Annotation of parser3/src/pcre/internal.h, revision 1.2
1.1 paf 1: /*************************************************
2: * Perl-Compatible Regular Expressions *
3: *************************************************/
4:
5:
6: /* This is a library of functions to support regular expressions whose syntax
7: and semantics are as close as possible to those of the Perl 5 language. See
8: the file Tech.Notes for some information on the internals.
9:
10: Written by: Philip Hazel <ph10@cam.ac.uk>
11:
12: Copyright (c) 1997-1999 University of Cambridge
13:
14: -----------------------------------------------------------------------------
15: Permission is granted to anyone to use this software for any purpose on any
16: computer system, and to redistribute it freely, subject to the following
17: restrictions:
18:
19: 1. This software is distributed in the hope that it will be useful,
20: but WITHOUT ANY WARRANTY; without even the implied warranty of
21: MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
22:
23: 2. The origin of this software must not be misrepresented, either by
24: explicit claim or by omission.
25:
26: 3. Altered versions must be plainly marked as such, and must not be
27: misrepresented as being the original software.
28:
29: 4. If PCRE is embedded in any software that is released under the GNU
30: General Purpose Licence (GPL), then the terms of that licence shall
31: supersede any condition above with which it is incompatible.
32: -----------------------------------------------------------------------------
33: */
34:
35: /* This header contains definitions that are shared between the different
36: modules, but which are not relevant to the outside. */
37:
38: /* To cope with SunOS4 and other systems that lack memmove() but have bcopy(),
39: define a macro for memmove() if USE_BCOPY is defined. */
40:
41: #ifdef USE_BCOPY
42: #undef memmove /* some systems may have a macro */
43: #define memmove(a, b, c) bcopy(b, a, c)
44: #endif
45:
46: /* Standard C headers plus the external interface definition */
47:
48: #include <ctype.h>
49: #include <limits.h>
50: #include <stddef.h>
51: #include <stdio.h>
52: #include <stdlib.h>
53: #include <string.h>
54: #include "pcre.h"
55:
56: /* In case there is no definition of offsetof() provided - though any proper
57: Standard C system should have one. */
58:
59: #ifndef offsetof
60: #define offsetof(p_type,field) ((size_t)&(((p_type *)0)->field))
61: #endif
62:
63: /* These are the public options that can change during matching. */
64:
65: #define PCRE_IMS (PCRE_CASELESS|PCRE_MULTILINE|PCRE_DOTALL)
66:
67: /* Private options flags start at the most significant end of the four bytes,
68: but skip the top bit so we can use ints for convenience without getting tangled
69: with negative values. The public options defined in pcre.h start at the least
70: significant end. Make sure they don't overlap, though now that we have expanded
71: to four bytes there is plenty of space. */
72:
73: #define PCRE_FIRSTSET 0x40000000 /* first_char is set */
74: #define PCRE_REQCHSET 0x20000000 /* req_char is set */
75: #define PCRE_STARTLINE 0x10000000 /* start after \n for multiline */
76: #define PCRE_INGROUP 0x08000000 /* compiling inside a group */
77: #define PCRE_ICHANGED 0x04000000 /* i option changes within regex */
78:
79: /* Options for the "extra" block produced by pcre_study(). */
80:
81: #define PCRE_STUDY_MAPPED 0x01 /* a map of starting chars exists */
82:
83: /* Masks for identifying the public options which are permitted at compile
84: time, run time or study time, respectively. */
85:
86: #define PUBLIC_OPTIONS \
87: (PCRE_CASELESS|PCRE_EXTENDED|PCRE_ANCHORED|PCRE_MULTILINE| \
