mirror of
				https://github.com/python/cpython.git
				synced 2025-10-31 10:26:02 +00:00 
			
		
		
		
	 deb921f851
			
		
	
	
		deb921f851
		
			
		
	
	
	
	
		
			
			This change gives a significant speedup, as the METH_FASTCALL calling convention is now used. The following bytes and bytearray methods are adapted: - count() - find() - index() - rfind() - rindex() Co-authored-by: Inada Naoki <songofacandy@gmail.com>
		
			
				
	
	
		
			2585 lines
		
	
	
	
		
			75 KiB
		
	
	
	
		
			C
		
	
	
	
	
	
			
		
		
	
	
			2585 lines
		
	
	
	
		
			75 KiB
		
	
	
	
		
			C
		
	
	
	
	
	
| /* PyByteArray (bytearray) implementation */
 | |
| 
 | |
| #include "Python.h"
 | |
| #include "pycore_abstract.h"      // _PyIndex_Check()
 | |
| #include "pycore_bytes_methods.h"
 | |
| #include "pycore_bytesobject.h"
 | |
| #include "pycore_ceval.h"         // _PyEval_GetBuiltin()
 | |
| #include "pycore_object.h"        // _PyObject_GC_UNTRACK()
 | |
| #include "pycore_strhex.h"        // _Py_strhex_with_sep()
 | |
| #include "pycore_long.h"          // _PyLong_FromUnsignedChar()
 | |
| #include "bytesobject.h"
 | |
| 
 | |
| /*[clinic input]
 | |
| class bytearray "PyByteArrayObject *" "&PyByteArray_Type"
 | |
| [clinic start generated code]*/
 | |
| /*[clinic end generated code: output=da39a3ee5e6b4b0d input=5535b77c37a119e0]*/
 | |
| 
 | |
| /* For PyByteArray_AS_STRING(). */
 | |
| char _PyByteArray_empty_string[] = "";
 | |
| 
 | |
| /* Helpers */
 | |
| 
 | |
| static int
 | |
| _getbytevalue(PyObject* arg, int *value)
 | |
| {
 | |
|     int overflow;
 | |
|     long face_value = PyLong_AsLongAndOverflow(arg, &overflow);
 | |
| 
 | |
|     if (face_value == -1 && PyErr_Occurred()) {
 | |
|         *value = -1;
 | |
|         return 0;
 | |
|     }
 | |
|     if (face_value < 0 || face_value >= 256) {
 | |
|         /* this includes an overflow in converting to C long */
 | |
|         PyErr_SetString(PyExc_ValueError, "byte must be in range(0, 256)");
 | |
|         *value = -1;
 | |
|         return 0;
 | |
|     }
 | |
| 
 | |
|     *value = face_value;
 | |
|     return 1;
 | |
| }
 | |
| 
 | |
| static int
 | |
| bytearray_getbuffer(PyByteArrayObject *obj, Py_buffer *view, int flags)
 | |
| {
 | |
|     void *ptr;
 | |
|     if (view == NULL) {
 | |
|         PyErr_SetString(PyExc_BufferError,
 | |
|             "bytearray_getbuffer: view==NULL argument is obsolete");
 | |
|         return -1;
 | |
|     }
 | |
|     ptr = (void *) PyByteArray_AS_STRING(obj);
 | |
|     /* cannot fail if view != NULL and readonly == 0 */
 | |
|     (void)PyBuffer_FillInfo(view, (PyObject*)obj, ptr, Py_SIZE(obj), 0, flags);
 | |
|     obj->ob_exports++;
 | |
|     return 0;
 | |
| }
 | |
| 
 | |
| static void
 | |
| bytearray_releasebuffer(PyByteArrayObject *obj, Py_buffer *view)
 | |
| {
 | |
|     obj->ob_exports--;
 | |
|     assert(obj->ob_exports >= 0);
 | |
| }
 | |
| 
 | |
| static int
 | |
| _canresize(PyByteArrayObject *self)
 | |
| {
 | |
|     if (self->ob_exports > 0) {
 | |
|         PyErr_SetString(PyExc_BufferError,
 | |
|                 "Existing exports of data: object cannot be re-sized");
 | |
|         return 0;
 | |
|     }
 | |
|     return 1;
 | |
| }
 | |
| 
 | |
| #include "clinic/bytearrayobject.c.h"
 | |
| 
 | |
| /* Direct API functions */
 | |
| 
 | |
| PyObject *
 | |
| PyByteArray_FromObject(PyObject *input)
 | |
| {
 | |
|     return PyObject_CallOneArg((PyObject *)&PyByteArray_Type, input);
 | |
| }
 | |
| 
 | |
| static PyObject *
 | |
| _PyByteArray_FromBufferObject(PyObject *obj)
 | |
| {
 | |
|     PyObject *result;
 | |
|     Py_buffer view;
 | |
| 
 | |
|     if (PyObject_GetBuffer(obj, &view, PyBUF_FULL_RO) < 0) {
 | |
|         return NULL;
 | |
|     }
 | |
|     result = PyByteArray_FromStringAndSize(NULL, view.len);
 | |
|     if (result != NULL &&
 | |
|         PyBuffer_ToContiguous(PyByteArray_AS_STRING(result),
 | |
|                               &view, view.len, 'C') < 0)
 | |
|     {
 | |
|         Py_CLEAR(result);
 | |
|     }
 | |
|     PyBuffer_Release(&view);
 | |
|     return result;
 | |
| }
 | |
| 
 | |
| PyObject *
 | |
| PyByteArray_FromStringAndSize(const char *bytes, Py_ssize_t size)
 | |
| {
 | |
|     PyByteArrayObject *new;
 | |
|     Py_ssize_t alloc;
 | |
| 
 | |
|     if (size < 0) {
 | |
|         PyErr_SetString(PyExc_SystemError,
 | |
|             "Negative size passed to PyByteArray_FromStringAndSize");
 | |
|         return NULL;
 | |
|     }
 | |
| 
 | |
|     /* Prevent buffer overflow when setting alloc to size+1. */
 | |
|     if (size == PY_SSIZE_T_MAX) {
 | |
|         return PyErr_NoMemory();
 | |
|     }
 | |
| 
 | |
|     new = PyObject_New(PyByteArrayObject, &PyByteArray_Type);
 | |
|     if (new == NULL)
 | |
|         return NULL;
 | |
| 
 | |
|     if (size == 0) {
 | |
|         new->ob_bytes = NULL;
 | |
|         alloc = 0;
 | |
|     }
 | |
|     else {
 | |
|         alloc = size + 1;
 | |
|         new->ob_bytes = PyMem_Malloc(alloc);
 | |
|         if (new->ob_bytes == NULL) {
 | |
|             Py_DECREF(new);
 | |
|             return PyErr_NoMemory();
 | |
|         }
 | |
|         if (bytes != NULL && size > 0)
 | |
|             memcpy(new->ob_bytes, bytes, size);
 | |
|         new->ob_bytes[size] = '\0';  /* Trailing null byte */
 | |
|     }
 | |
|     Py_SET_SIZE(new, size);
 | |
|     new->ob_alloc = alloc;
 | |
|     new->ob_start = new->ob_bytes;
 | |
|     new->ob_exports = 0;
 | |
| 
 | |
|     return (PyObject *)new;
 | |
| }
 | |
| 
 | |
| Py_ssize_t
 | |
| PyByteArray_Size(PyObject *self)
 | |
| {
 | |
|     assert(self != NULL);
 | |
|     assert(PyByteArray_Check(self));
 | |
| 
 | |
|     return PyByteArray_GET_SIZE(self);
 | |
| }
 | |
| 
 | |
| char  *
 | |
| PyByteArray_AsString(PyObject *self)
 | |
| {
 | |
|     assert(self != NULL);
 | |
|     assert(PyByteArray_Check(self));
 | |
| 
 | |
|     return PyByteArray_AS_STRING(self);
 | |
| }
 | |
| 
 | |
| int
 | |
| PyByteArray_Resize(PyObject *self, Py_ssize_t requested_size)
 | |
| {
 | |
|     void *sval;
 | |
|     PyByteArrayObject *obj = ((PyByteArrayObject *)self);
 | |
|     /* All computations are done unsigned to avoid integer overflows
 | |
|        (see issue #22335). */
 | |
|     size_t alloc = (size_t) obj->ob_alloc;
 | |
|     size_t logical_offset = (size_t) (obj->ob_start - obj->ob_bytes);
 | |
|     size_t size = (size_t) requested_size;
 | |
| 
 | |
|     assert(self != NULL);
 | |
|     assert(PyByteArray_Check(self));
 | |
|     assert(logical_offset <= alloc);
 | |
|     assert(requested_size >= 0);
 | |
| 
 | |
|     if (requested_size == Py_SIZE(self)) {
 | |
|         return 0;
 | |
|     }
 | |
|     if (!_canresize(obj)) {
 | |
|         return -1;
 | |
|     }
 | |
| 
 | |
|     if (size + logical_offset + 1 <= alloc) {
 | |
|         /* Current buffer is large enough to host the requested size,
 | |
|            decide on a strategy. */
 | |
|         if (size < alloc / 2) {
 | |
|             /* Major downsize; resize down to exact size */
 | |
|             alloc = size + 1;
 | |
|         }
 | |
|         else {
 | |
|             /* Minor downsize; quick exit */
 | |
|             Py_SET_SIZE(self, size);
 | |
|             PyByteArray_AS_STRING(self)[size] = '\0'; /* Trailing null */
 | |
|             return 0;
 | |
|         }
 | |
|     }
 | |
|     else {
 | |
|         /* Need growing, decide on a strategy */
 | |
|         if (size <= alloc * 1.125) {
 | |
|             /* Moderate upsize; overallocate similar to list_resize() */
 | |
|             alloc = size + (size >> 3) + (size < 9 ? 3 : 6);
 | |
|         }
 | |
|         else {
 | |
|             /* Major upsize; resize up to exact size */
 | |
|             alloc = size + 1;
 | |
|         }
 | |
|     }
 | |
|     if (alloc > PY_SSIZE_T_MAX) {
 | |
|         PyErr_NoMemory();
 | |
|         return -1;
 | |
|     }
 | |
| 
 | |
|     if (logical_offset > 0) {
 | |
|         sval = PyMem_Malloc(alloc);
 | |
|         if (sval == NULL) {
 | |
|             PyErr_NoMemory();
 | |
|             return -1;
 | |
|         }
 | |
|         memcpy(sval, PyByteArray_AS_STRING(self),
 | |
|                Py_MIN((size_t)requested_size, (size_t)Py_SIZE(self)));
 | |
|         PyMem_Free(obj->ob_bytes);
 | |
|     }
 | |
|     else {
 | |
|         sval = PyMem_Realloc(obj->ob_bytes, alloc);
 | |
|         if (sval == NULL) {
 | |
|             PyErr_NoMemory();
 | |
|             return -1;
 | |
|         }
 | |
|     }
 | |
| 
 | |
|     obj->ob_bytes = obj->ob_start = sval;
 | |
|     Py_SET_SIZE(self, size);
 | |
|     obj->ob_alloc = alloc;
 | |
|     obj->ob_bytes[size] = '\0'; /* Trailing null byte */
 | |
| 
 | |
|     return 0;
 | |
| }
 | |
| 
 | |
| PyObject *
 | |
| PyByteArray_Concat(PyObject *a, PyObject *b)
 | |
| {
 | |
|     Py_buffer va, vb;
 | |
|     PyByteArrayObject *result = NULL;
 | |
| 
 | |
|     va.len = -1;
 | |
|     vb.len = -1;
 | |
|     if (PyObject_GetBuffer(a, &va, PyBUF_SIMPLE) != 0 ||
 | |
|         PyObject_GetBuffer(b, &vb, PyBUF_SIMPLE) != 0) {
 | |
|             PyErr_Format(PyExc_TypeError, "can't concat %.100s to %.100s",
 | |
|                          Py_TYPE(b)->tp_name, Py_TYPE(a)->tp_name);
 | |
|             goto done;
 | |
|     }
 | |
| 
 | |
|     if (va.len > PY_SSIZE_T_MAX - vb.len) {
 | |
|         PyErr_NoMemory();
 | |
|         goto done;
 | |
|     }
 | |
| 
 | |
|     result = (PyByteArrayObject *) \
 | |
|         PyByteArray_FromStringAndSize(NULL, va.len + vb.len);
 | |
|     // result->ob_bytes is NULL if result is an empty bytearray:
 | |
|     // if va.len + vb.len equals zero.
 | |
|     if (result != NULL && result->ob_bytes != NULL) {
 | |
|         memcpy(result->ob_bytes, va.buf, va.len);
 | |
|         memcpy(result->ob_bytes + va.len, vb.buf, vb.len);
 | |
|     }
 | |
| 
 | |
|   done:
 | |
|     if (va.len != -1)
 | |
|         PyBuffer_Release(&va);
 | |
|     if (vb.len != -1)
 | |
|         PyBuffer_Release(&vb);
 | |
|     return (PyObject *)result;
 | |
| }
 | |
| 
 | |
| /* Functions stuffed into the type object */
 | |
| 
 | |
| static Py_ssize_t
 | |
| bytearray_length(PyByteArrayObject *self)
 | |
| {
 | |
|     return Py_SIZE(self);
 | |
| }
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_iconcat(PyByteArrayObject *self, PyObject *other)
 | |
| {
 | |
|     Py_ssize_t size;
 | |
|     Py_buffer vo;
 | |
| 
 | |
|     if (PyObject_GetBuffer(other, &vo, PyBUF_SIMPLE) != 0) {
 | |
|         PyErr_Format(PyExc_TypeError, "can't concat %.100s to %.100s",
 | |
|                      Py_TYPE(other)->tp_name, Py_TYPE(self)->tp_name);
 | |
|         return NULL;
 | |
|     }
 | |
| 
 | |
|     size = Py_SIZE(self);
 | |
|     if (size > PY_SSIZE_T_MAX - vo.len) {
 | |
|         PyBuffer_Release(&vo);
 | |
|         return PyErr_NoMemory();
 | |
|     }
 | |
|     if (PyByteArray_Resize((PyObject *)self, size + vo.len) < 0) {
 | |
|         PyBuffer_Release(&vo);
 | |
|         return NULL;
 | |
|     }
 | |
|     memcpy(PyByteArray_AS_STRING(self) + size, vo.buf, vo.len);
 | |
|     PyBuffer_Release(&vo);
 | |
|     return Py_NewRef(self);
 | |
| }
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_repeat(PyByteArrayObject *self, Py_ssize_t count)
 | |
| {
 | |
|     if (count < 0)
 | |
|         count = 0;
 | |
|     const Py_ssize_t mysize = Py_SIZE(self);
 | |
|     if (count > 0 && mysize > PY_SSIZE_T_MAX / count)
 | |
|         return PyErr_NoMemory();
 | |
|     Py_ssize_t size = mysize * count;
 | |
|     PyByteArrayObject* result = (PyByteArrayObject *)PyByteArray_FromStringAndSize(NULL, size);
 | |
|     const char* buf = PyByteArray_AS_STRING(self);
 | |
|     if (result != NULL && size != 0) {
 | |
|         _PyBytes_Repeat(result->ob_bytes, size, buf, mysize);
 | |
|     }
 | |
|     return (PyObject *)result;
 | |
| }
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_irepeat(PyByteArrayObject *self, Py_ssize_t count)
 | |
| {
 | |
|     if (count < 0)
