Mercurial > genshi > mirror
annotate genshi/_speedups.c @ 724:919809e55d16 experimental-match-fastpaths
update to trunk to track r847, fixing python 2.4 compatibility issues in speedup (and fixing copyrights, apparently :))
| author | aflett |
|---|---|
| date | Mon, 21 Apr 2008 19:36:53 +0000 |
| parents | d143dd73789b |
| children | ea46fb523485 |
| rev | line source |
|---|---|
| 541 | 1 /* |
|
724
919809e55d16
update to trunk to track r847, fixing python 2.4 compatibility issues in speedup (and fixing copyrights, apparently :))
aflett
parents:
718
diff
changeset
|
2 * Copyright (C) 2006-2008 Edgewall Software |
| 541 | 3 * All rights reserved. |
| 4 * | |
| 5 * This software is licensed as described in the file COPYING, which | |
| 6 * you should have received as part of this distribution. The terms | |
| 7 * are also available at http://genshi.edgewall.org/wiki/License. | |
| 8 * | |
| 9 * This software consists of voluntary contributions made by many | |
| 10 * individuals. For the exact contribution history, see the revision | |
| 11 * history and logs, available at http://genshi.edgewall.org/log/. | |
| 12 */ | |
| 13 | |
| 14 #include <Python.h> | |
| 15 #include <structmember.h> | |
| 16 | |
|
724
919809e55d16
update to trunk to track r847, fixing python 2.4 compatibility issues in speedup (and fixing copyrights, apparently :))
aflett
parents:
718
diff
changeset
|
17 #if PY_VERSION_HEX < 0x02050000 && !defined(PY_SSIZE_T_MIN) |
|
919809e55d16
update to trunk to track r847, fixing python 2.4 compatibility issues in speedup (and fixing copyrights, apparently :))
aflett
parents:
718
diff
changeset
|
18 typedef int Py_ssize_t; |
|
919809e55d16
update to trunk to track r847, fixing python 2.4 compatibility issues in speedup (and fixing copyrights, apparently :))
aflett
parents:
718
diff
changeset
|
19 #define PY_SSIZE_T_MAX INT_MAX |
|
919809e55d16
update to trunk to track r847, fixing python 2.4 compatibility issues in speedup (and fixing copyrights, apparently :))
aflett
parents:
718
diff
changeset
|
20 #define PY_SSIZE_T_MIN INT_MIN |
|
919809e55d16
update to trunk to track r847, fixing python 2.4 compatibility issues in speedup (and fixing copyrights, apparently :))
aflett
parents:
718
diff
changeset
|
21 #endif |
|
919809e55d16
update to trunk to track r847, fixing python 2.4 compatibility issues in speedup (and fixing copyrights, apparently :))
aflett
parents:
718
diff
changeset
|
22 |
| 541 | 23 static PyObject *amp1, *amp2, *lt1, *lt2, *gt1, *gt2, *qt1, *qt2; |
| 24 static PyObject *stripentities, *striptags; | |
| 25 | |
| 26 static void | |
| 27 init_constants(void) | |
| 28 { | |
| 29 PyObject *util = PyImport_ImportModule("genshi.util"); | |
| 30 stripentities = PyObject_GetAttrString(util, "stripentities"); | |
| 31 striptags = PyObject_GetAttrString(util, "striptags"); | |
| 32 Py_DECREF(util); | |
| 33 | |
| 34 amp1 = PyUnicode_DecodeASCII("&", 1, NULL); | |
| 35 amp2 = PyUnicode_DecodeASCII("&", 5, NULL); | |
| 36 lt1 = PyUnicode_DecodeASCII("<", 1, NULL); | |
| 37 lt2 = PyUnicode_DecodeASCII("<", 4, NULL); | |
| 38 gt1 = PyUnicode_DecodeASCII(">", 1, NULL); | |
| 39 gt2 = PyUnicode_DecodeASCII(">", 4, NULL); | |
| 40 qt1 = PyUnicode_DecodeASCII("\"", 1, NULL); | |
| 41 qt2 = PyUnicode_DecodeASCII(""", 5, NULL); | |
| 42 } | |
| 43 | |
| 44 /* Markup class */ | |
| 45 | |
| 46 PyAPI_DATA(PyTypeObject) MarkupType; | |
| 47 | |
