projects / mirror_edk2.git / blame
| Commit | Line | Data |
|---|---|---|
| 3257aa99 DM |
1 | # -*- coding: utf-8 -*-\r |
| 2 | #\r | |
| 3 | # Secret Labs' Regular Expression Engine\r | |
| 4 | #\r | |
| 5 | # convert template to internal format\r | |
| 6 | #\r | |
| 7 | # Copyright (c) 1997-2001 by Secret Labs AB. All rights reserved.\r | |
| 8 | #\r | |
| 9 | # See the sre.py file for information on usage and redistribution.\r | |
| 10 | #\r | |
| 11 | \r | |
| 12 | """Internal support module for sre"""\r | |
| 13 | \r | |
| 14 | import _sre, sys\r | |
| 15 | import sre_parse\r | |
| 16 | from sre_constants import *\r | |
| 17 | \r | |
| 18 | assert _sre.MAGIC == MAGIC, "SRE module mismatch"\r | |
| 19 | \r | |
| 20 | if _sre.CODESIZE == 2:\r | |
| 21 | MAXCODE = 65535\r | |
| 22 | else:\r | |
| 23 | MAXCODE = 0xFFFFFFFFL\r | |
| 24 | \r | |
| 25 | _LITERAL_CODES = set([LITERAL, NOT_LITERAL])\r | |
| 26 | _REPEATING_CODES = set([REPEAT, MIN_REPEAT, MAX_REPEAT])\r | |
| 27 | _SUCCESS_CODES = set([SUCCESS, FAILURE])\r | |
| 28 | _ASSERT_CODES = set([ASSERT, ASSERT_NOT])\r | |
| 29 | \r | |
| 30 | # Sets of lowercase characters which have the same uppercase.\r | |
| 31 | _equivalences = (\r | |
| 32 | # LATIN SMALL LETTER I, LATIN SMALL LETTER DOTLESS I\r | |
| 33 | (0x69, 0x131), # iı\r | |
| 34 | # LATIN SMALL LETTER S, LATIN SMALL LETTER LONG S\r | |
| 35 | (0x73, 0x17f), # sſ\r | |
| 36 | # MICRO SIGN, GREEK SMALL LETTER MU\r | |
| 37 | (0xb5, 0x3bc), # µμ\r | |
| 38 | # COMBINING GREEK YPOGEGRAMMENI, GREEK SMALL LETTER IOTA, GREEK PROSGEGRAMMENI\r | |
| 39 | (0x345, 0x3b9, 0x1fbe), # \u0345ιι\r | |
| 40 | # GREEK SMALL LETTER BETA, GREEK BETA SYMBOL\r | |
| 41 | (0x3b2, 0x3d0), # βϐ\r | |
| 42 | # GREEK SMALL LETTER EPSILON, GREEK LUNATE EPSILON SYMBOL\r | |
| 43 | (0x3b5, 0x3f5), # εϵ\r | |
| 44 | # GREEK SMALL LETTER THETA, GREEK THETA SYMBOL\r | |
| 45 | (0x3b8, 0x3d1), # θϑ\r | |
| 46 | # GREEK SMALL LETTER KAPPA, GREEK KAPPA SYMBOL\r | |
| 47 | (0x3ba, 0x3f0), # κϰ\r | |
| 48 | # GREEK SMALL LETTER PI, GREEK PI SYMBOL\r | |
| 49 | (0x3c0, 0x3d6), # πϖ\r | |
| 50 | # GREEK SMALL LETTER RHO, GREEK RHO SYMBOL\r | |
| 51 | (0x3c1, 0x3f1), # ρϱ\r | |
| 52 | # GREEK SMALL LETTER FINAL SIGMA, GREEK SMALL LETTER SIGMA\r | |
| 53 | (0x3c2, 0x3c3), # ςσ\r | |
| 54 | # GREEK SMALL LETTER PHI, GREEK PHI SYMBOL\r | |
| 55 | (0x3c6, 0x3d5), # φϕ\r | |
| 56 | # LATIN SMALL LETTER S WITH DOT ABOVE, LATIN SMALL LETTER LONG S WITH DOT ABOVE\r | |
| 57 | (0x1e61, 0x1e9b), # ṡẛ\r | |
| 58 | )\r | |
| 59 | \r | |
| 60 | # Maps the lowercase code to lowercase codes which have the same uppercase.\r | |
| 61 | _ignorecase_fixes = {i: tuple(j for j in t if i != j)\r | |
| 62 | for t in _equivalences for i in t}\r | |
| 63 | \r | |
