/[pcre]/code/trunk/pcre_internal.h
ViewVC logotype

Diff of /code/trunk/pcre_internal.h

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 598 by ph10, Sat May 7 15:37:31 2011 UTC revision 677 by ph10, Sun Aug 28 10:50:07 2011 UTC
# Line 7  Line 7 
7  and semantics are as close as possible to those of the Perl 5 language.  and semantics are as close as possible to those of the Perl 5 language.
8    
9                         Written by Philip Hazel                         Written by Philip Hazel
10             Copyright (c) 1997-2010 University of Cambridge             Copyright (c) 1997-2011 University of Cambridge
11    
12  -----------------------------------------------------------------------------  -----------------------------------------------------------------------------
13  Redistribution and use in source and binary forms, with or without  Redistribution and use in source and binary forms, with or without
# Line 582  Standard C system should have one. */ Line 582  Standard C system should have one. */
582  #endif  #endif
583    
584    
 /* These are the public options that can change during matching. */  
   
 #define PCRE_IMS (PCRE_CASELESS|PCRE_MULTILINE|PCRE_DOTALL)  
   
585  /* Private flags containing information about the compiled regex. They used to  /* Private flags containing information about the compiled regex. They used to
586  live at the top end of the options word, but that got almost full, so now they  live at the top end of the options word, but that got almost full, so now they
587  are in a 16-bit flags word. From release 8.00, PCRE_NOPARTIAL is unused, as  are in a 16-bit flags word. From release 8.00, PCRE_NOPARTIAL is unused, as
# Line 599  compatibility. */ Line 595  compatibility. */
595  #define PCRE_JCHANGED      0x0010  /* j option used in regex */  #define PCRE_JCHANGED      0x0010  /* j option used in regex */
596  #define PCRE_HASCRORLF     0x0020  /* explicit \r or \n in pattern */  #define PCRE_HASCRORLF     0x0020  /* explicit \r or \n in pattern */
597    
598  /* Options for the "extra" block produced by pcre_study(). */  /* Flags for the "extra" block produced by pcre_study(). */
599    
600  #define PCRE_STUDY_MAPPED   0x01     /* a map of starting chars exists */  #define PCRE_STUDY_MAPPED  0x0001  /* a map of starting chars exists */
601  #define PCRE_STUDY_MINLEN   0x02     /* a minimum length field exists */  #define PCRE_STUDY_MINLEN  0x0002  /* a minimum length field exists */
602    
603  /* Masks for identifying the public options that are permitted at compile  /* Masks for identifying the public options that are permitted at compile
604  time, run time, or study time, respectively. */  time, run time, or study time, respectively. */
# Line 628  time, run time, or study time, respectiv Line 624  time, run time, or study time, respectiv
624     PCRE_DFA_RESTART|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \     PCRE_DFA_RESTART|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \
625     PCRE_NO_START_OPTIMIZE)     PCRE_NO_START_OPTIMIZE)
626    
627  #define PUBLIC_STUDY_OPTIONS 0   /* None defined */  #define PUBLIC_STUDY_OPTIONS \
628       PCRE_STUDY_JIT_COMPILE
629    
630  /* Magic number to provide a small check against being handed junk. Also used  /* Magic number to provide a small check against being handed junk. Also used
631  to detect whether a pattern was compiled on a host of different endianness. */  to detect whether a pattern was compiled on a host of different endianness. */
# Line 1299  enum { Line 1296  enum {
1296    OP_WHITESPACE,         /*  9 \s */    OP_WHITESPACE,         /*  9 \s */
1297    OP_NOT_WORDCHAR,       /* 10 \W */    OP_NOT_WORDCHAR,       /* 10 \W */
1298    OP_WORDCHAR,           /* 11 \w */    OP_WORDCHAR,           /* 11 \w */
1299    
1300    OP_ANY,            /* 12 Match any character except newline */    OP_ANY,            /* 12 Match any character except newline */
1301    OP_ALLANY,         /* 13 Match any character */    OP_ALLANY,         /* 13 Match any character */
1302    OP_ANYBYTE,        /* 14 Match any byte (\C); different to OP_ANY for UTF-8 */    OP_ANYBYTE,        /* 14 Match any byte (\C); different to OP_ANY for UTF-8 */
# Line 1313  enum { Line 1311  enum {
1311    OP_EODN,           /* 23 End of data or \n at end of data: \Z. */    OP_EODN,           /* 23 End of data or \n at end of data: \Z. */
1312    OP_EOD,            /* 24 End of data: \z */    OP_EOD,            /* 24 End of data: \z */
1313    
1314    OP_OPT,            /* 25 Set runtime options */    OP_CIRC,           /* 25 Start of line - not multiline */
1315    OP_CIRC,           /* 26 Start of line - varies with multiline switch */    OP_CIRCM,          /* 26 Start of line - multiline */
1316    OP_DOLL,           /* 27 End of line - varies with multiline switch */    OP_DOLL,           /* 27 End of line - not multiline */
1317    OP_CHAR,           /* 28 Match one character, casefully */    OP_DOLLM,          /* 28 End of line - multiline */
1318    OP_CHARNC,         /* 29 Match one character, caselessly */    OP_CHAR,           /* 29 Match one character, casefully */
1319    OP_NOT,            /* 30 Match one character, not the following one */    OP_CHARI,          /* 30 Match one character, caselessly */
1320      OP_NOT,            /* 31 Match one character, not the given one, casefully */
1321    OP_STAR,           /* 31 The maximizing and minimizing versions of */    OP_NOTI,           /* 32 Match one character, not the given one, caselessly */
1322    OP_MINSTAR,        /* 32 these six opcodes must come in pairs, with */  
1323    OP_PLUS,           /* 33 the minimizing one second. */    /* The following sets of 13 opcodes must always be kept in step because
1324    OP_MINPLUS,        /* 34 This first set applies to single characters.*/    the offset from the first one is used to generate the others. */