| 48 PyDoc_STRVAR(Markup__doc__, | |
| 49 "Marks a string as being safe for inclusion in HTML/XML output without\n\ | |
| 50 needing to be escaped."); | |
| 51 | |
| 52 static PyObject * | |
| 53 escape(PyObject *text, int quotes) | |
| 54 { | |
| 55 PyObject *args, *ret; | |
| 56 PyUnicodeObject *in, *out; | |
| 57 Py_UNICODE *inp, *outp; | |
| 58 int len, inn, outn; | |
| 59 | |
| 60 if (PyObject_TypeCheck(text, &MarkupType)) { | |
| 61 Py_INCREF(text); | |
| 62 return text; | |
| 63 } | |
| 64 in = (PyUnicodeObject *) PyObject_Unicode(text); | |
| 65 if (in == NULL) { | |
| 66 return NULL; | |
| 67 } | |
| 68 /* First we need to figure out how long the escaped string will be */ | |
| 69 len = inn = 0; | |
| 70 inp = in->str; | |
| 71 while (*(inp) || in->length > inp - in->str) { | |
| 72 switch (*inp++) { | |
| 73 case '&': len += 5; inn++; break; | |
| 74 case '"': len += quotes ? 5 : 1; inn += quotes ? 1 : 0; break; | |
| 75 case '<': | |
| 76 case '>': len += 4; inn++; break; | |
| 77 default: len++; | |
| 78 } | |
| 79 } | |
| 80 | |
| 81 /* Do we need to escape anything at all? */ | |
| 82 if (!inn) { | |
| 83 args = PyTuple_New(1); | |
| 84 if (args == NULL) { | |
| 85 Py_DECREF((PyObject *) in); | |
| 86 return NULL; | |
| 87 } | |
| 88 PyTuple_SET_ITEM(args, 0, (PyObject *) in); | |
| 89 ret = MarkupType.tp_new(&MarkupType, args, NULL); | |
| 90 Py_DECREF(args); | |
| 91 return ret; | |
| 92 } | |
| 93 | |
| 94 out = (PyUnicodeObject*) PyUnicode_FromUnicode(NULL, len); | |
| 95 if (out == NULL) { | |
|
665
3ee92ec99ad9
Applied patch to fix a memory leak in the C implementation of the `Markup.escape()` function. Thanks to Christian Boos for reporting and figuring out the problem. Closes #166.
cmlenz
parents:
547
diff
changeset
|
96 Py_DECREF((PyObject *) in); |
| 541 | 97 return NULL; |
| 98 } | |
| 99 | |
| 100 outn = 0; | |
| 101 inp = in->str; | |
| 102 outp = out->str; | |
| 103 while (*(inp) || in->length > inp - in->str) { | |
| 104 if (outn == inn) { | |
| 105 /* copy rest of string if we have already replaced everything */ | |
| 106 Py_UNICODE_COPY(outp, inp, in->length - (inp - in->str)); | |
| 107 break; | |
| 108 } | |
| 109 switch (*inp) { | |
| 110 case '&': | |
| 111 Py_UNICODE_COPY(outp, ((PyUnicodeObject *) amp2)->str, 5); | |
| 112 outp += 5; | |
| 113 outn++; | |
| 114 break; | |
| 115 case '"': | |
| 116 if (quotes) { | |
| 117 Py_UNICODE_COPY(outp, ((PyUnicodeObject *) qt2)->str, 5); | |
| 118 outp += 5; | |
| 119 outn++; | |
| 120 } else { | |
| 121 *outp++ = *inp; | |
| 122 } | |
| 123 break; | |
| 124 case '<': | |
| 125 Py_UNICODE_COPY(outp, ((PyUnicodeObject *) lt2)->str, 4); | |
| 126 outp += 4; | |
| 127 outn++; | |
| 128 break; | |
| 129 case '>': | |
| 130 Py_UNICODE_COPY(outp, ((PyUnicodeObject *) gt2)->str, 4); | |
| 131 outp += 4; | |
| 132 outn++; | |
| 133 break; | |
| 134 default: | |
| 135 *outp++ = *inp; | |
| 136 } | |
| 137 inp++; | |
| 138 } | |
| 139 | |
|
665
3ee92ec99ad9
Applied patch to fix a memory leak in the C implementation of the `Markup.escape()` function. Thanks to Christian Boos for reporting and figuring out the problem. Closes #166.
cmlenz
parents:
547
diff
changeset
|
140 Py_DECREF((PyObject *) in); |
|
3ee92ec99ad9
Applied patch to fix a memory leak in the C implementation of the `Markup.escape()` function. Thanks to Christian Boos for reporting and figuring out the problem. Closes #166.
