mdw@git.distorted.org.uk Git - disorder/blob - lib/split.c

   1 /*
   2  * This file is part of DisOrder.
   3  * Copyright (C) 2004, 2006-2008 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
  19 #include "common.h"
  20
  21 #include <ctype.h>
  22 #include <errno.h>
  23
  24 #include "mem.h"
  25 #include "split.h"
  26 #include "log.h"
  27 #include "charset.h"
  28 #include "vector.h"
  29
  30 static inline int space(int c) {
  31   return (c == ' '
  32           || c == '\t'
  33           || c == '\n'
  34           || c == '\r');
  35 }
  36
  37 static void no_error_handler(const char attribute((unused)) *msg,
  38                              void attribute((unused)) *u) {
  39 }
  40
  41 /* TODO: handle combining characters attached to delimiters in some
  42  * sane way (might include reporting an error) */
  43
  44 char **split(const char *p,
  45              int *np,
  46              unsigned flags,
  47              void (*error_handler)(const char *msg, void *u),
  48              void *u) {
  49   char *f, *g;
  50   const char *q;
  51   struct vector v;
  52   size_t l;
  53   int qc;
  54
  55   if(!error_handler)
  56     error_handler = no_error_handler;
  57   vector_init(&v);
  58   while(*p && !(*p == '#' && (flags & SPLIT_COMMENTS))) {
  59     if(space(*p)) {
  60       ++p;
  61       continue;
  62     }
  63     if((flags & SPLIT_QUOTES) && (*p == '"' || *p == '\'')) {
  64       qc = *p++;
  65       l = 0;
  66       for(q = p; *q && *q != qc; ++q) {
  67         if(*q == '\\' && q[1])
  68           ++q;
  69         ++l;
  70       }
  71       if(!*q) {
  72         error_handler("unterminated quoted string", u);
  73         return 0;
  74       }
  75       f = g = xmalloc_noptr(l + 1);
  76       for(q = p; *q != qc;) {
  77         if(*q == '\\') {
  78           ++q;
  79           switch(*q) {
  80           case '\\':
  81           case '"':
  82           case '\'':
  83             *g++ = *q++;
  84             break;
  85           case 'n':
  86             ++q;
  87             *g++ = '\n';
  88             break;
  89           default:
  90             error_handler("illegal escape sequence", u);
  91             return 0;
  92           }
  93         } else
  94           *g++ = *q++;
  95       }
  96       *g = 0;
  97       p = q + 1;
  98     } else {
  99       for(q = p; *q && !space(*q); ++q)
 100         ;
 101       l = q - p;
 102       f = xstrndup(p, l);
 103       p = q;
 104     }
 105     vector_append(&v, f);
 106   }
 107   vector_terminate(&v);
 108   if(np)
 109     *np = v.nvec;
 110   return v.vec;
 111 }
 112
 113 /* TODO handle initial combining characters sanely */
 114
 115 const char *quoteutf8(const char *s) {
 116   size_t len = 3 + strlen(s);
 117   const char *t;
 118   char *r, *q;
 119
 120   /* see if we need to quote */
 121   if(*s) {
 122     for(t = s; *t; t++)
 123       if((unsigned char)*t <= ' '
 124          || *t == '"'
 125          || *t == '\\'
 126          || *t == '\''
 127          || *t == '#')
 128         break;
 129     if(!*t)
 130       return s;
 131   }
 132
 133   /* we rely on ASCII characters only ever representing themselves in UTF-8. */
 134   for(t = s; *t; t++) {
 135     switch(*t) {
 136     case '"':
 137     case '\\':
 138     case '\n':
 139       ++len;
 140       break;
 141     }
 142   }
 143   q = r = xmalloc_noptr(len);
 144   *q++ = '"';
 145   for(t = s; *t; t++) {
 146     switch(*t) {
 147     case '"':
 148     case '\\':
 149       *q++ = '\\';
 150       /* fall through */
 151     default:
 152       *q++ = *t;
 153       break;
 154     case '\n':
 155       *q++ = '\\';
 156       *q++ = 'n';
 157       break;
 158     }
 159   }
 160   *q++ = '"';
 161   *q = 0;
 162   return r;
 163 }
 164
 165 /*
 166 Local Variables:
 167 c-basic-offset:2
 168 comment-column:40
 169 End:
 170 */