1325    OP_QUERY,          /* 35 */  
1326    OP_MINQUERY,       /* 36 */    /**** Single characters, caseful, must precede the caseless ones ****/
1327    
1328    OP_UPTO,           /* 37 From 0 to n matches */    OP_STAR,           /* 33 The maximizing and minimizing versions of */
1329    OP_MINUPTO,        /* 38 */    OP_MINSTAR,        /* 34 these six opcodes must come in pairs, with */
1330    OP_EXACT,          /* 39 Exactly n matches */    OP_PLUS,           /* 35 the minimizing one second. */
1331      OP_MINPLUS,        /* 36 */
1332    OP_POSSTAR,        /* 40 Possessified star */    OP_QUERY,          /* 37 */
1333    OP_POSPLUS,        /* 41 Possessified plus */    OP_MINQUERY,       /* 38 */
1334    OP_POSQUERY,       /* 42 Posesssified query */  
1335    OP_POSUPTO,        /* 43 Possessified upto */    OP_UPTO,           /* 39 From 0 to n matches of one character, caseful*/
1336      OP_MINUPTO,        /* 40 */
1337    OP_NOTSTAR,        /* 44 The maximizing and minimizing versions of */    OP_EXACT,          /* 41 Exactly n matches */
1338    OP_NOTMINSTAR,     /* 45 these six opcodes must come in pairs, with */  
1339    OP_NOTPLUS,        /* 46 the minimizing one second. They must be in */    OP_POSSTAR,        /* 42 Possessified star, caseful */
1340    OP_NOTMINPLUS,     /* 47 exactly the same order as those above. */    OP_POSPLUS,        /* 43 Possessified plus, caseful */
1341    OP_NOTQUERY,       /* 48 This set applies to "not" single characters. */    OP_POSQUERY,       /* 44 Posesssified query, caseful */
1342    OP_NOTMINQUERY,    /* 49 */    OP_POSUPTO,        /* 45 Possessified upto, caseful */
1343    
1344    OP_NOTUPTO,        /* 50 From 0 to n matches */    /**** Single characters, caseless, must follow the caseful ones */
1345    OP_NOTMINUPTO,     /* 51 */  
1346    OP_NOTEXACT,       /* 52 Exactly n matches */    OP_STARI,          /* 46 */
1347      OP_MINSTARI,       /* 47 */
1348    OP_NOTPOSSTAR,     /* 53 Possessified versions */    OP_PLUSI,          /* 48 */
1349    OP_NOTPOSPLUS,     /* 54 */    OP_MINPLUSI,       /* 49 */
1350    OP_NOTPOSQUERY,    /* 55 */    OP_QUERYI,         /* 50 */
1351    OP_NOTPOSUPTO,     /* 56 */    OP_MINQUERYI,      /* 51 */
1352    
1353    OP_TYPESTAR,       /* 57 The maximizing and minimizing versions of */    OP_UPTOI,          /* 52 From 0 to n matches of one character, caseless */
1354    OP_TYPEMINSTAR,    /* 58 these six opcodes must come in pairs, with */    OP_MINUPTOI,       /* 53 */
1355    OP_TYPEPLUS,       /* 59 the minimizing one second. These codes must */    OP_EXACTI,         /* 54 */
1356    OP_TYPEMINPLUS,    /* 60 be in exactly the same order as those above. */  
1357    OP_TYPEQUERY,      /* 61 This set applies to character types such as \d */    OP_POSSTARI,       /* 55 Possessified star, caseless */
1358    OP_TYPEMINQUERY,   /* 62 */    OP_POSPLUSI,       /* 56 Possessified plus, caseless */
1359      OP_POSQUERYI,      /* 57 Posesssified query, caseless */
1360    OP_TYPEUPTO,       /* 63 From 0 to n matches */    OP_POSUPTOI,       /* 58 Possessified upto, caseless */
1361    OP_TYPEMINUPTO,    /* 64 */  
1362    OP_TYPEEXACT,      /* 65 Exactly n matches */    /**** The negated ones must follow the non-negated ones, and match them ****/
1363      /**** Negated single character, caseful; must precede the caseless ones ****/
1364    OP_TYPEPOSSTAR,    /* 66 Possessified versions */  
1365    OP_TYPEPOSPLUS,    /* 67 */    OP_NOTSTAR,        /* 59 The maximizing and minimizing versions of */
1366    OP_TYPEPOSQUERY,   /* 68 */    OP_NOTMINSTAR,     /* 60 these six opcodes must come in pairs, with */
1367    OP_TYPEPOSUPTO,    /* 69 */    OP_NOTPLUS,        /* 61 the minimizing one second. They must be in */
1368      OP_NOTMINPLUS,     /* 62 exactly the same order as those above. */
1369    OP_CRSTAR,         /* 70 The maximizing and minimizing versions of */    OP_NOTQUERY,       /* 63 */
1370    OP_CRMINSTAR,      /* 71 all these opcodes must come in pairs, with */    OP_NOTMINQUERY,    /* 64 */
1371    OP_CRPLUS,         /* 72 the minimizing one second. These codes must */  
1372    OP_CRMINPLUS,      /* 73 be in exactly the same order as those above. */    OP_NOTUPTO,        /* 65 From 0 to n matches, caseful */
1373    OP_CRQUERY,        /* 74 These are for character classes and back refs */    OP_NOTMINUPTO,     /* 66 */
1374    OP_CRMINQUERY,     /* 75 */    OP_NOTEXACT,       /* 67 Exactly n matches */
1375    OP_CRRANGE,        /* 76 These are different to the three sets above. */  
1376    OP_CRMINRANGE,     /* 77 */    OP_NOTPOSSTAR,     /* 68 Possessified versions, caseful */
1377      OP_NOTPOSPLUS,     /* 69 */
1378    OP_CLASS,          /* 78 Match a character class, chars < 256 only */    OP_NOTPOSQUERY,    /* 70 */
1379    OP_NCLASS,         /* 79 Same, but the bitmap was created from a negative    OP_NOTPOSUPTO,     /* 71 */
1380                             class - the difference is relevant only when a UTF-8  
1381                             character > 255 is encountered. */    /**** Negated single character, caseless; must follow the caseful ones ****/
1382    
1383    OP_XCLASS,         /* 80 Extended class for handling UTF-8 chars within the    OP_NOTSTARI,       /* 72 */
1384                             class. This does both positive and negative. */    OP_NOTMINSTARI,    /* 73 */
1385      OP_NOTPLUSI,       /* 74 */
1386    OP_REF,            /* 81 Match a back reference */    OP_NOTMINPLUSI,    /* 75 */
1387    OP_RECURSE,        /* 82 Match a numbered subpattern (possibly recursive) */    OP_NOTQUERYI,      /* 76 */
1388    OP_CALLOUT,        /* 83 Call out to external function if provided */    OP_NOTMINQUERYI,   /* 77 */
1389    
1390    OP_ALT,            /* 84 Start of alternation */    OP_NOTUPTOI,       /* 78 From 0 to n matches, caseless */