88: PCRE_DOTALL|PCRE_DOLLAR_ENDONLY|PCRE_EXTRA|PCRE_UNGREEDY)
89:
90: #define PUBLIC_EXEC_OPTIONS \
91: (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY)
92:
93: #define PUBLIC_STUDY_OPTIONS 0 /* None defined */
94:
95: /* Magic number to provide a small check against being handed junk. */
96:
97: #define MAGIC_NUMBER 0x50435245UL /* 'PCRE' */
98:
99: /* Miscellaneous definitions */
100:
101: typedef int BOOL;
102:
1.2 ! paf 103: #ifndef FALSE
1.1 paf 104: #define FALSE 0
1.2 ! paf 105: #endif
! 106: #ifndef TRUE
1.1 paf 107: #define TRUE 1
1.2 ! paf 108: #endif
1.1 paf 109:
110: /* These are escaped items that aren't just an encoding of a particular data
111: value such as \n. They must have non-zero values, as check_escape() returns
112: their negation. Also, they must appear in the same order as in the opcode
113: definitions below, up to ESC_z. The final one must be ESC_REF as subsequent
114: values are used for \1, \2, \3, etc. There is a test in the code for an escape
115: greater than ESC_b and less than ESC_X to detect the types that may be
116: repeated. If any new escapes are put in-between that don't consume a character,
117: that code will have to change. */
118:
119: enum { ESC_A = 1, ESC_B, ESC_b, ESC_D, ESC_d, ESC_S, ESC_s, ESC_W, ESC_w,
120: ESC_Z, ESC_z, ESC_REF };
121:
122: /* Opcode table: OP_BRA must be last, as all values >= it are used for brackets
123: that extract substrings. Starting from 1 (i.e. after OP_END), the values up to
124: OP_EOD must correspond in order to the list of escapes immediately above. */
125:
126: enum {
127: OP_END, /* End of pattern */
128:
129: /* Values corresponding to backslashed metacharacters */
130:
131: OP_SOD, /* Start of data: \A */
132: OP_NOT_WORD_BOUNDARY, /* \B */
133: OP_WORD_BOUNDARY, /* \b */
134: OP_NOT_DIGIT, /* \D */
135: OP_DIGIT, /* \d */
136: OP_NOT_WHITESPACE, /* \S */
137: OP_WHITESPACE, /* \s */
138: OP_NOT_WORDCHAR, /* \W */
139: OP_WORDCHAR, /* \w */
140: OP_EODN, /* End of data or \n at end of data: \Z. */
141: OP_EOD, /* End of data: \z */
142:
143: OP_OPT, /* Set runtime options */
144: OP_CIRC, /* Start of line - varies with multiline switch */
145: OP_DOLL, /* End of line - varies with multiline switch */
146: OP_ANY, /* Match any character */
147: OP_CHARS, /* Match string of characters */
148: OP_NOT, /* Match anything but the following char */
149:
150: OP_STAR, /* The maximizing and minimizing versions of */
151: OP_MINSTAR, /* all these opcodes must come in pairs, with */
152: OP_PLUS, /* the minimizing one second. */
153: OP_MINPLUS, /* This first set applies to single characters */
154: OP_QUERY,
155: OP_MINQUERY,
156: OP_UPTO, /* From 0 to n matches */
157: OP_MINUPTO,
158: OP_EXACT, /* Exactly n matches */
159:
160: OP_NOTSTAR, /* The maximizing and minimizing versions of */
161: OP_NOTMINSTAR, /* all these opcodes must come in pairs, with */
162: OP_NOTPLUS, /* the minimizing one second. */
163: OP_NOTMINPLUS, /* This first set applies to "not" single characters */
164: OP_NOTQUERY,
165: OP_NOTMINQUERY,
166: OP_NOTUPTO, /* From 0 to n matches */
167: OP_NOTMINUPTO,
168: OP_NOTEXACT, /* Exactly n matches */
169:
170: OP_TYPESTAR, /* The maximizing and minimizing versions of */
171: OP_TYPEMINSTAR, /* all these opcodes must come in pairs, with */
172: OP_TYPEPLUS, /* the minimizing one second. These codes must */
173: OP_TYPEMINPLUS, /* be in exactly the same order as those above. */
174: OP_TYPEQUERY, /* This set applies to character types such as \d */
175: OP_TYPEMINQUERY,
176: OP_TYPEUPTO, /* From 0 to n matches */
177: OP_TYPEMINUPTO,
178: OP_TYPEEXACT, /* Exactly n matches */
179:
180: OP_CRSTAR, /* The maximizing and minimizing versions of */