| 64 | def _compile(code, pattern, flags):\r | |
| 65 | # internal: compile a (sub)pattern\r | |
| 66 | emit = code.append\r | |
| 67 | _len = len\r | |
| 68 | LITERAL_CODES = _LITERAL_CODES\r | |
| 69 | REPEATING_CODES = _REPEATING_CODES\r | |
| 70 | SUCCESS_CODES = _SUCCESS_CODES\r | |
| 71 | ASSERT_CODES = _ASSERT_CODES\r | |
| 72 | if (flags & SRE_FLAG_IGNORECASE and\r | |
| 73 | not (flags & SRE_FLAG_LOCALE) and\r | |
| 74 | flags & SRE_FLAG_UNICODE):\r | |
| 75 | fixes = _ignorecase_fixes\r | |
| 76 | else:\r | |
| 77 | fixes = None\r | |
| 78 | for op, av in pattern:\r | |
| 79 | if op in LITERAL_CODES:\r | |
| 80 | if flags & SRE_FLAG_IGNORECASE:\r | |
| 81 | lo = _sre.getlower(av, flags)\r | |
| 82 | if fixes and lo in fixes:\r | |
| 83 | emit(OPCODES[IN_IGNORE])\r | |
| 84 | skip = _len(code); emit(0)\r | |
| 85 | if op is NOT_LITERAL:\r | |
| 86 | emit(OPCODES[NEGATE])\r | |
| 87 | for k in (lo,) + fixes[lo]:\r | |
| 88 | emit(OPCODES[LITERAL])\r | |
| 89 | emit(k)\r | |
| 90 | emit(OPCODES[FAILURE])\r | |
| 91 | code[skip] = _len(code) - skip\r | |
| 92 | else:\r | |
| 93 | emit(OPCODES[OP_IGNORE[op]])\r | |
| 94 | emit(lo)\r | |
| 95 | else:\r | |
| 96 | emit(OPCODES[op])\r | |
| 97 | emit(av)\r | |
| 98 | elif op is IN:\r | |
| 99 | if flags & SRE_FLAG_IGNORECASE:\r | |
| 100 | emit(OPCODES[OP_IGNORE[op]])\r | |
| 101 | def fixup(literal, flags=flags):\r | |
| 102 | return _sre.getlower(literal, flags)\r | |
| 103 | else:\r | |
| 104 | emit(OPCODES[op])\r | |
| 105 | fixup = None\r | |
| 106 | skip = _len(code); emit(0)\r | |
| 107 | _compile_charset(av, flags, code, fixup, fixes)\r | |
| 108 | code[skip] = _len(code) - skip\r | |
| 109 | elif op is ANY:\r | |
| 110 | if flags & SRE_FLAG_DOTALL:\r | |
| 111 | emit(OPCODES[ANY_ALL])\r | |
| 112 | else:\r | |
| 113 | emit(OPCODES[ANY])\r | |
| 114 | elif op in REPEATING_CODES:\r | |
| 115 | if flags & SRE_FLAG_TEMPLATE:\r | |
| 116 | raise error, "internal: unsupported template operator"\r | |
| 117 | emit(OPCODES[REPEAT])\r | |
| 118 | skip = _len(code); emit(0)\r | |
| 119 | emit(av[0])\r | |
| 120 | emit(av[1])\r | |
| 121 | _compile(code, av[2], flags)\r | |
| 122 | emit(OPCODES[SUCCESS])\r | |
| 123 | code[skip] = _len(code) - skip\r | |
| 124 | elif _simple(av) and op is not REPEAT:\r | |
| 125 | if op is MAX_REPEAT:\r | |
| 126 | emit(OPCODES[REPEAT_ONE])\r | |
| 127 | else:\r | |
| 128 | emit(OPCODES[MIN_REPEAT_ONE])\r | |
| 129 | skip = _len(code); emit(0)\r | |
| 130 | emit(av[0])\r | |
| 131 | emit(av[1])\r | |
| 132 | _compile(code, av[2], flags)\r | |
| 133 | emit(OPCODES[SUCCESS])\r | |
| 134 | code[skip] = _len(code) - skip\r | |
| 135 | else:\r | |
| 136 | emit(OPCODES[REPEAT])\r | |
| 137 | skip = _len(code); emit(0)\r | |
| 138 | emit(av[0])\r | |