cmlenz
parents:
547
diff
changeset
|
141 |
| 541 | 142 args = PyTuple_New(1); |
| 143 if (args == NULL) { | |
| 144 Py_DECREF((PyObject *) out); | |
| 145 return NULL; | |
| 146 } | |
| 147 PyTuple_SET_ITEM(args, 0, (PyObject *) out); | |
| 148 ret = MarkupType.tp_new(&MarkupType, args, NULL); | |
| 149 Py_DECREF(args); | |
| 150 return ret; | |
| 151 } | |
| 152 | |
| 153 PyDoc_STRVAR(escape__doc__, | |
| 154 "Create a Markup instance from a string and escape special characters\n\ | |
| 155 it may contain (<, >, & and \").\n\ | |
| 156 \n\ | |
| 547 | 157 >>> escape('\"1 < 2\"')\n\ |
| 158 <Markup u'"1 < 2"'>\n\ | |
| 159 \n\ | |
| 541 | 160 If the `quotes` parameter is set to `False`, the \" character is left\n\ |
| 161 as is. Escaping quotes is generally only required for strings that are\n\ | |
| 547 | 162 to be used in attribute values.\n\ |
| 163 \n\ | |
| 164 >>> escape('\"1 < 2\"', quotes=False)\n\ | |
| 165 <Markup u'\"1 < 2\"'>\n\ | |
| 166 \n\ | |
| 167 :param text: the text to escape\n\ | |
| 168 :param quotes: if ``True``, double quote characters are escaped in\n\ | |
| 169 addition to the other special characters\n\ | |
| 170 :return: the escaped `Markup` string\n\ | |
| 171 :rtype: `Markup`\n\ | |
| 172 "); | |
| 541 | 173 |
| 174 static PyObject * | |
| 175 Markup_escape(PyTypeObject* type, PyObject *args, PyObject *kwds) | |
| 176 { | |
| 177 static char *kwlist[] = {"text", "quotes", 0}; | |
| 178 PyObject *text = NULL; | |
| 179 char quotes = 1; | |
| 180 | |
| 181 if (!PyArg_ParseTupleAndKeywords(args, kwds, "O|b", kwlist, &text, "es)) { | |
| 182 return NULL; | |
| 183 } | |
| 184 if (PyObject_Not(text)) { | |
| 185 return type->tp_new(type, args, NULL); | |
| 186 } | |
| 187 if (PyObject_TypeCheck(text, type)) { | |
| 188 Py_INCREF(text); | |
| 189 return text; | |
| 190 } | |
| 191 return escape(text, quotes); | |
| 192 } | |
| 193 | |
| 547 | 194 PyDoc_STRVAR(join__doc__, |
| 195 "Return a `Markup` object which is the concatenation of the strings\n\ | |
| 196 in the given sequence, where this `Markup` object is the separator\n\ | |
| 197 between the joined elements.\n\ | |
| 198 \n\ | |
| 199 Any element in the sequence that is not a `Markup` instance is\n\ | |
| 200 automatically escaped.\n\ | |
| 201 \n\ | |
| 202 :param seq: the sequence of strings to join\n\ | |
| 203 :param escape_quotes: whether double quote characters in the elements\n\ | |
| 204 should be escaped\n\ | |
| 205 :return: the joined `Markup` object\n\ | |
| 206 :rtype: `Markup`\n\ | |
| 207 :see: `escape`\n\ | |
| 208 "); | |
| 209 | |
| 541 | 210 static PyObject * |
| 211 Markup_join(PyObject *self, PyObject *args, PyObject *kwds) | |
| 212 { | |
| 213 static char *kwlist[] = {"seq", "escape_quotes", 0}; | |
|
665
3ee92ec99ad9
Applied patch to fix a memory leak in the C implementation of the `Markup.escape()` function. Thanks to Christian Boos for reporting and figuring out the problem. Closes #166.
cmlenz
parents:
547
diff
changeset
|
214 PyObject *seq = NULL, *seq2, *tmp, *tmp2; |
| 541 | 215 char quotes = 1; |
| 216 int n, i; | |
| 217 | |
| 218 if (!PyArg_ParseTupleAndKeywords(args, kwds, "O|b", kwlist, &seq, "es)) { | |
| 219 return NULL; | |
| 220 } | |
| 221 if (!PySequence_Check(seq)) { | |
| 222 return NULL; | |
| 223 } | |
| 224 n = PySequence_Size(seq); | |
| 225 if (n < 0) { | |
| 226 return NULL; | |
| 227 } | |
| 228 seq2 = PyTuple_New(n); | |
| 229 if (seq2 == NULL) { | |
| 230 return NULL; | |
| 231 } | |
| 232 for (i = 0; i < n; i++) { | |
| 233 tmp = PySequence_GetItem(seq, i); | |
| 234 if (tmp == NULL) { | |
| 235 Py_DECREF(seq2); | |
| 236 return NULL; | |
| 237 } | |
|
665
3ee92ec99ad9
Applied patch to fix a memory leak in the C implementation of the `Markup.escape()` function. Thanks to Christian Boos for reporting and figuring out the problem. Closes #166.
