  #define mi_track_mem_undefined(p,size)
+  #define mi_track_mem_noaccess(p,size)
+
+-------------------------------------------------------------------------------------------------------*/
+
+#if MI_TRACK_VALGRIND
+// valgrind tool
+
+#define MI_TRACK_ENABLED      1
+#define MI_TRACK_HEAP_DESTROY 1           // track free of individual blocks on heap_destroy
+#define MI_TRACK_TOOL         "valgrind"
+
+#include <valgrind/valgrind.h>
+#include <valgrind/memcheck.h>
+
+#define mi_track_malloc_size(p,reqsize,size,zero) VALGRIND_MALLOCLIKE_BLOCK(p,size,MI_PADDING_SIZE /*red zone*/,zero)
+#define mi_track_free_size(p,_size)               VALGRIND_FREELIKE_BLOCK(p,MI_PADDING_SIZE /*red zone*/)
+#define mi_track_resize(p,oldsize,newsize)        VALGRIND_RESIZEINPLACE_BLOCK(p,oldsize,newsize,MI_PADDING_SIZE /*red zone*/)
+#define mi_track_mem_defined(p,size)              VALGRIND_MAKE_MEM_DEFINED(p,size)
+#define mi_track_mem_undefined(p,size)            VALGRIND_MAKE_MEM_UNDEFINED(p,size)
+#define mi_track_mem_noaccess(p,size)             VALGRIND_MAKE_MEM_NOACCESS(p,size)
+
+#elif MI_TRACK_ASAN
+// address sanitizer
+
+#define MI_TRACK_ENABLED      1
+#define MI_TRACK_HEAP_DESTROY 0
+#define MI_TRACK_TOOL         "asan"
+
+#include <sanitizer/asan_interface.h>
+
+#define mi_track_malloc_size(p,reqsize,size,zero) ASAN_UNPOISON_MEMORY_REGION(p,size)
+#define mi_track_free_size(p,size)                ASAN_POISON_MEMORY_REGION(p,size)
+#define mi_track_mem_defined(p,size)              ASAN_UNPOISON_MEMORY_REGION(p,size)
+#define mi_track_mem_undefined(p,size)            ASAN_UNPOISON_MEMORY_REGION(p,size)
+#define mi_track_mem_noaccess(p,size)             ASAN_POISON_MEMORY_REGION(p,size)
+
+#elif MI_TRACK_ETW
+// windows event tracing
+
+#define MI_TRACK_ENABLED      1
+#define MI_TRACK_HEAP_DESTROY 1
+#define MI_TRACK_TOOL         "ETW"
+
+#define WIN32_LEAN_AND_MEAN
+#include <windows.h>
+#include "../src/prim/windows/etw.h"
+
+#define mi_track_init()                           EventRegistermicrosoft_windows_mimalloc();
+#define mi_track_malloc_size(p,reqsize,size,zero) EventWriteETW_MI_ALLOC((UINT64)(p), size)
+#define mi_track_free_size(p,size)                EventWriteETW_MI_FREE((UINT64)(p), size)
+
+#else
+// no tracking
+
+#define MI_TRACK_ENABLED      0
+#define MI_TRACK_HEAP_DESTROY 0
+#define MI_TRACK_TOOL         "none"
+
+#define mi_track_malloc_size(p,reqsize,size,zero)
+#define mi_track_free_size(p,_size)
+
+#endif
+
+// -------------------
+// Utility definitions
+
+#ifndef mi_track_resize
+#define mi_track_resize(p,oldsize,newsize)      mi_track_free_size(p,oldsize); mi_track_malloc(p,newsize,false)
+#endif
+
+#ifndef mi_track_align
+#define mi_track_align(p,alignedp,offset,size)  mi_track_mem_noaccess(p,offset)
+#endif
+
+#ifndef mi_track_init
+#define mi_track_init()
+#endif
+
+#ifndef mi_track_mem_defined
+#define mi_track_mem_defined(p,size)
+#endif
+
+#ifndef mi_track_mem_undefined
+#define mi_track_mem_undefined(p,size)
+#endif
+
+#ifndef mi_track_mem_noaccess
+#define mi_track_mem_noaccess(p,size)
+#endif
+
+
+#if MI_PADDING
+#define mi_track_malloc(p,reqsize,zero) \
+  if ((p)!=NULL) { \
+    mi_assert_internal(mi_usable_size(p)==(reqsize)); \
+    mi_track_malloc_size(p,reqsize,reqsize,zero); \
+  }
+#else
+#define mi_track_malloc(p,reqsize,zero) \
+  if ((p)!=NULL) { \
+    mi_assert_internal(mi_usable_size(p)>=(reqsize)); \
+    mi_track_malloc_size(p,reqsize,mi_usable_size(p),zero); \
+  }
+#endif
+
+#endif