| 139 | emit(av[1])\r | |
| 140 | _compile(code, av[2], flags)\r | |
| 141 | code[skip] = _len(code) - skip\r | |
| 142 | if op is MAX_REPEAT:\r | |
| 143 | emit(OPCODES[MAX_UNTIL])\r | |
| 144 | else:\r | |
| 145 | emit(OPCODES[MIN_UNTIL])\r | |
| 146 | elif op is SUBPATTERN:\r | |
| 147 | if av[0]:\r | |
| 148 | emit(OPCODES[MARK])\r | |
| 149 | emit((av[0]-1)*2)\r | |
| 150 | # _compile_info(code, av[1], flags)\r | |
| 151 | _compile(code, av[1], flags)\r | |
| 152 | if av[0]:\r | |
| 153 | emit(OPCODES[MARK])\r | |
| 154 | emit((av[0]-1)*2+1)\r | |
| 155 | elif op in SUCCESS_CODES:\r | |
| 156 | emit(OPCODES[op])\r | |
| 157 | elif op in ASSERT_CODES:\r | |
| 158 | emit(OPCODES[op])\r | |
| 159 | skip = _len(code); emit(0)\r | |
| 160 | if av[0] >= 0:\r | |
| 161 | emit(0) # look ahead\r | |
| 162 | else:\r | |
| 163 | lo, hi = av[1].getwidth()\r | |
| 164 | if lo != hi:\r | |
| 165 | raise error, "look-behind requires fixed-width pattern"\r | |
| 166 | emit(lo) # look behind\r | |
| 167 | _compile(code, av[1], flags)\r | |
| 168 | emit(OPCODES[SUCCESS])\r | |
| 169 | code[skip] = _len(code) - skip\r | |
| 170 | elif op is CALL:\r | |
| 171 | emit(OPCODES[op])\r | |
| 172 | skip = _len(code); emit(0)\r | |
| 173 | _compile(code, av, flags)\r | |
| 174 | emit(OPCODES[SUCCESS])\r | |
| 175 | code[skip] = _len(code) - skip\r | |
| 176 | elif op is AT:\r | |
| 177 | emit(OPCODES[op])\r | |
| 178 | if flags & SRE_FLAG_MULTILINE:\r | |
| 179 | av = AT_MULTILINE.get(av, av)\r | |
| 180 | if flags & SRE_FLAG_LOCALE:\r | |
| 181 | av = AT_LOCALE.get(av, av)\r | |
| 182 | elif flags & SRE_FLAG_UNICODE:\r | |
| 183 | av = AT_UNICODE.get(av, av)\r | |
| 184 | emit(ATCODES[av])\r | |
| 185 | elif op is BRANCH:\r | |
| 186 | emit(OPCODES[op])\r | |
| 187 | tail = []\r | |
| 188 | tailappend = tail.append\r | |
| 189 | for av in av[1]:\r | |
| 190 | skip = _len(code); emit(0)\r | |
| 191 | # _compile_info(code, av, flags)\r | |
| 192 | _compile(code, av, flags)\r | |
| 193 | emit(OPCODES[JUMP])\r | |
| 194 | tailappend(_len(code)); emit(0)\r | |
| 195 | code[skip] = _len(code) - skip\r | |
| 196 | emit(0) # end of branch\r | |
| 197 | for tail in tail:\r | |
| 198 | code[tail] = _len(code) - tail\r | |
| 199 | elif op is CATEGORY:\r | |
| 200 | emit(OPCODES[op])\r | |
| 201 | if flags & SRE_FLAG_LOCALE:\r | |
| 202 | av = CH_LOCALE[av]\r | |
| 203 | elif flags & SRE_FLAG_UNICODE:\r | |
| 204 | av = CH_UNICODE[av]\r | |
| 205 | emit(CHCODES[av])\r | |
| 206 | elif op is GROUPREF:\r | |
| 207 | if flags & SRE_FLAG_IGNORECASE:\r | |
| 208 | emit(OPCODES[OP_IGNORE[op]])\r | |
| 209 | else:\r | |
| 210 | emit(OPCODES[op])\r | |
| 211 | emit(av-1)\r | |
| 212 | elif op is GROUPREF_EXISTS:\r | |
| 213 | emit(OPCODES[op])\r | |
| 214 | emit(av[0]-1)\r | |