 | |
|         count = 0;
 | |
|     else if (count == 1) {
 | |
|         return Py_NewRef(self);
 | |
|     }
 | |
| 
 | |
|     const Py_ssize_t mysize = Py_SIZE(self);
 | |
|     if (count > 0 && mysize > PY_SSIZE_T_MAX / count)
 | |
|         return PyErr_NoMemory();
 | |
|     const Py_ssize_t size = mysize * count;
 | |
|     if (PyByteArray_Resize((PyObject *)self, size) < 0)
 | |
|         return NULL;
 | |
| 
 | |
|     char* buf = PyByteArray_AS_STRING(self);
 | |
|     _PyBytes_Repeat(buf, size, buf, mysize);
 | |
| 
 | |
|     return Py_NewRef(self);
 | |
| }
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_getitem(PyByteArrayObject *self, Py_ssize_t i)
 | |
| {
 | |
|     if (i < 0 || i >= Py_SIZE(self)) {
 | |
|         PyErr_SetString(PyExc_IndexError, "bytearray index out of range");
 | |
|         return NULL;
 | |
|     }
 | |
|     return _PyLong_FromUnsignedChar((unsigned char)(self->ob_start[i]));
 | |
| }
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_subscript(PyByteArrayObject *self, PyObject *index)
 | |
| {
 | |
|     if (_PyIndex_Check(index)) {
 | |
|         Py_ssize_t i = PyNumber_AsSsize_t(index, PyExc_IndexError);
 | |
| 
 | |
|         if (i == -1 && PyErr_Occurred())
 | |
|             return NULL;
 | |
| 
 | |
|         if (i < 0)
 | |
|             i += PyByteArray_GET_SIZE(self);
 | |
| 
 | |
|         if (i < 0 || i >= Py_SIZE(self)) {
 | |
|             PyErr_SetString(PyExc_IndexError, "bytearray index out of range");
 | |
|             return NULL;
 | |
|         }
 | |
|         return _PyLong_FromUnsignedChar((unsigned char)(self->ob_start[i]));
 | |
|     }
 | |
|     else if (PySlice_Check(index)) {
 | |
|         Py_ssize_t start, stop, step, slicelength, i;
 | |
|         size_t cur;
 | |
|         if (PySlice_Unpack(index, &start, &stop, &step) < 0) {
 | |
|             return NULL;
 | |
|         }
 | |
|         slicelength = PySlice_AdjustIndices(PyByteArray_GET_SIZE(self),
 | |
|                                             &start, &stop, step);
 | |
| 
 | |
|         if (slicelength <= 0)
 | |
|             return PyByteArray_FromStringAndSize("", 0);
 | |
|         else if (step == 1) {
 | |
|             return PyByteArray_FromStringAndSize(
 | |
|                 PyByteArray_AS_STRING(self) + start, slicelength);
 | |
|         }
 | |
|         else {
 | |
|             char *source_buf = PyByteArray_AS_STRING(self);
 | |
|             char *result_buf;
 | |
|             PyObject *result;
 | |
| 
 | |
|             result = PyByteArray_FromStringAndSize(NULL, slicelength);
 | |
|             if (result == NULL)
 | |
|                 return NULL;
 | |
| 
 | |
|             result_buf = PyByteArray_AS_STRING(result);
 | |
|             for (cur = start, i = 0; i < slicelength;
 | |
|                  cur += step, i++) {
 | |
|                      result_buf[i] = source_buf[cur];
 | |
|             }
 | |
|             return result;
 | |
|         }
 | |
|     }
 | |
|     else {
 | |
|         PyErr_Format(PyExc_TypeError,
 | |
|                      "bytearray indices must be integers or slices, not %.200s",
 | |
|                      Py_TYPE(index)->tp_name);
 | |
|         return NULL;
 | |
|     }
 | |
| }
 | |
| 
 | |
| static int
 | |
| bytearray_setslice_linear(PyByteArrayObject *self,
 | |
|                           Py_ssize_t lo, Py_ssize_t hi,
 | |
|                           char *bytes, Py_ssize_t bytes_len)
 | |
| {
 | |
|     Py_ssize_t avail = hi - lo;
 | |
|     char *buf = PyByteArray_AS_STRING(self);
 | |
|     Py_ssize_t growth = bytes_len - avail;
 | |
|     int res = 0;
 | |
|     assert(avail >= 0);
 | |
| 
 | |
|     if (growth < 0) {
 | |
|         if (!_canresize(self))
 | |
|             return -1;
 | |
| 
 | |
|         if (lo == 0) {
 | |
|             /* Shrink the buffer by advancing its logical start */
 | |
|             self->ob_start -= growth;
 | |
|             /*
 | |
|               0   lo               hi             old_size
 | |
|               |   |<----avail----->|<-----tail------>|
 | |
|               |      |<-bytes_len->|<-----tail------>|
 | |
|               0    new_lo         new_hi          new_size
 | |
|             */
 | |
|         }
 | |
|         else {
 | |
|             /*
 | |
|               0   lo               hi               old_size
 | |
|               |   |<----avail----->|<-----tomove------>|
 | |
|               |   |<-bytes_len->|<-----tomove------>|
 | |
|               0   lo         new_hi              new_size
 | |
|             */
 | |
|             memmove(buf + lo + bytes_len, buf + hi,
 | |
|                     Py_SIZE(self) - hi);
 | |
|         }
 | |
|         if (PyByteArray_Resize((PyObject *)self,
 | |
|                                Py_SIZE(self) + growth) < 0) {
 | |
|             /* Issue #19578: Handling the memory allocation failure here is
 | |
|                tricky here because the bytearray object has already been
 | |
|                modified. Depending on growth and lo, the behaviour is
 | |
|                different.
 | |
| 
 | |
|                If growth < 0 and lo != 0, the operation is completed, but a
 | |
|                MemoryError is still raised and the memory block is not
 | |
|                shrunk. Otherwise, the bytearray is restored in its previous
 | |
|                state and a MemoryError is raised. */
 | |
|             if (lo == 0) {
 | |
|                 self->ob_start += growth;
 | |
|                 return -1;
 | |
|             }
 | |
|             /* memmove() removed bytes, the bytearray object cannot be
 | |
|                restored in its previous state. */
 | |
|             Py_SET_SIZE(self, Py_SIZE(self) + growth);
 | |
|             res = -1;
 | |
|         }
 | |
|         buf = PyByteArray_AS_STRING(self);
 | |
|     }
 | |
|     else if (growth > 0) {
 | |
|         if (Py_SIZE(self) > (Py_ssize_t)PY_SSIZE_T_MAX - growth) {
 | |
|             PyErr_NoMemory();
 | |
|             return -1;
 | |
|         }
 | |
| 
 | |
|         if (PyByteArray_Resize((PyObject *)self,
 | |
|                                Py_SIZE(self) + growth) < 0) {
 | |
|             return -1;
 | |
|         }
 | |
|         buf = PyByteArray_AS_STRING(self);
 | |
|         /* Make the place for the additional bytes */
 | |
|         /*
 | |
|           0   lo        hi               old_size
 | |
|           |   |<-avail->|<-----tomove------>|
 | |
|           |   |<---bytes_len-->|<-----tomove------>|
 | |
|           0   lo            new_hi              new_size
 | |
|          */
 | |
|         memmove(buf + lo + bytes_len, buf + hi,
 | |
|                 Py_SIZE(self) - lo - bytes_len);
 | |
|     }
 | |
| 
 | |
|     if (bytes_len > 0)
 | |
|         memcpy(buf + lo, bytes, bytes_len);
 | |
|     return res;
 | |
| }
 | |
| 
 | |
| static int
 | |
| bytearray_setslice(PyByteArrayObject *self, Py_ssize_t lo, Py_ssize_t hi,
 | |
|                PyObject *values)
 | |
| {
 | |
|     Py_ssize_t needed;
 | |
|     void *bytes;
 | |
|     Py_buffer vbytes;
 | |
|     int res = 0;
 | |
| 
 | |
|     vbytes.len = -1;
 | |
|     if (values == (PyObject *)self) {
 | |
|         /* Make a copy and call this function recursively */
 | |
|         int err;
 | |
|         values = PyByteArray_FromStringAndSize(PyByteArray_AS_STRING(values),
 | |
|                                                PyByteArray_GET_SIZE(values));
 | |
|         if (values == NULL)
 | |
|             return -1;
 | |
|         err = bytearray_setslice(self, lo, hi, values);
 | |
|         Py_DECREF(values);
 | |
|         return err;
 | |
|     }
 | |
|     if (values == NULL) {
 | |
|         /* del b[lo:hi] */
 | |
|         bytes = NULL;
 | |
|         needed = 0;
 | |
|     }
 | |
|     else {
 | |
|         if (PyObject_GetBuffer(values, &vbytes, PyBUF_SIMPLE) != 0) {
 | |
|             PyErr_Format(PyExc_TypeError,
 | |
|                          "can't set bytearray slice from %.100s",
 | |
|                          Py_TYPE(values)->tp_name);
 | |
|             return -1;
 | |
|         }
 | |
|         needed = vbytes.len;
 | |
|         bytes = vbytes.buf;
 | |
|     }
 | |
| 
 | |
|     if (lo < 0)
 | |
|         lo = 0;
 | |
|     if (hi < lo)
 | |
|         hi = lo;
 | |
|     if (hi > Py_SIZE(self))
 | |
|         hi = Py_SIZE(self);
 | |
| 
 | |
|     res = bytearray_setslice_linear(self, lo, hi, bytes, needed);
 | |
|     if (vbytes.len != -1)
 | |
|         PyBuffer_Release(&vbytes);
 | |
|     return res;
 | |
| }
 | |
| 
 | |
| static int
 | |
| bytearray_setitem(PyByteArrayObject *self, Py_ssize_t i, PyObject *value)
 | |
| {
 | |
|     int ival = -1;
 | |
| 
 | |
|     // GH-91153: We need to do this *before* the size check, in case value has a
 | |
|     // nasty __index__ method that changes the size of the bytearray:
 | |
|     if (value && !_getbytevalue(value, &ival)) {
 | |
|         return -1;
 | |
|     }
 | |
| 
 | |
|     if (i < 0) {
 | |
|         i += Py_SIZE(self);
 | |
|     }
 | |
| 
 | |
|     if (i < 0 || i >= Py_SIZE(self)) {
 | |
|         PyErr_SetString(PyExc_IndexError, "bytearray index out of range");
 | |
|         return -1;
 | |
|     }
 | |
| 
 | |
|     if (value == NULL) {
 | |
|         return bytearray_setslice(self, i, i+1, NULL);
 | |
|     }
 | |
| 
 | |
|     assert(0 <= ival && ival < 256);
 | |
|     PyByteArray_AS_STRING(self)[i] = ival;
 | |
|     return 0;
 | |
| }
 | |
| 
 | |
| static int
 | |
| bytearray_ass_subscript(PyByteArrayObject *self, PyObject *index, PyObject *values)
 | |
| {
 | |
|     Py_ssize_t start, stop, step, slicelen, needed;
 | |
|     char *buf, *bytes;
 | |
|     buf = PyByteArray_AS_STRING(self);
 | |
| 
 | |
|     if (_PyIndex_Check(index)) {
 | |
|         Py_ssize_t i = PyNumber_AsSsize_t(index, PyExc_IndexError);
 | |
| 
 | |
|         if (i == -1 && PyErr_Occurred()) {
 | |
|             return -1;
 | |
|         }
 | |
| 
 | |
|         int ival = -1;
 | |
| 
 | |
|         // GH-91153: We need to do this *before* the size check, in case values
 | |
|         // has a nasty __index__ method that changes the size of the bytearray:
 | |
|         if (values && !_getbytevalue(values, &ival)) {
 | |
|             return -1;
 | |
|         }
 | |
| 
 | |
|         if (i < 0) {
 | |
|             i += PyByteArray_GET_SIZE(self);
 | |
|         }
 | |
| 
 | |
|         if (i < 0 || i >= Py_SIZE(self)) {
 | |
|             PyErr_SetString(PyExc_IndexError, "bytearray index out of range");
 | |
|             return -1;
 | |
|         }
 | |
| 
 | |
|         if (values == NULL) {
 | |
|             /* Fall through to slice assignment */
 | |
|             start = i;
 | |
|             stop = i + 1;
 | |
|             step = 1;
 | |
|             slicelen = 1;
 | |
|         }
 | |
|         else {
 | |
|             assert(0 <= ival && ival < 256);
 | |
|             buf[i] = (char)ival;
 | |
|             return 0;
 | |
|         }
 | |
|     }
 | |
|     else if (PySlice_Check(index)) {
 | |
|         if (PySlice_Unpack(index, &start, &stop, &step) < 0) {
 | |
|             return -1;
 | |
|         }
 | |
|         slicelen = PySlice_AdjustIndices(PyByteArray_GET_SIZE(self), &start,
 | |
|                                          &stop, step);
 | |
|     }
 | |
|     else {
 | |
|         PyErr_Format(PyExc_TypeError,
 | |
|                      "bytearray indices must be integers or slices, not %.200s",
 | |
|                       Py_TYPE(index)->tp_name);
 | |
|         return -1;
 | |
|     }
 | |
| 
 | |
|     if (values == NULL) {
 | |
|         bytes = NULL;
 | |
|         needed = 0;
 | |
|     }
 | |
|     else if (values == (PyObject *)self || !PyByteArray_Check(values)) {
 | |
|         int err;
 | |
|         if (PyNumber_Check(values) || PyUnicode_Check(values)) {
 | |
|             PyErr_SetString(PyExc_TypeError,
 | |
|                             "can assign only bytes, buffers, or iterables "
 | |
