2 * Copyright (c) 1992 Diomidis Spinellis.
3 * Copyright (c) 1992, 1993, 1994
4 * The Regents of the University of California. All rights reserved.
6 * This code is derived from software contributed to Berkeley by
7 * Diomidis Spinellis of Imperial College, University of London.
9 * Redistribution and use in source and binary forms, with or without
10 * modification, are permitted provided that the following conditions
12 * 1. Redistributions of source code must retain the above copyright
13 * notice, this list of conditions and the following disclaimer.
14 * 2. Redistributions in binary form must reproduce the above copyright
15 * notice, this list of conditions and the following disclaimer in the
16 * documentation and/or other materials provided with the distribution.
17 * 4. Neither the name of the University nor the names of its contributors
18 * may be used to endorse or promote products derived from this software
19 * without specific prior written permission.
21 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
22 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
23 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
24 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
25 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
26 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
27 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
28 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
29 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
30 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
34 #include <sys/cdefs.h>
35 __FBSDID("$FreeBSD$");
38 static const char sccsid[] = "@(#)process.c 8.6 (Berkeley) 4/20/94";
41 #include <sys/types.h>
43 #include <sys/ioctl.h>
62 static SPACE HS, PS, SS, YS;
66 #define psanl PS.append_newline
70 static inline int applies(struct s_command *);
71 static void do_tr(struct s_tr *);
72 static void flush_appends(void);
73 static void lputs(char *, size_t);
74 static int regexec_e(regex_t *, const char *, int, int, size_t,
76 static void regsub(SPACE *, char *, char *);
77 static int substitute(struct s_command *);
79 struct s_appends *appends; /* Array of pointers to strings to append. */
80 static unsigned int appendx; /* Index into appends array. */
81 unsigned int appendnum; /* Size of appends array. */
83 static int lastaddr; /* Set by applies if last address of a range. */
84 static int sdone; /* If any substitutes since last line input. */
85 /* Iov structure for 'w' commands. */
86 static regex_t *defpreg;
91 fwrite(ps, 1, psl, outfile); \
92 if (psanl) fputc('\n', outfile); \
106 for (linenum = 0; mf_fgets(&PS, REPLACE);) {
121 if (appendx >= appendnum)
122 if ((appends = realloc(appends,
123 sizeof(struct s_appends) *
124 (appendnum *= 2))) == NULL)
126 appends[appendx].type = AP_STRING;
127 appends[appendx].s = cp->t;
128 appends[appendx].len = strlen(cp->t);
137 if (cp->a2 == NULL || lastaddr || lastline())
138 (void)fprintf(outfile, "%s", cp->t);
147 (p = memchr(ps, '\n', psl)) == NULL) {
152 memmove(ps, p + 1, psl);
156 cspace(&PS, hs, hsl, REPLACE);
159 cspace(&PS, "\n", 1, APPEND);
160 cspace(&PS, hs, hsl, APPEND);
163 cspace(&HS, ps, psl, REPLACE);
166 cspace(&HS, "\n", 1, APPEND);
167 cspace(&HS, ps, psl, APPEND);
170 (void)fprintf(outfile, "%s", cp->t);
179 if (!mf_fgets(&PS, REPLACE))
185 cspace(&PS, "\n", 1, APPEND);
186 if (!mf_fgets(&PS, APPEND))
197 if ((p = memchr(ps, '\n', psl)) != NULL) {
210 if (inplace == NULL) {
219 if (appendx >= appendnum)
220 if ((appends = realloc(appends,
221 sizeof(struct s_appends) *
222 (appendnum *= 2))) == NULL)
224 appends[appendx].type = AP_FILE;
225 appends[appendx].s = cp->t;
226 appends[appendx].len = strlen(cp->t);
230 sdone |= substitute(cp);
242 if (cp->u.fd == -1 && (cp->u.fd = open(cp->t,
243 O_WRONLY|O_APPEND|O_CREAT|O_TRUNC,
246 if (write(cp->u.fd, ps, psl) != (ssize_t)psl ||
247 write(cp->u.fd, "\n", 1) != 1)
252 * If the hold space is null, make it empty
253 * but not null. Otherwise the pattern space
254 * will become null after the swap, which is
255 * an abnormal condition.
258 cspace(&HS, "", 0, REPLACE);
261 psanl = tspace.append_newline;
273 (void)fprintf(outfile, "%lu\n", linenum);
278 new: if (!nflag && !pd)
281 } /* for all lines */
285 * TRUE if the address passed matches the current program state
286 * (lastline, linenumber, ps).
289 ((a)->type == AT_RE ? regexec_e((a)->u.r, ps, 0, 1, 0, psl) : \
290 (a)->type == AT_LINE ? linenum == (a)->u.l : lastline())
293 * Return TRUE if the command applies to the current line. Sets the start
294 * line for process ranges. Interprets the non-select (``!'') flag.
