| 1 |
#ifndef Py_CPYTHON_UNICODEOBJECT_H |
| 2 |
# error "this header file must not be included directly" |
| 3 |
#endif |
| 4 |
|
| 5 |
/* Py_UNICODE was the native Unicode storage format (code unit) used by |
| 6 |
Python and represents a single Unicode element in the Unicode type. |
| 7 |
With PEP 393, Py_UNICODE is deprecated and replaced with a |
| 8 |
typedef to wchar_t. */ |
| 9 |
Py_DEPRECATED(3.13) typedef wchar_t PY_UNICODE_TYPE; |
| 10 |
Py_DEPRECATED(3.13) typedef wchar_t Py_UNICODE; |
| 11 |
|
| 12 |
|
| 13 |
/* --- Internal Unicode Operations ---------------------------------------- */ |
| 14 |
|
| 15 |
// Static inline functions to work with surrogates |
| 16 |
static inline int Py_UNICODE_IS_SURROGATE(Py_UCS4 ch) { |
| 17 |
return (0xD800 <= ch && ch <= 0xDFFF); |
| 18 |
} |
| 19 |
static inline int Py_UNICODE_IS_HIGH_SURROGATE(Py_UCS4 ch) { |
| 20 |
return (0xD800 <= ch && ch <= 0xDBFF); |
| 21 |
} |
| 22 |
static inline int Py_UNICODE_IS_LOW_SURROGATE(Py_UCS4 ch) { |
| 23 |
return (0xDC00 <= ch && ch <= 0xDFFF); |
| 24 |
} |
| 25 |
|
| 26 |
// Join two surrogate characters and return a single Py_UCS4 value. |
| 27 |
static inline Py_UCS4 Py_UNICODE_JOIN_SURROGATES(Py_UCS4 high, Py_UCS4 low) { |
| 28 |
assert(Py_UNICODE_IS_HIGH_SURROGATE(high)); |
| 29 |
assert(Py_UNICODE_IS_LOW_SURROGATE(low)); |
| 30 |
return 0x10000 + (((high & 0x03FF) << 10) | (low & 0x03FF)); |
| 31 |
} |
| 32 |
|
| 33 |
// High surrogate = top 10 bits added to 0xD800. |
| 34 |
// The character must be in the range [U+10000; U+10ffff]. |
| 35 |
static inline Py_UCS4 Py_UNICODE_HIGH_SURROGATE(Py_UCS4 ch) { |
| 36 |
assert(0x10000 <= ch && ch <= 0x10ffff); |
| 37 |
return (0xD800 - (0x10000 >> 10) + (ch >> 10)); |
| 38 |
} |
| 39 |
|
| 40 |
// Low surrogate = bottom 10 bits added to 0xDC00. |
| 41 |
// The character must be in the range [U+10000; U+10ffff]. |
| 42 |
static inline Py_UCS4 Py_UNICODE_LOW_SURROGATE(Py_UCS4 ch) { |
| 43 |
assert(0x10000 <= ch && ch <= 0x10ffff); |
| 44 |
return (0xDC00 + (ch & 0x3FF)); |
| 45 |
} |
| 46 |
|
| 47 |
|
| 48 |
/* --- Unicode Type ------------------------------------------------------- */ |
| 49 |
|
| 50 |
/* ASCII-only strings created through PyUnicode_New use the PyASCIIObject |
| 51 |
structure. state.ascii and state.compact are set, and the data |
| 52 |
immediately follow the structure. utf8_length can be found |
| 53 |
in the length field; the utf8 pointer is equal to the data pointer. */ |
| 54 |
typedef struct { |
| 55 |
/* There are 4 forms of Unicode strings: |
| 56 |
|
| 57 |
- compact ascii: |
| 58 |
|
| 59 |
* structure = PyASCIIObject |
| 60 |
* test: PyUnicode_IS_COMPACT_ASCII(op) |
| 61 |
* kind = PyUnicode_1BYTE_KIND |
| 62 |
* compact = 1 |
| 63 |
* ascii = 1 |
| 64 |
* (length is the length of the utf8) |
| 65 |
* (data starts just after the structure) |
| 66 |
* (since ASCII is decoded from UTF-8, the utf8 string are the data) |
| 67 |
|
| 68 |
- compact: |
| 69 |
|
| 70 |
* structure = PyCompactUnicodeObject |
| 71 |
* test: PyUnicode_IS_COMPACT(op) && !PyUnicode_IS_ASCII(op) |
| 72 |
* kind = PyUnicode_1BYTE_KIND, PyUnicode_2BYTE_KIND or |
| 73 |
PyUnicode_4BYTE_KIND |
| 74 |
* compact = 1 |
| 75 |
* ascii = 0 |
| 76 |
* utf8 is not shared with data |
| 77 |
* utf8_length = 0 if utf8 is NULL |
| 78 |
* (data starts just after the structure) |
| 79 |
|
| 80 |
- legacy string: |
| 81 |
|
| 82 |
* structure = PyUnicodeObject structure |
| 83 |
* test: !PyUnicode_IS_COMPACT(op) |
| 84 |
* kind = PyUnicode_1BYTE_KIND, PyUnicode_2BYTE_KIND or |
| 85 |
PyUnicode_4BYTE_KIND |
| 86 |
* compact = 0 |
| 87 |
* data.any is not NULL |
| 88 |
* utf8 is shared and utf8_length = length with data.any if ascii = 1 |
| 89 |
* utf8_length = 0 if utf8 is NULL |
| 90 |
|
| 91 |
Compact strings use only one memory block (structure + characters), |
| 92 |
whereas legacy strings use one block for the structure and one block |
| 93 |
for characters. |
| 94 |
|
| 95 |
Legacy strings are created by subclasses of Unicode. |
| 96 |
|
