| 1 | /* |
| 2 | * This file is part of DisOrder |
| 3 | * Copyright (C) 2004, 2005, 2007, 20008 Richard Kettlewell |
| 4 | * |
| 5 | * This program is free software: you can redistribute it and/or modify |
| 6 | * it under the terms of the GNU General Public License as published by |
| 7 | * the Free Software Foundation, either version 3 of the License, or |
| 8 | * (at your option) any later version. |
| 9 | * |
| 10 | * This program is distributed in the hope that it will be useful, |
| 11 | * but WITHOUT ANY WARRANTY; without even the implied warranty of |
| 12 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
| 13 | * GNU General Public License for more details. |
| 14 | * |
| 15 | * You should have received a copy of the GNU General Public License |
| 16 | * along with this program. If not, see <http://www.gnu.org/licenses/>. |
| 17 | */ |
| 18 | /** @file lib/regsub.c |
| 19 | * @brief Regexp substitution |
| 20 | */ |
| 21 | #include "common.h" |
| 22 | |
| 23 | #include <pcre.h> |
| 24 | |
| 25 | #include "regsub.h" |
| 26 | #include "mem.h" |
| 27 | #include "vector.h" |
| 28 | #include "log.h" |
| 29 | |
| 30 | #define PREMATCH (-1) /* fictitious pre-match substring */ |
| 31 | #define POSTMATCH (-2) /* fictitious post-match substring */ |
| 32 | |
| 33 | static inline int substring_start(const char attribute((unused)) *subject, |
| 34 | const int *ovector, |
| 35 | int n) { |
| 36 | switch(n) { |
| 37 | case PREMATCH: return 0; |
| 38 | case POSTMATCH: return ovector[1]; |
| 39 | default: return ovector[2 * n]; |
| 40 | } |
| 41 | } |
| 42 | |
| 43 | static inline int substring_end(const char *subject, |
| 44 | const int *ovector, |
| 45 | int n) { |
| 46 | switch(n) { |
| 47 | case PREMATCH: return ovector[0]; |
| 48 | case POSTMATCH: return strlen(subject); |
| 49 | default: return ovector[2 * n + 1]; |
| 50 | } |
| 51 | } |
| 52 | |
| 53 | static void transform_append(struct dynstr *d, |
| 54 | const char *subject, |
| 55 | const int *ovector, |
| 56 | int n) { |
| 57 | int start = substring_start(subject, ovector, n); |
| 58 | int end = substring_end(subject, ovector, n); |
| 59 | |
| 60 | if(start != -1) |
| 61 | dynstr_append_bytes(d, subject + start, end - start); |
| 62 | } |
| 63 | |
| 64 | static void replace_core(struct dynstr *d, |
| 65 | const char *subject, |
| 66 | const char *replace, |
| 67 | int rc, |
| 68 | const int *ovector) { |
| 69 | int substr; |
| 70 | |
| 71 | while(*replace) { |
| 72 | if(*replace == '$') |
| 73 | switch(replace[1]) { |
| 74 | case '&': |
| 75 | transform_append(d, subject, ovector, 0); |
| 76 | replace += 2; |
| 77 | break; |
| 78 | case '1': case '2': case '3': |
| 79 | case '4': case '5': case '6': |
| 80 | case '7': case '8': case '9': |
| 81 | substr = replace[1] - '0'; |
| 82 | if(substr < rc) |
| 83 | transform_append(d, subject, ovector, substr); |
| 84 | replace += 2; |
| 85 | break; |
| 86 | case '$': |
| 87 | dynstr_append(d, '$'); |
| 88 | replace += 2; |
| 89 | break; |
| 90 | default: |
| 91 | dynstr_append(d, *replace++); |
| 92 | break; |
| 93 | } |
| 94 | else |
| 95 | dynstr_append(d, *replace++); |
| 96 | } |
| 97 | } |
| 98 | |
| 99 | unsigned regsub_flags(const char *flags) { |
| 100 | unsigned f = 0; |
| 101 | |
| 102 | while(*flags) { |
| 103 | switch(*flags++) { |
| 104 | case 'g': f |= REGSUB_GLOBAL; break; |
| 105 | case 'i': f |= REGSUB_CASE_INDEPENDENT; break; |
| 106 | default: break; |
| 107 | } |
| 108 | } |
| 109 | return f; |
| 110 | } |
| 111 | |
| 112 | int regsub_compile_options(unsigned flags) { |
| 113 | int options = 0; |
| 114 | |
| 115 | if(flags & REGSUB_CASE_INDEPENDENT) |
| 116 | options |= PCRE_CASELESS; |
| 117 | return options; |
| 118 | } |
| 119 | |
| 120 | const char *regsub(const pcre *re, const char *subject, const char *replace, |
| 121 | unsigned flags) { |
| 122 | int rc, ovector[99], matches; |
| 123 | struct dynstr d; |
| 124 | |
| 125 | dynstr_init(&d); |
| 126 | matches = 0; |
| 127 | /* find the next match */ |
| 128 | while((rc = pcre_exec(re, 0, subject, strlen(subject), 0, |
| 129 | 0, ovector, sizeof ovector / sizeof (int))) > 0) { |
| 130 | /* text just before the match */ |
| 131 | if(!(flags & REGSUB_REPLACE)) |
| 132 | transform_append(&d, subject, ovector, PREMATCH); |
| 133 | /* the replacement text */ |
| 134 | replace_core(&d, subject, replace, rc, ovector); |
| 135 | ++matches; |
| 136 | if(!*subject) /* end of subject */ |
| 137 | break; |
| 138 | if(flags & REGSUB_REPLACE) /* replace subject entirely */ |
| 139 | break; |
| 140 | /* step over the matched substring */ |
| 141 | subject += substring_start(subject, ovector, POSTMATCH); |
| 142 | if(!(flags & REGSUB_GLOBAL)) |
| 143 | break; |
| 144 | } |
| 145 | if(rc <= 0 && rc != PCRE_ERROR_NOMATCH) { |
| 146 | error(0, "pcre_exec returned %d, subject '%s'", rc, subject); |
| 147 | return 0; |
| 148 | } |
| 149 | if((flags & REGSUB_MUST_MATCH) && matches == 0) |
| 150 | return 0; |
| 151 | /* append the remainder of the subject */ |
| 152 | if(!(flags & REGSUB_REPLACE)) |
| 153 | dynstr_append_string(&d, subject); |
| 154 | dynstr_terminate(&d); |
| 155 | return d.vec; |
| 156 | } |
| 157 | |
| 158 | /* |
| 159 | Local Variables: |
| 160 | c-basic-offset:2 |
| 161 | comment-column:40 |
| 162 | End: |
| 163 | */ |