| 215 | skipyes = _len(code); emit(0)\r | |
| 216 | _compile(code, av[1], flags)\r | |
| 217 | if av[2]:\r | |
| 218 | emit(OPCODES[JUMP])\r | |
| 219 | skipno = _len(code); emit(0)\r | |
| 220 | code[skipyes] = _len(code) - skipyes + 1\r | |
| 221 | _compile(code, av[2], flags)\r | |
| 222 | code[skipno] = _len(code) - skipno\r | |
| 223 | else:\r | |
| 224 | code[skipyes] = _len(code) - skipyes + 1\r | |
| 225 | else:\r | |
| 226 | raise ValueError, ("unsupported operand type", op)\r | |
| 227 | \r | |
| 228 | def _compile_charset(charset, flags, code, fixup=None, fixes=None):\r | |
| 229 | # compile charset subprogram\r | |
| 230 | emit = code.append\r | |
| 231 | for op, av in _optimize_charset(charset, fixup, fixes,\r | |
| 232 | flags & SRE_FLAG_UNICODE):\r | |
| 233 | emit(OPCODES[op])\r | |
| 234 | if op is NEGATE:\r | |
| 235 | pass\r | |
| 236 | elif op is LITERAL:\r | |
| 237 | emit(av)\r | |
| 238 | elif op is RANGE:\r | |
| 239 | emit(av[0])\r | |
| 240 | emit(av[1])\r | |
| 241 | elif op is CHARSET:\r | |
| 242 | code.extend(av)\r | |
| 243 | elif op is BIGCHARSET:\r | |
| 244 | code.extend(av)\r | |
| 245 | elif op is CATEGORY:\r | |
| 246 | if flags & SRE_FLAG_LOCALE:\r | |
| 247 | emit(CHCODES[CH_LOCALE[av]])\r | |
| 248 | elif flags & SRE_FLAG_UNICODE:\r | |
| 249 | emit(CHCODES[CH_UNICODE[av]])\r | |
| 250 | else:\r | |
| 251 | emit(CHCODES[av])\r | |
| 252 | else:\r | |
| 253 | raise error, "internal: unsupported set operator"\r | |
| 254 | emit(OPCODES[FAILURE])\r | |
| 255 | \r | |
| 256 | def _optimize_charset(charset, fixup, fixes, isunicode):\r | |
| 257 | # internal: optimize character set\r | |
| 258 | out = []\r | |
| 259 | tail = []\r | |
| 260 | charmap = bytearray(256)\r | |
| 261 | for op, av in charset:\r | |
| 262 | while True:\r | |
| 263 | try:\r | |
| 264 | if op is LITERAL:\r | |
| 265 | if fixup:\r | |
| 266 | i = fixup(av)\r | |
| 267 | charmap[i] = 1\r | |
| 268 | if fixes and i in fixes:\r | |
| 269 | for k in fixes[i]:\r | |
| 270 | charmap[k] = 1\r | |
| 271 | else:\r | |
| 272 | charmap[av] = 1\r | |
| 273 | elif op is RANGE:\r | |
| 274 | r = range(av[0], av[1]+1)\r | |
| 275 | if fixup:\r | |
| 276 | r = map(fixup, r)\r | |
| 277 | if fixup and fixes:\r | |
| 278 | for i in r:\r | |
| 279 | charmap[i] = 1\r | |
| 280 | if i in fixes:\r | |
| 281 | for k in fixes[i]:\r | |
| 282 | charmap[k] = 1\r | |
| 283 | else:\r | |
| 284 | for i in r:\r | |
| 285 | charmap[i] = 1\r | |
| 286 | elif op is NEGATE:\r | |
| 287 | out.append((op, av))\r | |
| 288 | else:\r | |
| 289 | tail.append((op, av))\r | |
| 290 | except IndexError:\r | |
| 291 | if len(charmap) == 256:\r | |
| 292 | # character set contains non-UCS1 character codes\r | |
| 293 | charmap += b'\0' * 0xff00\r | |