Dependencies/Python/include/internal/mimalloc/mimalloc/types.h

@@ -0,0 +1,721 @@
+/* ----------------------------------------------------------------------------
+Copyright (c) 2018-2023, Microsoft Research, Daan Leijen
+This is free software; you can redistribute it and/or modify it under the
+terms of the MIT license. A copy of the license can be found in the file
+"LICENSE" at the root of this distribution.
+-----------------------------------------------------------------------------*/
+#pragma once
+#ifndef MIMALLOC_TYPES_H
+#define MIMALLOC_TYPES_H
+
+// --------------------------------------------------------------------------
+// This file contains the main type definitions for mimalloc:
+// mi_heap_t      : all data for a thread-local heap, contains
+//                  lists of all managed heap pages.
+// mi_segment_t   : a larger chunk of memory (32GiB) from where pages
+//                  are allocated.
+// mi_page_t      : a mimalloc page (usually 64KiB or 512KiB) from
+//                  where objects are allocated.
+// --------------------------------------------------------------------------
+
+
+#include <stddef.h>   // ptrdiff_t
+#include <stdint.h>   // uintptr_t, uint16_t, etc
+#include "atomic.h"   // _Atomic
+
+#ifdef _MSC_VER
+#pragma warning(disable:4214) // bitfield is not int
+#endif
+
+// Minimal alignment necessary. On most platforms 16 bytes are needed
+// due to SSE registers for example. This must be at least `sizeof(void*)`
+#ifndef MI_MAX_ALIGN_SIZE
+#define MI_MAX_ALIGN_SIZE  16   // sizeof(max_align_t)
+#endif
+
+#define MI_CACHE_LINE          64
+#if defined(_MSC_VER)
+#pragma warning(disable:4127)   // suppress constant conditional warning (due to MI_SECURE paths)
+#pragma warning(disable:26812)  // unscoped enum warning
+#define mi_decl_noinline        __declspec(noinline)
+#define mi_decl_thread          __declspec(thread)
+#define mi_decl_cache_align     __declspec(align(MI_CACHE_LINE))
+#elif (defined(__GNUC__) && (__GNUC__ >= 3)) || defined(__clang__) // includes clang and icc
+#define mi_decl_noinline        __attribute__((noinline))
+#define mi_decl_thread          __thread
+#define mi_decl_cache_align     __attribute__((aligned(MI_CACHE_LINE)))
+#else
+#define mi_decl_noinline
+#define mi_decl_thread          __thread        // hope for the best :-)
+#define mi_decl_cache_align
+#endif
+
+// ------------------------------------------------------
+// Variants
+// ------------------------------------------------------
+
+// Define NDEBUG in the release version to disable assertions.
+// #define NDEBUG
+
+// Define MI_TRACK_<tool> to enable tracking support
+// #define MI_TRACK_VALGRIND 1
+// #define MI_TRACK_ASAN     1
+// #define MI_TRACK_ETW      1
+
+// Define MI_STAT as 1 to maintain statistics; set it to 2 to have detailed statistics (but costs some performance).
+// #define MI_STAT 1
+
+// Define MI_SECURE to enable security mitigations
+// #define MI_SECURE 1  // guard page around metadata
+// #define MI_SECURE 2  // guard page around each mimalloc page
+// #define MI_SECURE 3  // encode free lists (detect corrupted free list (buffer overflow), and invalid pointer free)
+// #define MI_SECURE 4  // checks for double free. (may be more expensive)
+
+#if !defined(MI_SECURE)
+#define MI_SECURE 0
+#endif
+
+// Define MI_DEBUG for debug mode
+// #define MI_DEBUG 1  // basic assertion checks and statistics, check double free, corrupted free list, and invalid pointer free.
+// #define MI_DEBUG 2  // + internal assertion checks
+// #define MI_DEBUG 3  // + extensive internal invariant checking (cmake -DMI_DEBUG_FULL=ON)
+#if !defined(MI_DEBUG)
+#if !defined(NDEBUG) || defined(_DEBUG)
+#define MI_DEBUG 2
+#else
+#define MI_DEBUG 0
+#endif
+#endif
+
+// Reserve extra padding at the end of each block to be more resilient against heap block overflows.
+// The padding can detect buffer overflow on free.
+#if !defined(MI_PADDING) && (MI_SECURE>=3 || MI_DEBUG>=1 || (MI_TRACK_VALGRIND || MI_TRACK_ASAN || MI_TRACK_ETW))
+#define MI_PADDING  1
+#endif
+
+// Check padding bytes; allows byte-precise buffer overflow detection
+#if !defined(MI_PADDING_CHECK) && MI_PADDING && (MI_SECURE>=3 || MI_DEBUG>=1)
+#define MI_PADDING_CHECK 1
+#endif
+
+
+// Encoded free lists allow detection of corrupted free lists
+// and can detect buffer overflows, modify after free, and double `free`s.
+#if (MI_SECURE>=3 || MI_DEBUG>=1)
+#define MI_ENCODE_FREELIST  1
+#endif
+
+
+// We used to abandon huge pages but to eagerly deallocate if freed from another thread,
+// but that makes it not possible to visit them during a heap walk or include them in a
+// `mi_heap_destroy`. We therefore instead reset/decommit the huge blocks if freed from
+// another thread so most memory is available until it gets properly freed by the owning thread.
+// #define MI_HUGE_PAGE_ABANDON 1
+
+
+// ------------------------------------------------------
+// Platform specific values
+// ------------------------------------------------------
+
+// ------------------------------------------------------
+// Size of a pointer.
+// We assume that `sizeof(void*)==sizeof(intptr_t)`
+// and it holds for all platforms we know of.
+//
+// However, the C standard only requires that:
+//  p == (void*)((intptr_t)p))
+// but we also need:
+//  i == (intptr_t)((void*)i)
+// or otherwise one might define an intptr_t type that is larger than a pointer...
+// ------------------------------------------------------
+
+#if INTPTR_MAX > INT64_MAX
+# define MI_INTPTR_SHIFT (4)  // assume 128-bit  (as on arm CHERI for example)
+#elif INTPTR_MAX == INT64_MAX
+# define MI_INTPTR_SHIFT (3)
+#elif INTPTR_MAX == INT32_MAX
+# define MI_INTPTR_SHIFT (2)
+#else
+#error platform pointers must be 32, 64, or 128 bits
+#endif
+
+#if SIZE_MAX == UINT64_MAX
+# define MI_SIZE_SHIFT (3)
+typedef int64_t  mi_ssize_t;
+#elif SIZE_MAX == UINT32_MAX
+# define MI_SIZE_SHIFT (2)
+typedef int32_t  mi_ssize_t;
+#else
+#error platform objects must be 32 or 64 bits
+#endif
+
+#if (SIZE_MAX/2) > LONG_MAX
+# define MI_ZU(x)  x##ULL
+# define MI_ZI(x)  x##LL
+#else
+# define MI_ZU(x)  x##UL
+# define MI_ZI(x)  x##L
+#endif
+
+#define MI_INTPTR_SIZE  (1<<MI_INTPTR_SHIFT)
+#define MI_INTPTR_BITS  (MI_INTPTR_SIZE*8)
+
+#define MI_SIZE_SIZE  (1<<MI_SIZE_SHIFT)
+#define MI_SIZE_BITS  (MI_SIZE_SIZE*8)
+
+#define MI_KiB     (MI_ZU(1024))
+#define MI_MiB     (MI_KiB*MI_KiB)
+#define MI_GiB     (MI_MiB*MI_KiB)
+
+
+// ------------------------------------------------------
+// Main internal data-structures
+// ------------------------------------------------------
+
+// Main tuning parameters for segment and page sizes
+// Sizes for 64-bit (usually divide by two for 32-bit)
+#define MI_SEGMENT_SLICE_SHIFT            (13 + MI_INTPTR_SHIFT)         // 64KiB  (32KiB on 32-bit)
+
+#if MI_INTPTR_SIZE > 4
+#define MI_SEGMENT_SHIFT                  ( 9 + MI_SEGMENT_SLICE_SHIFT)  // 32MiB
+#else
+#define MI_SEGMENT_SHIFT                  ( 7 + MI_SEGMENT_SLICE_SHIFT)  // 4MiB on 32-bit
+#endif
+
+#define MI_SMALL_PAGE_SHIFT               (MI_SEGMENT_SLICE_SHIFT)       // 64KiB
+#define MI_MEDIUM_PAGE_SHIFT              ( 3 + MI_SMALL_PAGE_SHIFT)     // 512KiB
+
+
+// Derived constants
+#define MI_SEGMENT_SIZE                   (MI_ZU(1)<<MI_SEGMENT_SHIFT)
+#define MI_SEGMENT_ALIGN                  MI_SEGMENT_SIZE
+#define MI_SEGMENT_MASK                   ((uintptr_t)(MI_SEGMENT_ALIGN - 1))
+#define MI_SEGMENT_SLICE_SIZE             (MI_ZU(1)<< MI_SEGMENT_SLICE_SHIFT)
+#define MI_SLICES_PER_SEGMENT             (MI_SEGMENT_SIZE / MI_SEGMENT_SLICE_SIZE) // 1024
+
+#define MI_SMALL_PAGE_SIZE                (MI_ZU(1)<<MI_SMALL_PAGE_SHIFT)
+#define MI_MEDIUM_PAGE_SIZE               (MI_ZU(1)<<MI_MEDIUM_PAGE_SHIFT)
+
+#define MI_SMALL_OBJ_SIZE_MAX             (MI_SMALL_PAGE_SIZE/4)   // 8KiB on 64-bit
+#define MI_MEDIUM_OBJ_SIZE_MAX            (MI_MEDIUM_PAGE_SIZE/4)  // 128KiB on 64-bit
+#define MI_MEDIUM_OBJ_WSIZE_MAX           (MI_MEDIUM_OBJ_SIZE_MAX/MI_INTPTR_SIZE)
+#define MI_LARGE_OBJ_SIZE_MAX             (MI_SEGMENT_SIZE/2)      // 32MiB on 64-bit
+#define MI_LARGE_OBJ_WSIZE_MAX            (MI_LARGE_OBJ_SIZE_MAX/MI_INTPTR_SIZE)
+
+// Maximum number of size classes. (spaced exponentially in 12.5% increments)
+#define MI_BIN_HUGE  (73U)
+
+#if (MI_MEDIUM_OBJ_WSIZE_MAX >= 655360)
+#error "mimalloc internal: define more bins"
+#endif
+
+// Maximum slice offset (15)
+#define MI_MAX_SLICE_OFFSET               ((MI_ALIGNMENT_MAX / MI_SEGMENT_SLICE_SIZE) - 1)
+
+// Used as a special value to encode block sizes in 32 bits.
+#define MI_HUGE_BLOCK_SIZE                ((uint32_t)(2*MI_GiB))
+
+// blocks up to this size are always allocated aligned
+#define MI_MAX_ALIGN_GUARANTEE            (8*MI_MAX_ALIGN_SIZE)
+
+// Alignments over MI_ALIGNMENT_MAX are allocated in dedicated huge page segments
+#define MI_ALIGNMENT_MAX                  (MI_SEGMENT_SIZE >> 1)
+
+
+// ------------------------------------------------------
+// Mimalloc pages contain allocated blocks
+// ------------------------------------------------------
+
+// The free lists use encoded next fields
+// (Only actually encodes when MI_ENCODED_FREELIST is defined.)
+typedef uintptr_t  mi_encoded_t;
+
+// thread id's
+typedef size_t     mi_threadid_t;
+
+// free lists contain blocks
+typedef struct mi_block_s {
+  mi_encoded_t next;
+} mi_block_t;
+
+
+// The delayed flags are used for efficient multi-threaded free-ing
+typedef enum mi_delayed_e {
+  MI_USE_DELAYED_FREE   = 0, // push on the owning heap thread delayed list
+  MI_DELAYED_FREEING    = 1, // temporary: another thread is accessing the owning heap
+  MI_NO_DELAYED_FREE    = 2, // optimize: push on page local thread free queue if another block is already in the heap thread delayed free list
+  MI_NEVER_DELAYED_FREE = 3  // sticky, only resets on page reclaim
+} mi_delayed_t;
+
+
+// The `in_full` and `has_aligned` page flags are put in a union to efficiently
+// test if both are false (`full_aligned == 0`) in the `mi_free` routine.
+#if !MI_TSAN
+typedef union mi_page_flags_s {
+  uint8_t full_aligned;
+  struct {
+    uint8_t in_full : 1;
+    uint8_t has_aligned : 1;
+  } x;
+} mi_page_flags_t;
+#else
+// under thread sanitizer, use a byte for each flag to suppress warning, issue #130
+typedef union mi_page_flags_s {
+  uint16_t full_aligned;
+  struct {
+    uint8_t in_full;
+    uint8_t has_aligned;
+  } x;
+} mi_page_flags_t;
+#endif
+
+// Thread free list.
+// We use the bottom 2 bits of the pointer for mi_delayed_t flags
+typedef uintptr_t mi_thread_free_t;
+
+// A page contains blocks of one specific size (`block_size`).
+// Each page has three list of free blocks:
+// `free` for blocks that can be allocated,
+// `local_free` for freed blocks that are not yet available to `mi_malloc`
+// `thread_free` for freed blocks by other threads
+// The `local_free` and `thread_free` lists are migrated to the `free` list
+// when it is exhausted. The separate `local_free` list is necessary to
+// implement a monotonic heartbeat. The `thread_free` list is needed for
+// avoiding atomic operations in the common case.
+//
+//
+// `used - |thread_free|` == actual blocks that are in use (alive)
+// `used - |thread_free| + |free| + |local_free| == capacity`
+//
+// We don't count `freed` (as |free|) but use `used` to reduce
+// the number of memory accesses in the `mi_page_all_free` function(s).
+//
+// Notes:
+// - Access is optimized for `mi_free` and `mi_page_alloc` (in `alloc.c`)
+// - Using `uint16_t` does not seem to slow things down
+// - The size is 8 words on 64-bit which helps the page index calculations
+//   (and 10 words on 32-bit, and encoded free lists add 2 words. Sizes 10
+//    and 12 are still good for address calculation)
+// - To limit the structure size, the `xblock_size` is 32-bits only; for
+//   blocks > MI_HUGE_BLOCK_SIZE the size is determined from the segment page size
+// - `thread_free` uses the bottom bits as a delayed-free flags to optimize
+//   concurrent frees where only the first concurrent free adds to the owning
+//   heap `thread_delayed_free` list (see `alloc.c:mi_free_block_mt`).
+//   The invariant is that no-delayed-free is only set if there is
+//   at least one block that will be added, or as already been added, to
+//   the owning heap `thread_delayed_free` list. This guarantees that pages
+//   will be freed correctly even if only other threads free blocks.
+typedef struct mi_page_s {
+  // "owned" by the segment
+  uint32_t              slice_count;       // slices in this page (0 if not a page)
+  uint32_t              slice_offset;      // distance from the actual page data slice (0 if a page)
+  uint8_t               is_committed : 1;  // `true` if the page virtual memory is committed
+  uint8_t               is_zero_init : 1;  // `true` if the page was initially zero initialized
+  uint8_t               use_qsbr : 1;      // delay page freeing using qsbr
+  uint8_t               tag : 4;           // tag from the owning heap
+  uint8_t               debug_offset;      // number of bytes to preserve when filling freed or uninitialized memory
+
+  // layout like this to optimize access in `mi_malloc` and `mi_free`
+  uint16_t              capacity;          // number of blocks committed, must be the first field, see `segment.c:page_clear`
+  uint16_t              reserved;          // number of blocks reserved in memory
+  mi_page_flags_t       flags;             // `in_full` and `has_aligned` flags (8 bits)
+  uint8_t               free_is_zero : 1;  // `true` if the blocks in the free list are zero initialized
+  uint8_t               retire_expire : 7; // expiration count for retired blocks
+
+  mi_block_t*           free;              // list of available free blocks (`malloc` allocates from this list)
+  uint32_t              used;              // number of blocks in use (including blocks in `local_free` and `thread_free`)
+  uint32_t              xblock_size;       // size available in each block (always `>0`)
+  mi_block_t*           local_free;        // list of deferred free blocks by this thread (migrates to `free`)
+
+  #if (MI_ENCODE_FREELIST || MI_PADDING)
+  uintptr_t             keys[2];           // two random keys to encode the free lists (see `_mi_block_next`) or padding canary
+  #endif
+
+  _Atomic(mi_thread_free_t) xthread_free;  // list of deferred free blocks freed by other threads
+  _Atomic(uintptr_t)        xheap;
+
+  struct mi_page_s*     next;              // next page owned by this thread with the same `block_size`
+  struct mi_page_s*     prev;              // previous page owned by this thread with the same `block_size`
+
+#ifdef Py_GIL_DISABLED
+  struct llist_node     qsbr_node;
+  uint64_t              qsbr_goal;
+#endif
+
+  // 64-bit 9 words, 32-bit 12 words, (+2 for secure)
+  #if MI_INTPTR_SIZE==8 && !defined(Py_GIL_DISABLED)
+  uintptr_t padding[1];
+  #endif
+} mi_page_t;
+
+
+
+// ------------------------------------------------------
+// Mimalloc segments contain mimalloc pages
+// ------------------------------------------------------
+
+typedef enum mi_page_kind_e {
+  MI_PAGE_SMALL,    // small blocks go into 64KiB pages inside a segment
+  MI_PAGE_MEDIUM,   // medium blocks go into medium pages inside a segment
+  MI_PAGE_LARGE,    // larger blocks go into a page of just one block
+  MI_PAGE_HUGE,     // huge blocks (> 16 MiB) are put into a single page in a single segment.
+} mi_page_kind_t;
+
+typedef enum mi_segment_kind_e {
+  MI_SEGMENT_NORMAL, // MI_SEGMENT_SIZE size with pages inside.
+  MI_SEGMENT_HUGE,   // > MI_LARGE_SIZE_MAX segment with just one huge page inside.
+} mi_segment_kind_t;
+
+// ------------------------------------------------------
+// A segment holds a commit mask where a bit is set if
+// the corresponding MI_COMMIT_SIZE area is committed.
+// The MI_COMMIT_SIZE must be a multiple of the slice
+// size. If it is equal we have the most fine grained
+// decommit (but setting it higher can be more efficient).
+// The MI_MINIMAL_COMMIT_SIZE is the minimal amount that will
+// be committed in one go which can be set higher than
+// MI_COMMIT_SIZE for efficiency (while the decommit mask
+// is still tracked in fine-grained MI_COMMIT_SIZE chunks)
+// ------------------------------------------------------
+
+#define MI_MINIMAL_COMMIT_SIZE      (1*MI_SEGMENT_SLICE_SIZE)
+#define MI_COMMIT_SIZE              (MI_SEGMENT_SLICE_SIZE)              // 64KiB
+#define MI_COMMIT_MASK_BITS         (MI_SEGMENT_SIZE / MI_COMMIT_SIZE)
+#define MI_COMMIT_MASK_FIELD_BITS    MI_SIZE_BITS
+#define MI_COMMIT_MASK_FIELD_COUNT  (MI_COMMIT_MASK_BITS / MI_COMMIT_MASK_FIELD_BITS)
+
+#if (MI_COMMIT_MASK_BITS != (MI_COMMIT_MASK_FIELD_COUNT * MI_COMMIT_MASK_FIELD_BITS))
+#error "the segment size must be exactly divisible by the (commit size * size_t bits)"
+#endif
+
+typedef struct mi_commit_mask_s {
+  size_t mask[MI_COMMIT_MASK_FIELD_COUNT];
+} mi_commit_mask_t;
+
+typedef mi_page_t  mi_slice_t;
+typedef int64_t    mi_msecs_t;
+
+
+// Memory can reside in arena's, direct OS allocated, or statically allocated. The memid keeps track of this.
+typedef enum mi_memkind_e {
+  MI_MEM_NONE,      // not allocated
+  MI_MEM_EXTERNAL,  // not owned by mimalloc but provided externally (via `mi_manage_os_memory` for example)
+  MI_MEM_STATIC,    // allocated in a static area and should not be freed (for arena meta data for example)
+  MI_MEM_OS,        // allocated from the OS
+  MI_MEM_OS_HUGE,   // allocated as huge os pages
+  MI_MEM_OS_REMAP,  // allocated in a remapable area (i.e. using `mremap`)
+  MI_MEM_ARENA      // allocated from an arena (the usual case)
+} mi_memkind_t;
+
+static inline bool mi_memkind_is_os(mi_memkind_t memkind) {
+  return (memkind >= MI_MEM_OS && memkind <= MI_MEM_OS_REMAP);
+}
+
+typedef struct mi_memid_os_info {
+  void*         base;               // actual base address of the block (used for offset aligned allocations)
+  size_t        alignment;          // alignment at allocation
+} mi_memid_os_info_t;
+
+typedef struct mi_memid_arena_info {
+  size_t        block_index;        // index in the arena
+  mi_arena_id_t id;                 // arena id (>= 1)
+  bool          is_exclusive;       // the arena can only be used for specific arena allocations
+} mi_memid_arena_info_t;
+
+typedef struct mi_memid_s {
+  union {
+    mi_memid_os_info_t    os;       // only used for MI_MEM_OS
+    mi_memid_arena_info_t arena;    // only used for MI_MEM_ARENA
+  } mem;
+  bool          is_pinned;          // `true` if we cannot decommit/reset/protect in this memory (e.g. when allocated using large OS pages)
+  bool          initially_committed;// `true` if the memory was originally allocated as committed
+  bool          initially_zero;     // `true` if the memory was originally zero initialized
+  mi_memkind_t  memkind;
+} mi_memid_t;
+
+
+// Segments are large allocated memory blocks (8mb on 64 bit) from
+// the OS. Inside segments we allocated fixed size _pages_ that
+// contain blocks.
+typedef struct mi_segment_s {
+  // constant fields
+  mi_memid_t        memid;              // memory id for arena allocation
+  bool              allow_decommit;
+  bool              allow_purge;
+  size_t            segment_size;
+
+  // segment fields
+  mi_msecs_t        purge_expire;
+  mi_commit_mask_t  purge_mask;
+  mi_commit_mask_t  commit_mask;
+
+  _Atomic(struct mi_segment_s*) abandoned_next;
+
+  // from here is zero initialized
+  struct mi_segment_s* next;            // the list of freed segments in the cache (must be first field, see `segment.c:mi_segment_init`)
+
+  size_t            abandoned;          // abandoned pages (i.e. the original owning thread stopped) (`abandoned <= used`)
+  size_t            abandoned_visits;   // count how often this segment is visited in the abandoned list (to force reclaim it it is too long)
+  size_t            used;               // count of pages in use
+  uintptr_t         cookie;             // verify addresses in debug mode: `mi_ptr_cookie(segment) == segment->cookie`
+
+  size_t            segment_slices;      // for huge segments this may be different from `MI_SLICES_PER_SEGMENT`
+  size_t            segment_info_slices; // initial slices we are using segment info and possible guard pages.
+
+  // layout like this to optimize access in `mi_free`
+  mi_segment_kind_t kind;
+  size_t            slice_entries;       // entries in the `slices` array, at most `MI_SLICES_PER_SEGMENT`
+  _Atomic(mi_threadid_t) thread_id;      // unique id of the thread owning this segment
+
+  mi_slice_t        slices[MI_SLICES_PER_SEGMENT+1];  // one more for huge blocks with large alignment
+} mi_segment_t;
+
+typedef uintptr_t        mi_tagged_segment_t;
+
+// Segments unowned by any thread are put in a shared pool
+typedef struct mi_abandoned_pool_s {
+  // This is a list of visited abandoned pages that were full at the time.
+  // this list migrates to `abandoned` when that becomes NULL. The use of
+  // this list reduces contention and the rate at which segments are visited.
+  mi_decl_cache_align _Atomic(mi_segment_t*)       abandoned_visited; // = NULL
+
+  // The abandoned page list (tagged as it supports pop)
+  mi_decl_cache_align _Atomic(mi_tagged_segment_t) abandoned;         // = NULL
+
+  // Maintain these for debug purposes (these counts may be a bit off)
+  mi_decl_cache_align _Atomic(size_t)           abandoned_count;
+  mi_decl_cache_align _Atomic(size_t)           abandoned_visited_count;
+
+  // We also maintain a count of current readers of the abandoned list
+  // in order to prevent resetting/decommitting segment memory if it might
+  // still be read.
+  mi_decl_cache_align _Atomic(size_t)           abandoned_readers; // = 0
+} mi_abandoned_pool_t;
+
+
+// ------------------------------------------------------
+// Heaps
+// Provide first-class heaps to allocate from.
+// A heap just owns a set of pages for allocation and
+// can only be allocate/reallocate from the thread that created it.
+// Freeing blocks can be done from any thread though.
+// Per thread, the segments are shared among its heaps.
+// Per thread, there is always a default heap that is
+// used for allocation; it is initialized to statically
+// point to an empty heap to avoid initialization checks
+// in the fast path.
+// ------------------------------------------------------
+
+// Thread local data
+typedef struct mi_tld_s mi_tld_t;
+
+// Pages of a certain block size are held in a queue.
+typedef struct mi_page_queue_s {
+  mi_page_t* first;
+  mi_page_t* last;
+  size_t     block_size;
+} mi_page_queue_t;
+
+#define MI_BIN_FULL  (MI_BIN_HUGE+1)
+
+// Random context
+typedef struct mi_random_cxt_s {
+  uint32_t input[16];
+  uint32_t output[16];
+  int      output_available;
+  bool     weak;
+} mi_random_ctx_t;
+
+
+// In debug mode there is a padding structure at the end of the blocks to check for buffer overflows
+#if (MI_PADDING)
+typedef struct mi_padding_s {
+  uint32_t canary; // encoded block value to check validity of the padding (in case of overflow)
+  uint32_t delta;  // padding bytes before the block. (mi_usable_size(p) - delta == exact allocated bytes)
+} mi_padding_t;
+#define MI_PADDING_SIZE   (sizeof(mi_padding_t))
+#define MI_PADDING_WSIZE  ((MI_PADDING_SIZE + MI_INTPTR_SIZE - 1) / MI_INTPTR_SIZE)
+#else
+#define MI_PADDING_SIZE   0
+#define MI_PADDING_WSIZE  0
+#endif
+
+#define MI_PAGES_DIRECT   (MI_SMALL_WSIZE_MAX + MI_PADDING_WSIZE + 1)
+
+
+// A heap owns a set of pages.
+struct mi_heap_s {
+  mi_tld_t*             tld;
+  mi_page_t*            pages_free_direct[MI_PAGES_DIRECT];  // optimize: array where every entry points a page with possibly free blocks in the corresponding queue for that size.
+  mi_page_queue_t       pages[MI_BIN_FULL + 1];              // queue of pages for each size class (or "bin")
+  _Atomic(mi_block_t*)  thread_delayed_free;
+  mi_threadid_t         thread_id;                           // thread this heap belongs too
+  mi_arena_id_t         arena_id;                            // arena id if the heap belongs to a specific arena (or 0)
+  uintptr_t             cookie;                              // random cookie to verify pointers (see `_mi_ptr_cookie`)
+  uintptr_t             keys[2];                             // two random keys used to encode the `thread_delayed_free` list
+  mi_random_ctx_t       random;                              // random number context used for secure allocation
+  size_t                page_count;                          // total number of pages in the `pages` queues.
+  size_t                page_retired_min;                    // smallest retired index (retired pages are fully free, but still in the page queues)
+  size_t                page_retired_max;                    // largest retired index into the `pages` array.
+  mi_heap_t*            next;                                // list of heaps per thread
+  bool                  no_reclaim;                          // `true` if this heap should not reclaim abandoned pages
+  uint8_t               tag;                                 // custom identifier for this heap
+  uint8_t               debug_offset;                        // number of bytes to preserve when filling freed or uninitialized memory
+  bool                  page_use_qsbr;                       // should freeing pages be delayed using QSBR
+};
+
+
+
+// ------------------------------------------------------
+// Debug
+// ------------------------------------------------------
+
+#if !defined(MI_DEBUG_UNINIT)
+#define MI_DEBUG_UNINIT     (0xD0)
+#endif
+#if !defined(MI_DEBUG_FREED)
+#define MI_DEBUG_FREED      (0xDF)
+#endif
+#if !defined(MI_DEBUG_PADDING)
+#define MI_DEBUG_PADDING    (0xDE)
+#endif
+
+#if (MI_DEBUG)
+// use our own assertion to print without memory allocation
+void _mi_assert_fail(const char* assertion, const char* fname, unsigned int line, const char* func );
+#define mi_assert(expr)     ((expr) ? (void)0 : _mi_assert_fail(#expr,__FILE__,__LINE__,__func__))
+#else
+#define mi_assert(x)
+#endif
+
+#if (MI_DEBUG>1)
+#define mi_assert_internal    mi_assert
+#else
+#define mi_assert_internal(x)
+#endif
+
+#if (MI_DEBUG>2)
+#define mi_assert_expensive   mi_assert
+#else
+#define mi_assert_expensive(x)
+#endif
+
+// ------------------------------------------------------
+// Statistics
+// ------------------------------------------------------
+
+#ifndef MI_STAT
+#if (MI_DEBUG>0)
+#define MI_STAT 2
+#else
+#define MI_STAT 0
+#endif
+#endif
+
+typedef struct mi_stat_count_s {
+  int64_t allocated;
+  int64_t freed;
+  int64_t peak;
+  int64_t current;
+} mi_stat_count_t;
+
+typedef struct mi_stat_counter_s {
+  int64_t total;
+  int64_t count;
+} mi_stat_counter_t;
+
+typedef struct mi_stats_s {
+  mi_stat_count_t segments;
+  mi_stat_count_t pages;
+  mi_stat_count_t reserved;
+  mi_stat_count_t committed;
+  mi_stat_count_t reset;
+  mi_stat_count_t purged;
+  mi_stat_count_t page_committed;
+  mi_stat_count_t segments_abandoned;
+  mi_stat_count_t pages_abandoned;
+  mi_stat_count_t threads;
+  mi_stat_count_t normal;
+  mi_stat_count_t huge;
+  mi_stat_count_t large;
+  mi_stat_count_t malloc;
+  mi_stat_count_t segments_cache;
+  mi_stat_counter_t pages_extended;
+  mi_stat_counter_t mmap_calls;
+  mi_stat_counter_t commit_calls;
+  mi_stat_counter_t reset_calls;
+  mi_stat_counter_t purge_calls;
+  mi_stat_counter_t page_no_retire;
+  mi_stat_counter_t searches;
+  mi_stat_counter_t normal_count;
+  mi_stat_counter_t huge_count;
+  mi_stat_counter_t large_count;
+#if MI_STAT>1
+  mi_stat_count_t normal_bins[MI_BIN_HUGE+1];
+#endif
+} mi_stats_t;
+
+
+void _mi_stat_increase(mi_stat_count_t* stat, size_t amount);
+void _mi_stat_decrease(mi_stat_count_t* stat, size_t amount);
+void _mi_stat_counter_increase(mi_stat_counter_t* stat, size_t amount);
+
+#if (MI_STAT)
+#define mi_stat_increase(stat,amount)         _mi_stat_increase( &(stat), amount)
+#define mi_stat_decrease(stat,amount)         _mi_stat_decrease( &(stat), amount)
+#define mi_stat_counter_increase(stat,amount) _mi_stat_counter_increase( &(stat), amount)
+#else
+#define mi_stat_increase(stat,amount)         (void)0
+#define mi_stat_decrease(stat,amount)         (void)0
+#define mi_stat_counter_increase(stat,amount) (void)0
+#endif
+
+#define mi_heap_stat_counter_increase(heap,stat,amount)  mi_stat_counter_increase( (heap)->tld->stats.stat, amount)
+#define mi_heap_stat_increase(heap,stat,amount)  mi_stat_increase( (heap)->tld->stats.stat, amount)
+#define mi_heap_stat_decrease(heap,stat,amount)  mi_stat_decrease( (heap)->tld->stats.stat, amount)
+
+// ------------------------------------------------------
+// Thread Local data
+// ------------------------------------------------------
+
+// A "span" is is an available range of slices. The span queues keep
+// track of slice spans of at most the given `slice_count` (but more than the previous size class).
+typedef struct mi_span_queue_s {
+  mi_slice_t* first;
+  mi_slice_t* last;
+  size_t      slice_count;
+} mi_span_queue_t;
+
+#define MI_SEGMENT_BIN_MAX (35)     // 35 == mi_segment_bin(MI_SLICES_PER_SEGMENT)
+
+// OS thread local data
+typedef struct mi_os_tld_s {
+  size_t                region_idx;   // start point for next allocation
+  mi_stats_t*           stats;        // points to tld stats
+} mi_os_tld_t;
+
+
+// Segments thread local data
+typedef struct mi_segments_tld_s {
+  mi_span_queue_t     spans[MI_SEGMENT_BIN_MAX+1];  // free slice spans inside segments
+  size_t              count;        // current number of segments;
+  size_t              peak_count;   // peak number of segments
+  size_t              current_size; // current size of all segments
+  size_t              peak_size;    // peak size of all segments
+  mi_stats_t*         stats;        // points to tld stats
+  mi_os_tld_t*        os;           // points to os stats
+  mi_abandoned_pool_t* abandoned;   // pool of abandoned segments
+} mi_segments_tld_t;
+
+// Thread local data
+struct mi_tld_s {
+  unsigned long long  heartbeat;     // monotonic heartbeat count
+  bool                recurse;       // true if deferred was called; used to prevent infinite recursion.
+  mi_heap_t*          heap_backing;  // backing heap of this thread (cannot be deleted)
+  mi_heap_t*          heaps;         // list of heaps in this thread (so we can abandon all when the thread terminates)
+  mi_segments_tld_t   segments;      // segment tld
+  mi_os_tld_t         os;            // os tld
+  mi_stats_t          stats;         // statistics
+};
+
+#endif