cmlenz
parents:
547
diff
changeset
|
238 tmp2 = escape(tmp, quotes); |
|
3ee92ec99ad9
Applied patch to fix a memory leak in the C implementation of the `Markup.escape()` function. Thanks to Christian Boos for reporting and figuring out the problem. Closes #166.
cmlenz
parents:
547
diff
changeset
|
239 if (tmp2 == NULL) { |
| 541 | 240 Py_DECREF(seq2); |
| 241 return NULL; | |
| 242 } | |
|
665
3ee92ec99ad9
Applied patch to fix a memory leak in the C implementation of the `Markup.escape()` function. Thanks to Christian Boos for reporting and figuring out the problem. Closes #166.
cmlenz
parents:
547
diff
changeset
|
243 PyTuple_SET_ITEM(seq2, i, tmp2); |
|
3ee92ec99ad9
Applied patch to fix a memory leak in the C implementation of the `Markup.escape()` function. Thanks to Christian Boos for reporting and figuring out the problem. Closes #166.
cmlenz
parents:
547
diff
changeset
|
244 Py_DECREF(tmp); |
| 541 | 245 } |
| 246 tmp = PyUnicode_Join(self, seq2); | |
| 247 Py_DECREF(seq2); | |
| 248 if (tmp == NULL) | |
| 249 return NULL; | |
| 250 args = PyTuple_New(1); | |
| 251 if (args == NULL) { | |
| 252 Py_DECREF(tmp); | |
| 253 return NULL; | |
| 254 } | |
| 255 PyTuple_SET_ITEM(args, 0, tmp); | |
| 256 tmp = MarkupType.tp_new(&MarkupType, args, NULL); | |
| 257 Py_DECREF(args); | |
| 258 return tmp; | |
| 259 } | |
| 260 | |
| 261 static PyObject * | |
| 262 Markup_add(PyObject *self, PyObject *other) | |
| 263 { | |
| 264 PyObject *tmp, *tmp2, *args, *ret; | |
| 265 if (PyObject_TypeCheck(self, &MarkupType)) { | |
| 266 tmp = escape(other, 1); | |
| 267 if (tmp == NULL) | |
| 268 return NULL; | |
| 269 tmp2 = PyUnicode_Concat(self, tmp); | |
| 270 } else { // __radd__ | |
| 271 tmp = escape(self, 1); | |
| 272 if (tmp == NULL) | |
| 273 return NULL; | |
| 274 tmp2 = PyUnicode_Concat(tmp, other); | |
| 275 } | |
|
665
3ee92ec99ad9
Applied patch to fix a memory leak in the C implementation of the `Markup.escape()` function. Thanks to Christian Boos for reporting and figuring out the problem. Closes #166.
cmlenz
parents:
547
diff
changeset
|
276 Py_DECREF(tmp); |
|
3ee92ec99ad9
Applied patch to fix a memory leak in the C implementation of the `Markup.escape()` function. Thanks to Christian Boos for reporting and figuring out the problem. Closes #166.