1391    OP_KET,            /* 85 End of group that doesn't have an unbounded repeat */    OP_NOTMINUPTOI,    /* 79 */
1392    OP_KETRMAX,        /* 86 These two must remain together and in this */    OP_NOTEXACTI,      /* 80 Exactly n matches */
1393    OP_KETRMIN,        /* 87 order. They are for groups the repeat for ever. */  
1394      OP_NOTPOSSTARI,    /* 81 Possessified versions, caseless */
1395    /* The assertions must come before BRA, CBRA, ONCE, and COND.*/    OP_NOTPOSPLUSI,    /* 82 */
1396      OP_NOTPOSQUERYI,   /* 83 */
1397    OP_ASSERT,         /* 88 Positive lookahead */    OP_NOTPOSUPTOI,    /* 84 */
1398    OP_ASSERT_NOT,     /* 89 Negative lookahead */  
1399    OP_ASSERTBACK,     /* 90 Positive lookbehind */    /**** Character types ****/
1400    OP_ASSERTBACK_NOT, /* 91 Negative lookbehind */  
1401    OP_REVERSE,        /* 92 Move pointer back - used in lookbehind assertions */    OP_TYPESTAR,       /* 85 The maximizing and minimizing versions of */
1402      OP_TYPEMINSTAR,    /* 86 these six opcodes must come in pairs, with */
1403    /* ONCE, BRA, CBRA, and COND must come after the assertions, with ONCE first,    OP_TYPEPLUS,       /* 87 the minimizing one second. These codes must */
1404    as there's a test for >= ONCE for a subpattern that isn't an assertion. */    OP_TYPEMINPLUS,    /* 88 be in exactly the same order as those above. */
1405      OP_TYPEQUERY,      /* 89 */
1406    OP_ONCE,           /* 93 Atomic group */    OP_TYPEMINQUERY,   /* 90 */
1407    OP_BRA,            /* 94 Start of non-capturing bracket */  
1408    OP_CBRA,           /* 95 Start of capturing bracket */    OP_TYPEUPTO,       /* 91 From 0 to n matches */
1409    OP_COND,           /* 96 Conditional group */    OP_TYPEMINUPTO,    /* 92 */
1410      OP_TYPEEXACT,      /* 93 Exactly n matches */
1411    
1412      OP_TYPEPOSSTAR,    /* 94 Possessified versions */
1413      OP_TYPEPOSPLUS,    /* 95 */
1414      OP_TYPEPOSQUERY,   /* 96 */
1415      OP_TYPEPOSUPTO,    /* 97 */
1416    
1417      /* These are used for character classes and back references; only the
1418      first six are the same as the sets above. */
1419    
1420      OP_CRSTAR,         /* 98 The maximizing and minimizing versions of */
1421      OP_CRMINSTAR,      /* 99 all these opcodes must come in pairs, with */
1422      OP_CRPLUS,         /* 100 the minimizing one second. These codes must */
1423      OP_CRMINPLUS,      /* 101 be in exactly the same order as those above. */
1424      OP_CRQUERY,        /* 102 */
1425      OP_CRMINQUERY,     /* 103 */
1426    
1427      OP_CRRANGE,        /* 104 These are different to the three sets above. */
1428      OP_CRMINRANGE,     /* 105 */
1429    
1430      /* End of quantifier opcodes */
1431    
1432      OP_CLASS,          /* 106 Match a character class, chars < 256 only */
1433      OP_NCLASS,         /* 107 Same, but the bitmap was created from a negative
1434                                  class - the difference is relevant only when a
1435                                  UTF-8 character > 255 is encountered. */
1436      OP_XCLASS,         /* 108 Extended class for handling UTF-8 chars within the
1437                                  class. This does both positive and negative. */
1438      OP_REF,            /* 109 Match a back reference, casefully */
1439      OP_REFI,           /* 110 Match a back reference, caselessly */
1440      OP_RECURSE,        /* 111 Match a numbered subpattern (possibly recursive) */
1441      OP_CALLOUT,        /* 112 Call out to external function if provided */
1442    
1443      OP_ALT,            /* 113 Start of alternation */
1444      OP_KET,            /* 114 End of group that doesn't have an unbounded repeat */
1445      OP_KETRMAX,        /* 115 These two must remain together and in this */
1446      OP_KETRMIN,        /* 116 order. They are for groups the repeat for ever. */
1447      OP_KETRPOS,        /* 117 Possessive unlimited repeat. */
1448    
1449      /* The assertions must come before BRA, CBRA, ONCE, and COND, and the four
1450      asserts must remain in order. */
1451    
1452      OP_REVERSE,        /* 118 Move pointer back - used in lookbehind assertions */
1453      OP_ASSERT,         /* 119 Positive lookahead */
1454      OP_ASSERT_NOT,     /* 120 Negative lookahead */
1455      OP_ASSERTBACK,     /* 121 Positive lookbehind */
1456      OP_ASSERTBACK_NOT, /* 122 Negative lookbehind */
1457    
1458      /* ONCE, BRA, BRAPOS, CBRA, CBRAPOS, and COND must come immediately after the
1459      assertions, with ONCE first, as there's a test for >= ONCE for a subpattern
1460      that isn't an assertion. The POS versions must immediately follow the non-POS
1461      versions in each case. */
1462    
1463      OP_ONCE,           /* 123 Atomic group */
1464      OP_BRA,            /* 124 Start of non-capturing bracket */
1465      OP_BRAPOS,         /* 125 Ditto, with unlimited, possessive repeat */
1466      OP_CBRA,           /* 126 Start of capturing bracket */
1467      OP_CBRAPOS,        /* 127 Ditto, with unlimited, possessive repeat */
1468      OP_COND,           /* 128 Conditional group */
1469    
1470    /* These three must follow the previous three, in the same order. There's a    /* These five must follow the previous five, in the same order. There's a
1471    check for >= SBRA to distinguish the two sets. */    check for >= SBRA to distinguish the two sets. */
1472    
1473    OP_SBRA,           /* 97 Start of non-capturing bracket, check empty  */    OP_SBRA,           /* 129 Start of non-capturing bracket, check empty  */
1474    OP_SCBRA,          /* 98 Start of capturing bracket, check empty */    OP_SBRAPOS,        /* 130 Ditto, with unlimited, possessive repeat */
1475    OP_SCOND,          /* 99 Conditional group, check empty */    OP_SCBRA,          /* 131 Start of capturing bracket, check empty */