| 97 |
See also _PyUnicode_CheckConsistency(). |
| 98 |
*/ |
| 99 |
PyObject_HEAD |
| 100 |
Py_ssize_t length; /* Number of code points in the string */ |
| 101 |
Py_hash_t hash; /* Hash value; -1 if not set */ |
| 102 |
#ifdef Py_GIL_DISABLED |
| 103 |
/* Ensure 4 byte alignment for PyUnicode_DATA(), see gh-63736 on m68k. |
| 104 |
In the non-free-threaded build, we'll use explicit padding instead */ |
| 105 |
_Py_ALIGN_AS(4) |
| 106 |
#endif |
| 107 |
struct { |
| 108 |
/* If interned is non-zero, the two references from the |
| 109 |
dictionary to this object are *not* counted in ob_refcnt. |
| 110 |
The possible values here are: |
| 111 |
0: Not Interned |
| 112 |
1: Interned |
| 113 |
2: Interned and Immortal |
| 114 |
3: Interned, Immortal, and Static |
| 115 |
This categorization allows the runtime to determine the right |
| 116 |
cleanup mechanism at runtime shutdown. */ |
| 117 |
#ifdef Py_GIL_DISABLED |
| 118 |
// Needs to be accessed atomically, so can't be a bit field. |
| 119 |
unsigned char interned; |
| 120 |
#else |
| 121 |
unsigned int interned:2; |
| 122 |
#endif |
| 123 |
/* Character size: |
| 124 |
|
| 125 |
- PyUnicode_1BYTE_KIND (1): |
| 126 |
|
| 127 |
* character type = Py_UCS1 (8 bits, unsigned) |
| 128 |
* all characters are in the range U+0000-U+00FF (latin1) |
| 129 |
* if ascii is set, all characters are in the range U+0000-U+007F |
| 130 |
(ASCII), otherwise at least one character is in the range |
| 131 |
U+0080-U+00FF |
| 132 |
|
| 133 |
- PyUnicode_2BYTE_KIND (2): |
| 134 |
|
| 135 |
* character type = Py_UCS2 (16 bits, unsigned) |
| 136 |
* all characters are in the range U+0000-U+FFFF (BMP) |
| 137 |
* at least one character is in the range U+0100-U+FFFF |
| 138 |
|
| 139 |
- PyUnicode_4BYTE_KIND (4): |
| 140 |
|
| 141 |
* character type = Py_UCS4 (32 bits, unsigned) |
| 142 |
* all characters are in the range U+0000-U+10FFFF |
| 143 |
* at least one character is in the range U+10000-U+10FFFF |
| 144 |
*/ |
| 145 |
unsigned int kind:3; |
| 146 |
/* Compact is with respect to the allocation scheme. Compact unicode |
| 147 |
objects only require one memory block while non-compact objects use |
| 148 |
one block for the PyUnicodeObject struct and another for its data |
| 149 |
buffer. */ |
| 150 |
unsigned int compact:1; |
| 151 |
/* The string only contains characters in the range U+0000-U+007F (ASCII) |
| 152 |
and the kind is PyUnicode_1BYTE_KIND. If ascii is set and compact is |
| 153 |
set, use the PyASCIIObject structure. */ |
| 154 |
unsigned int ascii:1; |
| 155 |
/* The object is statically allocated. */ |
| 156 |
unsigned int statically_allocated:1; |
| 157 |
#ifndef Py_GIL_DISABLED |
| 158 |
/* Padding to ensure that PyUnicode_DATA() is always aligned to |
| 159 |
4 bytes (see issue gh-63736 on m68k) */ |
| 160 |
unsigned int :24; |
| 161 |
#endif |
| 162 |
} state; |
| 163 |
} PyASCIIObject; |
| 164 |
|
| 165 |
/* Non-ASCII strings allocated through PyUnicode_New use the |
| 166 |
PyCompactUnicodeObject structure. state.compact is set, and the data |
| 167 |
immediately follow the structure. */ |
| 168 |
typedef struct { |
| 169 |
PyASCIIObject _base; |
| 170 |
Py_ssize_t utf8_length; /* Number of bytes in utf8, excluding the |
| 171 |
* terminating \0. */ |
| 172 |
char *utf8; /* UTF-8 representation (null-terminated) */ |
| 173 |
} PyCompactUnicodeObject; |
| 174 |
|
| 175 |
/* Object format for Unicode subclasses. */ |
| 176 |
typedef struct { |
| 177 |
PyCompactUnicodeObject _base; |
| 178 |
union { |
| 179 |
void *any; |
| 180 |
Py_UCS1 *latin1; |
| 181 |
Py_UCS2 *ucs2; |
| 182 |
Py_UCS4 *ucs4; |
| 183 |
} data; /* Canonical, smallest-form Unicode buffer */ |
| 184 |
} PyUnicodeObject; |
| 185 |
|
| 186 |
|
| 187 |
#define _PyASCIIObject_CAST(op) \ |
| 188 |
(assert(PyUnicode_Check(op)), \ |
| 189 |
_Py_CAST(PyASCIIObject*, (op))) |
| 190 |
#define _PyCompactUnicodeObject_CAST(op) \ |
| 191 |
(assert(PyUnicode_Check(op)), \ |
| 192 |
_Py_CAST(PyCompactUnicodeObject*, (op))) |
| 193 |
#define _PyUnicodeObject_CAST(op) \ |