181: OP_CRMINSTAR, /* all these opcodes must come in pairs, with */
182: OP_CRPLUS, /* the minimizing one second. These codes must */
183: OP_CRMINPLUS, /* be in exactly the same order as those above. */
184: OP_CRQUERY, /* These are for character classes and back refs */
185: OP_CRMINQUERY,
186: OP_CRRANGE, /* These are different to the three seta above. */
187: OP_CRMINRANGE,
188:
189: OP_CLASS, /* Match a character class */
190: OP_REF, /* Match a back reference */
191:
192: OP_ALT, /* Start of alternation */
193: OP_KET, /* End of group that doesn't have an unbounded repeat */
194: OP_KETRMAX, /* These two must remain together and in this */
195: OP_KETRMIN, /* order. They are for groups the repeat for ever. */
196:
197: /* The assertions must come before ONCE and COND */
198:
199: OP_ASSERT, /* Positive lookahead */
200: OP_ASSERT_NOT, /* Negative lookahead */
201: OP_ASSERTBACK, /* Positive lookbehind */
202: OP_ASSERTBACK_NOT, /* Negative lookbehind */
203: OP_REVERSE, /* Move pointer back - used in lookbehind assertions */
204:
205: /* ONCE and COND must come after the assertions, with ONCE first, as there's
206: a test for >= ONCE for a subpattern that isn't an assertion. */
207:
208: OP_ONCE, /* Once matched, don't back up into the subpattern */
209: OP_COND, /* Conditional group */
210: OP_CREF, /* Used to hold an extraction string number */
211:
212: OP_BRAZERO, /* These two must remain together and in this */
213: OP_BRAMINZERO, /* order. */
214:
215: OP_BRA /* This and greater values are used for brackets that
216: extract substrings. */
217: };
218:
219: /* The highest extraction number. This is limited by the number of opcodes
220: left after OP_BRA, i.e. 255 - OP_BRA. We actually set it somewhat lower. */
221:
222: #define EXTRACT_MAX 99
223:
224: /* The texts of compile-time error messages are defined as macros here so that
225: they can be accessed by the POSIX wrapper and converted into error codes. Yes,
226: I could have used error codes in the first place, but didn't feel like changing
227: just to accommodate the POSIX wrapper. */
228:
229: #define ERR1 "\\ at end of pattern"
230: #define ERR2 "\\c at end of pattern"
231: #define ERR3 "unrecognized character follows \\"
232: #define ERR4 "numbers out of order in {} quantifier"
233: #define ERR5 "number too big in {} quantifier"
234: #define ERR6 "missing terminating ] for character class"
235: #define ERR7 "invalid escape sequence in character class"
236: #define ERR8 "range out of order in character class"
237: #define ERR9 "nothing to repeat"
238: #define ERR10 "operand of unlimited repeat could match the empty string"
239: #define ERR11 "internal error: unexpected repeat"
240: #define ERR12 "unrecognized character after (?"
241: #define ERR13 "too many capturing parenthesized sub-patterns"
242: #define ERR14 "missing )"
243: #define ERR15 "back reference to non-existent subpattern"
244: #define ERR16 "erroffset passed as NULL"
245: #define ERR17 "unknown option bit(s) set"
246: #define ERR18 "missing ) after comment"
247: #define ERR19 "too many sets of parentheses"
248: #define ERR20 "regular expression too large"
249: #define ERR21 "failed to get memory"
250: #define ERR22 "unmatched parentheses"
251: #define ERR23 "internal error: code overflow"
252: #define ERR24 "unrecognized character after (?<"
253: #define ERR25 "lookbehind assertion is not fixed length"
254: #define ERR26 "malformed number after (?("
255: #define ERR27 "conditional group contains more than two branches"
256: #define ERR28 "assertion expected after (?("
257:
258: /* All character handling must be done as unsigned characters. Otherwise there
259: are problems with top-bit-set characters and functions such as isspace().