1476      OP_SCBRAPOS,       /* 132 Ditto, with unlimited, possessive repeat */
1477      OP_SCOND,          /* 133 Conditional group, check empty */
1478    
1479    /* The next two pairs must (respectively) be kept together. */    /* The next two pairs must (respectively) be kept together. */
1480    
1481    OP_CREF,           /* 100 Used to hold a capture number as condition */    OP_CREF,           /* 134 Used to hold a capture number as condition */
1482    OP_NCREF,          /* 101 Same, but generated by a name reference*/    OP_NCREF,          /* 135 Same, but generated by a name reference*/
1483    OP_RREF,           /* 102 Used to hold a recursion number as condition */    OP_RREF,           /* 136 Used to hold a recursion number as condition */
1484    OP_NRREF,          /* 103 Same, but generated by a name reference*/    OP_NRREF,          /* 137 Same, but generated by a name reference*/
1485    OP_DEF,            /* 104 The DEFINE condition */    OP_DEF,            /* 138 The DEFINE condition */
1486    
1487    OP_BRAZERO,        /* 105 These two must remain together and in this */    OP_BRAZERO,        /* 139 These two must remain together and in this */
1488    OP_BRAMINZERO,     /* 106 order. */    OP_BRAMINZERO,     /* 140 order. */
1489      OP_BRAPOSZERO,     /* 141 */
1490    
1491    /* These are backtracking control verbs */    /* These are backtracking control verbs */
1492    
1493    OP_MARK,           /* 107 always has an argument */    OP_MARK,           /* 142 always has an argument */
1494    OP_PRUNE,          /* 108 */    OP_PRUNE,          /* 143 */
1495    OP_PRUNE_ARG,      /* 109 same, but with argument */    OP_PRUNE_ARG,      /* 144 same, but with argument */
1496    OP_SKIP,           /* 110 */    OP_SKIP,           /* 145 */
1497    OP_SKIP_ARG,       /* 111 same, but with argument */    OP_SKIP_ARG,       /* 146 same, but with argument */
1498    OP_THEN,           /* 112 */    OP_THEN,           /* 147 */
1499    OP_THEN_ARG,       /* 113 same, but with argument */    OP_THEN_ARG,       /* 148 same, but with argument */
1500    OP_COMMIT,         /* 114 */    OP_COMMIT,         /* 149 */
1501    
1502    /* These are forced failure and success verbs */    /* These are forced failure and success verbs */
1503    
1504    OP_FAIL,           /* 115 */    OP_FAIL,           /* 150 */
1505    OP_ACCEPT,         /* 116 */    OP_ACCEPT,         /* 151 */
1506    OP_CLOSE,          /* 117 Used before OP_ACCEPT to close open captures */    OP_ASSERT_ACCEPT,  /* 152 Used inside assertions */
1507      OP_CLOSE,          /* 153 Used before OP_ACCEPT to close open captures */
1508    
1509    /* This is used to skip a subpattern with a {0} quantifier */    /* This is used to skip a subpattern with a {0} quantifier */
1510    
1511    OP_SKIPZERO,       /* 118 */    OP_SKIPZERO,       /* 154 */
1512    
1513    /* This is not an opcode, but is used to check that tables indexed by opcode    /* This is not an opcode, but is used to check that tables indexed by opcode
1514    are the correct length, in order to catch updating errors - there have been    are the correct length, in order to catch updating errors - there have been
# Line 1462  called "coptable" and "poptable" in pcre Line 1523  called "coptable" and "poptable" in pcre
1523    
1524    
1525  /* This macro defines textual names for all the opcodes. These are used only  /* This macro defines textual names for all the opcodes. These are used only
1526  for debugging. The macro is referenced only in pcre_printint.c. */  for debugging, and some of them are only partial names. The macro is referenced
1527    only in pcre_printint.c, which fills out the full names in many cases (and in
1528    some cases doesn't actually use these names at all). */
1529    
1530  #define OP_NAME_LIST \  #define OP_NAME_LIST \
1531    "End", "\\A", "\\G", "\\K", "\\B", "\\b", "\\D", "\\d",         \    "End", "\\A", "\\G", "\\K", "\\B", "\\b", "\\D", "\\d",         \
1532    "\\S", "\\s", "\\W", "\\w", "Any", "AllAny", "Anybyte",         \    "\\S", "\\s", "\\W", "\\w", "Any", "AllAny", "Anybyte",         \
1533    "notprop", "prop", "\\R", "\\H", "\\h", "\\V", "\\v",           \    "notprop", "prop", "\\R", "\\H", "\\h", "\\V", "\\v",           \
1534    "extuni",  "\\Z", "\\z",                                        \    "extuni",  "\\Z", "\\z",                                        \
1535    "Opt", "^", "$", "char", "charnc", "not",                       \    "^", "^", "$", "$", "char", "chari", "not", "noti",             \
1536    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \    "*", "*?", "+", "+?", "?", "??",                                \
1537      "{", "{", "{",                                                  \
1538    "*+","++", "?+", "{",                                           \    "*+","++", "?+", "{",                                           \
1539    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \    "*", "*?", "+", "+?", "?", "??",                                \
1540      "{", "{", "{",                                                  \
1541      "*+","++", "?+", "{",                                           \
1542      "*", "*?", "+", "+?", "?", "??",                                \
1543      "{", "{", "{",                                                  \
1544      "*+","++", "?+", "{",                                           \
1545      "*", "*?", "+", "+?", "?", "??",                                \
1546      "{", "{", "{",                                                  \
1547    "*+","++", "?+", "{",                                           \    "*+","++", "?+", "{",                                           \
1548    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \
1549    "*+","++", "?+", "{",                                           \    "*+","++", "?+", "{",                                           \