| 294 | continue\r | |
| 295 | # character set contains non-BMP character codes\r | |
| 296 | if fixup and isunicode and op is RANGE:\r | |
| 297 | lo, hi = av\r | |
| 298 | ranges = [av]\r | |
| 299 | # There are only two ranges of cased astral characters:\r | |
| 300 | # 10400-1044F (Deseret) and 118A0-118DF (Warang Citi).\r | |
| 301 | _fixup_range(max(0x10000, lo), min(0x11fff, hi),\r | |
| 302 | ranges, fixup)\r | |
| 303 | for lo, hi in ranges:\r | |
| 304 | if lo == hi:\r | |
| 305 | tail.append((LITERAL, hi))\r | |
| 306 | else:\r | |
| 307 | tail.append((RANGE, (lo, hi)))\r | |
| 308 | else:\r | |
| 309 | tail.append((op, av))\r | |
| 310 | break\r | |
| 311 | \r | |
| 312 | # compress character map\r | |
| 313 | runs = []\r | |
| 314 | q = 0\r | |
| 315 | while True:\r | |
| 316 | p = charmap.find(b'\1', q)\r | |
| 317 | if p < 0:\r | |
| 318 | break\r | |
| 319 | if len(runs) >= 2:\r | |
| 320 | runs = None\r | |
| 321 | break\r | |
| 322 | q = charmap.find(b'\0', p)\r | |
| 323 | if q < 0:\r | |
| 324 | runs.append((p, len(charmap)))\r | |
| 325 | break\r | |
| 326 | runs.append((p, q))\r | |
| 327 | if runs is not None:\r | |
| 328 | # use literal/range\r | |
| 329 | for p, q in runs:\r | |
| 330 | if q - p == 1:\r | |
| 331 | out.append((LITERAL, p))\r | |
| 332 | else:\r | |
| 333 | out.append((RANGE, (p, q - 1)))\r | |
| 334 | out += tail\r | |
| 335 | # if the case was changed or new representation is more compact\r | |
| 336 | if fixup or len(out) < len(charset):\r | |
| 337 | return out\r | |
| 338 | # else original character set is good enough\r | |
| 339 | return charset\r | |
| 340 | \r | |
| 341 | # use bitmap\r | |
| 342 | if len(charmap) == 256:\r | |
| 343 | data = _mk_bitmap(charmap)\r | |
| 344 | out.append((CHARSET, data))\r | |
| 345 | out += tail\r | |
| 346 | return out\r | |
| 347 | \r | |
| 348 | # To represent a big charset, first a bitmap of all characters in the\r | |
| 349 | # set is constructed. Then, this bitmap is sliced into chunks of 256\r | |
| 350 | # characters, duplicate chunks are eliminated, and each chunk is\r | |
| 351 | # given a number. In the compiled expression, the charset is\r | |
| 352 | # represented by a 32-bit word sequence, consisting of one word for\r | |
| 353 | # the number of different chunks, a sequence of 256 bytes (64 words)\r | |
| 354 | # of chunk numbers indexed by their original chunk position, and a\r | |
| 355 | # sequence of 256-bit chunks (8 words each).\r | |
| 356 | \r | |
| 357 | # Compression is normally good: in a typical charset, large ranges of\r | |
| 358 | # Unicode will be either completely excluded (e.g. if only cyrillic\r | |
| 359 | # letters are to be matched), or completely included (e.g. if large\r | |
| 360 | # subranges of Kanji match). These ranges will be represented by\r | |
