1 /* $OpenBSD: gnum4.c,v 1.42 2011/11/06 12:25:43 espie Exp $ */
4 * Copyright (c) 1999 Marc Espie
6 * Redistribution and use in source and binary forms, with or without
7 * modification, are permitted provided that the following conditions
9 * 1. Redistributions of source code must retain the above copyright
10 * notice, this list of conditions and the following disclaimer.
11 * 2. Redistributions in binary form must reproduce the above copyright
12 * notice, this list of conditions and the following disclaimer in the
13 * documentation and/or other materials provided with the distribution.
15 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
16 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
17 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
18 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
19 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
20 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
21 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
22 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
23 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
24 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
27 #include <sys/cdefs.h>
28 __FBSDID("$FreeBSD$");
31 * functions needed to support gnu-m4 extensions, including a fake freezing
34 #include <sys/param.h>
35 #include <sys/types.h>
55 * Support for include path search
56 * First search in the current directory.
57 * If not found, and the path is not absolute, include path kicks in.
58 * First, -I options, in the order found on the command line.
59 * Then M4PATH env variable
62 static struct path_entry {
64 struct path_entry *next;
67 static struct path_entry *new_path_entry(const char *);
68 static void ensure_m4path(void);
69 static struct input_file *dopath(struct input_file *, const char *);
71 static struct path_entry *
72 new_path_entry(const char *dirname)
76 n = malloc(sizeof(struct path_entry));
78 errx(1, "out of memory");
79 n->name = strdup(dirname);
81 errx(1, "out of memory");
87 addtoincludepath(const char *dirname)
91 n = new_path_entry(dirname);
104 static int envpathdone = 0;
112 envpath = getenv("M4PATH");
115 /* for portability: getenv result is read-only */
116 envpath = strdup(envpath);
118 errx(1, "out of memory");
119 for (sweep = envpath;
120 (path = strsep(&sweep, ":")) != NULL;)
121 addtoincludepath(path);
127 dopath(struct input_file *i, const char *filename)
129 char path[MAXPATHLEN];
130 struct path_entry *pe;
133 for (pe = first; pe; pe = pe->next) {
134 snprintf(path, sizeof(path), "%s/%s", pe->name, filename);
135 if ((f = fopen(path, "r")) != 0) {
136 set_input(i, f, path);
144 fopen_trypath(struct input_file *i, const char *filename)
148 f = fopen(filename, "r");
150 set_input(i, f, filename);
153 if (filename[0] == '/')
158 return dopath(i, filename);
162 doindir(const char *argv[], int argc)
165 struct macro_definition *p = NULL;
168 if (n == NULL || (p = macro_getdef(n)) == NULL)
169 m4errx(1, "indir: undefined macro %s.", argv[2]);
172 eval(argv+1, argc-1, p->type, is_traced(n));
176 dobuiltin(const char *argv[], int argc)
181 p = macro_getbuiltin(argv[2]);
183 eval(argv+1, argc-1, macro_builtin_type(p), is_traced(p));
185 m4errx(1, "unknown builtin %s.", argv[2]);
189 /* We need some temporary buffer space, as pb pushes BACK and substitution
190 * proceeds forward... */
192 static size_t bufsize = 0;
193 static size_t current = 0;
195 static void addchars(const char *, size_t);
196 static void addchar(int);
197 static char *twiddle(const char *);
198 static char *getstring(void);
199 static void exit_regerror(int, regex_t *);
200 static void do_subst(const char *, regex_t *, const char *, regmatch_t *);
201 static void do_regexpindex(const char *, regex_t *, regmatch_t *);
202 static void do_regexp(const char *, regex_t *, const char *, regmatch_t *);
203 static void add_sub(int, const char *, regex_t *, regmatch_t *);
204 static void add_replace(const char *, regex_t *, const char *, regmatch_t *);
205 #define addconstantstring(s) addchars((s), sizeof(s)-1)
208 addchars(const char *c, size_t n)
212 while (current + n > bufsize) {
217 buffer = xrealloc(buffer, bufsize, NULL);
219 memcpy(buffer+current, c, n);
226 if (current +1 > bufsize) {
231 buffer = xrealloc(buffer, bufsize, NULL);
233 buffer[current++] = c;
246 exit_regerror(int er, regex_t *re)
251 errlen = regerror(er, re, NULL, 0);
252 errbuf = xalloc(errlen,
253 "malloc in regerror: %lu", (unsigned long)errlen);
254 regerror(er, re, errbuf, errlen);
255 m4errx(1, "regular expression error: %s.", errbuf);
259 add_sub(int n, const char *string, regex_t *re, regmatch_t *pm)
261 if (n > (int)re->re_nsub)
262 warnx("No subexpression %d", n);
263 /* Subexpressions that did not match are
265 else if (pm[n].rm_so != -1 &&
267 addchars(string + pm[n].rm_so,
268 pm[n].rm_eo - pm[n].rm_so);
272 /* Add replacement string to the output buffer, recognizing special
273 * constructs and replacing them with substrings of the original string.
