Skip to content
This repository

HTTPS clone URL

Subversion checkout URL

You can clone with HTTPS or Subversion.

Download ZIP
tree: 2fe6e177ff
Fetching contributors…

Cannot retrieve contributors at this time

file 330 lines (295 sloc) 8.286 kb
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330
/* regsub.c */

/* This file contains the regsub() function, which performs substitutions
* after a regexp match has been found.
*/

#include "elvis.h"
#ifdef FEATURE_RCSID
char id_regsub[] = "$Id: regsub.c,v 2.22 2003/10/17 17:41:23 steve Exp $";
#endif

/* Allocate a new copy of the replacement string, with all ~'s replaced by
* the previous replacement string.
*
* NOTE: The value returned by this function should never be freed from outside
* this function, because this function maintains an internal pointer to the
* same memory so it knows what value to substitute for ~ on next invocation.
*/
CHAR *regtilde(newp)
CHAR *newp; /* new text as supplied by user */
{
static CHAR *prev; /* previous replacement text */
static CHAR *willfree; /* previous replacement text if nosaveregex */
CHAR *ret; /* returned string */
CHAR *scan; /* used for stepping through chars of "prev" */

/* If "willfree" isn't NULL, then free it now. This is used when
* the saveregexp option is turned off, so we can leave prev unchanged.
*/
if (willfree)
{
safefree(willfree);
willfree = NULL;
}

/* copy new into ret, replacing the ~s by the previous text */
for (ret = NULL; *newp; )
{
if (o_magic && *newp == '~')
{
if (!prev) goto Fail;
for (scan = prev; *scan; scan++)
buildCHAR(&ret, *scan);
newp++;
}
else if (!o_magic && *newp == '\\' && *(newp + 1) == '~')
{
if (!prev) goto Fail;
for (scan = prev; *scan; scan++)
buildCHAR(&ret, *scan);
newp += 2;
}
else
{
if (*newp == '\\' && *(newp + 1))
{
buildCHAR(&ret, *newp++);
}
buildCHAR(&ret, *newp++);
}
}

/* if empty string, then allocate a single '\0' character */
if (!ret)
ret = (CHAR *)safealloc(1, sizeof(CHAR));

/* remember this as the "previous" for next time */
if (o_saveregexp)
{
if (prev)
safefree(prev);
prev = ret;
}
else
{
/* leave "prev" unchanged, but remember it somewhere else so
* we can free the text when regtilde() is called next time.
*/
willfree = ret;
}

return ret;

Fail:
msg(MSG_ERROR, "no previous text to substitute for ~");
if (ret)
safefree(ret);
return NULL;
}

/* Perform substitutions after a regexp match. "re" is the compiled regular
* expression which has been matched to a text string. "new" is a pointer to
* the replacement text string. Return the actual replacement text (after all
* metacharacters have been processed) if successful, or NULL if error. The
* calling function is responsible for calling safefree() on the returned
* string.
*/
CHAR *regsub(re, newp, doit)
regexp *re; /* a regular expression that has been matched */
REG CHAR *newp; /* the replacement text */
ELVBOOL doit; /* perform the substitution? (else just return string) */
{
MARKBUF cpy; /* start of text to copy */
long end; /* length of text to copy */
REG CHAR c; /* a character from "new" text */
long cval; /* numeric value of 'c', if 'c' is digit */
CHAR *inst; /* the new next, after processing escapes */
int mod = 0;/* used to track \U, \L, \u, \l, and \E */
int len; /* used to calculate length of subst string */
MARKBUF tmp; /* end of replacement region */
CHAR *scan; /* used for scanning a segment of orig text */
char lnum[12];/* line number */

/* initialize "cval" just to silence a compiler warning */
cval = 0;

/* for each character of the new text... */
for (inst = NULL, len = 0; (c = *newp++) != '\0'; )
{
/* recognize any meta characters */
if (c == '&' && o_magic)
{
(void)marktmp(cpy, re->buffer, re->startp[0]);
end = re->endp[0] - re->startp[0];
}
else if (c == '\\')
{
c = *newp++;
switch (c)
{
case '0':
/* Traditionally \0 has been a synonym for &,
* but we need a way to insert NUL so...
*/
len = buildCHAR(&inst, '\0');
continue;

case '1':
case '2':
case '3':
case '4':
case '5':
case '6':
case '7':
case '8':
case '9':
/* \0 thru \9 mean "copy subexpression" */
cval = c - '0';
(void)marktmp(cpy, re->buffer, re->startp[cval]);
end = re->endp[cval] - re->startp[cval];
break;