| 361 | # chunks of all one-bits or all zero-bits.\r | |
| 362 | \r | |
| 363 | # Matching can be also done efficiently: the more significant byte of\r | |
| 364 | # the Unicode character is an index into the chunk number, and the\r | |
| 365 | # less significant byte is a bit index in the chunk (just like the\r | |
| 366 | # CHARSET matching).\r | |
| 367 | \r | |
| 368 | # In UCS-4 mode, the BIGCHARSET opcode still supports only subsets\r | |
| 369 | # of the basic multilingual plane; an efficient representation\r | |
| 370 | # for all of Unicode has not yet been developed.\r | |
| 371 | \r | |
| 372 | charmap = bytes(charmap) # should be hashable\r | |
| 373 | comps = {}\r | |
| 374 | mapping = bytearray(256)\r | |
| 375 | block = 0\r | |
| 376 | data = bytearray()\r | |
| 377 | for i in range(0, 65536, 256):\r | |
| 378 | chunk = charmap[i: i + 256]\r | |
| 379 | if chunk in comps:\r | |
| 380 | mapping[i // 256] = comps[chunk]\r | |
| 381 | else:\r | |
| 382 | mapping[i // 256] = comps[chunk] = block\r | |
| 383 | block += 1\r | |
| 384 | data += chunk\r | |
| 385 | data = _mk_bitmap(data)\r | |
| 386 | data[0:0] = [block] + _bytes_to_codes(mapping)\r | |
| 387 | out.append((BIGCHARSET, data))\r | |
| 388 | out += tail\r | |
| 389 | return out\r | |
| 390 | \r | |
| 391 | def _fixup_range(lo, hi, ranges, fixup):\r | |
| 392 | for i in map(fixup, range(lo, hi+1)):\r | |
| 393 | for k, (lo, hi) in enumerate(ranges):\r | |
| 394 | if i < lo:\r | |
| 395 | if l == lo - 1:\r | |
| 396 | ranges[k] = (i, hi)\r | |
| 397 | else:\r | |
| 398 | ranges.insert(k, (i, i))\r | |
| 399 | break\r | |
| 400 | elif i > hi:\r | |
| 401 | if i == hi + 1:\r | |
| 402 | ranges[k] = (lo, i)\r | |
| 403 | break\r | |
| 404 | else:\r | |
| 405 | break\r | |
| 406 | else:\r | |
| 407 | ranges.append((i, i))\r | |
| 408 | \r | |
| 409 | _CODEBITS = _sre.CODESIZE * 8\r | |
| 410 | _BITS_TRANS = b'0' + b'1' * 255\r | |
| 411 | def _mk_bitmap(bits, _CODEBITS=_CODEBITS, _int=int):\r | |
| 412 | s = bytes(bits).translate(_BITS_TRANS)[::-1]\r | |
| 413 | return [_int(s[i - _CODEBITS: i], 2)\r | |
| 414 | for i in range(len(s), 0, -_CODEBITS)]\r | |
| 415 | \r | |
| 416 | def _bytes_to_codes(b):\r | |
| 417 | # Convert block indices to word array\r | |
| 418 | import array\r | |
| 419 | if _sre.CODESIZE == 2:\r | |
| 420 | code = 'H'\r | |
| 421 | else:\r | |
| 422 | code = 'I'\r | |
| 423 | a = array.array(code, bytes(b))\r | |
| 424 | assert a.itemsize == _sre.CODESIZE\r | |
| 425 | assert len(a) * a.itemsize == len(b)\r | |
| 426 | return a.tolist()\r | |
| 427 | \r | |
| 428 | def _simple(av):\r | |
| 429 | # check if av is a "simple" operator\r | |
| 430 | lo, hi = av[2].getwidth()\r | |