cmlenz
parents:
547
diff
changeset
|
277 if (tmp2 == NULL) |
| 541 | 278 return NULL; |
| 279 args = PyTuple_New(1); | |
| 280 if (args == NULL) { | |
| 281 Py_DECREF(tmp2); | |
| 282 return NULL; | |
| 283 } | |
| 284 PyTuple_SET_ITEM(args, 0, tmp2); | |
| 285 ret = MarkupType.tp_new(&MarkupType, args, NULL); | |
| 286 Py_DECREF(args); | |
| 287 return ret; | |
| 288 } | |
| 289 | |
| 290 static PyObject * | |
| 291 Markup_mod(PyObject *self, PyObject *args) | |
| 292 { | |
| 293 PyObject *tmp, *tmp2, *ret, *args2; | |
| 718 | 294 int i, nargs = 0; |
| 295 PyObject *kwds = NULL; | |
| 541 | 296 |
| 718 | 297 if (PyDict_Check(args)) { |
| 298 kwds = args; | |
| 299 } | |
| 300 if (kwds && PyDict_Size(kwds)) { | |
| 301 PyObject *kwcopy, *key, *value; | |
|
724
919809e55d16
update to trunk to track r847, fixing python 2.4 compatibility issues in speedup (and fixing copyrights, apparently :))
aflett
parents:
718
diff
changeset
|
302 #if PY_VERSION_HEX >= 0x02050000 |
| 718 | 303 Py_ssize_t pos = 0; |
|
724
919809e55d16
update to trunk to track r847, fixing python 2.4 compatibility issues in speedup (and fixing copyrights, apparently :))
aflett
parents:
718
diff
changeset
|
304 #else |
|
919809e55d16
update to trunk to track r847, fixing python 2.4 compatibility issues in speedup (and fixing copyrights, apparently :))
aflett
parents:
718
diff
changeset
|
305 int pos = 0; |
|
919809e55d16
update to trunk to track r847, fixing python 2.4 compatibility issues in speedup (and fixing copyrights, apparently :))
aflett
parents:
718
diff
changeset
|
306 #endif |
| 718 | 307 |
| 308 kwcopy = PyDict_Copy( kwds ); | |
| 309 if (kwcopy == NULL) { | |
| 310 return NULL; | |
| 311 } | |
| 312 while (PyDict_Next(kwcopy, &pos, &key, &value)) { | |
| 313 tmp = escape(value, 1); | |
| 314 if (tmp == NULL) { | |
| 315 Py_DECREF(kwcopy); | |
| 316 return NULL; | |
| 317 } | |
| 318 if (PyDict_SetItem(kwcopy, key, tmp) < 0) { | |
| 319 Py_DECREF(tmp); | |
| 320 Py_DECREF(kwcopy); | |
| 321 return NULL; | |
| 322 } | |
| 323 } | |
| 324 tmp = PyUnicode_Format(self, kwcopy); | |
| 325 Py_DECREF(kwcopy); | |
| 326 if (tmp == NULL) { | |
| 327 return NULL; | |
| 328 } | |
| 329 } else if (PyTuple_Check(args)) { | |
| 541 | 330 nargs = PyTuple_GET_SIZE(args); |
| 331 args2 = PyTuple_New(nargs); | |
| 332 if (args2 == NULL) { | |
| 333 return NULL; | |
| 334 } | |
| 335 for (i = 0; i < nargs; i++) { | |
| 336 tmp = escape(PyTuple_GET_ITEM(args, i), 1); | |
| 337 if (tmp == NULL) { | |
| 338 Py_DECREF(args2); | |
| 339 return NULL; | |
| 340 } | |
| 341 PyTuple_SET_ITEM(args2, i, tmp); | |
| 342 } | |
| 343 tmp = PyUnicode_Format(self, args2); | |
| 344 Py_DECREF(args2); | |
| 345 if (tmp == NULL) { | |
| 346 return NULL; | |
| 347 } | |
| 348 } else { | |
| 349 tmp2 = escape(args, 1); | |
| 350 if (tmp2 == NULL) { | |
| 351 return NULL; | |
| 352 } | |
| 353 tmp = PyUnicode_Format(self, tmp2); | |
| 354 Py_DECREF(tmp2); | |
| 355 if (tmp == NULL) { | |
| 356 return NULL; | |
| 357 } | |
| 358 } | |
| 359 args = PyTuple_New(1); | |
| 360 if (args == NULL) { | |
| 361 Py_DECREF(tmp); | |
| 362 return NULL; | |
| 363 } | |
| 364 PyTuple_SET_ITEM(args, 0, tmp); | |
| 365 ret = PyUnicode_Type.tp_new(&MarkupType, args, NULL); | |
| 366 Py_DECREF(args); | |
| 367 return ret; | |
| 368 } | |
| 369 | |
| 370 static PyObject * | |
| 371 Markup_mul(PyObject *self, PyObject *num) | |
| 372 { | |
| 373 PyObject *unicode, *result, *args; | |
| 374 | |
| 375 if (PyObject_TypeCheck(self, &MarkupType)) { | |
| 376 unicode = PyObject_Unicode(self); | |
| 377 if (unicode == NULL) return NULL; | |
| 378 result = PyNumber_Multiply(unicode, num); | |
| 379 } else { // __rmul__ | |
| 380 unicode = PyObject_Unicode(num); | |
| 381 if (unicode == NULL) return NULL; | |
| 382 result = PyNumber_Multiply(unicode, self); | |
| 383 } | |
|
665
3ee92ec99ad9
Applied patch to fix a memory leak in the C implementation of the `Markup.escape()` function. Thanks to Christian Boos for reporting and figuring out the problem. Closes #166.