297 applies(struct s_command *cp)
302 if (cp->a1 == NULL && cp->a2 == NULL)
305 if (cp->startline > 0) {
306 switch (cp->a2->type) {
308 if (linenum - cp->startline <= cp->a2->u.l)
320 } else if (cp->a2->type == AT_LINE &&
321 linenum > cp->a2->u.l) {
323 * We missed the 2nd address due to a
324 * branch, so just close the range and
332 } else if (cp->a1 && MATCH(cp->a1)) {
334 * If the second address is a number less than or
335 * equal to the line number first selected, only
336 * one line shall be selected.
338 * Likewise if the relative second line address is zero.
340 if ((cp->a2->type == AT_LINE &&
341 linenum >= cp->a2->u.l) ||
342 (cp->a2->type == AT_RELLINE && cp->a2->u.l == 0))
345 cp->startline = linenum;
352 return (cp->nonsel ? ! r : r);
356 * Reset the sed processor to its initial state.
361 struct s_command *cp;
364 * Reset all in-range markers.
366 for (cp = prog; cp; cp = cp->code == '{' ? cp->u.c : cp->next)
371 * Clear out the hold space.
373 cspace(&HS, "", 0, REPLACE);
378 * Do substitutions in the pattern space. Currently, we build a
379 * copy of the new pattern space in the substitute space structure
380 * and then swap them.
383 substitute(struct s_command *cp)
395 if (defpreg != NULL && cp->u.s->maxbref > defpreg->re_nsub) {
396 linenum = cp->u.s->linenum;
397 errx(1, "%lu: %s: \\%u not defined in the RE",
398 linenum, fname, cp->u.s->maxbref);
401 if (!regexec_e(re, ps, 0, 0, 0, psl))
404 SS.len = 0; /* Clean substitute space. */
410 /* Copy the leading retained string. */
411 if (n <= 1 && (match[0].rm_so > le))
412 cspace(&SS, s, match[0].rm_so - le, APPEND);
414 /* Skip zero-length matches right after other matches. */
415 if (lastempty || (match[0].rm_so - le) ||
416 match[0].rm_so != match[0].rm_eo) {
418 /* Want this match: append replacement. */
419 regsub(&SS, ps, cp->u.s->new);
423 /* Want a later match: append original. */
424 if (match[0].rm_eo - le)
425 cspace(&SS, s, match[0].rm_eo - le,
431 /* Move past this match. */
432 s = ps + match[0].rm_eo;
433 slen = psl - match[0].rm_eo;
437 * After a zero-length match, advance one byte,
438 * and at the end of the line, terminate.
440 if (match[0].rm_so == match[0].rm_eo) {
441 if (*s == '\0' || *s == '\n')
446 cspace(&SS, s++, 1, APPEND);
453 } while (n >= 0 && slen >= 0 &&
454 regexec_e(re, ps, REG_NOTBOL, 0, le, psl));
456 /* Did not find the requested number of matches. */
460 /* Copy the trailing retained string. */
462 cspace(&SS, s, slen, APPEND);
465 * Swap the substitute space and the pattern space, and make sure
466 * that any leftover pointers into stdio memory get lost.
470 psanl = tspace.append_newline;
474 /* Handle the 'p' flag. */
478 /* Handle the 'w' flag. */
479 if (cp->u.s->wfile && !pd) {
480 if (cp->u.s->wfd == -1 && (cp->u.s->wfd = open(cp->u.s->wfile,
481 O_WRONLY|O_APPEND|O_CREAT|O_TRUNC, DEFFILEMODE)) == -1)
482 err(1, "%s", cp->u.s->wfile);
483 if (write(cp->u.s->wfd, ps, psl) != (ssize_t)psl ||
484 write(cp->u.s->wfd, "\n", 1) != 1)
485 err(1, "%s", cp->u.s->wfile);
492 * Perform translation ('y' command) in the pattern space.
495 do_tr(struct s_tr *y)
502 if (MB_CUR_MAX == 1) {
504 * Single-byte encoding: perform in-place translation
505 * of the pattern space.
507 for (p = ps; p < &ps[psl]; p++)
508 *p = y->bytetab[(u_char)*p];
511 * Multi-byte encoding: perform translation into the
512 * translation space, then swap the translation and
515 /* Clean translation space. */
517 for (p = ps, left = psl; left > 0; p += clen, left -= clen) {
518 if ((c = y->bytetab[(u_char)*p]) != '\0') {
519 cspace(&YS, &c, 1, APPEND);
523 for (i = 0; i < y->nmultis; i++)
524 if (left >= y->multis[i].fromlen &&
525 memcmp(p, y->multis[i].from,
526 y->multis[i].fromlen) == 0)
528 if (i < y->nmultis) {
529 cspace(&YS, y->multis[i].to,
530 y->multis[i].tolen, APPEND);
531 clen = y->multis[i].fromlen;
533 cspace(&YS, p, 1, APPEND);
537 /* Swap the translation space and the pattern space. */
540 psanl = tmp.append_newline;
547 * Flush append requests. Always called before reading a line,
548 * therefore it also resets the substitution done (sdone) flag.