| 194 |
(assert(PyUnicode_Check(op)), \ |
| 195 |
_Py_CAST(PyUnicodeObject*, (op))) |
| 196 |
|
| 197 |
|
| 198 |
/* --- Flexible String Representation Helper Macros (PEP 393) -------------- */ |
| 199 |
|
| 200 |
/* Values for PyASCIIObject.state: */ |
| 201 |
|
| 202 |
/* Interning state. */ |
| 203 |
#define SSTATE_NOT_INTERNED 0 |
| 204 |
#define SSTATE_INTERNED_MORTAL 1 |
| 205 |
#define SSTATE_INTERNED_IMMORTAL 2 |
| 206 |
#define SSTATE_INTERNED_IMMORTAL_STATIC 3 |
| 207 |
|
| 208 |
/* Use only if you know it's a string */ |
| 209 |
static inline unsigned int PyUnicode_CHECK_INTERNED(PyObject *op) { |
| 210 |
#ifdef Py_GIL_DISABLED |
| 211 |
return _Py_atomic_load_uint8_relaxed(&_PyASCIIObject_CAST(op)->state.interned); |
| 212 |
#else |
| 213 |
return _PyASCIIObject_CAST(op)->state.interned; |
| 214 |
#endif |
| 215 |
} |
| 216 |
#define PyUnicode_CHECK_INTERNED(op) PyUnicode_CHECK_INTERNED(_PyObject_CAST(op)) |
| 217 |
|
| 218 |
/* For backward compatibility. Soft-deprecated. */ |
| 219 |
static inline unsigned int PyUnicode_IS_READY(PyObject* Py_UNUSED(op)) { |
| 220 |
return 1; |
| 221 |
} |
| 222 |
#define PyUnicode_IS_READY(op) PyUnicode_IS_READY(_PyObject_CAST(op)) |
| 223 |
|
| 224 |
/* Return true if the string contains only ASCII characters, or 0 if not. The |
| 225 |
string may be compact (PyUnicode_IS_COMPACT_ASCII) or not. */ |
| 226 |
static inline unsigned int PyUnicode_IS_ASCII(PyObject *op) { |
| 227 |
return _PyASCIIObject_CAST(op)->state.ascii; |
| 228 |
} |
| 229 |
#define PyUnicode_IS_ASCII(op) PyUnicode_IS_ASCII(_PyObject_CAST(op)) |
| 230 |
|
| 231 |
/* Return true if the string is compact or 0 if not. |
| 232 |
No type checks are performed. */ |
| 233 |
static inline unsigned int PyUnicode_IS_COMPACT(PyObject *op) { |
| 234 |
return _PyASCIIObject_CAST(op)->state.compact; |
| 235 |
} |
| 236 |
#define PyUnicode_IS_COMPACT(op) PyUnicode_IS_COMPACT(_PyObject_CAST(op)) |
| 237 |
|
| 238 |
/* Return true if the string is a compact ASCII string (use PyASCIIObject |
| 239 |
structure), or 0 if not. No type checks are performed. */ |
| 240 |
static inline int PyUnicode_IS_COMPACT_ASCII(PyObject *op) { |
| 241 |
return (_PyASCIIObject_CAST(op)->state.ascii && PyUnicode_IS_COMPACT(op)); |
| 242 |
} |
| 243 |
#define PyUnicode_IS_COMPACT_ASCII(op) PyUnicode_IS_COMPACT_ASCII(_PyObject_CAST(op)) |
| 244 |
|
| 245 |
enum PyUnicode_Kind { |
| 246 |
/* Return values of the PyUnicode_KIND() function: */ |
| 247 |
PyUnicode_1BYTE_KIND = 1, |
| 248 |
PyUnicode_2BYTE_KIND = 2, |
| 249 |
PyUnicode_4BYTE_KIND = 4 |
| 250 |
}; |
| 251 |
|
| 252 |
PyAPI_FUNC(int) PyUnicode_KIND(PyObject *op); |
| 253 |
|
| 254 |
// PyUnicode_KIND(): Return one of the PyUnicode_*_KIND values defined above. |
| 255 |
// |
| 256 |
// gh-89653: Converting this macro to a static inline function would introduce |
| 257 |
// new compiler warnings on "kind < PyUnicode_KIND(str)" (compare signed and |
| 258 |
// unsigned numbers) where kind type is an int or on |
| 259 |
// "unsigned int kind = PyUnicode_KIND(str)" (cast signed to unsigned). |
| 260 |
#define PyUnicode_KIND(op) _Py_RVALUE(_PyASCIIObject_CAST(op)->state.kind) |
| 261 |
|
| 262 |
/* Return a void pointer to the raw unicode buffer. */ |
| 263 |
static inline void* _PyUnicode_COMPACT_DATA(PyObject *op) { |
| 264 |
if (PyUnicode_IS_ASCII(op)) { |
| 265 |
return _Py_STATIC_CAST(void*, (_PyASCIIObject_CAST(op) + 1)); |
| 266 |
} |
| 267 |
return _Py_STATIC_CAST(void*, (_PyCompactUnicodeObject_CAST(op) + 1)); |
| 268 |
} |
| 269 |
|
| 270 |
static inline void* _PyUnicode_NONCOMPACT_DATA(PyObject *op) { |
| 271 |
void *data; |
| 272 |
assert(!PyUnicode_IS_COMPACT(op)); |
| 273 |
data = _PyUnicodeObject_CAST(op)->data.any; |
| 274 |
assert(data != NULL); |
| 275 |
return data; |
| 276 |
} |
| 277 |
|
| 278 |
PyAPI_FUNC(void*) PyUnicode_DATA(PyObject *op); |
| 279 |
|
| 280 |
static inline void* _PyUnicode_DATA(PyObject *op) { |
| 281 |
if (PyUnicode_IS_COMPACT(op)) { |
| 282 |
return _PyUnicode_COMPACT_DATA(op); |