1550    "*", "*?", "+", "+?", "?", "??", "{", "{",                      \    "*", "*?", "+", "+?", "?", "??", "{", "{",                      \
1551    "class", "nclass", "xclass", "Ref", "Recurse", "Callout",       \    "class", "nclass", "xclass", "Ref", "Refi",                     \
1552    "Alt", "Ket", "KetRmax", "KetRmin", "Assert", "Assert not",     \    "Recurse", "Callout",                                           \
1553    "AssertB", "AssertB not", "Reverse",                            \    "Alt", "Ket", "KetRmax", "KetRmin", "KetRpos",                  \
1554    "Once", "Bra", "CBra", "Cond", "SBra", "SCBra", "SCond",        \    "Reverse", "Assert", "Assert not", "AssertB", "AssertB not",    \
1555      "Once",                                                         \
1556      "Bra", "BraPos", "CBra", "CBraPos",                             \
1557      "Cond",                                                         \
1558      "SBra", "SBraPos", "SCBra", "SCBraPos",                         \
1559      "SCond",                                                        \
1560    "Cond ref", "Cond nref", "Cond rec", "Cond nrec", "Cond def",   \    "Cond ref", "Cond nref", "Cond rec", "Cond nrec", "Cond def",   \
1561    "Brazero", "Braminzero",                                        \    "Brazero", "Braminzero", "Braposzero",                          \
1562    "*MARK", "*PRUNE", "*PRUNE", "*SKIP", "*SKIP",                  \    "*MARK", "*PRUNE", "*PRUNE", "*SKIP", "*SKIP",                  \
1563    "*THEN", "*THEN", "*COMMIT", "*FAIL", "*ACCEPT",                \    "*THEN", "*THEN", "*COMMIT", "*FAIL",                           \
1564      "*ACCEPT", "*ASSERT_ACCEPT",                                    \
1565    "Close", "Skip zero"    "Close", "Skip zero"
1566    
1567    
# Line 1505  in UTF-8 mode. The code that uses this t Line 1582  in UTF-8 mode. The code that uses this t
1582    3, 3,                          /* \P, \p                                 */ \    3, 3,                          /* \P, \p                                 */ \
1583    1, 1, 1, 1, 1,                 /* \R, \H, \h, \V, \v                     */ \    1, 1, 1, 1, 1,                 /* \R, \H, \h, \V, \v                     */ \
1584    1,                             /* \X                                     */ \    1,                             /* \X                                     */ \
1585    1, 1, 2, 1, 1,                 /* \Z, \z, Opt, ^, $                      */ \    1, 1, 1, 1, 1, 1,              /* \Z, \z, ^, ^M, $, $M                   */ \
1586    2,                             /* Char  - the minimum length             */ \    2,                             /* Char  - the minimum length             */ \
1587    2,                             /* Charnc  - the minimum length           */ \    2,                             /* Chari  - the minimum length            */ \
1588    2,                             /* not                                    */ \    2,                             /* not                                    */ \
1589    /* Positive single-char repeats                            ** These are  */ \    2,                             /* noti                                   */ \
1590    2, 2, 2, 2, 2, 2,              /* *, *?, +, +?, ?, ??      ** minima in  */ \    /* Positive single-char repeats                             ** These are */ \
1591    4, 4, 4,                       /* upto, minupto, exact     ** UTF-8 mode */ \    2, 2, 2, 2, 2, 2,              /* *, *?, +, +?, ?, ??       ** minima in */ \
1592      4, 4, 4,                       /* upto, minupto, exact      ** mode      */ \
1593    2, 2, 2, 4,                    /* *+, ++, ?+, upto+                      */ \    2, 2, 2, 4,                    /* *+, ++, ?+, upto+                      */ \
1594      2, 2, 2, 2, 2, 2,              /* *I, *?I, +I, +?I, ?I, ??I ** UTF-8     */ \
1595      4, 4, 4,                       /* upto I, minupto I, exact I             */ \
1596      2, 2, 2, 4,                    /* *+I, ++I, ?+I, upto+I                  */ \
1597    /* Negative single-char repeats - only for chars < 256                   */ \    /* Negative single-char repeats - only for chars < 256                   */ \
1598    2, 2, 2, 2, 2, 2,              /* NOT *, *?, +, +?, ?, ??                */ \    2, 2, 2, 2, 2, 2,              /* NOT *, *?, +, +?, ?, ??                */ \
1599    4, 4, 4,                       /* NOT upto, minupto, exact               */ \    4, 4, 4,                       /* NOT upto, minupto, exact               */ \
1600    2, 2, 2, 4,                    /* Possessive *, +, ?, upto               */ \    2, 2, 2, 4,                    /* Possessive NOT *, +, ?, upto           */ \
1601      2, 2, 2, 2, 2, 2,              /* NOT *I, *?I, +I, +?I, ?I, ??I          */ \
1602      4, 4, 4,                       /* NOT upto I, minupto I, exact I         */ \
1603      2, 2, 2, 4,                    /* Possessive NOT *I, +I, ?I, upto I      */ \
1604    /* Positive type repeats                                                 */ \    /* Positive type repeats                                                 */ \
1605    2, 2, 2, 2, 2, 2,              /* Type *, *?, +, +?, ?, ??               */ \    2, 2, 2, 2, 2, 2,              /* Type *, *?, +, +?, ?, ??               */ \
1606    4, 4, 4,                       /* Type upto, minupto, exact              */ \    4, 4, 4,                       /* Type upto, minupto, exact              */ \
# Line 1528  in UTF-8 mode. The code that uses this t Line 1612  in UTF-8 mode. The code that uses this t
1612   33,                             /* NCLASS                                 */ \   33,                             /* NCLASS                                 */ \
1613    0,                             /* XCLASS - variable length               */ \    0,                             /* XCLASS - variable length               */ \
