| 1 | /* Definitions for data structures callers pass the regex library. |
| 2 | Copyright (C) 1985, 1989 Free Software Foundation, Inc. |
| 3 | |
| 4 | This program is free software; you can redistribute it and/or modify |
| 5 | it under the terms of the GNU General Public License as published by |
| 6 | the Free Software Foundation; either version 2 of the License, or |
| 7 | (at your option) any later version. |
| 8 | |
| 9 | This program is distributed in the hope that it will be useful, |
| 10 | but WITHOUT ANY WARRANTY; without even the implied warranty of |
| 11 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
| 12 | GNU General Public License for more details. |
| 13 | |
| 14 | You should have received a copy of the GNU General Public License |
| 15 | along with this program; if not, write to the Free Software |
| 16 | Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ |
| 17 | |
| 18 | /* Define number of parens for which we record the beginnings and ends. |
| 19 | This affects how much space the `struct re_registers' type takes up. */ |
| 20 | #ifndef RE_NREGS |
| 21 | #define RE_NREGS 10 |
| 22 | #endif |
| 23 | |
| 24 | /* These bits are used in the obscure_syntax variable to choose among |
| 25 | alternative regexp syntaxes. */ |
| 26 | |
| 27 | /* 1 means plain parentheses serve as grouping, and backslash |
| 28 | parentheses are needed for literal searching. |
| 29 | 0 means backslash-parentheses are grouping, and plain parentheses |
| 30 | are for literal searching. */ |
| 31 | #define RE_NO_BK_PARENS 1 |
| 32 | |
| 33 | /* 1 means plain | serves as the "or"-operator, and \| is a literal. |
| 34 | 0 means \| serves as the "or"-operator, and | is a literal. */ |
| 35 | #define RE_NO_BK_VBAR 2 |
| 36 | |
| 37 | /* 0 means plain + or ? serves as an operator, and \+, \? are literals. |
| 38 | 1 means \+, \? are operators and plain +, ? are literals. */ |
| 39 | #define RE_BK_PLUS_QM 4 |
| 40 | |
| 41 | /* 1 means | binds tighter than ^ or $. |
| 42 | 0 means the contrary. */ |
| 43 | #define RE_TIGHT_VBAR 8 |
| 44 | |
| 45 | /* 1 means treat \n as an _OR operator |
| 46 | 0 means treat it as a normal character */ |
| 47 | #define RE_NEWLINE_OR 16 |
| 48 | |
| 49 | /* 0 means that a special characters (such as *, ^, and $) always have |
| 50 | their special meaning regardless of the surrounding context. |
| 51 | 1 means that special characters may act as normal characters in some |
| 52 | contexts. Specifically, this applies to: |
| 53 | ^ - only special at the beginning, or after ( or | |
| 54 | $ - only special at the end, or before ) or | |
| 55 | *, +, ? - only special when not after the beginning, (, or | */ |
| 56 | #define RE_CONTEXT_INDEP_OPS 32 |
| 57 | |
| 58 | /* Now define combinations of bits for the standard possibilities. */ |
| 59 | #define RE_SYNTAX_AWK (RE_NO_BK_PARENS | RE_NO_BK_VBAR | RE_CONTEXT_INDEP_OPS) |
| 60 | #define RE_SYNTAX_EGREP (RE_SYNTAX_AWK | RE_NEWLINE_OR) |
| 61 | #define RE_SYNTAX_GREP (RE_BK_PLUS_QM | RE_NEWLINE_OR) |
| 62 | #define RE_SYNTAX_EMACS 0 |
| 63 | |
| 64 | /* This data structure is used to represent a compiled pattern. */ |
| 65 | |
| 66 | struct re_pattern_buffer |
| 67 | { |
| 68 | char *buffer; /* Space holding the compiled pattern commands. */ |
| 69 | int allocated; /* Size of space that buffer points to */ |
| 70 | int used; /* Length of portion of buffer actually occupied */ |
| 71 | char *fastmap; /* Pointer to fastmap, if any, or zero if none. */ |
| 72 | /* re_search uses the fastmap, if there is one, |
| 73 | to skip quickly over totally implausible characters */ |
| 74 | char *translate; /* Translate table to apply to all characters before comparing. |
| 75 | Or zero for no translation. |
| 76 | The translation is applied to a pattern when it is compiled |
| 77 | and to data when it is matched. */ |
| 78 | char fastmap_accurate; |
| 79 | /* Set to zero when a new pattern is stored, |
| 80 | set to one when the fastmap is updated from it. */ |
| 81 | char can_be_null; /* Set to one by compiling fastmap |
| 82 | if this pattern might match the null string. |
| 83 | It does not necessarily match the null string |
| 84 | in that case, but if this is zero, it cannot. |
| 85 | 2 as value means can match null string |
| 86 | but at end of range or before a character |
| 87 | listed in the fastmap. */ |
| 88 | }; |
| 89 | |
| 90 | /* Structure to store "register" contents data in. |
| 91 | |
| 92 | Pass the address of such a structure as an argument to re_match, etc., |