| 283 |
} |
| 284 |
return _PyUnicode_NONCOMPACT_DATA(op); |
| 285 |
} |
| 286 |
#define PyUnicode_DATA(op) _PyUnicode_DATA(_PyObject_CAST(op)) |
| 287 |
|
| 288 |
/* Return pointers to the canonical representation cast to unsigned char, |
| 289 |
Py_UCS2, or Py_UCS4 for direct character access. |
| 290 |
No checks are performed, use PyUnicode_KIND() before to ensure |
| 291 |
these will work correctly. */ |
| 292 |
|
| 293 |
#define PyUnicode_1BYTE_DATA(op) _Py_STATIC_CAST(Py_UCS1*, PyUnicode_DATA(op)) |
| 294 |
#define PyUnicode_2BYTE_DATA(op) _Py_STATIC_CAST(Py_UCS2*, PyUnicode_DATA(op)) |
| 295 |
#define PyUnicode_4BYTE_DATA(op) _Py_STATIC_CAST(Py_UCS4*, PyUnicode_DATA(op)) |
| 296 |
|
| 297 |
/* Returns the length of the unicode string. */ |
| 298 |
static inline Py_ssize_t PyUnicode_GET_LENGTH(PyObject *op) { |
| 299 |
return _PyASCIIObject_CAST(op)->length; |
| 300 |
} |
| 301 |
#define PyUnicode_GET_LENGTH(op) PyUnicode_GET_LENGTH(_PyObject_CAST(op)) |
| 302 |
|
| 303 |
/* Write into the canonical representation, this function does not do any sanity |
| 304 |
checks and is intended for usage in loops. The caller should cache the |
| 305 |
kind and data pointers obtained from other function calls. |
| 306 |
index is the index in the string (starts at 0) and value is the new |
| 307 |
code point value which should be written to that location. */ |
| 308 |
static inline void PyUnicode_WRITE(int kind, void *data, |
| 309 |
Py_ssize_t index, Py_UCS4 value) |
| 310 |
{ |
| 311 |
assert(index >= 0); |
| 312 |
if (kind == PyUnicode_1BYTE_KIND) { |
| 313 |
assert(value <= 0xffU); |
| 314 |
_Py_STATIC_CAST(Py_UCS1*, data)[index] = _Py_STATIC_CAST(Py_UCS1, value); |
| 315 |
} |
| 316 |
else if (kind == PyUnicode_2BYTE_KIND) { |
| 317 |
assert(value <= 0xffffU); |
| 318 |
_Py_STATIC_CAST(Py_UCS2*, data)[index] = _Py_STATIC_CAST(Py_UCS2, value); |
| 319 |
} |
| 320 |
else { |
| 321 |
assert(kind == PyUnicode_4BYTE_KIND); |
| 322 |
assert(value <= 0x10ffffU); |
| 323 |
_Py_STATIC_CAST(Py_UCS4*, data)[index] = value; |
| 324 |
} |
| 325 |
} |
| 326 |
#define PyUnicode_WRITE(kind, data, index, value) \ |
| 327 |
PyUnicode_WRITE(_Py_STATIC_CAST(int, kind), _Py_CAST(void*, data), \ |
| 328 |
(index), _Py_STATIC_CAST(Py_UCS4, value)) |
| 329 |
|
| 330 |
/* Read a code point from the string's canonical representation. No checks |
| 331 |
are performed. */ |
| 332 |
static inline Py_UCS4 PyUnicode_READ(int kind, |
| 333 |
const void *data, Py_ssize_t index) |
| 334 |
{ |
| 335 |
assert(index >= 0); |
| 336 |
if (kind == PyUnicode_1BYTE_KIND) { |
| 337 |
return _Py_STATIC_CAST(const Py_UCS1*, data)[index]; |
| 338 |
} |
| 339 |
if (kind == PyUnicode_2BYTE_KIND) { |
| 340 |
return _Py_STATIC_CAST(const Py_UCS2*, data)[index]; |
| 341 |
} |
| 342 |
assert(kind == PyUnicode_4BYTE_KIND); |
| 343 |
return _Py_STATIC_CAST(const Py_UCS4*, data)[index]; |
| 344 |
} |
| 345 |
#define PyUnicode_READ(kind, data, index) \ |
| 346 |
PyUnicode_READ(_Py_STATIC_CAST(int, kind), \ |
| 347 |
_Py_STATIC_CAST(const void*, data), \ |
| 348 |
(index)) |
| 349 |
|
| 350 |
/* PyUnicode_READ_CHAR() is less efficient than PyUnicode_READ() because it |
| 351 |
calls PyUnicode_KIND() and might call it twice. For single reads, use |
| 352 |
PyUnicode_READ_CHAR, for multiple consecutive reads callers should |
| 353 |
cache kind and use PyUnicode_READ instead. */ |
| 354 |
static inline Py_UCS4 PyUnicode_READ_CHAR(PyObject *unicode, Py_ssize_t index) |
| 355 |
{ |
| 356 |
int kind; |
| 357 |
|
| 358 |
assert(index >= 0); |
| 359 |
// Tolerate reading the NUL character at str[len(str)] |
| 360 |
assert(index <= PyUnicode_GET_LENGTH(unicode)); |
| 361 |
|
| 362 |
kind = PyUnicode_KIND(unicode); |
| 363 |
if (kind == PyUnicode_1BYTE_KIND) { |
| 364 |
return PyUnicode_1BYTE_DATA(unicode)[index]; |
| 365 |
} |
| 366 |
if (kind == PyUnicode_2BYTE_KIND) { |
| 367 |
return PyUnicode_2BYTE_DATA(unicode)[index]; |
| 368 |
} |
| 369 |
assert(kind == PyUnicode_4BYTE_KIND); |
| 370 |