| 431 | return lo == hi == 1 and av[2][0][0] != SUBPATTERN\r | |
| 432 | \r | |
| 433 | def _compile_info(code, pattern, flags):\r | |
| 434 | # internal: compile an info block. in the current version,\r | |
| 435 | # this contains min/max pattern width, and an optional literal\r | |
| 436 | # prefix or a character map\r | |
| 437 | lo, hi = pattern.getwidth()\r | |
| 438 | if lo == 0:\r | |
| 439 | return # not worth it\r | |
| 440 | # look for a literal prefix\r | |
| 441 | prefix = []\r | |
| 442 | prefixappend = prefix.append\r | |
| 443 | prefix_skip = 0\r | |
| 444 | charset = [] # not used\r | |
| 445 | charsetappend = charset.append\r | |
| 446 | if not (flags & SRE_FLAG_IGNORECASE):\r | |
| 447 | # look for literal prefix\r | |
| 448 | for op, av in pattern.data:\r | |
| 449 | if op is LITERAL:\r | |
| 450 | if len(prefix) == prefix_skip:\r | |
| 451 | prefix_skip = prefix_skip + 1\r | |
| 452 | prefixappend(av)\r | |
| 453 | elif op is SUBPATTERN and len(av[1]) == 1:\r | |
| 454 | op, av = av[1][0]\r | |
| 455 | if op is LITERAL:\r | |
| 456 | prefixappend(av)\r | |
| 457 | else:\r | |
| 458 | break\r | |
| 459 | else:\r | |
| 460 | break\r | |
| 461 | # if no prefix, look for charset prefix\r | |
| 462 | if not prefix and pattern.data:\r | |
| 463 | op, av = pattern.data[0]\r | |
| 464 | if op is SUBPATTERN and av[1]:\r | |
| 465 | op, av = av[1][0]\r | |
| 466 | if op is LITERAL:\r | |
| 467 | charsetappend((op, av))\r | |
| 468 | elif op is BRANCH:\r | |
| 469 | c = []\r | |
| 470 | cappend = c.append\r | |
| 471 | for p in av[1]:\r | |
| 472 | if not p:\r | |
| 473 | break\r | |
| 474 | op, av = p[0]\r | |
| 475 | if op is LITERAL:\r | |
| 476 | cappend((op, av))\r | |
| 477 | else:\r | |
| 478 | break\r | |
| 479 | else:\r | |
| 480 | charset = c\r | |
| 481 | elif op is BRANCH:\r | |
| 482 | c = []\r | |
| 483 | cappend = c.append\r | |
| 484 | for p in av[1]:\r | |
| 485 | if not p:\r | |
| 486 | break\r | |
| 487 | op, av = p[0]\r | |
| 488 | if op is LITERAL:\r | |
| 489 | cappend((op, av))\r | |
| 490 | else:\r | |
| 491 | break\r | |
| 492 | else:\r | |
| 493 | charset = c\r | |
| 494 | elif op is IN:\r | |
| 495 | charset = av\r | |
| 496 | ## if prefix:\r | |
| 497 | ## print "*** PREFIX", prefix, prefix_skip\r | |
| 498 | ## if charset:\r | |
| 499 | ## print "*** CHARSET", charset\r | |
| 500 | # add an info block\r | |
| 501 | emit = code.append\r | |
| 502 | emit(OPCODES[INFO])\r | |
| 503 | skip = len(code); emit(0)\r | |
| 504 | # literal flag\r | |
| 505 | mask = 0\r | |
| 506 | if prefix:\r | |
| 507 | mask = SRE_INFO_PREFIX\r | |
| 508 | if len(prefix) == prefix_skip == len(pattern.data):\r | |
| 509 | mask = mask + SRE_INFO_LITERAL\r | |
| 510 | elif charset:\r | |
| 511 | mask = mask + SRE_INFO_CHARSET\r | |