1614    3,                             /* REF                                    */ \    3,                             /* REF                                    */ \
1615      3,                             /* REFI                                   */ \
1616    1+LINK_SIZE,                   /* RECURSE                                */ \    1+LINK_SIZE,                   /* RECURSE                                */ \
1617    2+2*LINK_SIZE,                 /* CALLOUT                                */ \    2+2*LINK_SIZE,                 /* CALLOUT                                */ \
1618    1+LINK_SIZE,                   /* Alt                                    */ \    1+LINK_SIZE,                   /* Alt                                    */ \
1619    1+LINK_SIZE,                   /* Ket                                    */ \    1+LINK_SIZE,                   /* Ket                                    */ \
1620    1+LINK_SIZE,                   /* KetRmax                                */ \    1+LINK_SIZE,                   /* KetRmax                                */ \
1621    1+LINK_SIZE,                   /* KetRmin                                */ \    1+LINK_SIZE,                   /* KetRmin                                */ \
1622      1+LINK_SIZE,                   /* KetRpos                                */ \
1623      1+LINK_SIZE,                   /* Reverse                                */ \
1624    1+LINK_SIZE,                   /* Assert                                 */ \    1+LINK_SIZE,                   /* Assert                                 */ \
1625    1+LINK_SIZE,                   /* Assert not                             */ \    1+LINK_SIZE,                   /* Assert not                             */ \
1626    1+LINK_SIZE,                   /* Assert behind                          */ \    1+LINK_SIZE,                   /* Assert behind                          */ \
1627    1+LINK_SIZE,                   /* Assert behind not                      */ \    1+LINK_SIZE,                   /* Assert behind not                      */ \
   1+LINK_SIZE,                   /* Reverse                                */ \  
1628    1+LINK_SIZE,                   /* ONCE                                   */ \    1+LINK_SIZE,                   /* ONCE                                   */ \
1629    1+LINK_SIZE,                   /* BRA                                    */ \    1+LINK_SIZE,                   /* BRA                                    */ \
1630      1+LINK_SIZE,                   /* BRAPOS                                 */ \
1631    3+LINK_SIZE,                   /* CBRA                                   */ \    3+LINK_SIZE,                   /* CBRA                                   */ \
1632      3+LINK_SIZE,                   /* CBRAPOS                                */ \
1633    1+LINK_SIZE,                   /* COND                                   */ \    1+LINK_SIZE,                   /* COND                                   */ \
1634    1+LINK_SIZE,                   /* SBRA                                   */ \    1+LINK_SIZE,                   /* SBRA                                   */ \
1635      1+LINK_SIZE,                   /* SBRAPOS                                */ \
1636    3+LINK_SIZE,                   /* SCBRA                                  */ \    3+LINK_SIZE,                   /* SCBRA                                  */ \
1637      3+LINK_SIZE,                   /* SCBRAPOS                               */ \
1638    1+LINK_SIZE,                   /* SCOND                                  */ \    1+LINK_SIZE,                   /* SCOND                                  */ \
1639    3, 3,                          /* CREF, NCREF                            */ \    3, 3,                          /* CREF, NCREF                            */ \
1640    3, 3,                          /* RREF, NRREF                            */ \    3, 3,                          /* RREF, NRREF                            */ \
1641    1,                             /* DEF                                    */ \    1,                             /* DEF                                    */ \
1642    1, 1,                          /* BRAZERO, BRAMINZERO                    */ \    1, 1, 1,                       /* BRAZERO, BRAMINZERO, BRAPOSZERO        */ \
1643    3, 1, 3,                       /* MARK, PRUNE, PRUNE_ARG                 */ \    3, 1, 3,                       /* MARK, PRUNE, PRUNE_ARG                 */ \
1644    1, 3,                          /* SKIP, SKIP_ARG                         */ \    1, 3,                          /* SKIP, SKIP_ARG                         */ \
1645    1+LINK_SIZE, 3+LINK_SIZE,      /* THEN, THEN_ARG                         */ \    1+LINK_SIZE, 3+LINK_SIZE,      /* THEN, THEN_ARG                         */ \
1646    1, 1, 1, 3, 1                  /* COMMIT, FAIL, ACCEPT, CLOSE, SKIPZERO  */    1, 1, 1, 1,                    /* COMMIT, FAIL, ACCEPT, ASSERT_ACCEPT    */ \
1647      3, 1                           /* CLOSE, SKIPZERO  */
1648    
1649  /* A magic value for OP_RREF and OP_NRREF to indicate the "any recursion"  /* A magic value for OP_RREF and OP_NRREF to indicate the "any recursion"
1650  condition. */  condition. */
# Line 1571  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4, Line 1661  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4,
1661         ERR30, ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39,         ERR30, ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39,
1662         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,
1663         ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,         ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,