return PyUnicode_4BYTE_DATA(unicode)[index]; |
| 371 |
} |
| 372 |
#define PyUnicode_READ_CHAR(unicode, index) \ |
| 373 |
PyUnicode_READ_CHAR(_PyObject_CAST(unicode), (index)) |
| 374 |
|
| 375 |
/* Return a maximum character value which is suitable for creating another |
| 376 |
string based on op. This is always an approximation but more efficient |
| 377 |
than iterating over the string. */ |
| 378 |
static inline Py_UCS4 PyUnicode_MAX_CHAR_VALUE(PyObject *op) |
| 379 |
{ |
| 380 |
int kind; |
| 381 |
|
| 382 |
if (PyUnicode_IS_ASCII(op)) { |
| 383 |
return 0x7fU; |
| 384 |
} |
| 385 |
|
| 386 |
kind = PyUnicode_KIND(op); |
| 387 |
if (kind == PyUnicode_1BYTE_KIND) { |
| 388 |
return 0xffU; |
| 389 |
} |
| 390 |
if (kind == PyUnicode_2BYTE_KIND) { |
| 391 |
return 0xffffU; |
| 392 |
} |
| 393 |
assert(kind == PyUnicode_4BYTE_KIND); |
| 394 |
return 0x10ffffU; |
| 395 |
} |
| 396 |
#define PyUnicode_MAX_CHAR_VALUE(op) \ |
| 397 |
PyUnicode_MAX_CHAR_VALUE(_PyObject_CAST(op)) |
| 398 |
|
| 399 |
|
| 400 |
/* === Public API ========================================================= */ |
| 401 |
|
| 402 |
/* With PEP 393, this is the recommended way to allocate a new unicode object. |
| 403 |
This function will allocate the object and its buffer in a single memory |
| 404 |
block. Objects created using this function are not resizable. */ |
| 405 |
PyAPI_FUNC(PyObject*) PyUnicode_New( |
| 406 |
Py_ssize_t size, /* Number of code points in the new string */ |
| 407 |
Py_UCS4 maxchar /* maximum code point value in the string */ |
| 408 |
); |
| 409 |
|
| 410 |
/* For backward compatibility. Soft-deprecated. */ |
| 411 |
static inline int PyUnicode_READY(PyObject* Py_UNUSED(op)) |
| 412 |
{ |
| 413 |
return 0; |
| 414 |
} |
| 415 |
#define PyUnicode_READY(op) PyUnicode_READY(_PyObject_CAST(op)) |
| 416 |
|
| 417 |
/* Copy character from one unicode object into another, this function performs |
| 418 |
character conversion when necessary and falls back to memcpy() if possible. |
| 419 |
|
| 420 |
Fail if to is too small (smaller than *how_many* or smaller than |
| 421 |
len(from)-from_start), or if kind(from[from_start:from_start+how_many]) > |
| 422 |
kind(to), or if *to* has more than 1 reference. |
| 423 |
|
| 424 |
Return the number of written character, or return -1 and raise an exception |
| 425 |
on error. |
| 426 |
|
| 427 |
Pseudo-code: |
| 428 |
|
| 429 |
how_many = min(how_many, len(from) - from_start) |
| 430 |
to[to_start:to_start+how_many] = from[from_start:from_start+how_many] |
| 431 |
return how_many |
| 432 |
|
| 433 |
Note: The function doesn't write a terminating null character. |
| 434 |
*/ |
| 435 |
PyAPI_FUNC(Py_ssize_t) PyUnicode_CopyCharacters( |
| 436 |
PyObject *to, |
| 437 |
Py_ssize_t to_start, |
| 438 |
PyObject *from, |
| 439 |
Py_ssize_t from_start, |
| 440 |
Py_ssize_t how_many |
| 441 |
); |
| 442 |
|
| 443 |
/* Fill a string with a character: write fill_char into |
| 444 |
unicode[start:start+length]. |
| 445 |
|
| 446 |
Fail if fill_char is bigger than the string maximum character, or if the |
| 447 |
string has more than 1 reference. |
| 448 |
|
| 449 |
Return the number of written character, or return -1 and raise an exception |
| 450 |
on error. */ |
| 451 |
PyAPI_FUNC(Py_ssize_t) PyUnicode_Fill( |
| 452 |
PyObject *unicode, |
| 453 |
Py_ssize_t start, |
| 454 |
Py_ssize_t length, |
| 455 |
Py_UCS4 fill_char |
| 456 |
); |
| 457 |
|
| 458 |
/* Create a new string from a buffer of Py_UCS1, Py_UCS2 or Py_UCS4 characters. |
| 459 |
Scan the string to find the maximum character. */ |
| 460 |
PyAPI_FUNC(PyObject*) PyUnicode_FromKindAndData( |
| 461 |
int kind, |
| 462 |
const void *buffer, |
| 463 |
Py_ssize_t size); |
| 464 |
|
| 465 |
|
| 466 |
/* --- Public PyUnicodeWriter API ----------------------------------------- */ |
| 467 |
|
| 468 |
typedef struct PyUnicodeWriter PyUnicodeWriter; |
| 469 |
|
| 470 |
PyAPI_FUNC(PyUnicodeWriter*) PyUnicodeWriter_Create(Py_ssize_t length); |
| 471 |
PyAPI_FUNC(void) PyUnicodeWriter_Discard(PyUnicodeWriter *writer); |