| 512 | emit(mask)\r | |
| 513 | # pattern length\r | |
| 514 | if lo < MAXCODE:\r | |
| 515 | emit(lo)\r | |
| 516 | else:\r | |
| 517 | emit(MAXCODE)\r | |
| 518 | prefix = prefix[:MAXCODE]\r | |
| 519 | if hi < MAXCODE:\r | |
| 520 | emit(hi)\r | |
| 521 | else:\r | |
| 522 | emit(0)\r | |
| 523 | # add literal prefix\r | |
| 524 | if prefix:\r | |
| 525 | emit(len(prefix)) # length\r | |
| 526 | emit(prefix_skip) # skip\r | |
| 527 | code.extend(prefix)\r | |
| 528 | # generate overlap table\r | |
| 529 | table = [-1] + ([0]*len(prefix))\r | |
| 530 | for i in xrange(len(prefix)):\r | |
| 531 | table[i+1] = table[i]+1\r | |
| 532 | while table[i+1] > 0 and prefix[i] != prefix[table[i+1]-1]:\r | |
| 533 | table[i+1] = table[table[i+1]-1]+1\r | |
| 534 | code.extend(table[1:]) # don't store first entry\r | |
| 535 | elif charset:\r | |
| 536 | _compile_charset(charset, flags, code)\r | |
| 537 | code[skip] = len(code) - skip\r | |
| 538 | \r | |
| 539 | try:\r | |
| 540 | unicode\r | |
| 541 | except NameError:\r | |
| 542 | STRING_TYPES = (type(""),)\r | |
| 543 | else:\r | |
| 544 | STRING_TYPES = (type(""), type(unicode("")))\r | |
| 545 | \r | |
| 546 | def isstring(obj):\r | |
| 547 | for tp in STRING_TYPES:\r | |
| 548 | if isinstance(obj, tp):\r | |
| 549 | return 1\r | |
| 550 | return 0\r | |
| 551 | \r | |
| 552 | def _code(p, flags):\r | |
| 553 | \r | |
| 554 | flags = p.pattern.flags | flags\r | |
| 555 | code = []\r | |
| 556 | \r | |
| 557 | # compile info block\r | |
| 558 | _compile_info(code, p, flags)\r | |
| 559 | \r | |
| 560 | # compile the pattern\r | |
| 561 | _compile(code, p.data, flags)\r | |
| 562 | \r | |
| 563 | code.append(OPCODES[SUCCESS])\r | |
| 564 | \r | |
| 565 | return code\r | |
| 566 | \r | |
| 567 | def compile(p, flags=0):\r | |
| 568 | # internal: convert pattern list to internal format\r | |
| 569 | \r | |
| 570 | if isstring(p):\r | |
| 571 | pattern = p\r | |
| 572 | p = sre_parse.parse(p, flags)\r | |
| 573 | else:\r | |
| 574 | pattern = None\r | |
| 575 | \r | |
| 576 | code = _code(p, flags)\r | |
| 577 | \r | |
| 578 | # print code\r | |
| 579 | \r | |
| 580 | # XXX: <fl> get rid of this limitation!\r | |
| 581 | if p.pattern.groups > 100:\r | |
| 582 | raise AssertionError(\r | |
| 583 | "sorry, but this version only supports 100 named groups"\r | |
| 584 | )\r | |
| 585 | \r | |
| 586 | # map in either direction\r | |
| 587 | groupindex = p.pattern.groupdict\r | |
| 588 | indexgroup = [None] * p.pattern.groups\r | |
| 589 | for k, i in groupindex.items():\r | |
| 590 | indexgroup[i] = k\r | |
| 591 | \r | |
| 592 | return _sre.compile(\r | |
| 593 | pattern, flags | p.pattern.flags, code,\r | |
| 594 | p.pattern.groups-1,\r | |
| 595 | groupindex, indexgroup\r | |
| 596 | )\r |