276 add_replace(const char *string, regex_t *re, const char *replace, regmatch_t *pm)
280 for (p = replace; *p != '\0'; p++) {
281 if (*p == '&' && !mimic_gnu) {
282 add_sub(0, string, re, pm);
293 add_sub(0, string, re, pm);
300 add_sub(*(++p) - '0', string, re, pm);
309 do_subst(const char *string, regex_t *re, const char *replace, regmatch_t *pm)
313 const char *last_match = NULL;
315 while ((error = regexec(re, string, re->re_nsub+1, pm, flags)) == 0) {
316 if (pm[0].rm_eo != 0) {
317 if (string[pm[0].rm_eo-1] == '\n')
323 /* NULL length matches are special... We use the `vi-mode'
324 * rule: don't allow a NULL-match at the last match
327 if (pm[0].rm_so == pm[0].rm_eo &&
328 string + pm[0].rm_so == last_match) {
332 if (*string++ == '\n')
338 last_match = string + pm[0].rm_so;
339 addchars(string, pm[0].rm_so);
340 add_replace(string, re, replace, pm);
341 string += pm[0].rm_eo;
343 if (error != REG_NOMATCH)
344 exit_regerror(error, re);
349 do_regexp(const char *string, regex_t *re, const char *replace, regmatch_t *pm)
353 switch(error = regexec(re, string, re->re_nsub+1, pm, 0)) {
355 add_replace(string, re, replace, pm);
361 exit_regerror(error, re);
366 do_regexpindex(const char *string, regex_t *re, regmatch_t *pm)
370 switch(error = regexec(re, string, re->re_nsub+1, pm, 0)) {
372 pbunsigned(pm[0].rm_so);
378 exit_regerror(error, re);
382 /* In Gnu m4 mode, parentheses for backmatch don't work like POSIX 1003.2
383 * says. So we twiddle with the regexp before passing it to regcomp.
386 twiddle(const char *p)
388 /* + at start of regexp is a normal character for Gnu m4 */
396 /* This could use strcspn for speed... */
406 addconstantstring("[_a-zA-Z0-9]");
409 addconstantstring("[^_a-zA-Z0-9]");
412 addconstantstring("[[:<:]]");
415 addconstantstring("[[:>:]]");
424 if (*p == '(' || *p == ')' || *p == '|')
433 /* patsubst(string, regexp, opt replacement) */
439 dopatsubst(const char *argv[], int argc)
442 warnx("Too few arguments to patsubst");
445 /* special case: empty regexp */
446 if (argv[3][0] == '\0') {
449 if (argc > 4 && argv[4])
450 len = strlen(argv[4]);
453 for (s = argv[2]; *s != '\0'; s++) {
454 addchars(argv[4], len);
461 int mode = REG_EXTENDED;
462 size_t l = strlen(argv[3]);
465 (argv[3][0] == '^') ||
466 (l > 0 && argv[3][l-1] == '$'))
469 error = regcomp(&re, mimic_gnu ? twiddle(argv[3]) : argv[3],
472 exit_regerror(error, &re);
474 pmatch = xalloc(sizeof(regmatch_t) * (re.re_nsub+1), NULL);
475 do_subst(argv[2], &re,
476 argc > 4 && argv[4] != NULL ? argv[4] : "", pmatch);
484 doregexp(const char *argv[], int argc)
491 warnx("Too few arguments to regexp");
494 /* special gnu case */
495 if (argv[3][0] == '\0' && mimic_gnu) {
496 if (argc == 4 || argv[4] == NULL)
501 error = regcomp(&re, mimic_gnu ? twiddle(argv[3]) : argv[3],
502 REG_EXTENDED|REG_NEWLINE);
504 exit_regerror(error, &re);
506 pmatch = xalloc(sizeof(regmatch_t) * (re.re_nsub+1), NULL);
507 if (argc == 4 || argv[4] == NULL)
508 do_regexpindex(argv[2], &re, pmatch);
510 do_regexp(argv[2], &re, argv[4], pmatch);
516 doformat(const char *argv[], int argc)
518 const char *format = argv[2];
523 const char *thisarg = NULL;
527 while (*format != 0) {
528 if (*format != '%') {
534 if (*format == '%') {
543 if (*format == '*') {
547 "Format with too many format specifiers.");
548 width = strtol(argv[pos++], NULL, 10);
550 width = strtol(format, __DECONST(char **,&format), 10);
558 if (*format == '.') {
560 if (*format == '*') {
564 "Format with too many format specifiers.");
565 extra = strtol(argv[pos++], NULL, 10);
567 extra = strtol(format, __DECONST(char **, &format), 10);
573 m4errx(1, "Format with too many format specifiers.");
576 thisarg = argv[pos++];
579 temp[0] = strtoul(argv[pos++], NULL, 10);
584 m4errx(1, "Unsupported format specification: %s.",
592 while ((long)l < width--)
595 addchars(thisarg, l);
597 while ((long)l < width--)
605 doesyscmd(const char *cmd)
613 /* Follow gnu m4 documentation: first flush buffers. */
616 argv[0] = __DECONST(char *, "sh");
617 argv[1] = __DECONST(char *, "-c");
618 argv[2] = __DECONST(char *, cmd);
621 /* Just set up standard output, share stderr and stdin with m4 */
624 switch(cpid = fork()) {
630 (void) dup2(p[1], 1);
632 execv(_PATH_BSHELL, argv);
635 /* Read result in two stages, since m4's buffer is
639 char result[BUFSIZE];
640 cc = read(p[0], result, sizeof result);
642 addchars(result, cc);
643 } while (cc > 0 || (cc == -1 && errno == EINTR));
646 while ((pid = wait(&status)) != cpid && pid >= 0)
653 getdivfile(const char *name)
658 f = fopen(name, "r");
662 while ((c = getc(f))!= EOF)