| 472 |
PyAPI_FUNC(PyObject*) PyUnicodeWriter_Finish(PyUnicodeWriter *writer); |
| 473 |
|
| 474 |
PyAPI_FUNC(int) PyUnicodeWriter_WriteChar( |
| 475 |
PyUnicodeWriter *writer, |
| 476 |
Py_UCS4 ch); |
| 477 |
PyAPI_FUNC(int) PyUnicodeWriter_WriteUTF8( |
| 478 |
PyUnicodeWriter *writer, |
| 479 |
const char *str, |
| 480 |
Py_ssize_t size); |
| 481 |
PyAPI_FUNC(int) PyUnicodeWriter_WriteASCII( |
| 482 |
PyUnicodeWriter *writer, |
| 483 |
const char *str, |
| 484 |
Py_ssize_t size); |
| 485 |
PyAPI_FUNC(int) PyUnicodeWriter_WriteWideChar( |
| 486 |
PyUnicodeWriter *writer, |
| 487 |
const wchar_t *str, |
| 488 |
Py_ssize_t size); |
| 489 |
PyAPI_FUNC(int) PyUnicodeWriter_WriteUCS4( |
| 490 |
PyUnicodeWriter *writer, |
| 491 |
Py_UCS4 *str, |
| 492 |
Py_ssize_t size); |
| 493 |
|
| 494 |
PyAPI_FUNC(int) PyUnicodeWriter_WriteStr( |
| 495 |
PyUnicodeWriter *writer, |
| 496 |
PyObject *obj); |
| 497 |
PyAPI_FUNC(int) PyUnicodeWriter_WriteRepr( |
| 498 |
PyUnicodeWriter *writer, |
| 499 |
PyObject *obj); |
| 500 |
PyAPI_FUNC(int) PyUnicodeWriter_WriteSubstring( |
| 501 |
PyUnicodeWriter *writer, |
| 502 |
PyObject *str, |
| 503 |
Py_ssize_t start, |
| 504 |
Py_ssize_t end); |
| 505 |
PyAPI_FUNC(int) PyUnicodeWriter_Format( |
| 506 |
PyUnicodeWriter *writer, |
| 507 |
const char *format, |
| 508 |
...); |
| 509 |
PyAPI_FUNC(int) PyUnicodeWriter_DecodeUTF8Stateful( |
| 510 |
PyUnicodeWriter *writer, |
| 511 |
const char *string, /* UTF-8 encoded string */ |
| 512 |
Py_ssize_t length, /* size of string */ |
| 513 |
const char *errors, /* error handling */ |
| 514 |
Py_ssize_t *consumed); /* bytes consumed */ |
| 515 |
|
| 516 |
|
| 517 |
/* --- Private _PyUnicodeWriter API --------------------------------------- */ |
| 518 |
|
| 519 |
typedef struct { |
| 520 |
PyObject *buffer; |
| 521 |
void *data; |
| 522 |
int kind; |
| 523 |
Py_UCS4 maxchar; |
| 524 |
Py_ssize_t size; |
| 525 |
Py_ssize_t pos; |
| 526 |
|
| 527 |
/* minimum number of allocated characters (default: 0) */ |
| 528 |
Py_ssize_t min_length; |
| 529 |
|
| 530 |
/* minimum character (default: 127, ASCII) */ |
| 531 |
Py_UCS4 min_char; |
| 532 |
|
| 533 |
/* If non-zero, overallocate the buffer (default: 0). */ |
| 534 |
unsigned char overallocate; |
| 535 |
|
| 536 |
/* If readonly is 1, buffer is a shared string (cannot be modified) |
| 537 |
and size is set to 0. */ |
| 538 |
unsigned char readonly; |
| 539 |
} _PyUnicodeWriter; |
| 540 |
|
| 541 |
// Initialize a Unicode writer. |
| 542 |
// |
| 543 |
// By default, the minimum buffer size is 0 character and overallocation is |
| 544 |
// disabled. Set min_length, min_char and overallocate attributes to control |
| 545 |
// the allocation of the buffer. |
| 546 |
_Py_DEPRECATED_EXTERNALLY(3.14) PyAPI_FUNC(void) _PyUnicodeWriter_Init( |
| 547 |
_PyUnicodeWriter *writer); |
| 548 |
|
| 549 |
/* Prepare the buffer to write 'length' characters |
| 550 |
with the specified maximum character. |
| 551 |
|
| 552 |
Return 0 on success, raise an exception and return -1 on error. */ |
| 553 |
#define _PyUnicodeWriter_Prepare(WRITER, LENGTH, MAXCHAR) \ |
| 554 |
(((MAXCHAR) <= (WRITER)->maxchar \ |
| 555 |
&& (LENGTH) <= (WRITER)->size - (WRITER)->pos) \ |
| 556 |
? 0 \ |
| 557 |
: (((LENGTH) == 0) \ |
| 558 |
? 0 \ |
| 559 |
: _PyUnicodeWriter_PrepareInternal((WRITER), (LENGTH), (MAXCHAR)))) |
| 560 |
|
| 561 |
/* Don't call this function directly, use the _PyUnicodeWriter_Prepare() macro |
| 562 |
instead. */ |
| 563 |
_Py_DEPRECATED_EXTERNALLY(3.14) PyAPI_FUNC(int) _PyUnicodeWriter_PrepareInternal( |
| 564 |
_PyUnicodeWriter *writer, |
| 565 |
Py_ssize_t length, |
| 566 |
Py_UCS4 maxchar); |
| 567 |
|
| 568 |
/* Prepare the buffer to have at least the kind KIND. |
| 569 |
For example, kind=PyUnicode_2BYTE_KIND ensures that the writer will |
| 570 |
support characters in range U+000-U+FFFF. |
| 571 |
|
| 572 |
Return 0 on success, raise an exception and return -1 on error. */ |
| 573 |
#define _PyUnicodeWriter_PrepareKind(WRITER, KIND) \ |