Dependencies/Python/include/internal/pycore_abstract.h

@@ -0,0 +1,61 @@
+#ifndef Py_INTERNAL_ABSTRACT_H
+#define Py_INTERNAL_ABSTRACT_H
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+#ifndef Py_BUILD_CORE
+#  error "this header requires Py_BUILD_CORE define"
+#endif
+
+// Fast inlined version of PyIndex_Check()
+static inline int
+_PyIndex_Check(PyObject *obj)
+{
+    PyNumberMethods *tp_as_number = Py_TYPE(obj)->tp_as_number;
+    return (tp_as_number != NULL && tp_as_number->nb_index != NULL);
+}
+
+PyObject *_PyNumber_PowerNoMod(PyObject *lhs, PyObject *rhs);
+PyObject *_PyNumber_InPlacePowerNoMod(PyObject *lhs, PyObject *rhs);
+
+extern int _PyObject_HasLen(PyObject *o);
+
+/* === Sequence protocol ================================================ */
+
+#define PY_ITERSEARCH_COUNT    1
+#define PY_ITERSEARCH_INDEX    2
+#define PY_ITERSEARCH_CONTAINS 3
+
+/* Iterate over seq.
+
+   Result depends on the operation:
+
+   PY_ITERSEARCH_COUNT:  return # of times obj appears in seq; -1 if
+     error.
+   PY_ITERSEARCH_INDEX:  return 0-based index of first occurrence of
+     obj in seq; set ValueError and return -1 if none found;
+     also return -1 on error.
+   PY_ITERSEARCH_CONTAINS:  return 1 if obj in seq, else 0; -1 on
+     error. */
+extern Py_ssize_t _PySequence_IterSearch(PyObject *seq,
+                                         PyObject *obj, int operation);
+
+/* === Mapping protocol ================================================= */
+
+extern int _PyObject_RealIsInstance(PyObject *inst, PyObject *cls);
+
+extern int _PyObject_RealIsSubclass(PyObject *derived, PyObject *cls);
+
+// Convert Python int to Py_ssize_t. Do nothing if the argument is None.
+// Export for '_bisect' shared extension.
+PyAPI_FUNC(int) _Py_convert_optional_to_ssize_t(PyObject *, void *);
+
+// Same as PyNumber_Index() but can return an instance of a subclass of int.
+// Export for 'math' shared extension.
+PyAPI_FUNC(PyObject*) _PyNumber_Index(PyObject *o);
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* !Py_INTERNAL_ABSTRACT_H */

Dependencies/Python/include/internal/pycore_asdl.h

@@ -0,0 +1,112 @@
+#ifndef Py_INTERNAL_ASDL_H
+#define Py_INTERNAL_ASDL_H
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+#ifndef Py_BUILD_CORE
+#  error "this header requires Py_BUILD_CORE define"
+#endif
+
+#include "pycore_pyarena.h"       // _PyArena_Malloc()
+
+typedef PyObject * identifier;
+typedef PyObject * string;
+typedef PyObject * object;
+typedef PyObject * constant;
+
+/* It would be nice if the code generated by asdl_c.py was completely
+   independent of Python, but it is a goal the requires too much work
+   at this stage.  So, for example, I'll represent identifiers as
+   interned Python strings.
+*/
+
+#define _ASDL_SEQ_HEAD \
+    Py_ssize_t size;   \
+    void **elements;
+
+typedef struct {
+    _ASDL_SEQ_HEAD
+} asdl_seq;
+
+typedef struct {
+    _ASDL_SEQ_HEAD
+    void *typed_elements[1];
+} asdl_generic_seq;
+
+typedef struct {
+    _ASDL_SEQ_HEAD
+    PyObject *typed_elements[1];
+} asdl_identifier_seq;
+
+typedef struct {
+    _ASDL_SEQ_HEAD
+    int typed_elements[1];
+} asdl_int_seq;
+
+asdl_generic_seq *_Py_asdl_generic_seq_new(Py_ssize_t size, PyArena *arena);
+asdl_identifier_seq *_Py_asdl_identifier_seq_new(Py_ssize_t size, PyArena *arena);
+asdl_int_seq *_Py_asdl_int_seq_new(Py_ssize_t size, PyArena *arena);
+
+
+#define GENERATE_ASDL_SEQ_CONSTRUCTOR(NAME, TYPE) \
+asdl_ ## NAME ## _seq *_Py_asdl_ ## NAME ## _seq_new(Py_ssize_t size, PyArena *arena) \
+{ \
+    asdl_ ## NAME ## _seq *seq = NULL; \
+    size_t n; \
+    /* check size is sane */ \
+    if (size < 0 || \
+        (size && (((size_t)size - 1) > (SIZE_MAX / sizeof(void *))))) { \
+        PyErr_NoMemory(); \
+        return NULL; \
+    } \
+    n = (size ? (sizeof(TYPE *) * (size - 1)) : 0); \
+    /* check if size can be added safely */ \
+    if (n > SIZE_MAX - sizeof(asdl_ ## NAME ## _seq)) { \
+        PyErr_NoMemory(); \
+        return NULL; \
+    } \
+    n += sizeof(asdl_ ## NAME ## _seq); \
+    seq = (asdl_ ## NAME ## _seq *)_PyArena_Malloc(arena, n); \
+    if (!seq) { \
+        PyErr_NoMemory(); \
+        return NULL; \
+    } \
+    memset(seq, 0, n); \
+    seq->size = size; \
+    seq->elements = (void**)seq->typed_elements; \
+    return seq; \
+}
+
+#define asdl_seq_GET_UNTYPED(S, I) _Py_RVALUE((S)->elements[(I)])
+#define asdl_seq_GET(S, I) _Py_RVALUE((S)->typed_elements[(I)])
+#define asdl_seq_LEN(S) _Py_RVALUE(((S) == NULL ? 0 : (S)->size))
+
+#ifdef Py_DEBUG
+#  define asdl_seq_SET(S, I, V) \
+    do { \
+        Py_ssize_t _asdl_i = (I); \
+        assert((S) != NULL); \
+        assert(0 <= _asdl_i && _asdl_i < (S)->size); \
+        (S)->typed_elements[_asdl_i] = (V); \
+    } while (0)
+#else
+#  define asdl_seq_SET(S, I, V) _Py_RVALUE((S)->typed_elements[(I)] = (V))
+#endif
+
+#ifdef Py_DEBUG
+#  define asdl_seq_SET_UNTYPED(S, I, V) \
+    do { \
+        Py_ssize_t _asdl_i = (I); \
+        assert((S) != NULL); \
+        assert(0 <= _asdl_i && _asdl_i < (S)->size); \
+        (S)->elements[_asdl_i] = (V); \
+    } while (0)
+#else
+#  define asdl_seq_SET_UNTYPED(S, I, V) _Py_RVALUE((S)->elements[(I)] = (V))
+#endif
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* !Py_INTERNAL_ASDL_H */