| 93 | if you want this information back. |
| 94 | |
| 95 | start[i] and end[i] record the string matched by \( ... \) grouping i, |
| 96 | for i from 1 to RE_NREGS - 1. |
| 97 | start[0] and end[0] record the entire string matched. */ |
| 98 | |
| 99 | struct re_registers |
| 100 | { |
| 101 | int start[RE_NREGS]; |
| 102 | int end[RE_NREGS]; |
| 103 | }; |
| 104 | |
| 105 | /* These are the command codes that appear in compiled regular expressions, one per byte. |
| 106 | Some command codes are followed by argument bytes. |
| 107 | A command code can specify any interpretation whatever for its arguments. |
| 108 | Zero-bytes may appear in the compiled regular expression. */ |
| 109 | |
| 110 | enum regexpcode |
| 111 | { |
| 112 | unused, |
| 113 | exactn, /* followed by one byte giving n, and then by n literal bytes */ |
| 114 | begline, /* fails unless at beginning of line */ |
| 115 | endline, /* fails unless at end of line */ |
| 116 | jump, /* followed by two bytes giving relative address to jump to */ |
| 117 | on_failure_jump, /* followed by two bytes giving relative address of place |
| 118 | to resume at in case of failure. */ |
| 119 | finalize_jump, /* Throw away latest failure point and then jump to address. */ |
| 120 | maybe_finalize_jump, /* Like jump but finalize if safe to do so. |
| 121 | This is used to jump back to the beginning |
| 122 | of a repeat. If the command that follows |
| 123 | this jump is clearly incompatible with the |
| 124 | one at the beginning of the repeat, such that |
| 125 | we can be sure that there is no use backtracking |
| 126 | out of repetitions already completed, |
| 127 | then we finalize. */ |
| 128 | dummy_failure_jump, /* jump, and push a dummy failure point. |
| 129 | This failure point will be thrown away |
| 130 | if an attempt is made to use it for a failure. |
| 131 | A + construct makes this before the first repeat. */ |
| 132 | anychar, /* matches any one character */ |
| 133 | charset, /* matches any one char belonging to specified set. |
| 134 | First following byte is # bitmap bytes. |
| 135 | Then come bytes for a bit-map saying which chars are in. |
| 136 | Bits in each byte are ordered low-bit-first. |
| 137 | A character is in the set if its bit is 1. |
| 138 | A character too large to have a bit in the map |
| 139 | is automatically not in the set */ |
| 140 | charset_not, /* similar but match any character that is NOT one of those specified */ |
| 141 | start_memory, /* starts remembering the text that is matched |
| 142 | and stores it in a memory register. |
| 143 | followed by one byte containing the register number. |
| 144 | Register numbers must be in the range 0 through NREGS. */ |
| 145 | stop_memory, /* stops remembering the text that is matched |
| 146 | and stores it in a memory register. |
| 147 | followed by one byte containing the register number. |
| 148 | Register numbers must be in the range 0 through NREGS. */ |
| 149 | duplicate, /* match a duplicate of something remembered. |
| 150 | Followed by one byte containing the index of the memory register. */ |
| 151 | before_dot, /* Succeeds if before dot */ |
| 152 | at_dot, /* Succeeds if at dot */ |
| 153 | after_dot, /* Succeeds if after dot */ |
| 154 | begbuf, /* Succeeds if at beginning of buffer */ |
| 155 | endbuf, /* Succeeds if at end of buffer */ |
| 156 | wordchar, /* Matches any word-constituent character */ |
| 157 | notwordchar, /* Matches any char that is not a word-constituent */ |
| 158 | wordbeg, /* Succeeds if at word beginning */ |
| 159 | wordend, /* Succeeds if at word end */ |
| 160 | wordbound, /* Succeeds if at a word boundary */ |
| 161 | notwordbound, /* Succeeds if not at a word boundary */ |
| 162 | syntaxspec, /* Matches any character whose syntax is specified. |
| 163 | followed by a byte which contains a syntax code, Sword or such like */ |
| 164 | notsyntaxspec /* Matches any character whose syntax differs from the specified. */ |
| 165 | }; |
| 166 | \f |
| 167 | extern char *re_compile_pattern (); |
| 168 | /* Is this really advertised? */ |
| 169 | extern void re_compile_fastmap (); |
| 170 | extern int re_search (), re_search_2 (); |
| 171 | extern int re_match (), re_match_2 (); |
| 172 | |
| 173 | /* 4.2 bsd compatibility (yuck) */ |
| 174 | extern char *re_comp (); |
| 175 | extern int re_exec (); |
| 176 | |
| 177 | #ifdef SYNTAX_TABLE |
| 178 | extern char *re_syntax_table; |
| 179 | #endif |
| 180 | |
| 181 | extern int re_set_syntax (); |