| 574 |
((KIND) <= (WRITER)->kind \ |
| 575 |
? 0 \ |
| 576 |
: _PyUnicodeWriter_PrepareKindInternal((WRITER), (KIND))) |
| 577 |
|
| 578 |
/* Don't call this function directly, use the _PyUnicodeWriter_PrepareKind() |
| 579 |
macro instead. */ |
| 580 |
_Py_DEPRECATED_EXTERNALLY(3.14) PyAPI_FUNC(int) _PyUnicodeWriter_PrepareKindInternal( |
| 581 |
_PyUnicodeWriter *writer, |
| 582 |
int kind); |
| 583 |
|
| 584 |
/* Append a Unicode character. |
| 585 |
Return 0 on success, raise an exception and return -1 on error. */ |
| 586 |
_Py_DEPRECATED_EXTERNALLY(3.14) PyAPI_FUNC(int) _PyUnicodeWriter_WriteChar( |
| 587 |
_PyUnicodeWriter *writer, |
| 588 |
Py_UCS4 ch); |
| 589 |
|
| 590 |
/* Append a Unicode string. |
| 591 |
Return 0 on success, raise an exception and return -1 on error. */ |
| 592 |
_Py_DEPRECATED_EXTERNALLY(3.14) PyAPI_FUNC(int) _PyUnicodeWriter_WriteStr( |
| 593 |
_PyUnicodeWriter *writer, |
| 594 |
PyObject *str); /* Unicode string */ |
| 595 |
|
| 596 |
/* Append a substring of a Unicode string. |
| 597 |
Return 0 on success, raise an exception and return -1 on error. */ |
| 598 |
_Py_DEPRECATED_EXTERNALLY(3.14) PyAPI_FUNC(int) _PyUnicodeWriter_WriteSubstring( |
| 599 |
_PyUnicodeWriter *writer, |
| 600 |
PyObject *str, /* Unicode string */ |
| 601 |
Py_ssize_t start, |
| 602 |
Py_ssize_t end); |
| 603 |
|
| 604 |
/* Append an ASCII-encoded byte string. |
| 605 |
Return 0 on success, raise an exception and return -1 on error. */ |
| 606 |
_Py_DEPRECATED_EXTERNALLY(3.14) PyAPI_FUNC(int) _PyUnicodeWriter_WriteASCIIString( |
| 607 |
_PyUnicodeWriter *writer, |
| 608 |
const char *str, /* ASCII-encoded byte string */ |
| 609 |
Py_ssize_t len); /* number of bytes, or -1 if unknown */ |
| 610 |
|
| 611 |
/* Append a latin1-encoded byte string. |
| 612 |
Return 0 on success, raise an exception and return -1 on error. */ |
| 613 |
_Py_DEPRECATED_EXTERNALLY(3.14) PyAPI_FUNC(int) _PyUnicodeWriter_WriteLatin1String( |
| 614 |
_PyUnicodeWriter *writer, |
| 615 |
const char *str, /* latin1-encoded byte string */ |
| 616 |
Py_ssize_t len); /* length in bytes */ |
| 617 |
|
| 618 |
/* Get the value of the writer as a Unicode string. Clear the |
| 619 |
buffer of the writer. Raise an exception and return NULL |
| 620 |
on error. */ |
| 621 |
_Py_DEPRECATED_EXTERNALLY(3.14) PyAPI_FUNC(PyObject *) _PyUnicodeWriter_Finish( |
| 622 |
_PyUnicodeWriter *writer); |
| 623 |
|
| 624 |
/* Deallocate memory of a writer (clear its internal buffer). */ |
| 625 |
_Py_DEPRECATED_EXTERNALLY(3.14) PyAPI_FUNC(void) _PyUnicodeWriter_Dealloc( |
| 626 |
_PyUnicodeWriter *writer); |
| 627 |
|
| 628 |
|
| 629 |
/* --- Manage the default encoding ---------------------------------------- */ |
| 630 |
|
| 631 |
/* Returns a pointer to the default encoding (UTF-8) of the |
| 632 |
Unicode object unicode. |
| 633 |
|
| 634 |
Like PyUnicode_AsUTF8AndSize(), this also caches the UTF-8 representation |
| 635 |
in the unicodeobject. |
| 636 |
|
| 637 |
_PyUnicode_AsString is a #define for PyUnicode_AsUTF8 to |
| 638 |
support the previous internal function with the same behaviour. |
| 639 |
|
| 640 |
Use of this API is DEPRECATED since no size information can be |
| 641 |
extracted from the returned data. |
| 642 |
*/ |
| 643 |
|
| 644 |
PyAPI_FUNC(const char *) PyUnicode_AsUTF8(PyObject *unicode); |
| 645 |
|
| 646 |
// Deprecated alias kept for backward compatibility |
| 647 |
Py_DEPRECATED(3.14) static inline const char* |
| 648 |
_PyUnicode_AsString(PyObject *unicode) |
| 649 |
{ |
| 650 |
return PyUnicode_AsUTF8(unicode); |
| 651 |
} |
| 652 |
|
| 653 |
|
| 654 |
/* === Characters Type APIs =============================================== */ |
| 655 |
|
| 656 |
/* These should not be used directly. Use the Py_UNICODE_IS* and |
| 657 |
Py_UNICODE_TO* macros instead. |
| 658 |
|
| 659 |
These APIs are implemented in Objects/unicodectype.c. |
| 660 |
|
| 661 |
*/ |
| 662 |
|
| 663 |
PyAPI_FUNC(int) _PyUnicode_IsLowercase( |
| 664 |
Py_UCS4 ch /* Unicode character */ |
| 665 |
); |