Dependencies/Python/include/internal/pycore_ast.h

@@ -0,0 +1,926 @@
+// File automatically generated by Parser/asdl_c.py.
+
+#ifndef Py_INTERNAL_AST_H
+#define Py_INTERNAL_AST_H
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+#ifndef Py_BUILD_CORE
+#  error "this header requires Py_BUILD_CORE define"
+#endif
+
+#include "pycore_asdl.h"          // _ASDL_SEQ_HEAD
+
+typedef struct _mod *mod_ty;
+
+typedef struct _stmt *stmt_ty;
+
+typedef struct _expr *expr_ty;
+
+typedef enum _expr_context { Load=1, Store=2, Del=3 } expr_context_ty;
+
+typedef enum _boolop { And=1, Or=2 } boolop_ty;
+
+typedef enum _operator { Add=1, Sub=2, Mult=3, MatMult=4, Div=5, Mod=6, Pow=7,
+                         LShift=8, RShift=9, BitOr=10, BitXor=11, BitAnd=12,
+                         FloorDiv=13 } operator_ty;
+
+typedef enum _unaryop { Invert=1, Not=2, UAdd=3, USub=4 } unaryop_ty;
+
+typedef enum _cmpop { Eq=1, NotEq=2, Lt=3, LtE=4, Gt=5, GtE=6, Is=7, IsNot=8,
+                      In=9, NotIn=10 } cmpop_ty;
+
+typedef struct _comprehension *comprehension_ty;
+
+typedef struct _excepthandler *excepthandler_ty;
+
+typedef struct _arguments *arguments_ty;
+
+typedef struct _arg *arg_ty;
+
+typedef struct _keyword *keyword_ty;
+
+typedef struct _alias *alias_ty;
+
+typedef struct _withitem *withitem_ty;
+
+typedef struct _match_case *match_case_ty;
+
+typedef struct _pattern *pattern_ty;
+
+typedef struct _type_ignore *type_ignore_ty;
+
+typedef struct _type_param *type_param_ty;
+
+
+typedef struct {
+    _ASDL_SEQ_HEAD
+    mod_ty typed_elements[1];
+} asdl_mod_seq;
+
+asdl_mod_seq *_Py_asdl_mod_seq_new(Py_ssize_t size, PyArena *arena);
+
+typedef struct {
+    _ASDL_SEQ_HEAD
+    stmt_ty typed_elements[1];
+} asdl_stmt_seq;
+
+asdl_stmt_seq *_Py_asdl_stmt_seq_new(Py_ssize_t size, PyArena *arena);
+
+typedef struct {
+    _ASDL_SEQ_HEAD
+    expr_ty typed_elements[1];
+} asdl_expr_seq;
+
+asdl_expr_seq *_Py_asdl_expr_seq_new(Py_ssize_t size, PyArena *arena);
+
+typedef struct {
+    _ASDL_SEQ_HEAD
+    comprehension_ty typed_elements[1];
+} asdl_comprehension_seq;
+
+asdl_comprehension_seq *_Py_asdl_comprehension_seq_new(Py_ssize_t size, PyArena
+                                                       *arena);
+
+typedef struct {
+    _ASDL_SEQ_HEAD
+    excepthandler_ty typed_elements[1];
+} asdl_excepthandler_seq;
+
+asdl_excepthandler_seq *_Py_asdl_excepthandler_seq_new(Py_ssize_t size, PyArena
+                                                       *arena);
+
+typedef struct {
+    _ASDL_SEQ_HEAD
+    arguments_ty typed_elements[1];
+} asdl_arguments_seq;
+
+asdl_arguments_seq *_Py_asdl_arguments_seq_new(Py_ssize_t size, PyArena *arena);
+
+typedef struct {
+    _ASDL_SEQ_HEAD
+    arg_ty typed_elements[1];
+} asdl_arg_seq;
+
+asdl_arg_seq *_Py_asdl_arg_seq_new(Py_ssize_t size, PyArena *arena);
+
+typedef struct {
+    _ASDL_SEQ_HEAD
+    keyword_ty typed_elements[1];
+} asdl_keyword_seq;
+
+asdl_keyword_seq *_Py_asdl_keyword_seq_new(Py_ssize_t size, PyArena *arena);
+
+typedef struct {
+    _ASDL_SEQ_HEAD
+    alias_ty typed_elements[1];
+} asdl_alias_seq;
+
+asdl_alias_seq *_Py_asdl_alias_seq_new(Py_ssize_t size, PyArena *arena);
+
+typedef struct {
+    _ASDL_SEQ_HEAD
+    withitem_ty typed_elements[1];
+} asdl_withitem_seq;
+
+asdl_withitem_seq *_Py_asdl_withitem_seq_new(Py_ssize_t size, PyArena *arena);
+
+typedef struct {
+    _ASDL_SEQ_HEAD
+    match_case_ty typed_elements[1];
+} asdl_match_case_seq;
+
+asdl_match_case_seq *_Py_asdl_match_case_seq_new(Py_ssize_t size, PyArena
+                                                 *arena);
+
+typedef struct {
+    _ASDL_SEQ_HEAD
+    pattern_ty typed_elements[1];
+} asdl_pattern_seq;
+
+asdl_pattern_seq *_Py_asdl_pattern_seq_new(Py_ssize_t size, PyArena *arena);
+
+typedef struct {
+    _ASDL_SEQ_HEAD
+    type_ignore_ty typed_elements[1];
+} asdl_type_ignore_seq;
+
+asdl_type_ignore_seq *_Py_asdl_type_ignore_seq_new(Py_ssize_t size, PyArena
+                                                   *arena);
+
+typedef struct {
+    _ASDL_SEQ_HEAD
+    type_param_ty typed_elements[1];
+} asdl_type_param_seq;
+
+asdl_type_param_seq *_Py_asdl_type_param_seq_new(Py_ssize_t size, PyArena
+                                                 *arena);
+
+
+enum _mod_kind {Module_kind=1, Interactive_kind=2, Expression_kind=3,
+                 FunctionType_kind=4};
+struct _mod {
+    enum _mod_kind kind;
+    union {
+        struct {
+            asdl_stmt_seq *body;
+            asdl_type_ignore_seq *type_ignores;
+        } Module;
+
+        struct {
+            asdl_stmt_seq *body;
+        } Interactive;
+
+        struct {
+            expr_ty body;
+        } Expression;
+
+        struct {
+            asdl_expr_seq *argtypes;
+            expr_ty returns;
+        } FunctionType;
+
+    } v;
+};
+
+enum _stmt_kind {FunctionDef_kind=1, AsyncFunctionDef_kind=2, ClassDef_kind=3,
+                  Return_kind=4, Delete_kind=5, Assign_kind=6,
+                  TypeAlias_kind=7, AugAssign_kind=8, AnnAssign_kind=9,
+                  For_kind=10, AsyncFor_kind=11, While_kind=12, If_kind=13,
+                  With_kind=14, AsyncWith_kind=15, Match_kind=16,
+                  Raise_kind=17, Try_kind=18, TryStar_kind=19, Assert_kind=20,
+                  Import_kind=21, ImportFrom_kind=22, Global_kind=23,
+                  Nonlocal_kind=24, Expr_kind=25, Pass_kind=26, Break_kind=27,
+                  Continue_kind=28};
+struct _stmt {
+    enum _stmt_kind kind;
+    union {
+        struct {
+            identifier name;
+            arguments_ty args;
+            asdl_stmt_seq *body;
+            asdl_expr_seq *decorator_list;
+            expr_ty returns;
+            string type_comment;
+            asdl_type_param_seq *type_params;
+        } FunctionDef;
+
+        struct {
+            identifier name;
+            arguments_ty args;
+            asdl_stmt_seq *body;
+            asdl_expr_seq *decorator_list;
+            expr_ty returns;
+            string type_comment;
+            asdl_type_param_seq *type_params;
+        } AsyncFunctionDef;
+
+        struct {
+            identifier name;
+            asdl_expr_seq *bases;
+            asdl_keyword_seq *keywords;
+            asdl_stmt_seq *body;
+            asdl_expr_seq *decorator_list;
+            asdl_type_param_seq *type_params;
+        } ClassDef;
+
+        struct {
+            expr_ty value;
+        } Return;
+
+        struct {
+            asdl_expr_seq *targets;
+        } Delete;
+
+        struct {
+            asdl_expr_seq *targets;
+            expr_ty value;
+            string type_comment;
+        } Assign;
+
+        struct {
+            expr_ty name;
+            asdl_type_param_seq *type_params;
+            expr_ty value;
+        } TypeAlias;
+
+        struct {
+            expr_ty target;
+            operator_ty op;
+            expr_ty value;
+        } AugAssign;
+
+        struct {
+            expr_ty target;
+            expr_ty annotation;
+            expr_ty value;
+            int simple;
+        } AnnAssign;
+
+        struct {
+            expr_ty target;
+            expr_ty iter;
+            asdl_stmt_seq *body;
+            asdl_stmt_seq *orelse;
+            string type_comment;
+        } For;
+
+        struct {
+            expr_ty target;
+            expr_ty iter;
+            asdl_stmt_seq *body;
+            asdl_stmt_seq *orelse;
+            string type_comment;
+        } AsyncFor;
+
+        struct {
+            expr_ty test;
+            asdl_stmt_seq *body;
+            asdl_stmt_seq *orelse;
+        } While;
+
+        struct {
+            expr_ty test;
+            asdl_stmt_seq *body;
+            asdl_stmt_seq *orelse;
+        } If;
+
+        struct {
+            asdl_withitem_seq *items;
+            asdl_stmt_seq *body;
+            string type_comment;
+        } With;
+
+        struct {
+            asdl_withitem_seq *items;
+            asdl_stmt_seq *body;
+            string type_comment;
+        } AsyncWith;
+
+        struct {
+            expr_ty subject;
+            asdl_match_case_seq *cases;
+        } Match;
+
+        struct {
+            expr_ty exc;
+            expr_ty cause;
+        } Raise;
+
+        struct {
+            asdl_stmt_seq *body;
+            asdl_excepthandler_seq *handlers;
+            asdl_stmt_seq *orelse;
+            asdl_stmt_seq *finalbody;
+        } Try;
+
+        struct {
+            asdl_stmt_seq *body;
+            asdl_excepthandler_seq *handlers;
+            asdl_stmt_seq *orelse;
+            asdl_stmt_seq *finalbody;
+        } TryStar;
+
+        struct {
+            expr_ty test;
+            expr_ty msg;
+        } Assert;
+
+        struct {
+            asdl_alias_seq *names;
+        } Import;
+
+        struct {
+            identifier module;
+            asdl_alias_seq *names;
+            int level;
+        } ImportFrom;
+
+        struct {
+            asdl_identifier_seq *names;
+        } Global;
+
+        struct {
+            asdl_identifier_seq *names;
+        } Nonlocal;
+
+        struct {
+            expr_ty value;
+        } Expr;
+
+    } v;
+    int lineno;
+    int col_offset;
+    int end_lineno;
+    int end_col_offset;
+};
+
+enum _expr_kind {BoolOp_kind=1, NamedExpr_kind=2, BinOp_kind=3, UnaryOp_kind=4,
+                  Lambda_kind=5, IfExp_kind=6, Dict_kind=7, Set_kind=8,
+                  ListComp_kind=9, SetComp_kind=10, DictComp_kind=11,
+                  GeneratorExp_kind=12, Await_kind=13, Yield_kind=14,
+                  YieldFrom_kind=15, Compare_kind=16, Call_kind=17,
+                  FormattedValue_kind=18, JoinedStr_kind=19, Constant_kind=20,
+                  Attribute_kind=21, Subscript_kind=22, Starred_kind=23,
+                  Name_kind=24, List_kind=25, Tuple_kind=26, Slice_kind=27};
+struct _expr {
+    enum _expr_kind kind;
+    union {
+        struct {
+            boolop_ty op;
+            asdl_expr_seq *values;
+        } BoolOp;
+
+        struct {
+            expr_ty target;
+            expr_ty value;
+        } NamedExpr;
+
+        struct {
+            expr_ty left;
+            operator_ty op;
+            expr_ty right;
+        } BinOp;
+
+        struct {
+            unaryop_ty op;
+            expr_ty operand;
+        } UnaryOp;
+
+        struct {
+            arguments_ty args;
+            expr_ty body;
+        } Lambda;
+
+        struct {
+            expr_ty test;
+            expr_ty body;
+            expr_ty orelse;
+        } IfExp;
+
+        struct {
+            asdl_expr_seq *keys;
+            asdl_expr_seq *values;
+        } Dict;
+
+        struct {
+            asdl_expr_seq *elts;
+        } Set;
+
+        struct {
+            expr_ty elt;
+            asdl_comprehension_seq *generators;
+        } ListComp;
+
+        struct {
+            expr_ty elt;
+            asdl_comprehension_seq *generators;
+        } SetComp;
+
+        struct {
+            expr_ty key;
+            expr_ty value;
+            asdl_comprehension_seq *generators;
+        } DictComp;
+
+        struct {
+            expr_ty elt;
+            asdl_comprehension_seq *generators;
+        } GeneratorExp;
+
+        struct {
+            expr_ty value;
+        } Await;
+
+        struct {
+            expr_ty value;
+        } Yield;
+
+        struct {
+            expr_ty value;
+        } YieldFrom;
+
+        struct {
+            expr_ty left;
+            asdl_int_seq *ops;
+            asdl_expr_seq *comparators;
+        } Compare;
+
+        struct {
+            expr_ty func;
+            asdl_expr_seq *args;
+            asdl_keyword_seq *keywords;
+        } Call;
+
+        struct {
+            expr_ty value;
+            int conversion;
+            expr_ty format_spec;
+        } FormattedValue;
+
+        struct {
+            asdl_expr_seq *values;
+        } JoinedStr;
+
+        struct {
+            constant value;
+            string kind;
+        } Constant;
+
+        struct {
+            expr_ty value;
+            identifier attr;
+            expr_context_ty ctx;
+        } Attribute;
+
+        struct {
+            expr_ty value;
+            expr_ty slice;
+            expr_context_ty ctx;
+        } Subscript;
+
+        struct {
+            expr_ty value;
+            expr_context_ty ctx;
+        } Starred;
+
+        struct {
+            identifier id;
+            expr_context_ty ctx;
+        } Name;
+
+        struct {
+            asdl_expr_seq *elts;
+            expr_context_ty ctx;
+        } List;
+
+        struct {
+            asdl_expr_seq *elts;
+            expr_context_ty ctx;
+        } Tuple;
+
+        struct {
+            expr_ty lower;
+            expr_ty upper;
+            expr_ty step;
+        } Slice;
+
+    } v;
+    int lineno;
+    int col_offset;
+    int end_lineno;
+    int end_col_offset;
+};
+
+struct _comprehension {
+    expr_ty target;
+    expr_ty iter;
+    asdl_expr_seq *ifs;
+    int is_async;
+};
+
+enum _excepthandler_kind {ExceptHandler_kind=1};
+struct _excepthandler {
+    enum _excepthandler_kind kind;
+    union {
+        struct {
+            expr_ty type;
+            identifier name;
+            asdl_stmt_seq *body;
+        } ExceptHandler;
+
+    } v;
+    int lineno;
+    int col_offset;
+    int end_lineno;
+    int end_col_offset;
+};
+
+struct _arguments {
+    asdl_arg_seq *posonlyargs;
+    asdl_arg_seq *args;
+    arg_ty vararg;
+    asdl_arg_seq *kwonlyargs;
+    asdl_expr_seq *kw_defaults;
+    arg_ty kwarg;
+    asdl_expr_seq *defaults;
+};
+
+struct _arg {
+    identifier arg;
+    expr_ty annotation;
+    string type_comment;
+    int lineno;
+    int col_offset;
+    int end_lineno;
+    int end_col_offset;
+};
+
+struct _keyword {
+    identifier arg;
+    expr_ty value;
+    int lineno;
+    int col_offset;
+    int end_lineno;
+    int end_col_offset;
+};
+
+struct _alias {
+    identifier name;
+    identifier asname;
+    int lineno;
+    int col_offset;
+    int end_lineno;
+    int end_col_offset;
+};
+
+struct _withitem {
+    expr_ty context_expr;
+    expr_ty optional_vars;
+};
+
+struct _match_case {
+    pattern_ty pattern;
+    expr_ty guard;
+    asdl_stmt_seq *body;
+};
+
+enum _pattern_kind {MatchValue_kind=1, MatchSingleton_kind=2,
+                     MatchSequence_kind=3, MatchMapping_kind=4,
+                     MatchClass_kind=5, MatchStar_kind=6, MatchAs_kind=7,
+                     MatchOr_kind=8};
+struct _pattern {
+    enum _pattern_kind kind;
+    union {
+        struct {
+            expr_ty value;
+        } MatchValue;
+
+        struct {
+            constant value;
+        } MatchSingleton;
+
+        struct {
+            asdl_pattern_seq *patterns;
+        } MatchSequence;
+
+        struct {
+            asdl_expr_seq *keys;
+            asdl_pattern_seq *patterns;
+            identifier rest;
+        } MatchMapping;
+
+        struct {
+            expr_ty cls;
+            asdl_pattern_seq *patterns;
+            asdl_identifier_seq *kwd_attrs;
+            asdl_pattern_seq *kwd_patterns;
+        } MatchClass;
+
+        struct {
+            identifier name;
+        } MatchStar;
+
+        struct {
+            pattern_ty pattern;
+            identifier name;
+        } MatchAs;
+
+        struct {
+            asdl_pattern_seq *patterns;
+        } MatchOr;
+
+    } v;
+    int lineno;
+    int col_offset;
+    int end_lineno;
+    int end_col_offset;
+};
+
+enum _type_ignore_kind {TypeIgnore_kind=1};
+struct _type_ignore {
+    enum _type_ignore_kind kind;
+    union {
+        struct {
+            int lineno;
+            string tag;
+        } TypeIgnore;
+
+    } v;
+};
+
+enum _type_param_kind {TypeVar_kind=1, ParamSpec_kind=2, TypeVarTuple_kind=3};
+struct _type_param {
+    enum _type_param_kind kind;
+    union {
+        struct {
+            identifier name;
+            expr_ty bound;
+            expr_ty default_value;
+        } TypeVar;
+
+        struct {
+            identifier name;
+            expr_ty default_value;
+        } ParamSpec;
+
+        struct {
+            identifier name;
+            expr_ty default_value;
+        } TypeVarTuple;
+
+    } v;
+    int lineno;