cmlenz
parents:
547
diff
changeset
|
384 Py_DECREF(unicode); |
| 541 | 385 |
| 386 if (result == NULL) return NULL; | |
| 387 args = PyTuple_New(1); | |
| 388 if (args == NULL) { | |
| 389 Py_DECREF(result); | |
| 390 return NULL; | |
| 391 } | |
| 392 PyTuple_SET_ITEM(args, 0, result); | |
| 393 result = PyUnicode_Type.tp_new(&MarkupType, args, NULL); | |
| 394 Py_DECREF(args); | |
| 395 | |
| 396 return result; | |
| 397 } | |
| 398 | |
| 399 static PyObject * | |
| 400 Markup_repr(PyObject *self) | |
| 401 { | |
| 402 PyObject *format, *result, *args; | |
| 403 | |
| 404 format = PyString_FromString("<Markup %r>"); | |
| 405 if (format == NULL) return NULL; | |
| 406 result = PyObject_Unicode(self); | |
|
665
3ee92ec99ad9
Applied patch to fix a memory leak in the C implementation of the `Markup.escape()` function. Thanks to Christian Boos for reporting and figuring out the problem. Closes #166.
cmlenz
parents:
547
diff
changeset
|
407 if (result == NULL) { |
|
3ee92ec99ad9
Applied patch to fix a memory leak in the C implementation of the `Markup.escape()` function. Thanks to Christian Boos for reporting and figuring out the problem. Closes #166.
cmlenz
parents:
547
diff
changeset
|
408 Py_DECREF(format); |
|
3ee92ec99ad9
Applied patch to fix a memory leak in the C implementation of the `Markup.escape()` function. Thanks to Christian Boos for reporting and figuring out the problem. Closes #166.
cmlenz
parents:
547
diff
changeset
|
409 return NULL; |
|
3ee92ec99ad9
Applied patch to fix a memory leak in the C implementation of the `Markup.escape()` function. Thanks to Christian Boos for reporting and figuring out the problem. Closes #166.
cmlenz
parents:
547
diff
changeset
|
410 } |
| 541 | 411 args = PyTuple_New(1); |
| 412 if (args == NULL) { | |
|
665
3ee92ec99ad9
Applied patch to fix a memory leak in the C implementation of the `Markup.escape()` function. Thanks to Christian Boos for reporting and figuring out the problem. Closes #166.
cmlenz
parents:
547
diff
changeset
|
413 Py_DECREF(format); |
| 541 | 414 Py_DECREF(result); |
| 415 return NULL; | |
| 416 } | |
| 417 PyTuple_SET_ITEM(args, 0, result); | |
| 418 result = PyString_Format(format, args); | |
|
681
3e7cd32c9411
Fix another memory leak in the C speedups code. Thanks to Erik Bray for finding this one and providing a patch. Closes #166 (again).
cmlenz
parents:
665
diff
changeset
|
419 Py_DECREF(format); |
| 541 | 420 Py_DECREF(args); |
| 421 return result; | |
| 422 } | |
| 423 | |
| 424 PyDoc_STRVAR(unescape__doc__, | |
| 547 | 425 "Reverse-escapes &, <, >, and \" and returns a `unicode` object.\n\ |
| 426 \n\ | |
| 427 >>> Markup('1 < 2').unescape()\n\ | |
| 428 u'1 < 2'\n\ | |
| 429 \n\ | |
| 430 :return: the unescaped string\n\ | |
| 431 :rtype: `unicode`\n\ | |
| 432 :see: `genshi.core.unescape`\n\ | |
| 433 "); | |
| 541 | 434 |
| 435 static PyObject * | |
| 436 Markup_unescape(PyObject* self) | |
| 437 { | |
| 438 PyObject *tmp, *tmp2; | |
| 439 | |
| 440 tmp = PyUnicode_Replace(self, qt2, qt1, -1); | |
| 441 if (tmp == NULL) return NULL; | |
| 442 tmp2 = PyUnicode_Replace(tmp, gt2, gt1, -1); | |
| 443 Py_DECREF(tmp); | |
| 444 if (tmp2 == NULL) return NULL; | |
| 445 tmp = PyUnicode_Replace(tmp2, lt2, lt1, -1); | |