|                             "of ints in range(0, 256)");
 | |
|             return -1;
 | |
|         }
 | |
|         /* Make a copy and call this function recursively */
 | |
|         values = PyByteArray_FromObject(values);
 | |
|         if (values == NULL)
 | |
|             return -1;
 | |
|         err = bytearray_ass_subscript(self, index, values);
 | |
|         Py_DECREF(values);
 | |
|         return err;
 | |
|     }
 | |
|     else {
 | |
|         assert(PyByteArray_Check(values));
 | |
|         bytes = PyByteArray_AS_STRING(values);
 | |
|         needed = Py_SIZE(values);
 | |
|     }
 | |
|     /* Make sure b[5:2] = ... inserts before 5, not before 2. */
 | |
|     if ((step < 0 && start < stop) ||
 | |
|         (step > 0 && start > stop))
 | |
|         stop = start;
 | |
|     if (step == 1) {
 | |
|         return bytearray_setslice_linear(self, start, stop, bytes, needed);
 | |
|     }
 | |
|     else {
 | |
|         if (needed == 0) {
 | |
|             /* Delete slice */
 | |
|             size_t cur;
 | |
|             Py_ssize_t i;
 | |
| 
 | |
|             if (!_canresize(self))
 | |
|                 return -1;
 | |
| 
 | |
|             if (slicelen == 0)
 | |
|                 /* Nothing to do here. */
 | |
|                 return 0;
 | |
| 
 | |
|             if (step < 0) {
 | |
|                 stop = start + 1;
 | |
|                 start = stop + step * (slicelen - 1) - 1;
 | |
|                 step = -step;
 | |
|             }
 | |
|             for (cur = start, i = 0;
 | |
|                  i < slicelen; cur += step, i++) {
 | |
|                 Py_ssize_t lim = step - 1;
 | |
| 
 | |
|                 if (cur + step >= (size_t)PyByteArray_GET_SIZE(self))
 | |
|                     lim = PyByteArray_GET_SIZE(self) - cur - 1;
 | |
| 
 | |
|                 memmove(buf + cur - i,
 | |
|                         buf + cur + 1, lim);
 | |
|             }
 | |
|             /* Move the tail of the bytes, in one chunk */
 | |
|             cur = start + (size_t)slicelen*step;
 | |
|             if (cur < (size_t)PyByteArray_GET_SIZE(self)) {
 | |
|                 memmove(buf + cur - slicelen,
 | |
|                         buf + cur,
 | |
|                         PyByteArray_GET_SIZE(self) - cur);
 | |
|             }
 | |
|             if (PyByteArray_Resize((PyObject *)self,
 | |
|                                PyByteArray_GET_SIZE(self) - slicelen) < 0)
 | |
|                 return -1;
 | |
| 
 | |
|             return 0;
 | |
|         }
 | |
|         else {
 | |
|             /* Assign slice */
 | |
|             Py_ssize_t i;
 | |
|             size_t cur;
 | |
| 
 | |
|             if (needed != slicelen) {
 | |
|                 PyErr_Format(PyExc_ValueError,
 | |
|                              "attempt to assign bytes of size %zd "
 | |
|                              "to extended slice of size %zd",
 | |
|                              needed, slicelen);
 | |
|                 return -1;
 | |
|             }
 | |
|             for (cur = start, i = 0; i < slicelen; cur += step, i++)
 | |
|                 buf[cur] = bytes[i];
 | |
|             return 0;
 | |
|         }
 | |
|     }
 | |
| }
 | |
| 
 | |
| /*[clinic input]
 | |
| bytearray.__init__
 | |
| 
 | |
|     source as arg: object = NULL
 | |
|     encoding: str = NULL
 | |
|     errors: str = NULL
 | |
| 
 | |
| [clinic start generated code]*/
 | |
| 
 | |
| static int
 | |
| bytearray___init___impl(PyByteArrayObject *self, PyObject *arg,
 | |
|                         const char *encoding, const char *errors)
 | |
| /*[clinic end generated code: output=4ce1304649c2f8b3 input=1141a7122eefd7b9]*/
 | |
| {
 | |
|     Py_ssize_t count;
 | |
|     PyObject *it;
 | |
|     PyObject *(*iternext)(PyObject *);
 | |
| 
 | |
|     if (Py_SIZE(self) != 0) {
 | |
|         /* Empty previous contents (yes, do this first of all!) */
 | |
|         if (PyByteArray_Resize((PyObject *)self, 0) < 0)
 | |
|             return -1;
 | |
|     }
 | |
| 
 | |
|     /* Make a quick exit if no first argument */
 | |
|     if (arg == NULL) {
 | |
|         if (encoding != NULL || errors != NULL) {
 | |
|             PyErr_SetString(PyExc_TypeError,
 | |
|                             encoding != NULL ?
 | |
|                             "encoding without a string argument" :
 | |
|                             "errors without a string argument");
 | |
|             return -1;
 | |
|         }
 | |
|         return 0;
 | |
|     }
 | |
| 
 | |
|     if (PyUnicode_Check(arg)) {
 | |
|         /* Encode via the codec registry */
 | |
|         PyObject *encoded, *new;
 | |
|         if (encoding == NULL) {
 | |
|             PyErr_SetString(PyExc_TypeError,
 | |
|                             "string argument without an encoding");
 | |
|             return -1;
 | |
|         }
 | |
|         encoded = PyUnicode_AsEncodedString(arg, encoding, errors);
 | |
|         if (encoded == NULL)
 | |
|             return -1;
 | |
|         assert(PyBytes_Check(encoded));
 | |
|         new = bytearray_iconcat(self, encoded);
 | |
|         Py_DECREF(encoded);
 | |
|         if (new == NULL)
 | |
|             return -1;
 | |
|         Py_DECREF(new);
 | |
|         return 0;
 | |
|     }
 | |
| 
 | |
|     /* If it's not unicode, there can't be encoding or errors */
 | |
|     if (encoding != NULL || errors != NULL) {
 | |
|         PyErr_SetString(PyExc_TypeError,
 | |
|                         encoding != NULL ?
 | |
|                         "encoding without a string argument" :
 | |
|                         "errors without a string argument");
 | |
|         return -1;
 | |
|     }
 | |
| 
 | |
|     /* Is it an int? */
 | |
|     if (_PyIndex_Check(arg)) {
 | |
|         count = PyNumber_AsSsize_t(arg, PyExc_OverflowError);
 | |
|         if (count == -1 && PyErr_Occurred()) {
 | |
|             if (!PyErr_ExceptionMatches(PyExc_TypeError))
 | |
|                 return -1;
 | |
|             PyErr_Clear();  /* fall through */
 | |
|         }
 | |
|         else {
 | |
|             if (count < 0) {
 | |
|                 PyErr_SetString(PyExc_ValueError, "negative count");
 | |
|                 return -1;
 | |
|             }
 | |
|             if (count > 0) {
 | |
|                 if (PyByteArray_Resize((PyObject *)self, count))
 | |
|                     return -1;
 | |
|                 memset(PyByteArray_AS_STRING(self), 0, count);
 | |
|             }
 | |
|             return 0;
 | |
|         }
 | |
|     }
 | |
| 
 | |
|     /* Use the buffer API */
 | |
|     if (PyObject_CheckBuffer(arg)) {
 | |
|         Py_ssize_t size;
 | |
|         Py_buffer view;
 | |
|         if (PyObject_GetBuffer(arg, &view, PyBUF_FULL_RO) < 0)
 | |
|             return -1;
 | |
|         size = view.len;
 | |
|         if (PyByteArray_Resize((PyObject *)self, size) < 0) goto fail;
 | |
|         if (PyBuffer_ToContiguous(PyByteArray_AS_STRING(self),
 | |
|             &view, size, 'C') < 0)
 | |
|             goto fail;
 | |
|         PyBuffer_Release(&view);
 | |
|         return 0;
 | |
|     fail:
 | |
|         PyBuffer_Release(&view);
 | |
|         return -1;
 | |
|     }
 | |
| 
 | |
|     if (PyList_CheckExact(arg) || PyTuple_CheckExact(arg)) {
 | |
|         Py_ssize_t size = PySequence_Fast_GET_SIZE(arg);
 | |
|         if (PyByteArray_Resize((PyObject *)self, size) < 0) {
 | |
|             return -1;
 | |
|         }
 | |
|         PyObject **items = PySequence_Fast_ITEMS(arg);
 | |
|         char *s = PyByteArray_AS_STRING(self);
 | |
|         for (Py_ssize_t i = 0; i < size; i++) {
 | |
|             int value;
 | |
|             if (!PyLong_CheckExact(items[i])) {
 | |
|                 /* Resize to 0 and go through slowpath */
 | |
|                 if (Py_SIZE(self) != 0) {
 | |
|                    if (PyByteArray_Resize((PyObject *)self, 0) < 0) {
 | |
|                        return -1;
 | |
|                    }
 | |
|                 }
 | |
|                 goto slowpath;
 | |
|             }
 | |
|             int rc = _getbytevalue(items[i], &value);
 | |
|             if (!rc) {
 | |
|                 return -1;
 | |
|             }
 | |
|             s[i] = value;
 | |
|         }
 | |
|         return 0;
 | |
|     }
 | |
| slowpath:
 | |
|     /* Get the iterator */
 | |
|     it = PyObject_GetIter(arg);
 | |
|     if (it == NULL) {
 | |
|         if (PyErr_ExceptionMatches(PyExc_TypeError)) {
 | |
|             PyErr_Format(PyExc_TypeError,
 | |
|                          "cannot convert '%.200s' object to bytearray",
 | |
|                          Py_TYPE(arg)->tp_name);
 | |
|         }
 | |
|         return -1;
 | |
|     }
 | |
|     iternext = *Py_TYPE(it)->tp_iternext;
 | |
| 
 | |
|     /* Run the iterator to exhaustion */
 | |
|     for (;;) {
 | |
|         PyObject *item;
 | |
|         int rc, value;
 | |
| 
 | |
|         /* Get the next item */
 | |
|         item = iternext(it);
 | |
|         if (item == NULL) {
 | |
|             if (PyErr_Occurred()) {
 | |
|                 if (!PyErr_ExceptionMatches(PyExc_StopIteration))
 | |
|                     goto error;
 | |
|                 PyErr_Clear();
 | |
|             }
 | |
|             break;
 | |
|         }
 | |
| 
 | |
|         /* Interpret it as an int (__index__) */
 | |
|         rc = _getbytevalue(item, &value);
 | |
|         Py_DECREF(item);
 | |
|         if (!rc)
 | |
|             goto error;
 | |
| 
 | |
|         /* Append the byte */
 | |
|         if (Py_SIZE(self) + 1 < self->ob_alloc) {
 | |
|             Py_SET_SIZE(self, Py_SIZE(self) + 1);
 | |
|             PyByteArray_AS_STRING(self)[Py_SIZE(self)] = '\0';
 | |
|         }
 | |
|         else if (PyByteArray_Resize((PyObject *)self, Py_SIZE(self)+1) < 0)
 | |
|             goto error;
 | |
|         PyByteArray_AS_STRING(self)[Py_SIZE(self)-1] = value;
 | |
|     }
 | |
| 
 | |
|     /* Clean up and return success */
 | |
|     Py_DECREF(it);
 | |
|     return 0;
 | |
| 
 | |
|  error:
 | |
|     /* Error handling when it != NULL */
 | |
|     Py_DECREF(it);
 | |
|     return -1;
 | |
| }
 | |
| 
 | |
| /* Mostly copied from string_repr, but without the
 | |
|    "smart quote" functionality. */
 | |
| static PyObject *
 | |
| bytearray_repr(PyByteArrayObject *self)
 | |
| {
 | |
|     const char *className = _PyType_Name(Py_TYPE(self));
 | |
|     const char *quote_prefix = "(b";
 | |
|     const char *quote_postfix = ")";
 | |
|     Py_ssize_t length = Py_SIZE(self);
 | |
|     /* 6 == strlen(quote_prefix) + 2 + strlen(quote_postfix) + 1 */
 | |
|     Py_ssize_t newsize;
 | |
|     PyObject *v;
 | |
|     Py_ssize_t i;
 | |
|     char *bytes;
 | |
|     char c;
 | |
|     char *p;
 | |
|     int quote;
 | |
|     char *test, *start;
 | |
|     char *buffer;
 | |
| 
 | |
|     newsize = strlen(className);
 | |
|     if (length > (PY_SSIZE_T_MAX - 6 - newsize) / 4) {
 | |
|         PyErr_SetString(PyExc_OverflowError,
 | |
|             "bytearray object is too large to make repr");
 | |
|         return NULL;
 | |
|     }
 | |
| 
 | |
|     newsize += 6 + length * 4;
 | |
|     buffer = PyMem_Malloc(newsize);
 | |
|     if (buffer == NULL) {
 | |
|         PyErr_NoMemory();
 | |
|         return NULL;
 | |
|     }
 | |
| 
 | |
|     /* Figure out which quote to use; single is preferred */
 | |
|     quote = '\'';
 | |
|     start = PyByteArray_AS_STRING(self);
 | |
|     for (test = start; test < start+length; ++test) {
 | |
|         if (*test == '"') {
 | |
|             quote = '\''; /* back to single */
 | |
|             break;
 | |
|         }
 | |
|         else if (*test == '\'')
 | |
|             quote = '"';
 | |
|     }
 | |
| 
 | |
|     p = buffer;
 | |
|     while (*className)
 | |
|         *p++ = *className++;
 | |
|     while (*quote_prefix)
 | |
|         *p++ = *quote_prefix++;
 | |
|     *p++ = quote;
 | |
| 
 | |
|     bytes = PyByteArray_AS_STRING(self);
 | |
|     for (i = 0; i < length; i++) {
 | |
|         /* There's at least enough room for a hex escape