+    int col_offset;
+    int end_lineno;
+    int end_col_offset;
+};
+
+
+// Note: these macros affect function definitions, not only call sites.
+mod_ty _PyAST_Module(asdl_stmt_seq * body, asdl_type_ignore_seq * type_ignores,
+                     PyArena *arena);
+mod_ty _PyAST_Interactive(asdl_stmt_seq * body, PyArena *arena);
+mod_ty _PyAST_Expression(expr_ty body, PyArena *arena);
+mod_ty _PyAST_FunctionType(asdl_expr_seq * argtypes, expr_ty returns, PyArena
+                           *arena);
+stmt_ty _PyAST_FunctionDef(identifier name, arguments_ty args, asdl_stmt_seq *
+                           body, asdl_expr_seq * decorator_list, expr_ty
+                           returns, string type_comment, asdl_type_param_seq *
+                           type_params, int lineno, int col_offset, int
+                           end_lineno, int end_col_offset, PyArena *arena);
+stmt_ty _PyAST_AsyncFunctionDef(identifier name, arguments_ty args,
+                                asdl_stmt_seq * body, asdl_expr_seq *
+                                decorator_list, expr_ty returns, string
+                                type_comment, asdl_type_param_seq *
+                                type_params, int lineno, int col_offset, int
+                                end_lineno, int end_col_offset, PyArena *arena);
+stmt_ty _PyAST_ClassDef(identifier name, asdl_expr_seq * bases,
+                        asdl_keyword_seq * keywords, asdl_stmt_seq * body,
+                        asdl_expr_seq * decorator_list, asdl_type_param_seq *
+                        type_params, int lineno, int col_offset, int
+                        end_lineno, int end_col_offset, PyArena *arena);
+stmt_ty _PyAST_Return(expr_ty value, int lineno, int col_offset, int
+                      end_lineno, int end_col_offset, PyArena *arena);
+stmt_ty _PyAST_Delete(asdl_expr_seq * targets, int lineno, int col_offset, int
+                      end_lineno, int end_col_offset, PyArena *arena);
+stmt_ty _PyAST_Assign(asdl_expr_seq * targets, expr_ty value, string
+                      type_comment, int lineno, int col_offset, int end_lineno,
+                      int end_col_offset, PyArena *arena);
+stmt_ty _PyAST_TypeAlias(expr_ty name, asdl_type_param_seq * type_params,
+                         expr_ty value, int lineno, int col_offset, int
+                         end_lineno, int end_col_offset, PyArena *arena);
+stmt_ty _PyAST_AugAssign(expr_ty target, operator_ty op, expr_ty value, int
+                         lineno, int col_offset, int end_lineno, int
+                         end_col_offset, PyArena *arena);
+stmt_ty _PyAST_AnnAssign(expr_ty target, expr_ty annotation, expr_ty value, int
+                         simple, int lineno, int col_offset, int end_lineno,
+                         int end_col_offset, PyArena *arena);
+stmt_ty _PyAST_For(expr_ty target, expr_ty iter, asdl_stmt_seq * body,
+                   asdl_stmt_seq * orelse, string type_comment, int lineno, int
+                   col_offset, int end_lineno, int end_col_offset, PyArena
+                   *arena);
+stmt_ty _PyAST_AsyncFor(expr_ty target, expr_ty iter, asdl_stmt_seq * body,
+                        asdl_stmt_seq * orelse, string type_comment, int
+                        lineno, int col_offset, int end_lineno, int
+                        end_col_offset, PyArena *arena);
+stmt_ty _PyAST_While(expr_ty test, asdl_stmt_seq * body, asdl_stmt_seq *
+                     orelse, int lineno, int col_offset, int end_lineno, int
+                     end_col_offset, PyArena *arena);
+stmt_ty _PyAST_If(expr_ty test, asdl_stmt_seq * body, asdl_stmt_seq * orelse,
+                  int lineno, int col_offset, int end_lineno, int
+                  end_col_offset, PyArena *arena);
+stmt_ty _PyAST_With(asdl_withitem_seq * items, asdl_stmt_seq * body, string
+                    type_comment, int lineno, int col_offset, int end_lineno,
+                    int end_col_offset, PyArena *arena);
+stmt_ty _PyAST_AsyncWith(asdl_withitem_seq * items, asdl_stmt_seq * body,
+                         string type_comment, int lineno, int col_offset, int
+                         end_lineno, int end_col_offset, PyArena *arena);
+stmt_ty _PyAST_Match(expr_ty subject, asdl_match_case_seq * cases, int lineno,
+                     int col_offset, int end_lineno, int end_col_offset,
+                     PyArena *arena);
+stmt_ty _PyAST_Raise(expr_ty exc, expr_ty cause, int lineno, int col_offset,
+                     int end_lineno, int end_col_offset, PyArena *arena);
+stmt_ty _PyAST_Try(asdl_stmt_seq * body, asdl_excepthandler_seq * handlers,
+                   asdl_stmt_seq * orelse, asdl_stmt_seq * finalbody, int
+                   lineno, int col_offset, int end_lineno, int end_col_offset,
+                   PyArena *arena);
+stmt_ty _PyAST_TryStar(asdl_stmt_seq * body, asdl_excepthandler_seq * handlers,
+                       asdl_stmt_seq * orelse, asdl_stmt_seq * finalbody, int
+                       lineno, int col_offset, int end_lineno, int
+                       end_col_offset, PyArena *arena);
+stmt_ty _PyAST_Assert(expr_ty test, expr_ty msg, int lineno, int col_offset,
+                      int end_lineno, int end_col_offset, PyArena *arena);
+stmt_ty _PyAST_Import(asdl_alias_seq * names, int lineno, int col_offset, int
+                      end_lineno, int end_col_offset, PyArena *arena);
+stmt_ty _PyAST_ImportFrom(identifier module, asdl_alias_seq * names, int level,
+                          int lineno, int col_offset, int end_lineno, int
+                          end_col_offset, PyArena *arena);
+stmt_ty _PyAST_Global(asdl_identifier_seq * names, int lineno, int col_offset,
+                      int end_lineno, int end_col_offset, PyArena *arena);
+stmt_ty _PyAST_Nonlocal(asdl_identifier_seq * names, int lineno, int
+                        col_offset, int end_lineno, int end_col_offset, PyArena
+                        *arena);
+stmt_ty _PyAST_Expr(expr_ty value, int lineno, int col_offset, int end_lineno,
+                    int end_col_offset, PyArena *arena);
+stmt_ty _PyAST_Pass(int lineno, int col_offset, int end_lineno, int
+                    end_col_offset, PyArena *arena);
+stmt_ty _PyAST_Break(int lineno, int col_offset, int end_lineno, int
+                     end_col_offset, PyArena *arena);
+stmt_ty _PyAST_Continue(int lineno, int col_offset, int end_lineno, int
+                        end_col_offset, PyArena *arena);
+expr_ty _PyAST_BoolOp(boolop_ty op, asdl_expr_seq * values, int lineno, int
+                      col_offset, int end_lineno, int end_col_offset, PyArena
+                      *arena);
+expr_ty _PyAST_NamedExpr(expr_ty target, expr_ty value, int lineno, int
+                         col_offset, int end_lineno, int end_col_offset,
+                         PyArena *arena);
+expr_ty _PyAST_BinOp(expr_ty left, operator_ty op, expr_ty right, int lineno,
+                     int col_offset, int end_lineno, int end_col_offset,
+                     PyArena *arena);
+expr_ty _PyAST_UnaryOp(unaryop_ty op, expr_ty operand, int lineno, int
+                       col_offset, int end_lineno, int end_col_offset, PyArena
+                       *arena);
+expr_ty _PyAST_Lambda(arguments_ty args, expr_ty body, int lineno, int
+                      col_offset, int end_lineno, int end_col_offset, PyArena
+                      *arena);
+expr_ty _PyAST_IfExp(expr_ty test, expr_ty body, expr_ty orelse, int lineno,
+                     int col_offset, int end_lineno, int end_col_offset,
+                     PyArena *arena);
+expr_ty _PyAST_Dict(asdl_expr_seq * keys, asdl_expr_seq * values, int lineno,
+                    int col_offset, int end_lineno, int end_col_offset, PyArena
+                    *arena);
+expr_ty _PyAST_Set(asdl_expr_seq * elts, int lineno, int col_offset, int
+                   end_lineno, int end_col_offset, PyArena *arena);
+expr_ty _PyAST_ListComp(expr_ty elt, asdl_comprehension_seq * generators, int
+                        lineno, int col_offset, int end_lineno, int
+                        end_col_offset, PyArena *arena);
+expr_ty _PyAST_SetComp(expr_ty elt, asdl_comprehension_seq * generators, int
+                       lineno, int col_offset, int end_lineno, int
+                       end_col_offset, PyArena *arena);
+expr_ty _PyAST_DictComp(expr_ty key, expr_ty value, asdl_comprehension_seq *
+                        generators, int lineno, int col_offset, int end_lineno,
+                        int end_col_offset, PyArena *arena);
+expr_ty _PyAST_GeneratorExp(expr_ty elt, asdl_comprehension_seq * generators,
+                            int lineno, int col_offset, int end_lineno, int
+                            end_col_offset, PyArena *arena);
+expr_ty _PyAST_Await(expr_ty value, int lineno, int col_offset, int end_lineno,
+                     int end_col_offset, PyArena *arena);
+expr_ty _PyAST_Yield(expr_ty value, int lineno, int col_offset, int end_lineno,
+                     int end_col_offset, PyArena *arena);
+expr_ty _PyAST_YieldFrom(expr_ty value, int lineno, int col_offset, int
+                         end_lineno, int end_col_offset, PyArena *arena);
+expr_ty _PyAST_Compare(expr_ty left, asdl_int_seq * ops, asdl_expr_seq *
+                       comparators, int lineno, int col_offset, int end_lineno,
+                       int end_col_offset, PyArena *arena);
+expr_ty _PyAST_Call(expr_ty func, asdl_expr_seq * args, asdl_keyword_seq *
+                    keywords, int lineno, int col_offset, int end_lineno, int
+                    end_col_offset, PyArena *arena);
+expr_ty _PyAST_FormattedValue(expr_ty value, int conversion, expr_ty
+                              format_spec, int lineno, int col_offset, int
+                              end_lineno, int end_col_offset, PyArena *arena);
+expr_ty _PyAST_JoinedStr(asdl_expr_seq * values, int lineno, int col_offset,
+                         int end_lineno, int end_col_offset, PyArena *arena);
+expr_ty _PyAST_Constant(constant value, string kind, int lineno, int
+                        col_offset, int end_lineno, int end_col_offset, PyArena
+                        *arena);
+expr_ty _PyAST_Attribute(expr_ty value, identifier attr, expr_context_ty ctx,
+                         int lineno, int col_offset, int end_lineno, int
+                         end_col_offset, PyArena *arena);
+expr_ty _PyAST_Subscript(expr_ty value, expr_ty slice, expr_context_ty ctx, int
+                         lineno, int col_offset, int end_lineno, int
+                         end_col_offset, PyArena *arena);
+expr_ty _PyAST_Starred(expr_ty value, expr_context_ty ctx, int lineno, int
+                       col_offset, int end_lineno, int end_col_offset, PyArena
+                       *arena);
+expr_ty _PyAST_Name(identifier id, expr_context_ty ctx, int lineno, int
+                    col_offset, int end_lineno, int end_col_offset, PyArena
+                    *arena);
+expr_ty _PyAST_List(asdl_expr_seq * elts, expr_context_ty ctx, int lineno, int
+                    col_offset, int end_lineno, int end_col_offset, PyArena
+                    *arena);
+expr_ty _PyAST_Tuple(asdl_expr_seq * elts, expr_context_ty ctx, int lineno, int
+                     col_offset, int end_lineno, int end_col_offset, PyArena
+                     *arena);
+expr_ty _PyAST_Slice(expr_ty lower, expr_ty upper, expr_ty step, int lineno,
+                     int col_offset, int end_lineno, int end_col_offset,
+                     PyArena *arena);
+comprehension_ty _PyAST_comprehension(expr_ty target, expr_ty iter,
+                                      asdl_expr_seq * ifs, int is_async,
+                                      PyArena *arena);
+excepthandler_ty _PyAST_ExceptHandler(expr_ty type, identifier name,
+                                      asdl_stmt_seq * body, int lineno, int
+                                      col_offset, int end_lineno, int
+                                      end_col_offset, PyArena *arena);
+arguments_ty _PyAST_arguments(asdl_arg_seq * posonlyargs, asdl_arg_seq * args,
+                              arg_ty vararg, asdl_arg_seq * kwonlyargs,
+                              asdl_expr_seq * kw_defaults, arg_ty kwarg,
+                              asdl_expr_seq * defaults, PyArena *arena);
+arg_ty _PyAST_arg(identifier arg, expr_ty annotation, string type_comment, int
+                  lineno, int col_offset, int end_lineno, int end_col_offset,
+                  PyArena *arena);
+keyword_ty _PyAST_keyword(identifier arg, expr_ty value, int lineno, int
+                          col_offset, int end_lineno, int end_col_offset,
+                          PyArena *arena);
+alias_ty _PyAST_alias(identifier name, identifier asname, int lineno, int
+                      col_offset, int end_lineno, int end_col_offset, PyArena
+                      *arena);
+withitem_ty _PyAST_withitem(expr_ty context_expr, expr_ty optional_vars,
+                            PyArena *arena);
+match_case_ty _PyAST_match_case(pattern_ty pattern, expr_ty guard,
+                                asdl_stmt_seq * body, PyArena *arena);
+pattern_ty _PyAST_MatchValue(expr_ty value, int lineno, int col_offset, int
+                             end_lineno, int end_col_offset, PyArena *arena);
+pattern_ty _PyAST_MatchSingleton(constant value, int lineno, int col_offset,
+                                 int end_lineno, int end_col_offset, PyArena
+                                 *arena);
+pattern_ty _PyAST_MatchSequence(asdl_pattern_seq * patterns, int lineno, int
+                                col_offset, int end_lineno, int end_col_offset,
+                                PyArena *arena);
+pattern_ty _PyAST_MatchMapping(asdl_expr_seq * keys, asdl_pattern_seq *
+                               patterns, identifier rest, int lineno, int
+                               col_offset, int end_lineno, int end_col_offset,
+                               PyArena *arena);
+pattern_ty _PyAST_MatchClass(expr_ty cls, asdl_pattern_seq * patterns,
+                             asdl_identifier_seq * kwd_attrs, asdl_pattern_seq
+                             * kwd_patterns, int lineno, int col_offset, int
+                             end_lineno, int end_col_offset, PyArena *arena);
+pattern_ty _PyAST_MatchStar(identifier name, int lineno, int col_offset, int
+                            end_lineno, int end_col_offset, PyArena *arena);
+pattern_ty _PyAST_MatchAs(pattern_ty pattern, identifier name, int lineno, int
+                          col_offset, int end_lineno, int end_col_offset,
+                          PyArena *arena);
+pattern_ty _PyAST_MatchOr(asdl_pattern_seq * patterns, int lineno, int
+                          col_offset, int end_lineno, int end_col_offset,
+                          PyArena *arena);
+type_ignore_ty _PyAST_TypeIgnore(int lineno, string tag, PyArena *arena);
+type_param_ty _PyAST_TypeVar(identifier name, expr_ty bound, expr_ty
+                             default_value, int lineno, int col_offset, int
+                             end_lineno, int end_col_offset, PyArena *arena);
+type_param_ty _PyAST_ParamSpec(identifier name, expr_ty default_value, int
+                               lineno, int col_offset, int end_lineno, int
+                               end_col_offset, PyArena *arena);
+type_param_ty _PyAST_TypeVarTuple(identifier name, expr_ty default_value, int
+                                  lineno, int col_offset, int end_lineno, int
+                                  end_col_offset, PyArena *arena);
+
+
+PyObject* PyAST_mod2obj(mod_ty t);
+mod_ty PyAST_obj2mod(PyObject* ast, PyArena* arena, int mode);
+int PyAST_Check(PyObject* obj);
+
+extern int _PyAST_Validate(mod_ty);
+
+/* _PyAST_ExprAsUnicode is defined in ast_unparse.c */
+extern PyObject* _PyAST_ExprAsUnicode(expr_ty);
+
+/* Return the borrowed reference to the first literal string in the
+   sequence of statements or NULL if it doesn't start from a literal string.
+   Doesn't set exception. */
+extern PyObject* _PyAST_GetDocString(asdl_stmt_seq *);
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* !Py_INTERNAL_AST_H */