| 446 Py_DECREF(tmp2); | |
| 447 if (tmp == NULL) return NULL; | |
| 448 tmp2 = PyUnicode_Replace(tmp, amp2, amp1, -1); | |
| 449 Py_DECREF(tmp); | |
| 450 return tmp2; | |
| 451 } | |
| 452 | |
| 453 PyDoc_STRVAR(stripentities__doc__, | |
| 454 "Return a copy of the text with any character or numeric entities\n\ | |
| 455 replaced by the equivalent UTF-8 characters.\n\ | |
| 456 \n\ | |
| 457 If the `keepxmlentities` parameter is provided and evaluates to `True`,\n\ | |
| 547 | 458 the core XML entities (``&``, ``'``, ``>``, ``<`` and\n\ |
| 459 ``"``) are not stripped.\n\ | |
| 460 \n\ | |
| 461 :return: a `Markup` instance with entities removed\n\ | |
| 462 :rtype: `Markup`\n\ | |
| 463 :see: `genshi.util.stripentities`\n\ | |
| 464 "); | |
| 541 | 465 |
| 466 static PyObject * | |
| 467 Markup_stripentities(PyObject* self, PyObject *args, PyObject *kwds) | |
| 468 { | |
| 469 static char *kwlist[] = {"keepxmlentities", 0}; | |
| 470 PyObject *result, *args2; | |
| 471 char keepxml = 0; | |
| 472 | |
| 473 if (!PyArg_ParseTupleAndKeywords(args, kwds, "|b", kwlist, &keepxml)) { | |
| 474 return NULL; | |
| 475 } | |
| 476 | |
| 477 if (stripentities == NULL) return NULL; | |
| 478 result = PyObject_CallFunction(stripentities, "Ob", self, keepxml); | |
| 479 if (result == NULL) return NULL; | |
| 480 args2 = PyTuple_New(1); | |
| 481 if (args2 == NULL) { | |
| 482 Py_DECREF(result); | |
| 483 return NULL; | |
| 484 } | |
| 485 PyTuple_SET_ITEM(args2, 0, result); | |
| 486 result = MarkupType.tp_new(&MarkupType, args2, NULL); | |
| 487 Py_DECREF(args2); | |
| 488 return result; | |
| 489 } | |
| 490 | |
| 491 PyDoc_STRVAR(striptags__doc__, | |
| 547 | 492 """Return a copy of the text with all XML/HTML tags removed.\n\ |
| 493 \n\ | |
| 494 :return: a `Markup` instance with all tags removed\n\ | |
| 495 :rtype: `Markup`\n\ | |
| 496 :see: `genshi.util.striptags`\n\ | |
| 497 "); | |
| 541 | 498 |
| 499 static PyObject * | |
| 500 Markup_striptags(PyObject* self) | |
| 501 { | |
| 502 PyObject *result, *args; | |
| 503 | |
| 504 if (striptags == NULL) return NULL; | |
| 505 result = PyObject_CallFunction(striptags, "O", self); | |
| 506 if (result == NULL) return NULL; | |
| 507 args = PyTuple_New(1); | |
| 508 if (args == NULL) { | |
| 509 Py_DECREF(result); | |
| 510 return NULL; | |
| 511 } | |
| 512 PyTuple_SET_ITEM(args, 0, result); | |
| 513 result = MarkupType.tp_new(&MarkupType, args, NULL); | |
| 514 Py_DECREF(args); | |
| 515 return result; | |
| 516 } | |
| 517 | |
| 518 typedef struct { | |
| 519 PyUnicodeObject HEAD; | |
| 520 } MarkupObject; | |
| 521 | |
| 522 static PyMethodDef Markup_methods[] = { | |
| 523 {"escape", (PyCFunction) Markup_escape, | |
| 547 | 524 METH_VARARGS|METH_CLASS|METH_KEYWORDS, escape__doc__}, |
| 525 {"join", (PyCFunction)Markup_join, METH_VARARGS|METH_KEYWORDS, join__doc__}, | |
| 541 | 526 {"unescape", (PyCFunction)Markup_unescape, METH_NOARGS, unescape__doc__}, |
| 527 {"stripentities", (PyCFunction) Markup_stripentities, | |
| 528 METH_VARARGS|METH_KEYWORDS, stripentities__doc__}, | |
| 529 {"striptags", (PyCFunction) Markup_striptags, METH_NOARGS, | |
| 530 striptags__doc__}, | |
| 531 {NULL} /* Sentinel */ | |
| 532 }; | |