 | |
|            and a closing quote. */
 | |
|         assert(newsize - (p - buffer) >= 5);
 | |
|         c = bytes[i];
 | |
|         if (c == '\'' || c == '\\')
 | |
|             *p++ = '\\', *p++ = c;
 | |
|         else if (c == '\t')
 | |
|             *p++ = '\\', *p++ = 't';
 | |
|         else if (c == '\n')
 | |
|             *p++ = '\\', *p++ = 'n';
 | |
|         else if (c == '\r')
 | |
|             *p++ = '\\', *p++ = 'r';
 | |
|         else if (c == 0)
 | |
|             *p++ = '\\', *p++ = 'x', *p++ = '0', *p++ = '0';
 | |
|         else if (c < ' ' || c >= 0x7f) {
 | |
|             *p++ = '\\';
 | |
|             *p++ = 'x';
 | |
|             *p++ = Py_hexdigits[(c & 0xf0) >> 4];
 | |
|             *p++ = Py_hexdigits[c & 0xf];
 | |
|         }
 | |
|         else
 | |
|             *p++ = c;
 | |
|     }
 | |
|     assert(newsize - (p - buffer) >= 1);
 | |
|     *p++ = quote;
 | |
|     while (*quote_postfix) {
 | |
|        *p++ = *quote_postfix++;
 | |
|     }
 | |
| 
 | |
|     v = PyUnicode_FromStringAndSize(buffer, p - buffer);
 | |
|     PyMem_Free(buffer);
 | |
|     return v;
 | |
| }
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_str(PyObject *op)
 | |
| {
 | |
|     if (_Py_GetConfig()->bytes_warning) {
 | |
|         if (PyErr_WarnEx(PyExc_BytesWarning,
 | |
|                          "str() on a bytearray instance", 1)) {
 | |
|                 return NULL;
 | |
|         }
 | |
|     }
 | |
|     return bytearray_repr((PyByteArrayObject*)op);
 | |
| }
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_richcompare(PyObject *self, PyObject *other, int op)
 | |
| {
 | |
|     Py_ssize_t self_size, other_size;
 | |
|     Py_buffer self_bytes, other_bytes;
 | |
|     int cmp;
 | |
| 
 | |
|     if (!PyObject_CheckBuffer(self) || !PyObject_CheckBuffer(other)) {
 | |
|         if (PyUnicode_Check(self) || PyUnicode_Check(other)) {
 | |
|             if (_Py_GetConfig()->bytes_warning && (op == Py_EQ || op == Py_NE)) {
 | |
|                 if (PyErr_WarnEx(PyExc_BytesWarning,
 | |
|                                 "Comparison between bytearray and string", 1))
 | |
|                     return NULL;
 | |
|             }
 | |
|         }
 | |
|         Py_RETURN_NOTIMPLEMENTED;
 | |
|     }
 | |
| 
 | |
|     /* Bytearrays can be compared to anything that supports the buffer API. */
 | |
|     if (PyObject_GetBuffer(self, &self_bytes, PyBUF_SIMPLE) != 0) {
 | |
|         PyErr_Clear();
 | |
|         Py_RETURN_NOTIMPLEMENTED;
 | |
|     }
 | |
|     self_size = self_bytes.len;
 | |
| 
 | |
|     if (PyObject_GetBuffer(other, &other_bytes, PyBUF_SIMPLE) != 0) {
 | |
|         PyErr_Clear();
 | |
|         PyBuffer_Release(&self_bytes);
 | |
|         Py_RETURN_NOTIMPLEMENTED;
 | |
|     }
 | |
|     other_size = other_bytes.len;
 | |
| 
 | |
|     if (self_size != other_size && (op == Py_EQ || op == Py_NE)) {
 | |
|         /* Shortcut: if the lengths differ, the objects differ */
 | |
|         PyBuffer_Release(&self_bytes);
 | |
|         PyBuffer_Release(&other_bytes);
 | |
|         return PyBool_FromLong((op == Py_NE));
 | |
|     }
 | |
|     else {
 | |
|         cmp = memcmp(self_bytes.buf, other_bytes.buf,
 | |
|                      Py_MIN(self_size, other_size));
 | |
|         /* In ISO C, memcmp() guarantees to use unsigned bytes! */
 | |
| 
 | |
|         PyBuffer_Release(&self_bytes);
 | |
|         PyBuffer_Release(&other_bytes);
 | |
| 
 | |
|         if (cmp != 0) {
 | |
|             Py_RETURN_RICHCOMPARE(cmp, 0, op);
 | |
|         }
 | |
| 
 | |
|         Py_RETURN_RICHCOMPARE(self_size, other_size, op);
 | |
|     }
 | |
| 
 | |
| }
 | |
| 
 | |
| static void
 | |
| bytearray_dealloc(PyByteArrayObject *self)
 | |
| {
 | |
|     if (self->ob_exports > 0) {
 | |
|         PyErr_SetString(PyExc_SystemError,
 | |
|                         "deallocated bytearray object has exported buffers");
 | |
|         PyErr_Print();
 | |
|     }
 | |
|     if (self->ob_bytes != 0) {
 | |
|         PyMem_Free(self->ob_bytes);
 | |
|     }
 | |
|     Py_TYPE(self)->tp_free((PyObject *)self);
 | |
| }
 | |
| 
 | |
| 
 | |
| /* -------------------------------------------------------------------- */
 | |
| /* Methods */
 | |
| 
 | |
| #define STRINGLIB_IS_UNICODE 0
 | |
| #define FASTSEARCH fastsearch
 | |
| #define STRINGLIB(F) stringlib_##F
 | |
| #define STRINGLIB_CHAR char
 | |
| #define STRINGLIB_SIZEOF_CHAR 1
 | |
| #define STRINGLIB_LEN PyByteArray_GET_SIZE
 | |
| #define STRINGLIB_STR PyByteArray_AS_STRING
 | |
| #define STRINGLIB_NEW PyByteArray_FromStringAndSize
 | |
| #define STRINGLIB_ISSPACE Py_ISSPACE
 | |
| #define STRINGLIB_ISLINEBREAK(x) ((x == '\n') || (x == '\r'))
 | |
| #define STRINGLIB_CHECK_EXACT PyByteArray_CheckExact
 | |
| #define STRINGLIB_FAST_MEMCHR memchr
 | |
| #define STRINGLIB_MUTABLE 1
 | |
| 
 | |
| #include "stringlib/fastsearch.h"
 | |
| #include "stringlib/count.h"
 | |
| #include "stringlib/find.h"
 | |
| #include "stringlib/join.h"
 | |
| #include "stringlib/partition.h"
 | |
| #include "stringlib/split.h"
 | |
| #include "stringlib/ctype.h"
 | |
| #include "stringlib/transmogrify.h"
 | |
| 
 | |
| 
 | |
| /*[clinic input]
 | |
| @text_signature "($self, sub[, start[, end]], /)"
 | |
| bytearray.find
 | |
| 
 | |
|     sub: object
 | |
|     start: slice_index(accept={int, NoneType}, c_default='0') = None
 | |
|          Optional start position. Default: start of the bytes.
 | |
|     end: slice_index(accept={int, NoneType}, c_default='PY_SSIZE_T_MAX') = None
 | |
|          Optional stop position. Default: end of the bytes.
 | |
|     /
 | |
| 
 | |
| Return the lowest index in B where subsection 'sub' is found, such that 'sub' is contained within B[start:end].
 | |
| 
 | |
| Return -1 on failure.
 | |
| [clinic start generated code]*/
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_find_impl(PyByteArrayObject *self, PyObject *sub, Py_ssize_t start,
 | |
|                     Py_ssize_t end)
 | |
| /*[clinic end generated code: output=413e1cab2ae87da0 input=793dfad803e2952f]*/
 | |
| {
 | |
|     return _Py_bytes_find(PyByteArray_AS_STRING(self), PyByteArray_GET_SIZE(self),
 | |
|                           sub, start, end);
 | |
| }
 | |
| 
 | |
| /*[clinic input]
 | |
| bytearray.count = bytearray.find
 | |
| 
 | |
| Return the number of non-overlapping occurrences of subsection 'sub' in bytes B[start:end].
 | |
| [clinic start generated code]*/
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_count_impl(PyByteArrayObject *self, PyObject *sub,
 | |
|                      Py_ssize_t start, Py_ssize_t end)
 | |
| /*[clinic end generated code: output=a21ee2692e4f1233 input=4deb529db38deda8]*/
 | |
| {
 | |
|     return _Py_bytes_count(PyByteArray_AS_STRING(self), PyByteArray_GET_SIZE(self),
 | |
|                            sub, start, end);
 | |
| }
 | |
| 
 | |
| /*[clinic input]
 | |
| bytearray.clear
 | |
| 
 | |
| Remove all items from the bytearray.
 | |
| [clinic start generated code]*/
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_clear_impl(PyByteArrayObject *self)
 | |
| /*[clinic end generated code: output=85c2fe6aede0956c input=ed6edae9de447ac4]*/
 | |
| {
 | |
|     if (PyByteArray_Resize((PyObject *)self, 0) < 0)
 | |
|         return NULL;
 | |
|     Py_RETURN_NONE;
 | |
| }
 | |
| 
 | |
| /*[clinic input]
 | |
| bytearray.copy
 | |
| 
 | |
| Return a copy of B.
 | |
| [clinic start generated code]*/
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_copy_impl(PyByteArrayObject *self)
 | |
| /*[clinic end generated code: output=68cfbcfed484c132 input=6597b0c01bccaa9e]*/
 | |
| {
 | |
|     return PyByteArray_FromStringAndSize(PyByteArray_AS_STRING((PyObject *)self),
 | |
|                                          PyByteArray_GET_SIZE(self));
 | |
| }
 | |
| 
 | |
| /*[clinic input]
 | |
| bytearray.index = bytearray.find
 | |
| 
 | |
| Return the lowest index in B where subsection 'sub' is found, such that 'sub' is contained within B[start:end].
 | |
| 
 | |
| Raise ValueError if the subsection is not found.
 | |
| [clinic start generated code]*/
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_index_impl(PyByteArrayObject *self, PyObject *sub,
 | |
|                      Py_ssize_t start, Py_ssize_t end)
 | |
| /*[clinic end generated code: output=067a1e78efc672a7 input=8cbaf6836dbd2a9a]*/
 | |
| {
 | |
|     return _Py_bytes_index(PyByteArray_AS_STRING(self), PyByteArray_GET_SIZE(self),
 | |
|                            sub, start, end);
 | |
| }
 | |
| 
 | |
| /*[clinic input]
 | |
| bytearray.rfind = bytearray.find
 | |
| 
 | |
| Return the highest index in B where subsection 'sub' is found, such that 'sub' is contained within B[start:end].
 | |
| 
 | |
| Return -1 on failure.
 | |
| [clinic start generated code]*/
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_rfind_impl(PyByteArrayObject *self, PyObject *sub,
 | |
|                      Py_ssize_t start, Py_ssize_t end)
 | |
| /*[clinic end generated code: output=51bf886f932b283c input=eaa107468a158423]*/
 | |
| {
 | |
|     return _Py_bytes_rfind(PyByteArray_AS_STRING(self), PyByteArray_GET_SIZE(self),
 | |
|                            sub, start, end);
 | |
| }
 | |
| 
 | |
| /*[clinic input]
 | |
| bytearray.rindex = bytearray.find
 | |
| 
 | |
| Return the highest index in B where subsection 'sub' is found, such that 'sub' is contained within B[start:end].
 | |
| 
 | |
| Raise ValueError if the subsection is not found.
 | |
| [clinic start generated code]*/
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_rindex_impl(PyByteArrayObject *self, PyObject *sub,
 | |
|                       Py_ssize_t start, Py_ssize_t end)
 | |
| /*[clinic end generated code: output=38e1cf66bafb08b9 input=81cf49d0af4d5bd0]*/
 | |
| {
 | |
|     return _Py_bytes_rindex(PyByteArray_AS_STRING(self), PyByteArray_GET_SIZE(self),
 | |
|                             sub, start, end);
 | |
| }
 | |
| 
 | |
| static int
 | |
| bytearray_contains(PyObject *self, PyObject *arg)
 | |
| {
 | |
|     return _Py_bytes_contains(PyByteArray_AS_STRING(self), PyByteArray_GET_SIZE(self), arg);
 | |
| }
 | |
| 
 | |
| /*[clinic input]
 | |
| @text_signature "($self, prefix[, start[, end]], /)"
 | |
| bytearray.startswith
 | |
| 
 | |
|     prefix as subobj: object
 | |
|         A bytes or a tuple of bytes to try.
 | |
|     start: slice_index(accept={int, NoneType}, c_default='0') = None
 | |
|         Optional start position. Default: start of the bytearray.
 | |
|     end: slice_index(accept={int, NoneType}, c_default='PY_SSIZE_T_MAX') = None
 | |
|         Optional stop position. Default: end of the bytearray.
 | |
|     /
 | |
| 
 | |
| Return True if the bytearray starts with the specified prefix, False otherwise.
 | |
| [clinic start generated code]*/
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_startswith_impl(PyByteArrayObject *self, PyObject *subobj,
 | |
|                           Py_ssize_t start, Py_ssize_t end)
 | |
| /*[clinic end generated code: output=a3d9b6d44d3662a6 input=76385e0b376b45c1]*/
 | |
| {
 | |
|     return _Py_bytes_startswith(PyByteArray_AS_STRING(self), PyByteArray_GET_SIZE(self),
 | |
|                                 subobj, start, end);
 | |
| }
 | |
| 
 | |
| /*[clinic input]
 | |
| @text_signature "($self, suffix[, start[, end]], /)"
 | |
| bytearray.endswith
 | |
| 
 | |
|     suffix as subobj: object
 | |
|         A bytes or a tuple of bytes to try.
 | |
|     start: slice_index(accept={int, NoneType}, c_default='0') = None
 | |
|          Optional start position. Default: start of the bytearray.
 | |
|     end: slice_index(accept={int, NoneType}, c_default='PY_SSIZE_T_MAX') = None
 | |
|          Optional stop position. Default: end of the bytearray.
 | |
|     /
 | |
| 
 | |
| Return True if the bytearray ends with the specified suffix, False otherwise.
 | |
| [clinic start generated code]*/
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_endswith_impl(PyByteArrayObject *self, PyObject *subobj,