Dependencies/Python/include/internal/pycore_ast_state.h

@@ -0,0 +1,268 @@
+// File automatically generated by Parser/asdl_c.py.
+
+#ifndef Py_INTERNAL_AST_STATE_H
+#define Py_INTERNAL_AST_STATE_H
+
+#include "pycore_lock.h"    // _PyOnceFlag
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+#ifndef Py_BUILD_CORE
+#  error "this header requires Py_BUILD_CORE define"
+#endif
+
+struct ast_state {
+    _PyOnceFlag once;
+    int finalized;
+    PyObject *AST_type;
+    PyObject *Add_singleton;
+    PyObject *Add_type;
+    PyObject *And_singleton;
+    PyObject *And_type;
+    PyObject *AnnAssign_type;
+    PyObject *Assert_type;
+    PyObject *Assign_type;
+    PyObject *AsyncFor_type;
+    PyObject *AsyncFunctionDef_type;
+    PyObject *AsyncWith_type;
+    PyObject *Attribute_type;
+    PyObject *AugAssign_type;
+    PyObject *Await_type;
+    PyObject *BinOp_type;
+    PyObject *BitAnd_singleton;
+    PyObject *BitAnd_type;
+    PyObject *BitOr_singleton;
+    PyObject *BitOr_type;
+    PyObject *BitXor_singleton;
+    PyObject *BitXor_type;
+    PyObject *BoolOp_type;
+    PyObject *Break_type;
+    PyObject *Call_type;
+    PyObject *ClassDef_type;
+    PyObject *Compare_type;
+    PyObject *Constant_type;
+    PyObject *Continue_type;
+    PyObject *Del_singleton;
+    PyObject *Del_type;
+    PyObject *Delete_type;
+    PyObject *DictComp_type;
+    PyObject *Dict_type;
+    PyObject *Div_singleton;
+    PyObject *Div_type;
+    PyObject *Eq_singleton;
+    PyObject *Eq_type;
+    PyObject *ExceptHandler_type;
+    PyObject *Expr_type;
+    PyObject *Expression_type;
+    PyObject *FloorDiv_singleton;
+    PyObject *FloorDiv_type;
+    PyObject *For_type;
+    PyObject *FormattedValue_type;
+    PyObject *FunctionDef_type;
+    PyObject *FunctionType_type;
+    PyObject *GeneratorExp_type;
+    PyObject *Global_type;
+    PyObject *GtE_singleton;
+    PyObject *GtE_type;
+    PyObject *Gt_singleton;
+    PyObject *Gt_type;
+    PyObject *IfExp_type;
+    PyObject *If_type;
+    PyObject *ImportFrom_type;
+    PyObject *Import_type;
+    PyObject *In_singleton;
+    PyObject *In_type;
+    PyObject *Interactive_type;
+    PyObject *Invert_singleton;
+    PyObject *Invert_type;
+    PyObject *IsNot_singleton;
+    PyObject *IsNot_type;
+    PyObject *Is_singleton;
+    PyObject *Is_type;
+    PyObject *JoinedStr_type;
+    PyObject *LShift_singleton;
+    PyObject *LShift_type;
+    PyObject *Lambda_type;
+    PyObject *ListComp_type;
+    PyObject *List_type;
+    PyObject *Load_singleton;
+    PyObject *Load_type;
+    PyObject *LtE_singleton;
+    PyObject *LtE_type;
+    PyObject *Lt_singleton;
+    PyObject *Lt_type;
+    PyObject *MatMult_singleton;
+    PyObject *MatMult_type;
+    PyObject *MatchAs_type;
+    PyObject *MatchClass_type;
+    PyObject *MatchMapping_type;
+    PyObject *MatchOr_type;
+    PyObject *MatchSequence_type;
+    PyObject *MatchSingleton_type;
+    PyObject *MatchStar_type;
+    PyObject *MatchValue_type;
+    PyObject *Match_type;
+    PyObject *Mod_singleton;
+    PyObject *Mod_type;
+    PyObject *Module_type;
+    PyObject *Mult_singleton;
+    PyObject *Mult_type;
+    PyObject *Name_type;
+    PyObject *NamedExpr_type;
+    PyObject *Nonlocal_type;
+    PyObject *NotEq_singleton;
+    PyObject *NotEq_type;
+    PyObject *NotIn_singleton;
+    PyObject *NotIn_type;
+    PyObject *Not_singleton;
+    PyObject *Not_type;
+    PyObject *Or_singleton;
+    PyObject *Or_type;
+    PyObject *ParamSpec_type;
+    PyObject *Pass_type;
+    PyObject *Pow_singleton;
+    PyObject *Pow_type;
+    PyObject *RShift_singleton;
+    PyObject *RShift_type;
+    PyObject *Raise_type;
+    PyObject *Return_type;
+    PyObject *SetComp_type;
+    PyObject *Set_type;
+    PyObject *Slice_type;
+    PyObject *Starred_type;
+    PyObject *Store_singleton;
+    PyObject *Store_type;
+    PyObject *Sub_singleton;
+    PyObject *Sub_type;
+    PyObject *Subscript_type;
+    PyObject *TryStar_type;
+    PyObject *Try_type;
+    PyObject *Tuple_type;
+    PyObject *TypeAlias_type;
+    PyObject *TypeIgnore_type;
+    PyObject *TypeVarTuple_type;
+    PyObject *TypeVar_type;
+    PyObject *UAdd_singleton;
+    PyObject *UAdd_type;
+    PyObject *USub_singleton;
+    PyObject *USub_type;
+    PyObject *UnaryOp_type;
+    PyObject *While_type;
+    PyObject *With_type;
+    PyObject *YieldFrom_type;
+    PyObject *Yield_type;
+    PyObject *__dict__;
+    PyObject *__doc__;
+    PyObject *__match_args__;
+    PyObject *__module__;
+    PyObject *_attributes;
+    PyObject *_fields;
+    PyObject *alias_type;
+    PyObject *annotation;
+    PyObject *arg;
+    PyObject *arg_type;
+    PyObject *args;
+    PyObject *argtypes;
+    PyObject *arguments_type;
+    PyObject *asname;
+    PyObject *ast;
+    PyObject *attr;
+    PyObject *bases;
+    PyObject *body;
+    PyObject *boolop_type;
+    PyObject *bound;
+    PyObject *cases;
+    PyObject *cause;
+    PyObject *cls;
+    PyObject *cmpop_type;
+    PyObject *col_offset;
+    PyObject *comparators;
+    PyObject *comprehension_type;
+    PyObject *context_expr;
+    PyObject *conversion;
+    PyObject *ctx;
+    PyObject *decorator_list;
+    PyObject *default_value;
+    PyObject *defaults;
+    PyObject *elt;
+    PyObject *elts;
+    PyObject *end_col_offset;
+    PyObject *end_lineno;
+    PyObject *exc;
+    PyObject *excepthandler_type;
+    PyObject *expr_context_type;
+    PyObject *expr_type;
+    PyObject *finalbody;
+    PyObject *format_spec;
+    PyObject *func;
+    PyObject *generators;
+    PyObject *guard;
+    PyObject *handlers;
+    PyObject *id;
+    PyObject *ifs;
+    PyObject *is_async;
+    PyObject *items;
+    PyObject *iter;
+    PyObject *key;
+    PyObject *keys;
+    PyObject *keyword_type;
+    PyObject *keywords;
+    PyObject *kind;
+    PyObject *kw_defaults;
+    PyObject *kwarg;
+    PyObject *kwd_attrs;
+    PyObject *kwd_patterns;
+    PyObject *kwonlyargs;
+    PyObject *left;
+    PyObject *level;
+    PyObject *lineno;
+    PyObject *lower;
+    PyObject *match_case_type;
+    PyObject *mod_type;
+    PyObject *module;
+    PyObject *msg;
+    PyObject *name;
+    PyObject *names;
+    PyObject *op;
+    PyObject *operand;
+    PyObject *operator_type;
+    PyObject *ops;
+    PyObject *optional_vars;
+    PyObject *orelse;
+    PyObject *pattern;
+    PyObject *pattern_type;
+    PyObject *patterns;
+    PyObject *posonlyargs;
+    PyObject *rest;
+    PyObject *returns;
+    PyObject *right;
+    PyObject *simple;
+    PyObject *slice;
+    PyObject *step;
+    PyObject *stmt_type;
+    PyObject *subject;
+    PyObject *tag;
+    PyObject *target;
+    PyObject *targets;
+    PyObject *test;
+    PyObject *type;
+    PyObject *type_comment;
+    PyObject *type_ignore_type;
+    PyObject *type_ignores;
+    PyObject *type_param_type;
+    PyObject *type_params;
+    PyObject *unaryop_type;
+    PyObject *upper;
+    PyObject *value;
+    PyObject *values;
+    PyObject *vararg;
+    PyObject *withitem_type;
+};
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* !Py_INTERNAL_AST_STATE_H */
+

Dependencies/Python/include/internal/pycore_atexit.h

@@ -0,0 +1,67 @@
+#ifndef Py_INTERNAL_ATEXIT_H
+#define Py_INTERNAL_ATEXIT_H
+
+#include "pycore_lock.h"        // PyMutex
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+#ifndef Py_BUILD_CORE
+#  error "this header requires Py_BUILD_CORE define"
+#endif
+
+
+//###############
+// runtime atexit
+
+typedef void (*atexit_callbackfunc)(void);
+
+struct _atexit_runtime_state {
+    PyMutex mutex;
+#define NEXITFUNCS 32
+    atexit_callbackfunc callbacks[NEXITFUNCS];
+    int ncallbacks;
+};
+
+
+//###################
+// interpreter atexit
+
+typedef void (*atexit_datacallbackfunc)(void *);
+
+typedef struct atexit_callback {
+    atexit_datacallbackfunc func;
+    void *data;
+    struct atexit_callback *next;
+} atexit_callback;
+
+typedef struct {
+    PyObject *func;
+    PyObject *args;
+    PyObject *kwargs;
+} atexit_py_callback;
+
+struct atexit_state {
+    atexit_callback *ll_callbacks;
+    // Kept for ABI compatibility--do not use! (See GH-127791.)
+    atexit_callback *last_ll_callback;
+
+    // XXX The rest of the state could be moved to the atexit module state
+    // and a low-level callback added for it during module exec.
+    // For the moment we leave it here.
+    atexit_py_callback **callbacks;
+    int ncallbacks;
+    int callback_len;
+};
+
+// Export for '_interpchannels' shared extension
+PyAPI_FUNC(int) _Py_AtExit(
+    PyInterpreterState *interp,
+    atexit_datacallbackfunc func,
+    void *data);
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* !Py_INTERNAL_ATEXIT_H */

Dependencies/Python/include/internal/pycore_backoff.h

@@ -0,0 +1,145 @@
+
+#ifndef Py_INTERNAL_BACKOFF_H
+#define Py_INTERNAL_BACKOFF_H
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+#ifndef Py_BUILD_CORE
+#  error "this header requires Py_BUILD_CORE define"
+#endif
+
+#include <assert.h>
+#include <stdbool.h>
+#include <stdint.h>
+
+
+typedef struct {
+    union {
+        struct {
+            uint16_t backoff : 4;
+            uint16_t value : 12;
+        };
+        uint16_t as_counter;  // For printf("%#x", ...)
+    };
+} _Py_BackoffCounter;
+
+
+/* 16-bit countdown counters using exponential backoff.
+
+   These are used by the adaptive specializer to count down until
+   it is time to specialize an instruction. If specialization fails
+   the counter is reset using exponential backoff.
+
+   Another use is for the Tier 2 optimizer to decide when to create
+   a new Tier 2 trace (executor). Again, exponential backoff is used.
+
+   The 16-bit counter is structured as a 12-bit unsigned 'value'
+   and a 4-bit 'backoff' field. When resetting the counter, the
+   backoff field is incremented (until it reaches a limit) and the
+   value is set to a bit mask representing the value 2**backoff - 1.
+   The maximum backoff is 12 (the number of value bits).
+
+   There is an exceptional value which must not be updated, 0xFFFF.
+*/
+
+#define UNREACHABLE_BACKOFF 0xFFFF
+
+static inline bool
+is_unreachable_backoff_counter(_Py_BackoffCounter counter)
+{
+    return counter.as_counter == UNREACHABLE_BACKOFF;
+}
+
+static inline _Py_BackoffCounter
+make_backoff_counter(uint16_t value, uint16_t backoff)
+{
+    assert(backoff <= 15);
+    assert(value <= 0xFFF);
+    _Py_BackoffCounter result;
+    result.value = value;
+    result.backoff = backoff;
+    return result;
+}
+
+static inline _Py_BackoffCounter
+forge_backoff_counter(uint16_t counter)
+{
+    _Py_BackoffCounter result;
+    result.as_counter = counter;
+    return result;
+}
+
+static inline _Py_BackoffCounter
+restart_backoff_counter(_Py_BackoffCounter counter)
+{
+    assert(!is_unreachable_backoff_counter(counter));
+    if (counter.backoff < 12) {
+        return make_backoff_counter((1 << (counter.backoff + 1)) - 1, counter.backoff + 1);
+    }
+    else {
+        return make_backoff_counter((1 << 12) - 1, 12);
+    }
+}
+
+static inline _Py_BackoffCounter
+pause_backoff_counter(_Py_BackoffCounter counter)
+{
+    return make_backoff_counter(counter.value | 1, counter.backoff);
+}
+
+static inline _Py_BackoffCounter
+advance_backoff_counter(_Py_BackoffCounter counter)
+{
+    if (!is_unreachable_backoff_counter(counter)) {
+        return make_backoff_counter((counter.value - 1) & 0xFFF, counter.backoff);
+    }
+    else {
+        return counter;
+    }
+}
+
+static inline bool
+backoff_counter_triggers(_Py_BackoffCounter counter)
+{
+    return counter.value == 0;
+}
+
+/* Initial JUMP_BACKWARD counter.
+ * This determines when we create a trace for a loop.
+* Backoff sequence 16, 32, 64, 128, 256, 512, 1024, 2048, 4096. */
+#define JUMP_BACKWARD_INITIAL_VALUE 16
+#define JUMP_BACKWARD_INITIAL_BACKOFF 4
+static inline _Py_BackoffCounter
+initial_jump_backoff_counter(void)
+{
+    return make_backoff_counter(JUMP_BACKWARD_INITIAL_VALUE,
+                                JUMP_BACKWARD_INITIAL_BACKOFF);
+}
+
+/* Initial exit temperature.
+ * Must be larger than ADAPTIVE_COOLDOWN_VALUE,
+ * otherwise when a side exit warms up we may construct
+ * a new trace before the Tier 1 code has properly re-specialized.
+ * Backoff sequence 64, 128, 256, 512, 1024, 2048, 4096. */
+#define COLD_EXIT_INITIAL_VALUE 64
+#define COLD_EXIT_INITIAL_BACKOFF 6
+
+static inline _Py_BackoffCounter
+initial_temperature_backoff_counter(void)
+{
+    return make_backoff_counter(COLD_EXIT_INITIAL_VALUE,
+                                COLD_EXIT_INITIAL_BACKOFF);
+}
+
+/* Unreachable backoff counter. */
+static inline _Py_BackoffCounter
+initial_unreachable_backoff_counter(void)
+{
+    return forge_backoff_counter(UNREACHABLE_BACKOFF);
+}
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* !Py_INTERNAL_BACKOFF_H */