554 unsigned int count, idx;
557 for (idx = 0; idx < appendx; idx++)
558 switch (appends[idx].type) {
560 fwrite(appends[idx].s, sizeof(char), appends[idx].len,
565 * Read files probably shouldn't be cached. Since
566 * it's not an error to read a non-existent file,
567 * it's possible that another program is interacting
568 * with the sed script through the filesystem. It
569 * would be truly bizarre, but possible. It's probably
570 * not that big a performance win, anyhow.
572 if ((f = fopen(appends[idx].s, "r")) == NULL)
574 while ((count = fread(buf, sizeof(char), sizeof(buf), f)))
575 (void)fwrite(buf, sizeof(char), count, outfile);
580 errx(1, "%s: %s", outfname, strerror(errno ? errno : EIO));
585 lputs(char *s, size_t len)
587 static const char escapes[] = "\\\a\b\f\r\t\v";
591 static int termwidth = -1;
596 if (outfile != stdout)
598 if (termwidth == -1) {
599 if ((p = getenv("COLUMNS")) && *p != '\0')
601 else if (ioctl(STDOUT_FILENO, TIOCGWINSZ, &win) == 0 &&
603 termwidth = win.ws_col;
610 memset(&mbs, 0, sizeof(mbs));
613 clen = mbrtowc(&wc, s, len, &mbs);
616 if (clen == (size_t)-1 || clen == (size_t)-2) {
617 wc = (unsigned char)*s;
619 memset(&mbs, 0, sizeof(mbs));
622 if (col + 1 >= termwidth)
623 fprintf(outfile, "\\\n");
625 fputc('\n', outfile);
627 } else if (iswprint(wc)) {
629 if (col + width >= termwidth) {
630 fprintf(outfile, "\\\n");
633 fwrite(s, 1, clen, outfile);
635 } else if (wc != L'\0' && (c = wctob(wc)) != EOF &&
636 (p = strchr(escapes, c)) != NULL) {
637 if (col + 2 >= termwidth) {
638 fprintf(outfile, "\\\n");
641 fprintf(outfile, "\\%c", "\\abfrtv"[p - escapes]);
644 if (col + 4 * clen >= (unsigned)termwidth) {
645 fprintf(outfile, "\\\n");
648 for (i = 0; i < clen; i++)
649 fprintf(outfile, "\\%03o",
650 (int)(unsigned char)s[i]);
656 if (col + 1 >= termwidth)
657 fprintf(outfile, "\\\n");
658 (void)fputc('$', outfile);
659 (void)fputc('\n', outfile);
661 errx(1, "%s: %s", outfname, strerror(errno ? errno : EIO));
665 regexec_e(regex_t *preg, const char *string, int eflags, int nomatch,
666 size_t start, size_t stop)
672 errx(1, "first RE may not be empty");
677 match[0].rm_so = start;
678 match[0].rm_eo = stop;
680 eval = regexec(defpreg, string,
681 nomatch ? 0 : maxnsub + 1, match, eflags | REG_STARTEND);
688 errx(1, "RE error: %s", strregerror(eval, defpreg));
693 * regsub - perform substitutions after a regexp match
694 * Based on a routine by Henry Spencer
697 regsub(SPACE *sp, char *string, char *src)
702 #define NEEDSP(reqlen) \
703 /* XXX What is the +1 for? */ \
704 if (sp->len + (reqlen) + 1 >= sp->blen) { \
705 sp->blen += (reqlen) + 1024; \
706 if ((sp->space = sp->back = realloc(sp->back, sp->blen)) \
709 dst = sp->space + sp->len; \
712 dst = sp->space + sp->len;
713 while ((c = *src++) != '\0') {
716 else if (c == '\\' && isdigit((unsigned char)*src))
720 if (no < 0) { /* Ordinary character. */
721 if (c == '\\' && (*src == '\\' || *src == '&'))
726 } else if (match[no].rm_so != -1 && match[no].rm_eo != -1) {
727 len = match[no].rm_eo - match[no].rm_so;
729 memmove(dst, string + match[no].rm_so, len);
740 * Concatenate space: append the source space to the destination space,
741 * allocating new space as necessary.
744 cspace(SPACE *sp, const char *p, size_t len, enum e_spflag spflag)
748 /* Make sure SPACE has enough memory and ramp up quickly. */
749 tlen = sp->len + len + 1;
750 if (tlen > sp->blen) {
751 sp->blen = tlen + 1024;
752 if ((sp->space = sp->back = realloc(sp->back, sp->blen)) ==
757 if (spflag == REPLACE)
760 memmove(sp->space + sp->len, p, len);
762 sp->space[sp->len += len] = '\0';
766 * Close all cached opened files and report any errors
769 cfclose(struct s_command *cp, struct s_command *end)
772 for (; cp != end; cp = cp->next)
775 if (cp->u.s->wfd != -1 && close(cp->u.s->wfd))
776 err(1, "%s", cp->u.s->wfile);
780 if (cp->u.fd != -1 && close(cp->u.fd))
785 cfclose(cp->u.c, cp->next);