 | |
|                         Py_ssize_t start, Py_ssize_t end)
 | |
| /*[clinic end generated code: output=e75ea8c227954caa input=9b8baa879aa3d74b]*/
 | |
| {
 | |
|     return _Py_bytes_endswith(PyByteArray_AS_STRING(self), PyByteArray_GET_SIZE(self),
 | |
|                               subobj, start, end);
 | |
| }
 | |
| 
 | |
| /*[clinic input]
 | |
| bytearray.removeprefix as bytearray_removeprefix
 | |
| 
 | |
|     prefix: Py_buffer
 | |
|     /
 | |
| 
 | |
| Return a bytearray with the given prefix string removed if present.
 | |
| 
 | |
| If the bytearray starts with the prefix string, return
 | |
| bytearray[len(prefix):].  Otherwise, return a copy of the original
 | |
| bytearray.
 | |
| [clinic start generated code]*/
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_removeprefix_impl(PyByteArrayObject *self, Py_buffer *prefix)
 | |
| /*[clinic end generated code: output=6cabc585e7f502e0 input=968aada38aedd262]*/
 | |
| {
 | |
|     const char *self_start = PyByteArray_AS_STRING(self);
 | |
|     Py_ssize_t self_len = PyByteArray_GET_SIZE(self);
 | |
|     const char *prefix_start = prefix->buf;
 | |
|     Py_ssize_t prefix_len = prefix->len;
 | |
| 
 | |
|     if (self_len >= prefix_len
 | |
|         && memcmp(self_start, prefix_start, prefix_len) == 0)
 | |
|     {
 | |
|         return PyByteArray_FromStringAndSize(self_start + prefix_len,
 | |
|                                              self_len - prefix_len);
 | |
|     }
 | |
| 
 | |
|     return PyByteArray_FromStringAndSize(self_start, self_len);
 | |
| }
 | |
| 
 | |
| /*[clinic input]
 | |
| bytearray.removesuffix as bytearray_removesuffix
 | |
| 
 | |
|     suffix: Py_buffer
 | |
|     /
 | |
| 
 | |
| Return a bytearray with the given suffix string removed if present.
 | |
| 
 | |
| If the bytearray ends with the suffix string and that suffix is not
 | |
| empty, return bytearray[:-len(suffix)].  Otherwise, return a copy of
 | |
| the original bytearray.
 | |
| [clinic start generated code]*/
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_removesuffix_impl(PyByteArrayObject *self, Py_buffer *suffix)
 | |
| /*[clinic end generated code: output=2bc8cfb79de793d3 input=c1827e810b2f6b99]*/
 | |
| {
 | |
|     const char *self_start = PyByteArray_AS_STRING(self);
 | |
|     Py_ssize_t self_len = PyByteArray_GET_SIZE(self);
 | |
|     const char *suffix_start = suffix->buf;
 | |
|     Py_ssize_t suffix_len = suffix->len;
 | |
| 
 | |
|     if (self_len >= suffix_len
 | |
|         && memcmp(self_start + self_len - suffix_len,
 | |
|                   suffix_start, suffix_len) == 0)
 | |
|     {
 | |
|         return PyByteArray_FromStringAndSize(self_start,
 | |
|                                              self_len - suffix_len);
 | |
|     }
 | |
| 
 | |
|     return PyByteArray_FromStringAndSize(self_start, self_len);
 | |
| }
 | |
| 
 | |
| 
 | |
| /*[clinic input]
 | |
| bytearray.translate
 | |
| 
 | |
|     table: object
 | |
|         Translation table, which must be a bytes object of length 256.
 | |
|     /
 | |
|     delete as deletechars: object(c_default="NULL") = b''
 | |
| 
 | |
| Return a copy with each character mapped by the given translation table.
 | |
| 
 | |
| All characters occurring in the optional argument delete are removed.
 | |
| The remaining characters are mapped through the given translation table.
 | |
| [clinic start generated code]*/
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_translate_impl(PyByteArrayObject *self, PyObject *table,
 | |
|                          PyObject *deletechars)
 | |
| /*[clinic end generated code: output=b6a8f01c2a74e446 input=cfff956d4d127a9b]*/
 | |
| {
 | |
|     char *input, *output;
 | |
|     const char *table_chars;
 | |
|     Py_ssize_t i, c;
 | |
|     PyObject *input_obj = (PyObject*)self;
 | |
|     const char *output_start;
 | |
|     Py_ssize_t inlen;
 | |
|     PyObject *result = NULL;
 | |
|     int trans_table[256];
 | |
|     Py_buffer vtable, vdel;
 | |
| 
 | |
|     if (table == Py_None) {
 | |
|         table_chars = NULL;
 | |
|         table = NULL;
 | |
|     } else if (PyObject_GetBuffer(table, &vtable, PyBUF_SIMPLE) != 0) {
 | |
|         return NULL;
 | |
|     } else {
 | |
|         if (vtable.len != 256) {
 | |
|             PyErr_SetString(PyExc_ValueError,
 | |
|                             "translation table must be 256 characters long");
 | |
|             PyBuffer_Release(&vtable);
 | |
|             return NULL;
 | |
|         }
 | |
|         table_chars = (const char*)vtable.buf;
 | |
|     }
 | |
| 
 | |
|     if (deletechars != NULL) {
 | |
|         if (PyObject_GetBuffer(deletechars, &vdel, PyBUF_SIMPLE) != 0) {
 | |
|             if (table != NULL)
 | |
|                 PyBuffer_Release(&vtable);
 | |
|             return NULL;
 | |
|         }
 | |
|     }
 | |
|     else {
 | |
|         vdel.buf = NULL;
 | |
|         vdel.len = 0;
 | |
|     }
 | |
| 
 | |
|     inlen = PyByteArray_GET_SIZE(input_obj);
 | |
|     result = PyByteArray_FromStringAndSize((char *)NULL, inlen);
 | |
|     if (result == NULL)
 | |
|         goto done;
 | |
|     output_start = output = PyByteArray_AS_STRING(result);
 | |
|     input = PyByteArray_AS_STRING(input_obj);
 | |
| 
 | |
|     if (vdel.len == 0 && table_chars != NULL) {
 | |
|         /* If no deletions are required, use faster code */
 | |
|         for (i = inlen; --i >= 0; ) {
 | |
|             c = Py_CHARMASK(*input++);
 | |
|             *output++ = table_chars[c];
 | |
|         }
 | |
|         goto done;
 | |
|     }
 | |
| 
 | |
|     if (table_chars == NULL) {
 | |
|         for (i = 0; i < 256; i++)
 | |
|             trans_table[i] = Py_CHARMASK(i);
 | |
|     } else {
 | |
|         for (i = 0; i < 256; i++)
 | |
|             trans_table[i] = Py_CHARMASK(table_chars[i]);
 | |
|     }
 | |
| 
 | |
|     for (i = 0; i < vdel.len; i++)
 | |
|         trans_table[(int) Py_CHARMASK( ((unsigned char*)vdel.buf)[i] )] = -1;
 | |
| 
 | |
|     for (i = inlen; --i >= 0; ) {
 | |
|         c = Py_CHARMASK(*input++);
 | |
|         if (trans_table[c] != -1)
 | |
|             *output++ = (char)trans_table[c];
 | |
|     }
 | |
|     /* Fix the size of the resulting bytearray */
 | |
|     if (inlen > 0)
 | |
|         if (PyByteArray_Resize(result, output - output_start) < 0) {
 | |
|             Py_CLEAR(result);
 | |
|             goto done;
 | |
|         }
 | |
| 
 | |
| done:
 | |
|     if (table != NULL)
 | |
|         PyBuffer_Release(&vtable);
 | |
|     if (deletechars != NULL)
 | |
|         PyBuffer_Release(&vdel);
 | |
|     return result;
 | |
| }
 | |
| 
 | |
| 
 | |
| /*[clinic input]
 | |
| 
 | |
| @staticmethod
 | |
| bytearray.maketrans
 | |
| 
 | |
|     frm: Py_buffer
 | |
|     to: Py_buffer
 | |
|     /
 | |
| 
 | |
| Return a translation table usable for the bytes or bytearray translate method.
 | |
| 
 | |
| The returned table will be one where each byte in frm is mapped to the byte at
 | |
| the same position in to.
 | |
| 
 | |
| The bytes objects frm and to must be of the same length.
 | |
| [clinic start generated code]*/
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_maketrans_impl(Py_buffer *frm, Py_buffer *to)
 | |
| /*[clinic end generated code: output=1df267d99f56b15e input=b10de38c85950a63]*/
 | |
| {
 | |
|     return _Py_bytes_maketrans(frm, to);
 | |
| }
 | |
| 
 | |
| 
 | |
| /*[clinic input]
 | |
| bytearray.replace
 | |
| 
 | |
|     old: Py_buffer
 | |
|     new: Py_buffer
 | |
|     count: Py_ssize_t = -1
 | |
|         Maximum number of occurrences to replace.
 | |
|         -1 (the default value) means replace all occurrences.
 | |
|     /
 | |
| 
 | |
| Return a copy with all occurrences of substring old replaced by new.
 | |
| 
 | |
| If the optional argument count is given, only the first count occurrences are
 | |
| replaced.
 | |
| [clinic start generated code]*/
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_replace_impl(PyByteArrayObject *self, Py_buffer *old,
 | |
|                        Py_buffer *new, Py_ssize_t count)
 | |
| /*[clinic end generated code: output=d39884c4dc59412a input=aa379d988637c7fb]*/
 | |
| {
 | |
|     return stringlib_replace((PyObject *)self,
 | |
|                              (const char *)old->buf, old->len,
 | |
|                              (const char *)new->buf, new->len, count);
 | |
| }
 | |
| 
 | |
| /*[clinic input]
 | |
| bytearray.split
 | |
| 
 | |
|     sep: object = None
 | |
|         The delimiter according which to split the bytearray.
 | |
|         None (the default value) means split on ASCII whitespace characters
 | |
|         (space, tab, return, newline, formfeed, vertical tab).
 | |
|     maxsplit: Py_ssize_t = -1
 | |
|         Maximum number of splits to do.
 | |
|         -1 (the default value) means no limit.
 | |
| 
 | |
| Return a list of the sections in the bytearray, using sep as the delimiter.
 | |
| [clinic start generated code]*/
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_split_impl(PyByteArrayObject *self, PyObject *sep,
 | |
|                      Py_ssize_t maxsplit)
 | |
| /*[clinic end generated code: output=833e2cf385d9a04d input=24f82669f41bf523]*/
 | |
| {
 | |
|     Py_ssize_t len = PyByteArray_GET_SIZE(self), n;
 | |
|     const char *s = PyByteArray_AS_STRING(self), *sub;
 | |
|     PyObject *list;
 | |
|     Py_buffer vsub;
 | |
| 
 | |
|     if (maxsplit < 0)
 | |
|         maxsplit = PY_SSIZE_T_MAX;
 | |
| 
 | |
|     if (sep == Py_None)
 | |
|         return stringlib_split_whitespace((PyObject*) self, s, len, maxsplit);
 | |
| 
 | |
|     if (PyObject_GetBuffer(sep, &vsub, PyBUF_SIMPLE) != 0)
 | |
|         return NULL;
 | |
|     sub = vsub.buf;
 | |
|     n = vsub.len;
 | |
| 
 | |
|     list = stringlib_split(
 | |
|         (PyObject*) self, s, len, sub, n, maxsplit
 | |
|         );
 | |
|     PyBuffer_Release(&vsub);
 | |
|     return list;
 | |
| }
 | |
| 
 | |
| /*[clinic input]
 | |
| bytearray.partition
 | |
| 
 | |
|     sep: object
 | |
|     /
 | |
| 
 | |
| Partition the bytearray into three parts using the given separator.
 | |
| 
 | |
| This will search for the separator sep in the bytearray. If the separator is
 | |
| found, returns a 3-tuple containing the part before the separator, the
 | |
| separator itself, and the part after it as new bytearray objects.
 | |
| 
 | |
| If the separator is not found, returns a 3-tuple containing the copy of the
 | |
| original bytearray object and two empty bytearray objects.
 | |
| [clinic start generated code]*/
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_partition(PyByteArrayObject *self, PyObject *sep)
 | |
| /*[clinic end generated code: output=45d2525ddd35f957 input=8f644749ee4fc83a]*/
 | |
| {
 | |
|     PyObject *bytesep, *result;
 | |
| 
 | |
|     bytesep = _PyByteArray_FromBufferObject(sep);
 | |
|     if (! bytesep)
 | |
|         return NULL;
 | |
| 
 | |
|     result = stringlib_partition(
 | |
|             (PyObject*) self,
 | |
|             PyByteArray_AS_STRING(self), PyByteArray_GET_SIZE(self),
 | |
|             bytesep,
 | |
|             PyByteArray_AS_STRING(bytesep), PyByteArray_GET_SIZE(bytesep)
 | |
|             );
 | |
| 
 | |
|     Py_DECREF(bytesep);
 | |
|     return result;
 | |
| }
 | |
| 
 | |
| /*[clinic input]
 | |
| bytearray.rpartition
 | |
| 
 | |
|     sep: object
 | |
|     /
 | |
| 
 | |
| Partition the bytearray into three parts using the given separator.
 | |
| 
 | |
| This will search for the separator sep in the bytearray, starting at the end.
 | |
| If the separator is found, returns a 3-tuple containing the part before the
 | |
| separator, the separator itself, and the part after it as new bytearray
 | |
| objects.
 | |
| 
 | |
| If the separator is not found, returns a 3-tuple containing two empty bytearray
 | |
| objects and the copy of the original bytearray object.
 | |
| [clinic start generated code]*/