Dependencies/Python/include/internal/pycore_bitutils.h

@@ -0,0 +1,186 @@
+/* Bit and bytes utilities.
+
+   Bytes swap functions, reverse order of bytes:
+
+   - _Py_bswap16(uint16_t)
+   - _Py_bswap32(uint32_t)
+   - _Py_bswap64(uint64_t)
+*/
+
+#ifndef Py_INTERNAL_BITUTILS_H
+#define Py_INTERNAL_BITUTILS_H
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+#ifndef Py_BUILD_CORE
+#  error "this header requires Py_BUILD_CORE define"
+#endif
+
+#if defined(__GNUC__) \
+      && ((__GNUC__ >= 5) || (__GNUC__ == 4) && (__GNUC_MINOR__ >= 8))
+   /* __builtin_bswap16() is available since GCC 4.8,
+      __builtin_bswap32() is available since GCC 4.3,
+      __builtin_bswap64() is available since GCC 4.3. */
+#  define _PY_HAVE_BUILTIN_BSWAP
+#endif
+
+#ifdef _MSC_VER
+#  include <intrin.h>             // _byteswap_uint64()
+#endif
+
+
+static inline uint16_t
+_Py_bswap16(uint16_t word)
+{
+#if defined(_PY_HAVE_BUILTIN_BSWAP) || _Py__has_builtin(__builtin_bswap16)
+    return __builtin_bswap16(word);
+#elif defined(_MSC_VER)
+    Py_BUILD_ASSERT(sizeof(word) == sizeof(unsigned short));
+    return _byteswap_ushort(word);
+#else
+    // Portable implementation which doesn't rely on circular bit shift
+    return ( ((word & UINT16_C(0x00FF)) << 8)
+           | ((word & UINT16_C(0xFF00)) >> 8));
+#endif
+}
+
+static inline uint32_t
+_Py_bswap32(uint32_t word)
+{
+#if defined(_PY_HAVE_BUILTIN_BSWAP) || _Py__has_builtin(__builtin_bswap32)
+    return __builtin_bswap32(word);
+#elif defined(_MSC_VER)
+    Py_BUILD_ASSERT(sizeof(word) == sizeof(unsigned long));
+    return _byteswap_ulong(word);
+#else
+    // Portable implementation which doesn't rely on circular bit shift
+    return ( ((word & UINT32_C(0x000000FF)) << 24)
+           | ((word & UINT32_C(0x0000FF00)) <<  8)
+           | ((word & UINT32_C(0x00FF0000)) >>  8)
+           | ((word & UINT32_C(0xFF000000)) >> 24));
+#endif
+}
+
+static inline uint64_t
+_Py_bswap64(uint64_t word)
+{
+#if defined(_PY_HAVE_BUILTIN_BSWAP) || _Py__has_builtin(__builtin_bswap64)
+    return __builtin_bswap64(word);
+#elif defined(_MSC_VER)
+    return _byteswap_uint64(word);
+#else
+    // Portable implementation which doesn't rely on circular bit shift
+    return ( ((word & UINT64_C(0x00000000000000FF)) << 56)
+           | ((word & UINT64_C(0x000000000000FF00)) << 40)
+           | ((word & UINT64_C(0x0000000000FF0000)) << 24)
+           | ((word & UINT64_C(0x00000000FF000000)) <<  8)
+           | ((word & UINT64_C(0x000000FF00000000)) >>  8)
+           | ((word & UINT64_C(0x0000FF0000000000)) >> 24)
+           | ((word & UINT64_C(0x00FF000000000000)) >> 40)
+           | ((word & UINT64_C(0xFF00000000000000)) >> 56));
+#endif
+}
+
+
+// Population count: count the number of 1's in 'x'
+// (number of bits set to 1), also known as the hamming weight.
+//
+// Implementation note. CPUID is not used, to test if x86 POPCNT instruction
+// can be used, to keep the implementation simple. For example, Visual Studio
+// __popcnt() is not used this reason. The clang and GCC builtin function can
+// use the x86 POPCNT instruction if the target architecture has SSE4a or
+// newer.
+static inline int
+_Py_popcount32(uint32_t x)
+{
+#if (defined(__clang__) || defined(__GNUC__))
+
+#if SIZEOF_INT >= 4
+    Py_BUILD_ASSERT(sizeof(x) <= sizeof(unsigned int));
+    return __builtin_popcount(x);
+#else
+    // The C standard guarantees that unsigned long will always be big enough
+    // to hold a uint32_t value without losing information.
+    Py_BUILD_ASSERT(sizeof(x) <= sizeof(unsigned long));
+    return __builtin_popcountl(x);
+#endif
+
+#else
+    // 32-bit SWAR (SIMD Within A Register) popcount
+
+    // Binary: 0 1 0 1 ...
+    const uint32_t M1 = 0x55555555;
+    // Binary: 00 11 00 11. ..
+    const uint32_t M2 = 0x33333333;
+    // Binary: 0000 1111 0000 1111 ...
+    const uint32_t M4 = 0x0F0F0F0F;
+
+    // Put count of each 2 bits into those 2 bits
+    x = x - ((x >> 1) & M1);
+    // Put count of each 4 bits into those 4 bits
+    x = (x & M2) + ((x >> 2) & M2);
+    // Put count of each 8 bits into those 8 bits
+    x = (x + (x >> 4)) & M4;
+    // Sum of the 4 byte counts.
+    // Take care when considering changes to the next line. Portability and
+    // correctness are delicate here, thanks to C's "integer promotions" (C99
+    // §6.3.1.1p2). On machines where the `int` type has width greater than 32
+    // bits, `x` will be promoted to an `int`, and following C's "usual
+    // arithmetic conversions" (C99 §6.3.1.8), the multiplication will be
+    // performed as a multiplication of two `unsigned int` operands. In this
+    // case it's critical that we cast back to `uint32_t` in order to keep only
+    // the least significant 32 bits. On machines where the `int` type has
+    // width no greater than 32, the multiplication is of two 32-bit unsigned
+    // integer types, and the (uint32_t) cast is a no-op. In both cases, we
+    // avoid the risk of undefined behaviour due to overflow of a
+    // multiplication of signed integer types.
+    return (uint32_t)(x * 0x01010101U) >> 24;
+#endif
+}
+
+
+// Return the index of the most significant 1 bit in 'x'. This is the smallest
+// integer k such that x < 2**k. Equivalent to floor(log2(x)) + 1 for x != 0.
+static inline int
+_Py_bit_length(unsigned long x)
+{
+#if (defined(__clang__) || defined(__GNUC__))
+    if (x != 0) {
+        // __builtin_clzl() is available since GCC 3.4.
+        // Undefined behavior for x == 0.
+        return (int)sizeof(unsigned long) * 8 - __builtin_clzl(x);
+    }
+    else {
+        return 0;
+    }
+#elif defined(_MSC_VER)
+    // _BitScanReverse() is documented to search 32 bits.
+    Py_BUILD_ASSERT(sizeof(unsigned long) <= 4);
+    unsigned long msb;
+    if (_BitScanReverse(&msb, x)) {
+        return (int)msb + 1;
+    }
+    else {
+        return 0;
+    }
+#else
+    const int BIT_LENGTH_TABLE[32] = {
+        0, 1, 2, 2, 3, 3, 3, 3, 4, 4, 4, 4, 4, 4, 4, 4,
+        5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5
+    };
+    int msb = 0;
+    while (x >= 32) {
+        msb += 6;
+        x >>= 6;
+    }
+    msb += BIT_LENGTH_TABLE[x];
+    return msb;
+#endif
+}
+
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* !Py_INTERNAL_BITUTILS_H */

Dependencies/Python/include/internal/pycore_blocks_output_buffer.h

@@ -0,0 +1,321 @@
+/*
+   _BlocksOutputBuffer is used to maintain an output buffer
+   that has unpredictable size. Suitable for compression/decompression
+   API (bz2/lzma/zlib) that has stream->next_out and stream->avail_out:
+
+        stream->next_out:  point to the next output position.
+        stream->avail_out: the number of available bytes left in the buffer.
+
+   It maintains a list of bytes object, so there is no overhead of resizing
+   the buffer.
+
+   Usage:
+
+   1, Initialize the struct instance like this:
+        _BlocksOutputBuffer buffer = {.list = NULL};
+      Set .list to NULL for _BlocksOutputBuffer_OnError()
+
+   2, Initialize the buffer use one of these functions:
+        _BlocksOutputBuffer_InitAndGrow()
+        _BlocksOutputBuffer_InitWithSize()
+
+   3, If (avail_out == 0), grow the buffer:
+        _BlocksOutputBuffer_Grow()
+
+   4, Get the current outputted data size:
+        _BlocksOutputBuffer_GetDataSize()
+
+   5, Finish the buffer, and return a bytes object:
+        _BlocksOutputBuffer_Finish()
+
+   6, Clean up the buffer when an error occurred:
+        _BlocksOutputBuffer_OnError()
+*/
+
+#ifndef Py_INTERNAL_BLOCKS_OUTPUT_BUFFER_H
+#define Py_INTERNAL_BLOCKS_OUTPUT_BUFFER_H
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+#include "Python.h"
+
+#ifndef Py_BUILD_CORE
+#  error "this header requires Py_BUILD_CORE define"
+#endif
+
+typedef struct {
+    // List of bytes objects
+    PyObject *list;
+    // Number of whole allocated size
+    Py_ssize_t allocated;
+    // Max length of the buffer, negative number means unlimited length.
+    Py_ssize_t max_length;
+} _BlocksOutputBuffer;
+
+static const char unable_allocate_msg[] = "Unable to allocate output buffer.";
+
+/* In 32-bit build, the max block size should <= INT32_MAX. */
+#define OUTPUT_BUFFER_MAX_BLOCK_SIZE (256*1024*1024)
+
+/* Block size sequence */
+#define KB (1024)
+#define MB (1024*1024)
+static const Py_ssize_t BUFFER_BLOCK_SIZE[] =
+    { 32*KB, 64*KB, 256*KB, 1*MB, 4*MB, 8*MB, 16*MB, 16*MB,
+      32*MB, 32*MB, 32*MB, 32*MB, 64*MB, 64*MB, 128*MB, 128*MB,
+      OUTPUT_BUFFER_MAX_BLOCK_SIZE };
+#undef KB
+#undef MB
+
+/* According to the block sizes defined by BUFFER_BLOCK_SIZE, the whole
+   allocated size growth step is:
+    1   32 KB       +32 KB
+    2   96 KB       +64 KB
+    3   352 KB      +256 KB
+    4   1.34 MB     +1 MB
+    5   5.34 MB     +4 MB
+    6   13.34 MB    +8 MB
+    7   29.34 MB    +16 MB
+    8   45.34 MB    +16 MB
+    9   77.34 MB    +32 MB
+    10  109.34 MB   +32 MB
+    11  141.34 MB   +32 MB
+    12  173.34 MB   +32 MB
+    13  237.34 MB   +64 MB
+    14  301.34 MB   +64 MB
+    15  429.34 MB   +128 MB
+    16  557.34 MB   +128 MB
+    17  813.34 MB   +256 MB
+    18  1069.34 MB  +256 MB
+    19  1325.34 MB  +256 MB
+    20  1581.34 MB  +256 MB
+    21  1837.34 MB  +256 MB
+    22  2093.34 MB  +256 MB
+    ...
+*/
+
+/* Initialize the buffer, and grow the buffer.
+
+   max_length: Max length of the buffer, -1 for unlimited length.
+
+   On success, return allocated size (>=0)
+   On failure, return -1
+*/
+static inline Py_ssize_t
+_BlocksOutputBuffer_InitAndGrow(_BlocksOutputBuffer *buffer,
+                                const Py_ssize_t max_length,
+                                void **next_out)
+{
+    PyObject *b;
+    Py_ssize_t block_size;
+
+    // ensure .list was set to NULL
+    assert(buffer->list == NULL);
+
+    // get block size
+    if (0 <= max_length && max_length < BUFFER_BLOCK_SIZE[0]) {
+        block_size = max_length;
+    } else {
+        block_size = BUFFER_BLOCK_SIZE[0];
+    }
+
+    // the first block
+    b = PyBytes_FromStringAndSize(NULL, block_size);
+    if (b == NULL) {
+        return -1;
+    }
+
+    // create the list
+    buffer->list = PyList_New(1);
+    if (buffer->list == NULL) {
+        Py_DECREF(b);
+        return -1;
+    }
+    PyList_SET_ITEM(buffer->list, 0, b);
+
+    // set variables
+    buffer->allocated = block_size;
+    buffer->max_length = max_length;
+
+    *next_out = PyBytes_AS_STRING(b);
+    return block_size;
+}
+
+/* Initialize the buffer, with an initial size.
+
+   Check block size limit in the outer wrapper function. For example, some libs
+   accept UINT32_MAX as the maximum block size, then init_size should <= it.
+
+   On success, return allocated size (>=0)
+   On failure, return -1
+*/
+static inline Py_ssize_t
+_BlocksOutputBuffer_InitWithSize(_BlocksOutputBuffer *buffer,
+                                 const Py_ssize_t init_size,
+                                 void **next_out)
+{
+    PyObject *b;
+
+    // ensure .list was set to NULL
+    assert(buffer->list == NULL);
+
+    // the first block
+    b = PyBytes_FromStringAndSize(NULL, init_size);
+    if (b == NULL) {
+        PyErr_SetString(PyExc_MemoryError, unable_allocate_msg);
+        return -1;
+    }
+
+    // create the list
+    buffer->list = PyList_New(1);
+    if (buffer->list == NULL) {
+        Py_DECREF(b);
+        return -1;
+    }
+    PyList_SET_ITEM(buffer->list, 0, b);
+
+    // set variables
+    buffer->allocated = init_size;
+    buffer->max_length = -1;
+
+    *next_out = PyBytes_AS_STRING(b);
+    return init_size;
+}
+
+/* Grow the buffer. The avail_out must be 0, please check it before calling.
+
+   On success, return allocated size (>=0)
+   On failure, return -1
+*/
+static inline Py_ssize_t
+_BlocksOutputBuffer_Grow(_BlocksOutputBuffer *buffer,
+                         void **next_out,
+                         const Py_ssize_t avail_out)
+{
+    PyObject *b;
+    const Py_ssize_t list_len = Py_SIZE(buffer->list);
+    Py_ssize_t block_size;
+
+    // ensure no gaps in the data
+    if (avail_out != 0) {
+        PyErr_SetString(PyExc_SystemError,
+                        "avail_out is non-zero in _BlocksOutputBuffer_Grow().");
+        return -1;
+    }
+
+    // get block size
+    if (list_len < (Py_ssize_t) Py_ARRAY_LENGTH(BUFFER_BLOCK_SIZE)) {
+        block_size = BUFFER_BLOCK_SIZE[list_len];
+    } else {
+        block_size = BUFFER_BLOCK_SIZE[Py_ARRAY_LENGTH(BUFFER_BLOCK_SIZE) - 1];
+    }
+
+    // check max_length
+    if (buffer->max_length >= 0) {
+        // if (rest == 0), should not grow the buffer.
+        Py_ssize_t rest = buffer->max_length - buffer->allocated;
+        assert(rest > 0);
+
+        // block_size of the last block
+        if (block_size > rest) {
+            block_size = rest;
+        }
+    }
+
+    // check buffer->allocated overflow
+    if (block_size > PY_SSIZE_T_MAX - buffer->allocated) {
+        PyErr_SetString(PyExc_MemoryError, unable_allocate_msg);
+        return -1;
+    }
+
+    // create the block
+    b = PyBytes_FromStringAndSize(NULL, block_size);
+    if (b == NULL) {
+        PyErr_SetString(PyExc_MemoryError, unable_allocate_msg);
+        return -1;
+    }
+    if (PyList_Append(buffer->list, b) < 0) {
+        Py_DECREF(b);
+        return -1;
+    }
+    Py_DECREF(b);
+
+    // set variables
+    buffer->allocated += block_size;
+
+    *next_out = PyBytes_AS_STRING(b);
+    return block_size;
+}
+
+/* Return the current outputted data size. */
+static inline Py_ssize_t
+_BlocksOutputBuffer_GetDataSize(_BlocksOutputBuffer *buffer,
+                                const Py_ssize_t avail_out)
+{
+    return buffer->allocated - avail_out;
+}
+
+/* Finish the buffer.
+
+   Return a bytes object on success
+   Return NULL on failure
+*/
+static inline PyObject *
+_BlocksOutputBuffer_Finish(_BlocksOutputBuffer *buffer,
+                           const Py_ssize_t avail_out)
+{
+    PyObject *result, *block;
+    const Py_ssize_t list_len = Py_SIZE(buffer->list);
+
+    // fast path for single block
+    if ((list_len == 1 && avail_out == 0) ||
+        (list_len == 2 && Py_SIZE(PyList_GET_ITEM(buffer->list, 1)) == avail_out))
+    {
+        block = PyList_GET_ITEM(buffer->list, 0);
+        Py_INCREF(block);
+
+        Py_CLEAR(buffer->list);
+        return block;
+    }
+
+    // final bytes object
+    result = PyBytes_FromStringAndSize(NULL, buffer->allocated - avail_out);
+    if (result == NULL) {
+        PyErr_SetString(PyExc_MemoryError, unable_allocate_msg);
+        return NULL;
+    }
+
+    // memory copy
+    if (list_len > 0) {
+        char *posi = PyBytes_AS_STRING(result);
+
+        // blocks except the last one
+        Py_ssize_t i = 0;
+        for (; i < list_len-1; i++) {
+            block = PyList_GET_ITEM(buffer->list, i);
+            memcpy(posi, PyBytes_AS_STRING(block), Py_SIZE(block));
+            posi += Py_SIZE(block);
+        }
+        // the last block
+        block = PyList_GET_ITEM(buffer->list, i);
+        memcpy(posi, PyBytes_AS_STRING(block), Py_SIZE(block) - avail_out);
+    } else {
+        assert(Py_SIZE(result) == 0);
+    }
+
+    Py_CLEAR(buffer->list);
+    return result;
+}
+
+/* Clean up the buffer when an error occurred. */
+static inline void
+_BlocksOutputBuffer_OnError(_BlocksOutputBuffer *buffer)
+{
+    Py_CLEAR(buffer->list);
+}
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* Py_INTERNAL_BLOCKS_OUTPUT_BUFFER_H */