1664         ERR60, ERR61, ERR62, ERR63, ERR64, ERR65, ERR66, ERR67, ERR68,         ERR60, ERR61, ERR62, ERR63, ERR64, ERR65, ERR66, ERR67, ERR68, ERR69,
1665         ERRCOUNT };         ERRCOUNT };
1666    
1667  /* The real format of the start of the pcre block; the index of names and the  /* The real format of the start of the pcre block; the index of names and the
# Line 1652  typedef struct compile_data { Line 1742  typedef struct compile_data {
1742    int  final_bracount;          /* Saved value after first pass */    int  final_bracount;          /* Saved value after first pass */
1743    int  top_backref;             /* Maximum back reference */    int  top_backref;             /* Maximum back reference */
1744    unsigned int backref_map;     /* Bitmap of low back refs */    unsigned int backref_map;     /* Bitmap of low back refs */
1745      int  assert_depth;            /* Depth of nested assertions */
1746    int  external_options;        /* External (initial) options */    int  external_options;        /* External (initial) options */
1747    int  external_flags;          /* External flag bits to be set */    int  external_flags;          /* External flag bits to be set */
1748    int  req_varyopt;             /* "After variable item" flag for reqbyte */    int  req_varyopt;             /* "After variable item" flag for reqbyte */
# Line 1663  typedef struct compile_data { Line 1754  typedef struct compile_data {
1754  } compile_data;  } compile_data;
1755    
1756  /* Structure for maintaining a chain of pointers to the currently incomplete  /* Structure for maintaining a chain of pointers to the currently incomplete
1757  branches, for testing for left recursion. */  branches, for testing for left recursion while compiling. */
1758    
1759  typedef struct branch_chain {  typedef struct branch_chain {
1760    struct branch_chain *outer;    struct branch_chain *outer;
# Line 1671  typedef struct branch_chain { Line 1762  typedef struct branch_chain {
1762  } branch_chain;  } branch_chain;
1763    
1764  /* Structure for items in a linked list that represents an explicit recursive  /* Structure for items in a linked list that represents an explicit recursive
1765  call within the pattern. */  call within the pattern; used by pcre_exec(). */
1766    
1767  typedef struct recursion_info {  typedef struct recursion_info {
1768    struct recursion_info *prevrec; /* Previous recursion record (or NULL) */    struct recursion_info *prevrec; /* Previous recursion record (or NULL) */
1769    int group_num;                /* Number of group that was called */    int group_num;                  /* Number of group that was called */
1770    const uschar *after_call;     /* "Return value": points after the call in the expr */    int *offset_save;               /* Pointer to start of saved offsets */
1771    int *offset_save;             /* Pointer to start of saved offsets */    int saved_max;                  /* Number of saved offsets */
1772    int saved_max;                /* Number of saved offsets */    USPTR subject_position;         /* Position at start of recursion */
   int save_offset_top;          /* Current value of offset_top */  
1773  } recursion_info;  } recursion_info;
1774    
1775    /* A similar structure for pcre_dfa_exec(). */
1776    
1777    typedef struct dfa_recursion_info {
1778      struct dfa_recursion_info *prevrec;
1779      int group_num;
1780      USPTR subject_position;
1781    } dfa_recursion_info;
1782    
1783  /* Structure for building a chain of data for holding the values of the subject  /* Structure for building a chain of data for holding the values of the subject
1784  pointer at the start of each subpattern, so as to detect when an empty string  pointer at the start of each subpattern, so as to detect when an empty string
1785  has been matched by a subpattern - to break infinite loops. */  has been matched by a subpattern - to break infinite loops; used by
1786    pcre_exec(). */
1787    
1788  typedef struct eptrblock {  typedef struct eptrblock {
1789    struct eptrblock *epb_prev;    struct eptrblock *epb_prev;
# Line 1708  typedef struct match_data { Line 1807  typedef struct match_data {
1807    int    name_entry_size;       /* Size of entry in names table */    int    name_entry_size;       /* Size of entry in names table */
1808    uschar *name_table;           /* Table of names */    uschar *name_table;           /* Table of names */
1809    uschar nl[4];                 /* Newline string when fixed */    uschar nl[4];                 /* Newline string when fixed */
1810    const uschar *lcc;            /* Points to lower casing table */    const  uschar *lcc;           /* Points to lower casing table */
1811    const uschar *ctypes;         /* Points to table of type maps */    const  uschar *ctypes;        /* Points to table of type maps */
1812    BOOL   offset_overflow;       /* Set if too many extractions */    BOOL   offset_overflow;       /* Set if too many extractions */
1813    BOOL   notbol;                /* NOTBOL flag */    BOOL   notbol;                /* NOTBOL flag */
1814    BOOL   noteol;                /* NOTEOL flag */    BOOL   noteol;                /* NOTEOL flag */
# Line 1721  typedef struct match_data { Line 1820  typedef struct match_data {
1820    BOOL   notempty_atstart;      /* Empty string match at start not wanted */    BOOL   notempty_atstart;      /* Empty string match at start not wanted */
1821    BOOL   hitend;                /* Hit the end of the subject at some point */    BOOL   hitend;                /* Hit the end of the subject at some point */
1822    BOOL   bsr_anycrlf;           /* \R is just any CRLF, not full Unicode */    BOOL   bsr_anycrlf;           /* \R is just any CRLF, not full Unicode */
1823    const uschar *start_code;     /* For use when recursing */    const  uschar *start_code;    /* For use when recursing */