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_rpartition(PyByteArrayObject *self, PyObject *sep)
 | |
| /*[clinic end generated code: output=440de3c9426115e8 input=7e3df3e6cb8fa0ac]*/
 | |
| {
 | |
|     PyObject *bytesep, *result;
 | |
| 
 | |
|     bytesep = _PyByteArray_FromBufferObject(sep);
 | |
|     if (! bytesep)
 | |
|         return NULL;
 | |
| 
 | |
|     result = stringlib_rpartition(
 | |
|             (PyObject*) self,
 | |
|             PyByteArray_AS_STRING(self), PyByteArray_GET_SIZE(self),
 | |
|             bytesep,
 | |
|             PyByteArray_AS_STRING(bytesep), PyByteArray_GET_SIZE(bytesep)
 | |
|             );
 | |
| 
 | |
|     Py_DECREF(bytesep);
 | |
|     return result;
 | |
| }
 | |
| 
 | |
| /*[clinic input]
 | |
| bytearray.rsplit = bytearray.split
 | |
| 
 | |
| Return a list of the sections in the bytearray, using sep as the delimiter.
 | |
| 
 | |
| Splitting is done starting at the end of the bytearray and working to the front.
 | |
| [clinic start generated code]*/
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_rsplit_impl(PyByteArrayObject *self, PyObject *sep,
 | |
|                       Py_ssize_t maxsplit)
 | |
| /*[clinic end generated code: output=a55e0b5a03cb6190 input=a68286e4dd692ffe]*/
 | |
| {
 | |
|     Py_ssize_t len = PyByteArray_GET_SIZE(self), n;
 | |
|     const char *s = PyByteArray_AS_STRING(self), *sub;
 | |
|     PyObject *list;
 | |
|     Py_buffer vsub;
 | |
| 
 | |
|     if (maxsplit < 0)
 | |
|         maxsplit = PY_SSIZE_T_MAX;
 | |
| 
 | |
|     if (sep == Py_None)
 | |
|         return stringlib_rsplit_whitespace((PyObject*) self, s, len, maxsplit);
 | |
| 
 | |
|     if (PyObject_GetBuffer(sep, &vsub, PyBUF_SIMPLE) != 0)
 | |
|         return NULL;
 | |
|     sub = vsub.buf;
 | |
|     n = vsub.len;
 | |
| 
 | |
|     list = stringlib_rsplit(
 | |
|         (PyObject*) self, s, len, sub, n, maxsplit
 | |
|         );
 | |
|     PyBuffer_Release(&vsub);
 | |
|     return list;
 | |
| }
 | |
| 
 | |
| /*[clinic input]
 | |
| bytearray.reverse
 | |
| 
 | |
| Reverse the order of the values in B in place.
 | |
| [clinic start generated code]*/
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_reverse_impl(PyByteArrayObject *self)
 | |
| /*[clinic end generated code: output=9f7616f29ab309d3 input=543356319fc78557]*/
 | |
| {
 | |
|     char swap, *head, *tail;
 | |
|     Py_ssize_t i, j, n = Py_SIZE(self);
 | |
| 
 | |
|     j = n / 2;
 | |
|     head = PyByteArray_AS_STRING(self);
 | |
|     tail = head + n - 1;
 | |
|     for (i = 0; i < j; i++) {
 | |
|         swap = *head;
 | |
|         *head++ = *tail;
 | |
|         *tail-- = swap;
 | |
|     }
 | |
| 
 | |
|     Py_RETURN_NONE;
 | |
| }
 | |
| 
 | |
| 
 | |
| /*[python input]
 | |
| class bytesvalue_converter(CConverter):
 | |
|     type = 'int'
 | |
|     converter = '_getbytevalue'
 | |
| [python start generated code]*/
 | |
| /*[python end generated code: output=da39a3ee5e6b4b0d input=29c2e7c26c212812]*/
 | |
| 
 | |
| 
 | |
| /*[clinic input]
 | |
| bytearray.insert
 | |
| 
 | |
|     index: Py_ssize_t
 | |
|         The index where the value is to be inserted.
 | |
|     item: bytesvalue
 | |
|         The item to be inserted.
 | |
|     /
 | |
| 
 | |
| Insert a single item into the bytearray before the given index.
 | |
| [clinic start generated code]*/
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_insert_impl(PyByteArrayObject *self, Py_ssize_t index, int item)
 | |
| /*[clinic end generated code: output=76c775a70e7b07b7 input=b2b5d07e9de6c070]*/
 | |
| {
 | |
|     Py_ssize_t n = Py_SIZE(self);
 | |
|     char *buf;
 | |
| 
 | |
|     if (n == PY_SSIZE_T_MAX) {
 | |
|         PyErr_SetString(PyExc_OverflowError,
 | |
|                         "cannot add more objects to bytearray");
 | |
|         return NULL;
 | |
|     }
 | |
|     if (PyByteArray_Resize((PyObject *)self, n + 1) < 0)
 | |
|         return NULL;
 | |
|     buf = PyByteArray_AS_STRING(self);
 | |
| 
 | |
|     if (index < 0) {
 | |
|         index += n;
 | |
|         if (index < 0)
 | |
|             index = 0;
 | |
|     }
 | |
|     if (index > n)
 | |
|         index = n;
 | |
|     memmove(buf + index + 1, buf + index, n - index);
 | |
|     buf[index] = item;
 | |
| 
 | |
|     Py_RETURN_NONE;
 | |
| }
 | |
| 
 | |
| /*[clinic input]
 | |
| bytearray.append
 | |
| 
 | |
|     item: bytesvalue
 | |
|         The item to be appended.
 | |
|     /
 | |
| 
 | |
| Append a single item to the end of the bytearray.
 | |
| [clinic start generated code]*/
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_append_impl(PyByteArrayObject *self, int item)
 | |
| /*[clinic end generated code: output=a154e19ed1886cb6 input=20d6bec3d1340593]*/
 | |
| {
 | |
|     Py_ssize_t n = Py_SIZE(self);
 | |
| 
 | |
|     if (n == PY_SSIZE_T_MAX) {
 | |
|         PyErr_SetString(PyExc_OverflowError,
 | |
|                         "cannot add more objects to bytearray");
 | |
|         return NULL;
 | |
|     }
 | |
|     if (PyByteArray_Resize((PyObject *)self, n + 1) < 0)
 | |
|         return NULL;
 | |
| 
 | |
|     PyByteArray_AS_STRING(self)[n] = item;
 | |
| 
 | |
|     Py_RETURN_NONE;
 | |
| }
 | |
| 
 | |
| /*[clinic input]
 | |
| bytearray.extend
 | |
| 
 | |
|     iterable_of_ints: object
 | |
|         The iterable of items to append.
 | |
|     /
 | |
| 
 | |
| Append all the items from the iterator or sequence to the end of the bytearray.
 | |
| [clinic start generated code]*/
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_extend(PyByteArrayObject *self, PyObject *iterable_of_ints)
 | |
| /*[clinic end generated code: output=98155dbe249170b1 input=c617b3a93249ba28]*/
 | |
| {
 | |
|     PyObject *it, *item, *bytearray_obj;
 | |
|     Py_ssize_t buf_size = 0, len = 0;
 | |
|     int value;
 | |
|     char *buf;
 | |
| 
 | |
|     /* bytearray_setslice code only accepts something supporting PEP 3118. */
 | |
|     if (PyObject_CheckBuffer(iterable_of_ints)) {
 | |
|         if (bytearray_setslice(self, Py_SIZE(self), Py_SIZE(self), iterable_of_ints) == -1)
 | |
|             return NULL;
 | |
| 
 | |
|         Py_RETURN_NONE;
 | |
|     }
 | |
| 
 | |
|     it = PyObject_GetIter(iterable_of_ints);
 | |
|     if (it == NULL) {
 | |
|         if (PyErr_ExceptionMatches(PyExc_TypeError)) {
 | |
|             PyErr_Format(PyExc_TypeError,
 | |
|                          "can't extend bytearray with %.100s",
 | |
|                          Py_TYPE(iterable_of_ints)->tp_name);
 | |
|         }
 | |
|         return NULL;
 | |
|     }
 | |
| 
 | |
|     /* Try to determine the length of the argument. 32 is arbitrary. */
 | |
|     buf_size = PyObject_LengthHint(iterable_of_ints, 32);
 | |
|     if (buf_size == -1) {
 | |
|         Py_DECREF(it);
 | |
|         return NULL;
 | |
|     }
 | |
| 
 | |
|     bytearray_obj = PyByteArray_FromStringAndSize(NULL, buf_size);
 | |
|     if (bytearray_obj == NULL) {
 | |
|         Py_DECREF(it);
 | |
|         return NULL;
 | |
|     }
 | |
|     buf = PyByteArray_AS_STRING(bytearray_obj);
 | |
| 
 | |
|     while ((item = PyIter_Next(it)) != NULL) {
 | |
|         if (! _getbytevalue(item, &value)) {
 | |
|             if (PyErr_ExceptionMatches(PyExc_TypeError) && PyUnicode_Check(iterable_of_ints)) {
 | |
|                 PyErr_Format(PyExc_TypeError,
 | |
|                              "expected iterable of integers; got: 'str'");
 | |
|             }
 | |
|             Py_DECREF(item);
 | |
|             Py_DECREF(it);
 | |
|             Py_DECREF(bytearray_obj);
 | |
|             return NULL;
 | |
|         }
 | |
|         buf[len++] = value;
 | |
|         Py_DECREF(item);
 | |
| 
 | |
|         if (len >= buf_size) {
 | |
|             Py_ssize_t addition;
 | |
|             if (len == PY_SSIZE_T_MAX) {
 | |
|                 Py_DECREF(it);
 | |
|                 Py_DECREF(bytearray_obj);
 | |
|                 return PyErr_NoMemory();
 | |
|             }
 | |
|             addition = len >> 1;
 | |
|             if (addition > PY_SSIZE_T_MAX - len - 1)
 | |
|                 buf_size = PY_SSIZE_T_MAX;
 | |
|             else
 | |
|                 buf_size = len + addition + 1;
 | |
|             if (PyByteArray_Resize((PyObject *)bytearray_obj, buf_size) < 0) {
 | |
|                 Py_DECREF(it);
 | |
|                 Py_DECREF(bytearray_obj);
 | |
|                 return NULL;
 | |
|             }
 | |
|             /* Recompute the `buf' pointer, since the resizing operation may
 | |
|                have invalidated it. */
 | |
|             buf = PyByteArray_AS_STRING(bytearray_obj);
 | |
|         }
 | |
|     }
 | |
|     Py_DECREF(it);
 | |
| 
 | |
|     if (PyErr_Occurred()) {
 | |
|         Py_DECREF(bytearray_obj);
 | |
|         return NULL;
 | |
|     }
 | |
| 
 | |
|     /* Resize down to exact size. */
 | |
|     if (PyByteArray_Resize((PyObject *)bytearray_obj, len) < 0) {
 | |
|         Py_DECREF(bytearray_obj);
 | |
|         return NULL;
 | |
|     }
 | |
| 
 | |
|     if (bytearray_setslice(self, Py_SIZE(self), Py_SIZE(self), bytearray_obj) == -1) {
 | |
|         Py_DECREF(bytearray_obj);
 | |
|         return NULL;
 | |
|     }
 | |
|     Py_DECREF(bytearray_obj);
 | |
| 
 | |
|     assert(!PyErr_Occurred());
 | |
|     Py_RETURN_NONE;
 | |
| }
 | |
| 
 | |
| /*[clinic input]
 | |
| bytearray.pop
 | |
| 
 | |
|     index: Py_ssize_t = -1
 | |
|         The index from where to remove the item.
 | |
|         -1 (the default value) means remove the last item.
 | |
|     /
 | |
| 
 | |
| Remove and return a single item from B.
 | |
| 
 | |
| If no index argument is given, will pop the last item.
 | |
| [clinic start generated code]*/
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_pop_impl(PyByteArrayObject *self, Py_ssize_t index)
 | |
| /*[clinic end generated code: output=e0ccd401f8021da8 input=3591df2d06c0d237]*/
 | |
| {
 | |
|     int value;
 | |
|     Py_ssize_t n = Py_SIZE(self);
 | |
|     char *buf;
 | |
| 
 | |
|     if (n == 0) {
 | |
|         PyErr_SetString(PyExc_IndexError,
 | |
|                         "pop from empty bytearray");
 | |
|         return NULL;
 | |
|     }
 | |
|     if (index < 0)
 | |
|         index += Py_SIZE(self);
 | |
|     if (index < 0 || index >= Py_SIZE(self)) {
 | |
|         PyErr_SetString(PyExc_IndexError, "pop index out of range");
 | |
|         return NULL;
 | |
|     }
 | |
|     if (!_canresize(self))
 | |
|         return NULL;
 | |
| 
 | |
|     buf = PyByteArray_AS_STRING(self);
 | |
|     value = buf[index];
 | |
|     memmove(buf + index, buf + index + 1, n - index);
 | |
|     if (PyByteArray_Resize((PyObject *)self, n - 1) < 0)
 | |
|         return NULL;
 | |
| 
 | |
|     return _PyLong_FromUnsignedChar((unsigned char)value);
 | |
| }
 | |
| 
 | |
| /*[clinic input]
 | |
| bytearray.remove
 | |
| 
 | |
|     value: bytesvalue
 | |
|         The value to remove.
 | |
|     /
 | |
| 
 | |
| Remove the first occurrence of a value in the bytearray.
 | |
| [clinic start generated code]*/
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_remove_impl(PyByteArrayObject *self, int value)
 | |
| /*[clinic end generated code: output=d659e37866709c13 input=121831240cd51ddf]*/
 | |
| {
 | |
|     Py_ssize_t where, n = Py_SIZE(self);
 | |
|     char *buf = PyByteArray_AS_STRING(self);
 | |
| 
 | |
|     where = stringlib_find_char(buf, n, value);
 | |
|     if (where < 0) {
 | |
|         PyErr_SetString(PyExc_ValueError, "value not found in bytearray");
 | |
|         return NULL;
 | |
|     }
 | |
|     if (!_canresize(self))
 | |
|         return NULL;
 | |
| 
 | |
|     memmove(buf + where, buf + where + 1, n - where);
 | |
|     if (PyByteArray_Resize((PyObject *)self, n - 1) < 0)
 | |
|         return NULL;
 | |
| 
 | |
|     Py_RETURN_NONE;
 | |
| }
 | |
| 
 | |
| #define LEFTSTRIP 0
 | |
| #define RIGHTSTRIP 1
 | |
| #define BOTHSTRIP 2
 | |
| 
 | |
| static PyObject*
 | |
| bytearray_strip_impl_helper(PyByteArrayObject* self, PyObject* bytes, int striptype)