Dependencies/Python/include/internal/pycore_brc.h

@@ -0,0 +1,74 @@
+#ifndef Py_INTERNAL_BRC_H
+#define Py_INTERNAL_BRC_H
+
+#include <stdint.h>
+#include "pycore_llist.h"           // struct llist_node
+#include "pycore_lock.h"            // PyMutex
+#include "pycore_object_stack.h"    // _PyObjectStack
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+#ifndef Py_BUILD_CORE
+#  error "this header requires Py_BUILD_CORE define"
+#endif
+
+#ifdef Py_GIL_DISABLED
+
+// Prime number to avoid correlations with memory addresses.
+#define _Py_BRC_NUM_BUCKETS 257
+
+// Hash table bucket
+struct _brc_bucket {
+    // Mutex protects both the bucket and thread state queues in this bucket.
+    PyMutex mutex;
+
+    // Linked list of _PyThreadStateImpl objects hashed to this bucket.
+    struct llist_node root;
+};
+
+// Per-interpreter biased reference counting state
+struct _brc_state {
+    // Hash table of thread states by thread-id. Thread states within a bucket
+    // are chained using a doubly-linked list.
+    struct _brc_bucket table[_Py_BRC_NUM_BUCKETS];
+};
+
+// Per-thread biased reference counting state
+struct _brc_thread_state {
+    // Linked-list of thread states per hash bucket
+    struct llist_node bucket_node;
+
+    // Thread-id as determined by _PyThread_Id()
+    uintptr_t tid;
+
+    // Objects with refcounts to be merged (protected by bucket mutex)
+    _PyObjectStack objects_to_merge;
+
+    // Local stack of objects to be merged (not accessed by other threads)
+    _PyObjectStack local_objects_to_merge;
+};
+
+// Initialize/finalize the per-thread biased reference counting state
+void _Py_brc_init_thread(PyThreadState *tstate);
+void _Py_brc_remove_thread(PyThreadState *tstate);
+
+// Initialize per-interpreter state
+void _Py_brc_init_state(PyInterpreterState *interp);
+
+void _Py_brc_after_fork(PyInterpreterState *interp);
+
+// Enqueues an object to be merged by it's owning thread (tid). This
+// steals a reference to the object.
+void _Py_brc_queue_object(PyObject *ob);
+
+// Merge the refcounts of queued objects for the current thread.
+void _Py_brc_merge_refcounts(PyThreadState *tstate);
+
+#endif /* Py_GIL_DISABLED */
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* !Py_INTERNAL_BRC_H */

Dependencies/Python/include/internal/pycore_bytes_methods.h

@@ -0,0 +1,82 @@
+#ifndef Py_LIMITED_API
+#ifndef Py_BYTES_CTYPE_H
+#define Py_BYTES_CTYPE_H
+
+#ifndef Py_BUILD_CORE
+#  error "this header requires Py_BUILD_CORE define"
+#endif
+
+/*
+ * The internal implementation behind PyBytes (bytes) and PyByteArray (bytearray)
+ * methods of the given names, they operate on ASCII byte strings.
+ */
+extern PyObject* _Py_bytes_isspace(const char *cptr, Py_ssize_t len);
+extern PyObject* _Py_bytes_isalpha(const char *cptr, Py_ssize_t len);
+extern PyObject* _Py_bytes_isalnum(const char *cptr, Py_ssize_t len);
+extern PyObject* _Py_bytes_isascii(const char *cptr, Py_ssize_t len);
+extern PyObject* _Py_bytes_isdigit(const char *cptr, Py_ssize_t len);
+extern PyObject* _Py_bytes_islower(const char *cptr, Py_ssize_t len);
+extern PyObject* _Py_bytes_isupper(const char *cptr, Py_ssize_t len);
+extern PyObject* _Py_bytes_istitle(const char *cptr, Py_ssize_t len);
+
+/* These store their len sized answer in the given preallocated *result arg. */
+extern void _Py_bytes_lower(char *result, const char *cptr, Py_ssize_t len);
+extern void _Py_bytes_upper(char *result, const char *cptr, Py_ssize_t len);
+extern void _Py_bytes_title(char *result, const char *s, Py_ssize_t len);
+extern void _Py_bytes_capitalize(char *result, const char *s, Py_ssize_t len);
+extern void _Py_bytes_swapcase(char *result, const char *s, Py_ssize_t len);
+
+extern PyObject *_Py_bytes_find(const char *str, Py_ssize_t len, PyObject *sub,
+                                Py_ssize_t start, Py_ssize_t end);
+extern PyObject *_Py_bytes_index(const char *str, Py_ssize_t len, PyObject *sub,
+                                 Py_ssize_t start, Py_ssize_t end);
+extern PyObject *_Py_bytes_rfind(const char *str, Py_ssize_t len, PyObject *sub,
+                                 Py_ssize_t start, Py_ssize_t end);
+extern PyObject *_Py_bytes_rindex(const char *str, Py_ssize_t len, PyObject *sub,
+                                 Py_ssize_t start, Py_ssize_t end);
+extern PyObject *_Py_bytes_count(const char *str, Py_ssize_t len, PyObject *sub,
+                                 Py_ssize_t start, Py_ssize_t end);
+extern int _Py_bytes_contains(const char *str, Py_ssize_t len, PyObject *arg);
+extern PyObject *_Py_bytes_startswith(const char *str, Py_ssize_t len,
+                                      PyObject *subobj, Py_ssize_t start,
+                                      Py_ssize_t end);
+extern PyObject *_Py_bytes_endswith(const char *str, Py_ssize_t len,
+                                    PyObject *subobj, Py_ssize_t start,
+                                    Py_ssize_t end);
+
+/* The maketrans() static method. */
+extern PyObject* _Py_bytes_maketrans(Py_buffer *frm, Py_buffer *to);
+
+/* Shared __doc__ strings. */
+extern const char _Py_isspace__doc__[];
+extern const char _Py_isalpha__doc__[];
+extern const char _Py_isalnum__doc__[];
+extern const char _Py_isascii__doc__[];
+extern const char _Py_isdigit__doc__[];
+extern const char _Py_islower__doc__[];
+extern const char _Py_isupper__doc__[];
+extern const char _Py_istitle__doc__[];
+extern const char _Py_lower__doc__[];
+extern const char _Py_upper__doc__[];
+extern const char _Py_title__doc__[];
+extern const char _Py_capitalize__doc__[];
+extern const char _Py_swapcase__doc__[];
+extern const char _Py_count__doc__[];
+extern const char _Py_find__doc__[];
+extern const char _Py_index__doc__[];
+extern const char _Py_rfind__doc__[];
+extern const char _Py_rindex__doc__[];
+extern const char _Py_startswith__doc__[];
+extern const char _Py_endswith__doc__[];
+extern const char _Py_maketrans__doc__[];
+extern const char _Py_expandtabs__doc__[];
+extern const char _Py_ljust__doc__[];
+extern const char _Py_rjust__doc__[];
+extern const char _Py_center__doc__[];
+extern const char _Py_zfill__doc__[];
+
+/* this is needed because some docs are shared from the .o, not static */
+#define PyDoc_STRVAR_shared(name,str) const char name[] = PyDoc_STR(str)
+
+#endif /* !Py_BYTES_CTYPE_H */
+#endif /* !Py_LIMITED_API */

Dependencies/Python/include/internal/pycore_bytesobject.h

@@ -0,0 +1,148 @@
+#ifndef Py_INTERNAL_BYTESOBJECT_H
+#define Py_INTERNAL_BYTESOBJECT_H
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+#ifndef Py_BUILD_CORE
+#  error "this header requires Py_BUILD_CORE define"
+#endif
+
+extern PyObject* _PyBytes_FormatEx(
+    const char *format,
+    Py_ssize_t format_len,
+    PyObject *args,
+    int use_bytearray);
+
+extern PyObject* _PyBytes_FromHex(
+    PyObject *string,
+    int use_bytearray);
+
+// Helper for PyBytes_DecodeEscape that detects invalid escape chars.
+// Export for test_peg_generator.
+PyAPI_FUNC(PyObject*) _PyBytes_DecodeEscape(const char *, Py_ssize_t,
+                                            const char *, const char **);
+
+
+// Substring Search.
+//
+// Returns the index of the first occurrence of
+// a substring ("needle") in a larger text ("haystack").
+// If the needle is not found, return -1.
+// If the needle is found, add offset to the index.
+//
+// Export for 'mmap' shared extension.
+PyAPI_FUNC(Py_ssize_t)
+_PyBytes_Find(const char *haystack, Py_ssize_t len_haystack,
+              const char *needle, Py_ssize_t len_needle,
+              Py_ssize_t offset);
+
+// Same as above, but search right-to-left.
+// Export for 'mmap' shared extension.
+PyAPI_FUNC(Py_ssize_t)
+_PyBytes_ReverseFind(const char *haystack, Py_ssize_t len_haystack,
+                     const char *needle, Py_ssize_t len_needle,
+                     Py_ssize_t offset);
+
+
+// Helper function to implement the repeat and inplace repeat methods on a
+// buffer.
+//
+// len_dest is assumed to be an integer multiple of len_src.
+// If src equals dest, then assume the operation is inplace.
+//
+// This method repeately doubles the number of bytes copied to reduce
+// the number of invocations of memcpy.
+//
+// Export for 'array' shared extension.
+PyAPI_FUNC(void)
+_PyBytes_Repeat(char* dest, Py_ssize_t len_dest,
+    const char* src, Py_ssize_t len_src);
+
+/* --- _PyBytesWriter ----------------------------------------------------- */
+
+/* The _PyBytesWriter structure is big: it contains an embedded "stack buffer".
+   A _PyBytesWriter variable must be declared at the end of variables in a
+   function to optimize the memory allocation on the stack. */
+typedef struct {
+    /* bytes, bytearray or NULL (when the small buffer is used) */
+    PyObject *buffer;
+
+    /* Number of allocated size. */
+    Py_ssize_t allocated;
+
+    /* Minimum number of allocated bytes,
+       incremented by _PyBytesWriter_Prepare() */
+    Py_ssize_t min_size;
+
+    /* If non-zero, use a bytearray instead of a bytes object for buffer. */
+    int use_bytearray;
+
+    /* If non-zero, overallocate the buffer (default: 0).
+       This flag must be zero if use_bytearray is non-zero. */
+    int overallocate;
+
+    /* Stack buffer */
+    int use_small_buffer;
+    char small_buffer[512];
+} _PyBytesWriter;
+
+/* Initialize a bytes writer
+
+   By default, the overallocation is disabled. Set the overallocate attribute
+   to control the allocation of the buffer.
+
+   Export _PyBytesWriter API for '_pickle' shared extension. */
+PyAPI_FUNC(void) _PyBytesWriter_Init(_PyBytesWriter *writer);
+
+/* Get the buffer content and reset the writer.
+   Return a bytes object, or a bytearray object if use_bytearray is non-zero.
+   Raise an exception and return NULL on error. */
+PyAPI_FUNC(PyObject *) _PyBytesWriter_Finish(_PyBytesWriter *writer,
+    void *str);
+
+/* Deallocate memory of a writer (clear its internal buffer). */
+PyAPI_FUNC(void) _PyBytesWriter_Dealloc(_PyBytesWriter *writer);
+
+/* Allocate the buffer to write size bytes.
+   Return the pointer to the beginning of buffer data.
+   Raise an exception and return NULL on error. */
+PyAPI_FUNC(void*) _PyBytesWriter_Alloc(_PyBytesWriter *writer,
+    Py_ssize_t size);
+
+/* Ensure that the buffer is large enough to write *size* bytes.
+   Add size to the writer minimum size (min_size attribute).
+
+   str is the current pointer inside the buffer.
+   Return the updated current pointer inside the buffer.
+   Raise an exception and return NULL on error. */
+PyAPI_FUNC(void*) _PyBytesWriter_Prepare(_PyBytesWriter *writer,
+    void *str,
+    Py_ssize_t size);
+
+/* Resize the buffer to make it larger.
+   The new buffer may be larger than size bytes because of overallocation.
+   Return the updated current pointer inside the buffer.
+   Raise an exception and return NULL on error.
+
+   Note: size must be greater than the number of allocated bytes in the writer.
+
+   This function doesn't use the writer minimum size (min_size attribute).
+
+   See also _PyBytesWriter_Prepare().
+   */
+PyAPI_FUNC(void*) _PyBytesWriter_Resize(_PyBytesWriter *writer,
+    void *str,
+    Py_ssize_t size);
+
+/* Write bytes.
+   Raise an exception and return NULL on error. */
+PyAPI_FUNC(void*) _PyBytesWriter_WriteBytes(_PyBytesWriter *writer,
+    void *str,
+    const void *bytes,
+    Py_ssize_t size);
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* !Py_INTERNAL_BYTESOBJECT_H */