| 533 | |
| 534 static PyNumberMethods Markup_as_number = { | |
| 547 | 535 Markup_add, /*nb_add*/ |
| 536 0, /*nb_subtract*/ | |
| 537 Markup_mul, /*nb_multiply*/ | |
| 538 0, /*nb_divide*/ | |
| 539 Markup_mod, /*nb_remainder*/ | |
| 541 | 540 }; |
| 541 | |
| 542 PyTypeObject MarkupType = { | |
| 543 PyObject_HEAD_INIT(NULL) | |
| 544 0, | |
| 545 "genshi._speedups.Markup", | |
| 546 sizeof(MarkupObject), | |
| 547 0, | |
| 548 0, /*tp_dealloc*/ | |
|
724
919809e55d16
update to trunk to track r847, fixing python 2.4 compatibility issues in speedup (and fixing copyrights, apparently :))
aflett
parents:
718
diff
changeset
|
549 0, /*tp_print*/ |
| 541 | 550 0, /*tp_getattr*/ |
| 551 0, /*tp_setattr*/ | |
| 552 0, /*tp_compare*/ | |
| 553 Markup_repr, /*tp_repr*/ | |
| 554 &Markup_as_number, /*tp_as_number*/ | |
| 555 0, /*tp_as_sequence*/ | |
| 556 0, /*tp_as_mapping*/ | |
| 557 0, /*tp_hash */ | |
| 558 | |
| 559 0, /*tp_call*/ | |
| 560 0, /*tp_str*/ | |
| 561 0, /*tp_getattro*/ | |
| 562 0, /*tp_setattro*/ | |
| 563 0, /*tp_as_buffer*/ | |
| 564 | |
| 565 Py_TPFLAGS_DEFAULT | Py_TPFLAGS_BASETYPE | Py_TPFLAGS_CHECKTYPES, /*tp_flags*/ | |
| 566 Markup__doc__,/*tp_doc*/ | |
|
724
919809e55d16
update to trunk to track r847, fixing python 2.4 compatibility issues in speedup (and fixing copyrights, apparently :))
aflett
parents:
718
diff
changeset
|
567 |
| 541 | 568 0, /*tp_traverse*/ |
| 569 0, /*tp_clear*/ | |
| 570 | |
| 571 0, /*tp_richcompare*/ | |
| 572 0, /*tp_weaklistoffset*/ | |
| 573 | |
| 574 0, /*tp_iter*/ | |
| 575 0, /*tp_iternext*/ | |
| 576 | |
| 577 /* Attribute descriptor and subclassing stuff */ | |
| 578 | |
| 579 Markup_methods,/*tp_methods*/ | |
| 580 0, /*tp_members*/ | |
| 581 0, /*tp_getset*/ | |
| 582 0, /*tp_base*/ | |
| 583 0, /*tp_dict*/ | |
|
724
919809e55d16
update to trunk to track r847, fixing python 2.4 compatibility issues in speedup (and fixing copyrights, apparently :))
aflett
parents:
718
diff
changeset
|
584 |
| 541 | 585 0, /*tp_descr_get*/ |
| 586 0, /*tp_descr_set*/ | |
| 587 0, /*tp_dictoffset*/ | |
|
724
919809e55d16
update to trunk to track r847, fixing python 2.4 compatibility issues in speedup (and fixing copyrights, apparently :))
aflett
parents:
718
diff
changeset
|
588 |
| 541 | 589 0, /*tp_init*/ |
| 590 0, /*tp_alloc will be set to PyType_GenericAlloc in module init*/ | |
| 718 | 591 0, /*tp_new*/ |
| 541 | 592 0, /*tp_free Low-level free-memory routine */ |
| 593 0, /*tp_is_gc For PyObject_IS_GC */ | |
| 594 0, /*tp_bases*/ | |
| 595 0, /*tp_mro method resolution order */ | |
| 596 0, /*tp_cache*/ | |
| 597 0, /*tp_subclasses*/ | |
| 598 0 /*tp_weaklist*/ | |
| 599 }; | |
| 600 | |
| 601 PyMODINIT_FUNC | |
| 602 init_speedups(void) | |
| 603 { | |
| 604 PyObject *module; | |
| 605 | |
| 606 /* Workaround for quirk in Visual Studio, see | |
| 607 <http://www.python.it/faq/faq-3.html#3.24> */ | |
| 608 MarkupType.tp_base = &PyUnicode_Type; | |
| 609 | |
| 610 if (PyType_Ready(&MarkupType) < 0) | |
| 611 return; | |
| 612 | |
| 613 init_constants(); | |
| 614 | |
| 615 module = Py_InitModule("_speedups", NULL); | |
| 616 Py_INCREF(&MarkupType); | |
| 617 PyModule_AddObject(module, "Markup", (PyObject *) &MarkupType); | |
| 618 } |