| 666 |
|
| 667 |
PyAPI_FUNC(int) _PyUnicode_IsUppercase( |
| 668 |
Py_UCS4 ch /* Unicode character */ |
| 669 |
); |
| 670 |
|
| 671 |
PyAPI_FUNC(int) _PyUnicode_IsTitlecase( |
| 672 |
Py_UCS4 ch /* Unicode character */ |
| 673 |
); |
| 674 |
|
| 675 |
PyAPI_FUNC(int) _PyUnicode_IsWhitespace( |
| 676 |
const Py_UCS4 ch /* Unicode character */ |
| 677 |
); |
| 678 |
|
| 679 |
PyAPI_FUNC(int) _PyUnicode_IsLinebreak( |
| 680 |
const Py_UCS4 ch /* Unicode character */ |
| 681 |
); |
| 682 |
|
| 683 |
PyAPI_FUNC(Py_UCS4) _PyUnicode_ToLowercase( |
| 684 |
Py_UCS4 ch /* Unicode character */ |
| 685 |
); |
| 686 |
|
| 687 |
PyAPI_FUNC(Py_UCS4) _PyUnicode_ToUppercase( |
| 688 |
Py_UCS4 ch /* Unicode character */ |
| 689 |
); |
| 690 |
|
| 691 |
PyAPI_FUNC(Py_UCS4) _PyUnicode_ToTitlecase( |
| 692 |
Py_UCS4 ch /* Unicode character */ |
| 693 |
); |
| 694 |
|
| 695 |
PyAPI_FUNC(int) _PyUnicode_ToDecimalDigit( |
| 696 |
Py_UCS4 ch /* Unicode character */ |
| 697 |
); |
| 698 |
|
| 699 |
PyAPI_FUNC(int) _PyUnicode_ToDigit( |
| 700 |
Py_UCS4 ch /* Unicode character */ |
| 701 |
); |
| 702 |
|
| 703 |
PyAPI_FUNC(double) _PyUnicode_ToNumeric( |
| 704 |
Py_UCS4 ch /* Unicode character */ |
| 705 |
); |
| 706 |
|
| 707 |
PyAPI_FUNC(int) _PyUnicode_IsDecimalDigit( |
| 708 |
Py_UCS4 ch /* Unicode character */ |
| 709 |
); |
| 710 |
|
| 711 |
PyAPI_FUNC(int) _PyUnicode_IsDigit( |
| 712 |
Py_UCS4 ch /* Unicode character */ |
| 713 |
); |
| 714 |
|
| 715 |
PyAPI_FUNC(int) _PyUnicode_IsNumeric( |
| 716 |
Py_UCS4 ch /* Unicode character */ |
| 717 |
); |
| 718 |
|
| 719 |
PyAPI_FUNC(int) _PyUnicode_IsPrintable( |
| 720 |
Py_UCS4 ch /* Unicode character */ |
| 721 |
); |
| 722 |
|
| 723 |
PyAPI_FUNC(int) _PyUnicode_IsAlpha( |
| 724 |
Py_UCS4 ch /* Unicode character */ |
| 725 |
); |
| 726 |
|
| 727 |
// Helper array used by Py_UNICODE_ISSPACE(). |
| 728 |
PyAPI_DATA(const unsigned char) _Py_ascii_whitespace[]; |
| 729 |
|
| 730 |
// Since splitting on whitespace is an important use case, and |
| 731 |
// whitespace in most situations is solely ASCII whitespace, we |
| 732 |
// optimize for the common case by using a quick look-up table |
| 733 |
// _Py_ascii_whitespace (see below) with an inlined check. |
| 734 |
static inline int Py_UNICODE_ISSPACE(Py_UCS4 ch) { |
| 735 |
if (ch < 128) { |
| 736 |
return _Py_ascii_whitespace[ch]; |
| 737 |
} |
| 738 |
return _PyUnicode_IsWhitespace(ch); |
| 739 |
} |
| 740 |
|
| 741 |
#define Py_UNICODE_ISLOWER(ch) _PyUnicode_IsLowercase(ch) |
| 742 |
#define Py_UNICODE_ISUPPER(ch) _PyUnicode_IsUppercase(ch) |
| 743 |
#define Py_UNICODE_ISTITLE(ch) _PyUnicode_IsTitlecase(ch) |
| 744 |
#define Py_UNICODE_ISLINEBREAK(ch) _PyUnicode_IsLinebreak(ch) |
| 745 |
|
| 746 |
#define Py_UNICODE_TOLOWER(ch) _PyUnicode_ToLowercase(ch) |
| 747 |
#define Py_UNICODE_TOUPPER(ch) _PyUnicode_ToUppercase(ch) |
| 748 |
#define Py_UNICODE_TOTITLE(ch) _PyUnicode_ToTitlecase(ch) |
| 749 |
|
| 750 |
#define Py_UNICODE_ISDECIMAL(ch) _PyUnicode_IsDecimalDigit(ch) |
| 751 |
#define Py_UNICODE_ISDIGIT(ch) _PyUnicode_IsDigit(ch) |
| 752 |
#define Py_UNICODE_ISNUMERIC(ch) _PyUnicode_IsNumeric(ch) |
| 753 |
#define Py_UNICODE_ISPRINTABLE(ch) _PyUnicode_IsPrintable(ch) |
| 754 |
|
| 755 |
#define Py_UNICODE_TODECIMAL(ch) _PyUnicode_ToDecimalDigit(ch) |
| 756 |
#define Py_UNICODE_TODIGIT(ch) _PyUnicode_ToDigit(ch) |
| 757 |
#define Py_UNICODE_TONUMERIC(ch) _PyUnicode_ToNumeric(ch) |
| 758 |
|
| 759 |
#define Py_UNICODE_ISALPHA(ch) _PyUnicode_IsAlpha(ch) |
| 760 |
|
| 761 |
static inline int Py_UNICODE_ISALNUM(Py_UCS4 ch) { |
| 762 |
return (Py_UNICODE_ISALPHA(ch) |
| 763 |
|| Py_UNICODE_ISDECIMAL(ch) |
| 764 |
|| Py_UNICODE_ISDIGIT(ch) |
| 765 |
|| Py_UNICODE_ISNUMERIC(ch)); |
| 766 |
} |
| 767 |
|
| 768 |
|
| 769 |
/* === Misc functions ===================================================== */ |
| 770 |
|
| 771 |
// Return an interned Unicode object for an Identifier; may fail if there is no |
| 772 |
// memory. |
| 773 |
PyAPI_FUNC(PyObject*) _PyUnicode_FromId(_Py_Identifier*); |