1824    USPTR  start_subject;         /* Start of the subject string */    USPTR  start_subject;         /* Start of the subject string */
1825    USPTR  end_subject;           /* End of the subject string */    USPTR  end_subject;           /* End of the subject string */
1826    USPTR  start_match_ptr;       /* Start of matched string */    USPTR  start_match_ptr;       /* Start of matched string */
# Line 1731  typedef struct match_data { Line 1830  typedef struct match_data {
1830    int    end_offset_top;        /* Highwater mark at end of match */    int    end_offset_top;        /* Highwater mark at end of match */
1831    int    capture_last;          /* Most recent capture number */    int    capture_last;          /* Most recent capture number */
1832    int    start_offset;          /* The start offset value */    int    start_offset;          /* The start offset value */
1833      int    match_function_type;   /* Set for certain special calls of MATCH() */
1834    eptrblock *eptrchain;         /* Chain of eptrblocks for tail recursions */    eptrblock *eptrchain;         /* Chain of eptrblocks for tail recursions */
1835    int    eptrn;                 /* Next free eptrblock */    int    eptrn;                 /* Next free eptrblock */
1836    recursion_info *recursive;    /* Linked list of recursion data */    recursion_info *recursive;    /* Linked list of recursion data */
1837    void  *callout_data;          /* To pass back to callouts */    void  *callout_data;          /* To pass back to callouts */
1838    const uschar *mark;           /* Mark pointer to pass back */    const  uschar *mark;          /* Mark pointer to pass back */
1839      const  uschar *once_target;   /* Where to back up to for atomic groups */
1840  } match_data;  } match_data;
1841    
1842  /* A similar structure is used for the same purpose by the DFA matching  /* A similar structure is used for the same purpose by the DFA matching
1843  functions. */  functions. */
1844    
1845  typedef struct dfa_match_data {  typedef struct dfa_match_data {
1846    const uschar *start_code;     /* Start of the compiled pattern */    const uschar *start_code;      /* Start of the compiled pattern */
1847    const uschar *start_subject;  /* Start of the subject string */    const uschar *start_subject;   /* Start of the subject string */
1848    const uschar *end_subject;    /* End of subject string */    const uschar *end_subject;     /* End of subject string */
1849    const uschar *start_used_ptr; /* Earliest consulted character */    const uschar *start_used_ptr;  /* Earliest consulted character */
1850    const uschar *tables;         /* Character tables */    const uschar *tables;          /* Character tables */
1851    int   start_offset;           /* The start offset value */    int   start_offset;            /* The start offset value */
1852    int   moptions;               /* Match options */    int   moptions;                /* Match options */
1853    int   poptions;               /* Pattern options */    int   poptions;                /* Pattern options */
1854    int    nltype;                /* Newline type */    int    nltype;                 /* Newline type */
1855    int    nllen;                 /* Newline string length */    int    nllen;                  /* Newline string length */
1856    uschar nl[4];                 /* Newline string when fixed */    uschar nl[4];                  /* Newline string when fixed */
1857    void  *callout_data;          /* To pass back to callouts */    void  *callout_data;           /* To pass back to callouts */
1858      dfa_recursion_info *recursive; /* Linked list of recursion data */
1859  } dfa_match_data;  } dfa_match_data;
1860    
1861  /* Bit definitions for entries in the pcre_ctypes table. */  /* Bit definitions for entries in the pcre_ctypes table. */
# Line 1811  extern const int    _pcre_utf8_table2[]; Line 1913  extern const int    _pcre_utf8_table2[];
1913  extern const int    _pcre_utf8_table3[];  extern const int    _pcre_utf8_table3[];
1914  extern const uschar _pcre_utf8_table4[];  extern const uschar _pcre_utf8_table4[];
1915    
1916    #ifdef SUPPORT_JIT
1917    extern const uschar _pcre_utf8_char_sizes[];
1918    #endif
1919    
1920  extern const int    _pcre_utf8_table1_size;  extern const int    _pcre_utf8_table1_size;
1921    
1922  extern const char   _pcre_utt_names[];  extern const char   _pcre_utt_names[];
# Line 1835  extern int           _pcre_valid_utf8(US Line 1941  extern int           _pcre_valid_utf8(US
1941  extern BOOL          _pcre_was_newline(USPTR, int, USPTR, int *, BOOL);  extern BOOL          _pcre_was_newline(USPTR, int, USPTR, int *, BOOL);
1942  extern BOOL          _pcre_xclass(int, const uschar *);  extern BOOL          _pcre_xclass(int, const uschar *);
1943    
1944    #ifdef SUPPORT_JIT
1945    extern void          _pcre_jit_compile(const real_pcre *, pcre_extra *);
1946    extern int           _pcre_jit_exec(const real_pcre *, void *, PCRE_SPTR,
1947                            int, int, int, int, int *, int);
1948    extern void          _pcre_jit_free(void *);
1949    #endif
1950    
1951  /* Unicode character database (UCD) */  /* Unicode character database (UCD) */
1952    
# Line 1848  extern const ucd_record  _pcre_ucd_recor Line 1960  extern const ucd_record  _pcre_ucd_recor
1960  extern const uschar      _pcre_ucd_stage1[];  extern const uschar      _pcre_ucd_stage1[];
1961  extern const pcre_uint16 _pcre_ucd_stage2[];  extern const pcre_uint16 _pcre_ucd_stage2[];
1962  extern const int         _pcre_ucp_gentype[];  extern const int         _pcre_ucp_gentype[];
1963    #ifdef SUPPORT_JIT
1964    extern const int         _pcre_ucp_typerange[];
1965    #endif
1966    
1967  /* UCD access macros */  /* UCD access macros */
1968    

Legend:
Removed from v.598  
changed lines
  Added in v.677

  ViewVC Help
Powered by ViewVC 1.1.5