case 'U':
case 'u':
case 'L':
case 'l':
/* \U and \L mean "convert to upper/lowercase" */
mod = c;
continue;

case 'E':
case 'e':
/* \E ends the \U or \L */
mod = 0;
continue;

case '&':
/* "\&" means "original text" */
if (o_magic)
{
len = buildCHAR(&inst, c);
continue;
}
(void)marktmp(cpy, re->buffer, re->startp[0]);
end = re->endp[0] - re->startp[0];
break;

case '#': /* "\#" means "line number" */
sprintf(lnum, "%ld", markline(marktmp(tmp, re->buffer, re->startp[0])));
len = buildstr(&inst, lnum);
continue;

case 'a': /* \a => ^G, <BEL> */
len = buildCHAR(&inst, '\007');
continue;

case 'b': /* \b => ^H, <BS> */
len = buildCHAR(&inst, '\b');
continue;

  case 'f': /* \f => ^L, <FF> */
len = buildCHAR(&inst, '\f');
continue;

  case 'n': /* \n => ^J, <NL> */
len = buildCHAR(&inst, '\n');
continue;

  case 'r': /* \r => ^M, <CR> */
len = buildCHAR(&inst, '\r');
continue;

case 't': /* \t => ^I, <TAB> */
len = buildCHAR(&inst, '\t');
continue;

#if 0
/* \e is already taken for ending of \U and \L.
* Though it's a shame both \E and \e are well
* documented to ex/vi here we' like \E to
* suffice, so we would use \e for <Esc>.
*/
case 'e': /* \e => ^[, <ESC> */
len = buildCHAR(&inst, '\033');
continue;
#endif

default:
/* ordinary char preceded by backslash */
len = buildCHAR(&inst, c);
continue;
}
}
# if OSK
else if (c == '\l')
# else
else if (c == '\r')
# endif
{
/* transliterate ^M into newline */
len = buildCHAR(&inst, '\n');
continue;
}
else
{
/* ordinary character, so just copy it */
if (!mod)
len = buildCHAR(&inst, c);
else if (elvtolower(mod) == 'l')
len = buildCHAR(&inst, elvtolower(c));
else
len = buildCHAR(&inst, elvtoupper(c));
if (elvlower(mod))
mod = 0;
continue;
}

/* Note: to reach this point in the code, we have evaded
* all "continue" statements. To do that, we must have hit
* a metacharacter that involves copying.
*/

/* if there is nothing to copy, loop */
if (markoffset(&cpy) < 0)
{
msg(MSG_ERROR, "[d]too few \\\\\\(\\\\\\)s to use \\\\$1", cval);
if (inst)
safefree(inst);
return NULL;
}

/* copy over a portion of the original */
for (scanalloc(&scan, &cpy);
scan && end > 0;
scannext(&scan), end--)
{
switch (mod)
{
case 'U':
case 'u':
/* convert to uppercase */
len = buildCHAR(&inst, (_CHAR_)elvtoupper(*scan));
break;

case 'L':
case 'l':
/* convert to lowercase */
len = buildCHAR(&inst, (_CHAR_)elvtolower(*scan));
break;

default:
/* copy without any conversion */
len = buildCHAR(&inst, *scan);
}

/* \u and \l end automatically after the first char */
if (mod == 'u' || mod == 'l')
{
mod = 0;
}
}
scanfree(&scan);
}

/* if we're supposed to perform the substitution, then do it */
if (doit)
{
/* replace the old text with the new text in the buffer */
bufreplace(marktmp(cpy, re->buffer, re->startp[0]),
marktmp(tmp, re->buffer, re->endp[0]), inst, len);

/* Adjust the offset of the end of the whole expression
* to compensate for the change in the length of text.
* Also, if this regexp could conceivably match a
* zero-length string, then require at least 1 unmatched
* character between matches.
*/
re->endp[0] = re->startp[0] + len;
if (re->minlen == 0
&& re->endp[0] < o_bufchars(re->buffer)
&& scanchar(marktmp(tmp, re->buffer, re->endp[0])) != '\n')
{
re->endp[0]++;
}
}

/* At this point, we know we were successful but the "inst" pointer
* will be NULL if the replacement text is 0 characters long. We don't
* want to return NULL for a successful substitution, so allocate
* a string which contains only a '\0' character and return that.
*/
if (!inst)
{
assert(len == 0);
buildCHAR(&inst, (_CHAR_)'\0');
assert(inst != NULL);
}

return inst;
}
Something went wrong with that request. Please try again.