 | |
| {
 | |
|     Py_ssize_t mysize, byteslen;
 | |
|     const char* myptr;
 | |
|     const char* bytesptr;
 | |
|     Py_buffer vbytes;
 | |
| 
 | |
|     if (bytes == Py_None) {
 | |
|         bytesptr = "\t\n\r\f\v ";
 | |
|         byteslen = 6;
 | |
|     }
 | |
|     else {
 | |
|         if (PyObject_GetBuffer(bytes, &vbytes, PyBUF_SIMPLE) != 0)
 | |
|             return NULL;
 | |
|         bytesptr = (const char*)vbytes.buf;
 | |
|         byteslen = vbytes.len;
 | |
|     }
 | |
|     myptr = PyByteArray_AS_STRING(self);
 | |
|     mysize = Py_SIZE(self);
 | |
| 
 | |
|     Py_ssize_t left = 0;
 | |
|     if (striptype != RIGHTSTRIP) {
 | |
|         while (left < mysize && memchr(bytesptr, (unsigned char)myptr[left], byteslen))
 | |
|             left++;
 | |
|     }
 | |
|     Py_ssize_t right = mysize;
 | |
|     if (striptype != LEFTSTRIP) {
 | |
|         do {
 | |
|             right--;
 | |
|         } while (right >= left && memchr(bytesptr, (unsigned char)myptr[right], byteslen));
 | |
|         right++;
 | |
|     }
 | |
|     if (bytes != Py_None)
 | |
|         PyBuffer_Release(&vbytes);
 | |
|     return PyByteArray_FromStringAndSize(myptr + left, right - left);
 | |
| }
 | |
| 
 | |
| /*[clinic input]
 | |
| bytearray.strip
 | |
| 
 | |
|     bytes: object = None
 | |
|     /
 | |
| 
 | |
| Strip leading and trailing bytes contained in the argument.
 | |
| 
 | |
| If the argument is omitted or None, strip leading and trailing ASCII whitespace.
 | |
| [clinic start generated code]*/
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_strip_impl(PyByteArrayObject *self, PyObject *bytes)
 | |
| /*[clinic end generated code: output=760412661a34ad5a input=ef7bb59b09c21d62]*/
 | |
| {
 | |
|     return bytearray_strip_impl_helper(self, bytes, BOTHSTRIP);
 | |
| }
 | |
| 
 | |
| /*[clinic input]
 | |
| bytearray.lstrip
 | |
| 
 | |
|     bytes: object = None
 | |
|     /
 | |
| 
 | |
| Strip leading bytes contained in the argument.
 | |
| 
 | |
| If the argument is omitted or None, strip leading ASCII whitespace.
 | |
| [clinic start generated code]*/
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_lstrip_impl(PyByteArrayObject *self, PyObject *bytes)
 | |
| /*[clinic end generated code: output=d005c9d0ab909e66 input=80843f975dd7c480]*/
 | |
| {
 | |
|     return bytearray_strip_impl_helper(self, bytes, LEFTSTRIP);
 | |
| }
 | |
| 
 | |
| /*[clinic input]
 | |
| bytearray.rstrip
 | |
| 
 | |
|     bytes: object = None
 | |
|     /
 | |
| 
 | |
| Strip trailing bytes contained in the argument.
 | |
| 
 | |
| If the argument is omitted or None, strip trailing ASCII whitespace.
 | |
| [clinic start generated code]*/
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_rstrip_impl(PyByteArrayObject *self, PyObject *bytes)
 | |
| /*[clinic end generated code: output=030e2fbd2f7276bd input=e728b994954cfd91]*/
 | |
| {
 | |
|     return bytearray_strip_impl_helper(self, bytes, RIGHTSTRIP);
 | |
| }
 | |
| 
 | |
| /*[clinic input]
 | |
| bytearray.decode
 | |
| 
 | |
|     encoding: str(c_default="NULL") = 'utf-8'
 | |
|         The encoding with which to decode the bytearray.
 | |
|     errors: str(c_default="NULL") = 'strict'
 | |
|         The error handling scheme to use for the handling of decoding errors.
 | |
|         The default is 'strict' meaning that decoding errors raise a
 | |
|         UnicodeDecodeError. Other possible values are 'ignore' and 'replace'
 | |
|         as well as any other name registered with codecs.register_error that
 | |
|         can handle UnicodeDecodeErrors.
 | |
| 
 | |
| Decode the bytearray using the codec registered for encoding.
 | |
| [clinic start generated code]*/
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_decode_impl(PyByteArrayObject *self, const char *encoding,
 | |
|                       const char *errors)
 | |
| /*[clinic end generated code: output=f57d43f4a00b42c5 input=f28d8f903020257b]*/
 | |
| {
 | |
|     if (encoding == NULL)
 | |
|         encoding = PyUnicode_GetDefaultEncoding();
 | |
|     return PyUnicode_FromEncodedObject((PyObject*)self, encoding, errors);
 | |
| }
 | |
| 
 | |
| PyDoc_STRVAR(alloc_doc,
 | |
| "B.__alloc__() -> int\n\
 | |
| \n\
 | |
| Return the number of bytes actually allocated.");
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_alloc(PyByteArrayObject *self, PyObject *Py_UNUSED(ignored))
 | |
| {
 | |
|     return PyLong_FromSsize_t(self->ob_alloc);
 | |
| }
 | |
| 
 | |
| /*[clinic input]
 | |
| bytearray.join
 | |
| 
 | |
|     iterable_of_bytes: object
 | |
|     /
 | |
| 
 | |
| Concatenate any number of bytes/bytearray objects.
 | |
| 
 | |
| The bytearray whose method is called is inserted in between each pair.
 | |
| 
 | |
| The result is returned as a new bytearray object.
 | |
| [clinic start generated code]*/
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_join(PyByteArrayObject *self, PyObject *iterable_of_bytes)
 | |
| /*[clinic end generated code: output=a8516370bf68ae08 input=aba6b1f9b30fcb8e]*/
 | |
| {
 | |
|     self->ob_exports++; // this protects `self` from being cleared/resized if `iterable_of_bytes` is a custom iterator
 | |
|     PyObject* ret = stringlib_bytes_join((PyObject*)self, iterable_of_bytes);
 | |
|     self->ob_exports--; // unexport `self`
 | |
|     return ret;
 | |
| }
 | |
| 
 | |
| /*[clinic input]
 | |
| bytearray.splitlines
 | |
| 
 | |
|     keepends: bool = False
 | |
| 
 | |
| Return a list of the lines in the bytearray, breaking at line boundaries.
 | |
| 
 | |
| Line breaks are not included in the resulting list unless keepends is given and
 | |
| true.
 | |
| [clinic start generated code]*/
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_splitlines_impl(PyByteArrayObject *self, int keepends)
 | |
| /*[clinic end generated code: output=4223c94b895f6ad9 input=66b2dcdea8d093bf]*/
 | |
| {
 | |
|     return stringlib_splitlines(
 | |
|         (PyObject*) self, PyByteArray_AS_STRING(self),
 | |
|         PyByteArray_GET_SIZE(self), keepends
 | |
|         );
 | |
| }
 | |
| 
 | |
| /*[clinic input]
 | |
| @classmethod
 | |
| bytearray.fromhex
 | |
| 
 | |
|     string: unicode
 | |
|     /
 | |
| 
 | |
| Create a bytearray object from a string of hexadecimal numbers.
 | |
| 
 | |
| Spaces between two numbers are accepted.
 | |
| Example: bytearray.fromhex('B9 01EF') -> bytearray(b'\\xb9\\x01\\xef')
 | |
| [clinic start generated code]*/
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_fromhex_impl(PyTypeObject *type, PyObject *string)
 | |
| /*[clinic end generated code: output=8f0f0b6d30fb3ba0 input=f033a16d1fb21f48]*/
 | |
| {
 | |
|     PyObject *result = _PyBytes_FromHex(string, type == &PyByteArray_Type);
 | |
|     if (type != &PyByteArray_Type && result != NULL) {
 | |
|         Py_SETREF(result, PyObject_CallOneArg((PyObject *)type, result));
 | |
|     }
 | |
|     return result;
 | |
| }
 | |
| 
 | |
| /*[clinic input]
 | |
| bytearray.hex
 | |
| 
 | |
|     sep: object = NULL
 | |
|         An optional single character or byte to separate hex bytes.
 | |
|     bytes_per_sep: int = 1
 | |
|         How many bytes between separators.  Positive values count from the
 | |
|         right, negative values count from the left.
 | |
| 
 | |
| Create a string of hexadecimal numbers from a bytearray object.
 | |
| 
 | |
| Example:
 | |
| >>> value = bytearray([0xb9, 0x01, 0xef])
 | |
| >>> value.hex()
 | |
| 'b901ef'
 | |
| >>> value.hex(':')
 | |
| 'b9:01:ef'
 | |
| >>> value.hex(':', 2)
 | |
| 'b9:01ef'
 | |
| >>> value.hex(':', -2)
 | |
| 'b901:ef'
 | |
| [clinic start generated code]*/
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_hex_impl(PyByteArrayObject *self, PyObject *sep, int bytes_per_sep)
 | |
| /*[clinic end generated code: output=29c4e5ef72c565a0 input=808667e49bcccb54]*/
 | |
| {
 | |
|     char* argbuf = PyByteArray_AS_STRING(self);
 | |
|     Py_ssize_t arglen = PyByteArray_GET_SIZE(self);
 | |
|     return _Py_strhex_with_sep(argbuf, arglen, sep, bytes_per_sep);
 | |
| }
 | |
| 
 | |
| static PyObject *
 | |
| _common_reduce(PyByteArrayObject *self, int proto)
 | |
| {
 | |
|     PyObject *state;
 | |
|     const char *buf;
 | |
| 
 | |
|     state = _PyObject_GetState((PyObject *)self);
 | |
|     if (state == NULL) {
 | |
|         return NULL;
 | |
|     }
 | |
| 
 | |
|     if (!Py_SIZE(self)) {
 | |
|         return Py_BuildValue("(O()N)", Py_TYPE(self), state);
 | |
|     }
 | |
|     buf = PyByteArray_AS_STRING(self);
 | |
|     if (proto < 3) {
 | |
|         /* use str based reduction for backwards compatibility with Python 2.x */
 | |
|         PyObject *latin1 = PyUnicode_DecodeLatin1(buf, Py_SIZE(self), NULL);
 | |
|         return Py_BuildValue("(O(Ns)N)", Py_TYPE(self), latin1, "latin-1", state);
 | |
|     }
 | |
|     else {
 | |
|         /* use more efficient byte based reduction */
 | |
|         return Py_BuildValue("(O(y#)N)", Py_TYPE(self), buf, Py_SIZE(self), state);
 | |
|     }
 | |
| }
 | |
| 
 | |
| /*[clinic input]
 | |
| bytearray.__reduce__ as bytearray_reduce
 | |
| 
 | |
| Return state information for pickling.
 | |
| [clinic start generated code]*/
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_reduce_impl(PyByteArrayObject *self)
 | |
| /*[clinic end generated code: output=52bf304086464cab input=44b5737ada62dd3f]*/
 | |
| {
 | |
|     return _common_reduce(self, 2);
 | |
| }
 | |
| 
 | |
| /*[clinic input]
 | |
| bytearray.__reduce_ex__ as bytearray_reduce_ex
 | |
| 
 | |
|     proto: int = 0
 | |
|     /
 | |
| 
 | |
| Return state information for pickling.
 | |
| [clinic start generated code]*/
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_reduce_ex_impl(PyByteArrayObject *self, int proto)
 | |
| /*[clinic end generated code: output=52eac33377197520 input=f129bc1a1aa151ee]*/
 | |
| {
 | |
|     return _common_reduce(self, proto);
 | |
| }
 | |
| 
 | |
| /*[clinic input]
 | |
| bytearray.__sizeof__ as bytearray_sizeof
 | |
| 
 | |
| Returns the size of the bytearray object in memory, in bytes.
 | |
| [clinic start generated code]*/
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_sizeof_impl(PyByteArrayObject *self)
 | |
| /*[clinic end generated code: output=738abdd17951c427 input=e27320fd98a4bc5a]*/
 | |
| {
 | |
|     size_t res = _PyObject_SIZE(Py_TYPE(self));
 | |
|     res += (size_t)self->ob_alloc * sizeof(char);
 | |
|     return PyLong_FromSize_t(res);
 | |
| }
 | |
| 
 | |
| static PySequenceMethods bytearray_as_sequence = {
 | |
|     (lenfunc)bytearray_length,              /* sq_length */
 | |
|     (binaryfunc)PyByteArray_Concat,         /* sq_concat */
 | |
|     (ssizeargfunc)bytearray_repeat,         /* sq_repeat */
 | |
|     (ssizeargfunc)bytearray_getitem,        /* sq_item */
 | |
|     0,                                      /* sq_slice */
 | |
|     (ssizeobjargproc)bytearray_setitem,     /* sq_ass_item */
 | |
|     0,                                      /* sq_ass_slice */
 | |
|     (objobjproc)bytearray_contains,         /* sq_contains */
 | |
|     (binaryfunc)bytearray_iconcat,          /* sq_inplace_concat */
 | |
|     (ssizeargfunc)bytearray_irepeat,        /* sq_inplace_repeat */
 | |
| };
 | |
| 
 | |
| static PyMappingMethods bytearray_as_mapping = {
 | |
|     (lenfunc)bytearray_length,
 | |
|     (binaryfunc)bytearray_subscript,
 | |
|     (objobjargproc)bytearray_ass_subscript,
 | |
| };
 | |
| 
 | |
| static PyBufferProcs bytearray_as_buffer = {
 | |
|     (getbufferproc)bytearray_getbuffer,
 | |
|     (releasebufferproc)bytearray_releasebuffer,
 | |
| };
 | |
| 
 | |
| static PyMethodDef
 | |
| bytearray_methods[] = {
 | |
|     {"__alloc__", (PyCFunction)bytearray_alloc, METH_NOARGS, alloc_doc},
 | |
|     BYTEARRAY_REDUCE_METHODDEF
 | |
|     BYTEARRAY_REDUCE_EX_METHODDEF
 | |
|     BYTEARRAY_SIZEOF_METHODDEF
 | |
|     BYTEARRAY_APPEND_METHODDEF
 | |
|     {"capitalize", stringlib_capitalize, METH_NOARGS,
 | |
|      _Py_capitalize__doc__},
 | |