260: However, we leave the interface to the outside world as char *, because that
261: should make things easier for callers. We define a short type for unsigned char
262: to save lots of typing. I tried "uchar", but it causes problems on Digital
263: Unix, where it is defined in sys/types, so use "uschar" instead. */
264:
265: typedef unsigned char uschar;
266:
267: /* The real format of the start of the pcre block; the actual code vector
268: runs on as long as necessary after the end. */
269:
270: typedef struct real_pcre {
271: unsigned long int magic_number;
272: const unsigned char *tables;
273: unsigned long int options;
274: uschar top_bracket;
275: uschar top_backref;
276: uschar first_char;
277: uschar req_char;
278: uschar code[1];
279: } real_pcre;
280:
281: /* The real format of the extra block returned by pcre_study(). */
282:
283: typedef struct real_pcre_extra {
284: uschar options;
285: uschar start_bits[32];
286: } real_pcre_extra;
287:
288:
289: /* Structure for passing "static" information around between the functions
290: doing the compiling, so that they are thread-safe. */
291:
292: typedef struct compile_data {
293: const uschar *lcc; /* Points to lower casing table */
294: const uschar *fcc; /* Points to case-flipping table */
295: const uschar *cbits; /* Points to character type table */
296: const uschar *ctypes; /* Points to table of type maps */
297: } compile_data;
298:
299: /* Structure for passing "static" information around between the functions
300: doing the matching, so that they are thread-safe. */
301:
302: typedef struct match_data {
303: int errorcode; /* As it says */
304: int *offset_vector; /* Offset vector */
305: int offset_end; /* One past the end */
306: int offset_max; /* The maximum usable for return data */
307: const uschar *lcc; /* Points to lower casing table */
308: const uschar *ctypes; /* Points to table of type maps */
309: BOOL offset_overflow; /* Set if too many extractions */
310: BOOL notbol; /* NOTBOL flag */
311: BOOL noteol; /* NOTEOL flag */
312: BOOL endonly; /* Dollar not before final \n */
313: BOOL notempty; /* Empty string match not wanted */
314: const uschar *start_subject; /* Start of the subject string */
315: const uschar *end_subject; /* End of the subject string */
316: const uschar *start_match; /* Start of this match attempt */
317: const uschar *end_match_ptr; /* Subject position at end match */
318: int end_offset_top; /* Highwater mark at end of match */
319: } match_data;
320:
321: /* Bit definitions for entries in the pcre_ctypes table. */
322:
323: #define ctype_space 0x01
324: #define ctype_letter 0x02
325: #define ctype_digit 0x04
326: #define ctype_xdigit 0x08
327: #define ctype_word 0x10 /* alphameric or '_' */
328: #define ctype_meta 0x80 /* regexp meta char or zero (end pattern) */
329:
330: /* Offsets for the bitmap tables in pcre_cbits. Each table contains a set
331: of bits for a class map. */
332:
333: #define cbit_digit 0 /* for \d */
334: #define cbit_word 32 /* for \w */
335: #define cbit_space 64 /* for \s */
336: #define cbit_length 96 /* Length of the cbits table */
337:
338: /* Offsets of the various tables from the base tables pointer, and
339: total length. */
340:
341: #define lcc_offset 0
342: #define fcc_offset 256
343: #define cbits_offset 512
344: #define ctypes_offset (cbits_offset + cbit_length)
345: #define tables_length (ctypes_offset + 256)
346:
347: /* End of internal.h */
E-mail:
![[BACK]](/cvsweb/icons/back.gif){kind=icon}
Return to internal.h CVS log ![[TXT]](/cvsweb/icons/text.gif){kind=icon}
![[DIR]](/cvsweb/icons/dir.gif){kind=icon}
Up to [parser3project] / parser3 / src / pcre