|     STRINGLIB_CENTER_METHODDEF
 | |
|     BYTEARRAY_CLEAR_METHODDEF
 | |
|     BYTEARRAY_COPY_METHODDEF
 | |
|     BYTEARRAY_COUNT_METHODDEF
 | |
|     BYTEARRAY_DECODE_METHODDEF
 | |
|     BYTEARRAY_ENDSWITH_METHODDEF
 | |
|     STRINGLIB_EXPANDTABS_METHODDEF
 | |
|     BYTEARRAY_EXTEND_METHODDEF
 | |
|     BYTEARRAY_FIND_METHODDEF
 | |
|     BYTEARRAY_FROMHEX_METHODDEF
 | |
|     BYTEARRAY_HEX_METHODDEF
 | |
|     BYTEARRAY_INDEX_METHODDEF
 | |
|     BYTEARRAY_INSERT_METHODDEF
 | |
|     {"isalnum", stringlib_isalnum, METH_NOARGS,
 | |
|      _Py_isalnum__doc__},
 | |
|     {"isalpha", stringlib_isalpha, METH_NOARGS,
 | |
|      _Py_isalpha__doc__},
 | |
|     {"isascii", stringlib_isascii, METH_NOARGS,
 | |
|      _Py_isascii__doc__},
 | |
|     {"isdigit", stringlib_isdigit, METH_NOARGS,
 | |
|      _Py_isdigit__doc__},
 | |
|     {"islower", stringlib_islower, METH_NOARGS,
 | |
|      _Py_islower__doc__},
 | |
|     {"isspace", stringlib_isspace, METH_NOARGS,
 | |
|      _Py_isspace__doc__},
 | |
|     {"istitle", stringlib_istitle, METH_NOARGS,
 | |
|      _Py_istitle__doc__},
 | |
|     {"isupper", stringlib_isupper, METH_NOARGS,
 | |
|      _Py_isupper__doc__},
 | |
|     BYTEARRAY_JOIN_METHODDEF
 | |
|     STRINGLIB_LJUST_METHODDEF
 | |
|     {"lower", stringlib_lower, METH_NOARGS, _Py_lower__doc__},
 | |
|     BYTEARRAY_LSTRIP_METHODDEF
 | |
|     BYTEARRAY_MAKETRANS_METHODDEF
 | |
|     BYTEARRAY_PARTITION_METHODDEF
 | |
|     BYTEARRAY_POP_METHODDEF
 | |
|     BYTEARRAY_REMOVE_METHODDEF
 | |
|     BYTEARRAY_REPLACE_METHODDEF
 | |
|     BYTEARRAY_REMOVEPREFIX_METHODDEF
 | |
|     BYTEARRAY_REMOVESUFFIX_METHODDEF
 | |
|     BYTEARRAY_REVERSE_METHODDEF
 | |
|     BYTEARRAY_RFIND_METHODDEF
 | |
|     BYTEARRAY_RINDEX_METHODDEF
 | |
|     STRINGLIB_RJUST_METHODDEF
 | |
|     BYTEARRAY_RPARTITION_METHODDEF
 | |
|     BYTEARRAY_RSPLIT_METHODDEF
 | |
|     BYTEARRAY_RSTRIP_METHODDEF
 | |
|     BYTEARRAY_SPLIT_METHODDEF
 | |
|     BYTEARRAY_SPLITLINES_METHODDEF
 | |
|     BYTEARRAY_STARTSWITH_METHODDEF
 | |
|     BYTEARRAY_STRIP_METHODDEF
 | |
|     {"swapcase", stringlib_swapcase, METH_NOARGS,
 | |
|      _Py_swapcase__doc__},
 | |
|     {"title", stringlib_title, METH_NOARGS, _Py_title__doc__},
 | |
|     BYTEARRAY_TRANSLATE_METHODDEF
 | |
|     {"upper", stringlib_upper, METH_NOARGS, _Py_upper__doc__},
 | |
|     STRINGLIB_ZFILL_METHODDEF
 | |
|     {NULL}
 | |
| };
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_mod(PyObject *v, PyObject *w)
 | |
| {
 | |
|     if (!PyByteArray_Check(v))
 | |
|         Py_RETURN_NOTIMPLEMENTED;
 | |
|     return _PyBytes_FormatEx(PyByteArray_AS_STRING(v), PyByteArray_GET_SIZE(v), w, 1);
 | |
| }
 | |
| 
 | |
| static PyNumberMethods bytearray_as_number = {
 | |
|     0,              /*nb_add*/
 | |
|     0,              /*nb_subtract*/
 | |
|     0,              /*nb_multiply*/
 | |
|     bytearray_mod,  /*nb_remainder*/
 | |
| };
 | |
| 
 | |
| PyDoc_STRVAR(bytearray_doc,
 | |
| "bytearray(iterable_of_ints) -> bytearray\n\
 | |
| bytearray(string, encoding[, errors]) -> bytearray\n\
 | |
| bytearray(bytes_or_buffer) -> mutable copy of bytes_or_buffer\n\
 | |
| bytearray(int) -> bytes array of size given by the parameter initialized with null bytes\n\
 | |
| bytearray() -> empty bytes array\n\
 | |
| \n\
 | |
| Construct a mutable bytearray object from:\n\
 | |
|   - an iterable yielding integers in range(256)\n\
 | |
|   - a text string encoded using the specified encoding\n\
 | |
|   - a bytes or a buffer object\n\
 | |
|   - any object implementing the buffer API.\n\
 | |
|   - an integer");
 | |
| 
 | |
| 
 | |
| static PyObject *bytearray_iter(PyObject *seq);
 | |
| 
 | |
| PyTypeObject PyByteArray_Type = {
 | |
|     PyVarObject_HEAD_INIT(&PyType_Type, 0)
 | |
|     "bytearray",
 | |
|     sizeof(PyByteArrayObject),
 | |
|     0,
 | |
|     (destructor)bytearray_dealloc,       /* tp_dealloc */
 | |
|     0,                                  /* tp_vectorcall_offset */
 | |
|     0,                                  /* tp_getattr */
 | |
|     0,                                  /* tp_setattr */
 | |
|     0,                                  /* tp_as_async */
 | |
|     (reprfunc)bytearray_repr,           /* tp_repr */
 | |
|     &bytearray_as_number,               /* tp_as_number */
 | |
|     &bytearray_as_sequence,             /* tp_as_sequence */
 | |
|     &bytearray_as_mapping,              /* tp_as_mapping */
 | |
|     0,                                  /* tp_hash */
 | |
|     0,                                  /* tp_call */
 | |
|     bytearray_str,                      /* tp_str */
 | |
|     PyObject_GenericGetAttr,            /* tp_getattro */
 | |
|     0,                                  /* tp_setattro */
 | |
|     &bytearray_as_buffer,               /* tp_as_buffer */
 | |
|     Py_TPFLAGS_DEFAULT | Py_TPFLAGS_BASETYPE |
 | |
|         _Py_TPFLAGS_MATCH_SELF,       /* tp_flags */
 | |
|     bytearray_doc,                      /* tp_doc */
 | |
|     0,                                  /* tp_traverse */
 | |
|     0,                                  /* tp_clear */
 | |
|     (richcmpfunc)bytearray_richcompare, /* tp_richcompare */
 | |
|     0,                                  /* tp_weaklistoffset */
 | |
|     bytearray_iter,                     /* tp_iter */
 | |
|     0,                                  /* tp_iternext */
 | |
|     bytearray_methods,                  /* tp_methods */
 | |
|     0,                                  /* tp_members */
 | |
|     0,                                  /* tp_getset */
 | |
|     0,                                  /* tp_base */
 | |
|     0,                                  /* tp_dict */
 | |
|     0,                                  /* tp_descr_get */
 | |
|     0,                                  /* tp_descr_set */
 | |
|     0,                                  /* tp_dictoffset */
 | |
|     (initproc)bytearray___init__,       /* tp_init */
 | |
|     PyType_GenericAlloc,                /* tp_alloc */
 | |
|     PyType_GenericNew,                  /* tp_new */
 | |
|     PyObject_Del,                       /* tp_free */
 | |
| };
 | |
| 
 | |
| /*********************** Bytearray Iterator ****************************/
 | |
| 
 | |
| typedef struct {
 | |
|     PyObject_HEAD
 | |
|     Py_ssize_t it_index;
 | |
|     PyByteArrayObject *it_seq; /* Set to NULL when iterator is exhausted */
 | |
| } bytesiterobject;
 | |
| 
 | |
| static void
 | |
| bytearrayiter_dealloc(bytesiterobject *it)
 | |
| {
 | |
|     _PyObject_GC_UNTRACK(it);
 | |
|     Py_XDECREF(it->it_seq);
 | |
|     PyObject_GC_Del(it);
 | |
| }
 | |
| 
 | |
| static int
 | |
| bytearrayiter_traverse(bytesiterobject *it, visitproc visit, void *arg)
 | |
| {
 | |
|     Py_VISIT(it->it_seq);
 | |
|     return 0;
 | |
| }
 | |
| 
 | |
| static PyObject *
 | |
| bytearrayiter_next(bytesiterobject *it)
 | |
| {
 | |
|     PyByteArrayObject *seq;
 | |
| 
 | |
|     assert(it != NULL);
 | |
|     seq = it->it_seq;
 | |
|     if (seq == NULL)
 | |
|         return NULL;
 | |
|     assert(PyByteArray_Check(seq));
 | |
| 
 | |
|     if (it->it_index < PyByteArray_GET_SIZE(seq)) {
 | |
|         return _PyLong_FromUnsignedChar(
 | |
|             (unsigned char)PyByteArray_AS_STRING(seq)[it->it_index++]);
 | |
|     }
 | |
| 
 | |
|     it->it_seq = NULL;
 | |
|     Py_DECREF(seq);
 | |
|     return NULL;
 | |
| }
 | |
| 
 | |
| static PyObject *
 | |
| bytearrayiter_length_hint(bytesiterobject *it, PyObject *Py_UNUSED(ignored))
 | |
| {
 | |
|     Py_ssize_t len = 0;
 | |
|     if (it->it_seq) {
 | |
|         len = PyByteArray_GET_SIZE(it->it_seq) - it->it_index;
 | |
|         if (len < 0) {
 | |
|             len = 0;
 | |
|         }
 | |
|     }
 | |
|     return PyLong_FromSsize_t(len);
 | |
| }
 | |
| 
 | |
| PyDoc_STRVAR(length_hint_doc,
 | |
|     "Private method returning an estimate of len(list(it)).");
 | |
| 
 | |
| static PyObject *
 | |
| bytearrayiter_reduce(bytesiterobject *it, PyObject *Py_UNUSED(ignored))
 | |
| {
 | |
|     PyObject *iter = _PyEval_GetBuiltin(&_Py_ID(iter));
 | |
| 
 | |
|     /* _PyEval_GetBuiltin can invoke arbitrary code,
 | |
|      * call must be before access of iterator pointers.
 | |
|      * see issue #101765 */
 | |
| 
 | |
|     if (it->it_seq != NULL) {
 | |
|         return Py_BuildValue("N(O)n", iter, it->it_seq, it->it_index);
 | |
|     } else {
 | |
|         return Py_BuildValue("N(())", iter);
 | |
|     }
 | |
| }
 | |
| 
 | |
| static PyObject *
 | |
| bytearrayiter_setstate(bytesiterobject *it, PyObject *state)
 | |
| {
 | |
|     Py_ssize_t index = PyLong_AsSsize_t(state);
 | |
|     if (index == -1 && PyErr_Occurred())
 | |
|         return NULL;
 | |
|     if (it->it_seq != NULL) {
 | |
|         if (index < 0)
 | |
|             index = 0;
 | |
|         else if (index > PyByteArray_GET_SIZE(it->it_seq))
 | |
|             index = PyByteArray_GET_SIZE(it->it_seq); /* iterator exhausted */
 | |
|         it->it_index = index;
 | |
|     }
 | |
|     Py_RETURN_NONE;
 | |
| }
 | |
| 
 | |
| PyDoc_STRVAR(setstate_doc, "Set state information for unpickling.");
 | |
| 
 | |
| static PyMethodDef bytearrayiter_methods[] = {
 | |
|     {"__length_hint__", (PyCFunction)bytearrayiter_length_hint, METH_NOARGS,
 | |
|      length_hint_doc},
 | |
|      {"__reduce__",      (PyCFunction)bytearrayiter_reduce, METH_NOARGS,
 | |
|      bytearray_reduce__doc__},
 | |
|     {"__setstate__",    (PyCFunction)bytearrayiter_setstate, METH_O,
 | |
|      setstate_doc},
 | |
|     {NULL, NULL} /* sentinel */
 | |
| };
 | |
| 
 | |
| PyTypeObject PyByteArrayIter_Type = {
 | |
|     PyVarObject_HEAD_INIT(&PyType_Type, 0)
 | |
|     "bytearray_iterator",              /* tp_name */
 | |
|     sizeof(bytesiterobject),           /* tp_basicsize */
 | |
|     0,                                 /* tp_itemsize */
 | |
|     /* methods */
 | |
|     (destructor)bytearrayiter_dealloc, /* tp_dealloc */
 | |
|     0,                                 /* tp_vectorcall_offset */
 | |
|     0,                                 /* tp_getattr */
 | |
|     0,                                 /* tp_setattr */
 | |
|     0,                                 /* tp_as_async */
 | |
|     0,                                 /* tp_repr */
 | |
|     0,                                 /* tp_as_number */
 | |
|     0,                                 /* tp_as_sequence */
 | |
|     0,                                 /* tp_as_mapping */
 | |
|     0,                                 /* tp_hash */
 | |
|     0,                                 /* tp_call */
 | |
|     0,                                 /* tp_str */
 | |
|     PyObject_GenericGetAttr,           /* tp_getattro */
 | |
|     0,                                 /* tp_setattro */
 | |
|     0,                                 /* tp_as_buffer */
 | |
|     Py_TPFLAGS_DEFAULT | Py_TPFLAGS_HAVE_GC, /* tp_flags */
 | |
|     0,                                 /* tp_doc */
 | |
|     (traverseproc)bytearrayiter_traverse,  /* tp_traverse */
 | |
|     0,                                 /* tp_clear */
 | |
|     0,                                 /* tp_richcompare */
 | |
|     0,                                 /* tp_weaklistoffset */
 | |
|     PyObject_SelfIter,                 /* tp_iter */
 | |
|     (iternextfunc)bytearrayiter_next,  /* tp_iternext */
 | |
|     bytearrayiter_methods,             /* tp_methods */
 | |
|     0,
 | |
| };
 | |
| 
 | |
| static PyObject *
 | |
| bytearray_iter(PyObject *seq)
 | |
| {
 | |
|     bytesiterobject *it;
 | |
| 
 | |
|     if (!PyByteArray_Check(seq)) {
 | |
|         PyErr_BadInternalCall();
 | |
|         return NULL;
 | |
|     }
 | |
|     it = PyObject_GC_New(bytesiterobject, &PyByteArrayIter_Type);
 | |
|     if (it == NULL)
 | |
|         return NULL;
 | |
|     it->it_index = 0;
 | |
|     it->it_seq = (PyByteArrayObject *)Py_NewRef(seq);
 | |
|     _PyObject_GC_TRACK(it);
 | |
|     return (PyObject *)it;
 | |
| }
 |