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 int appendx
; /* Index into appends array. */
81 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
) {
215 if (appendx
>= appendnum
)
216 if ((appends
= realloc(appends
,
217 sizeof(struct s_appends
) *
218 (appendnum
*= 2))) == NULL
)
220 appends
[appendx
].type
= AP_FILE
;
221 appends
[appendx
].s
= cp
->t
;
222 appends
[appendx
].len
= strlen(cp
->t
);
226 sdone
|= substitute(cp
);
238 if (cp
->u
.fd
== -1 && (cp
->u
.fd
= open(cp
->t
,
239 O_WRONLY
|O_APPEND
|O_CREAT
|O_TRUNC
,
242 if (write(cp
->u
.fd
, ps
, psl
) != (ssize_t
)psl
||
243 write(cp
->u
.fd
, "\n", 1) != 1)
248 * If the hold space is null, make it empty
249 * but not null. Otherwise the pattern space
250 * will become null after the swap, which is
251 * an abnormal condition.
254 cspace(&HS
, "", 0, REPLACE
);
257 psanl
= tspace
.append_newline
;
269 (void)fprintf(outfile
, "%lu\n", linenum
);
274 new: if (!nflag
&& !pd
)
277 } /* for all lines */
281 * TRUE if the address passed matches the current program state
282 * (lastline, linenumber, ps).
285 ((a)->type == AT_RE ? regexec_e((a)->u.r, ps, 0, 1, 0, psl) : \
286 (a)->type == AT_LINE ? linenum == (a)->u.l : lastline())
289 * Return TRUE if the command applies to the current line. Sets the start
290 * line for process ranges. Interprets the non-select (``!'') flag.
293 applies(struct s_command
*cp
)
298 if (cp
->a1
== NULL
&& cp
->a2
== NULL
)
301 if (cp
->startline
> 0) {
302 switch (cp
->a2
->type
) {
304 if (linenum
- cp
->startline
<= cp
->a2
->u
.l
)
316 } else if (cp
->a2
->type
== AT_LINE
&&
317 linenum
> cp
->a2
->u
.l
) {
319 * We missed the 2nd address due to a
320 * branch, so just close the range and
328 } else if (cp
->a1
&& MATCH(cp
->a1
)) {
330 * If the second address is a number less than or
331 * equal to the line number first selected, only
332 * one line shall be selected.
334 * Likewise if the relative second line address is zero.
336 if ((cp
->a2
->type
== AT_LINE
&&
337 linenum
>= cp
->a2
->u
.l
) ||
338 (cp
->a2
->type
== AT_RELLINE
&& cp
->a2
->u
.l
== 0))
341 cp
->startline
= linenum
;
348 return (cp
->nonsel
? ! r
: r
);
352 * Reset the sed processor to its initial state.
357 struct s_command
*cp
;
360 * Reset all in-range markers.
362 for (cp
= prog
; cp
; cp
= cp
->code
== '{' ? cp
->u
.c
: cp
->next
)
367 * Clear out the hold space.
369 cspace(&HS
, "", 0, REPLACE
);
374 * Do substitutions in the pattern space. Currently, we build a
375 * copy of the new pattern space in the substitute space structure
376 * and then swap them.
379 substitute(struct s_command
*cp
)
391 if (defpreg
!= NULL
&& cp
->u
.s
->maxbref
> defpreg
->re_nsub
) {
392 linenum
= cp
->u
.s
->linenum
;
393 errx(1, "%lu: %s: \\%u not defined in the RE",
394 linenum
, fname
, cp
->u
.s
->maxbref
);
397 if (!regexec_e(re
, s
, 0, 0, 0, psl
))
400 SS
.len
= 0; /* Clean substitute space. */
406 /* Copy the leading retained string. */
407 if (n
<= 1 && match
[0].rm_so
- le
)
408 cspace(&SS
, s
, match
[0].rm_so
- le
, APPEND
);
410 /* Skip zero-length matches right after other matches. */
411 if (lastempty
|| (match
[0].rm_so
- le
) ||
412 match
[0].rm_so
!= match
[0].rm_eo
) {
414 /* Want this match: append replacement. */
415 regsub(&SS
, ps
, cp
->u
.s
->new);
419 /* Want a later match: append original. */
420 if (match
[0].rm_eo
- le
)
421 cspace(&SS
, s
, match
[0].rm_eo
- le
,
427 /* Move past this match. */
428 s
+= (match
[0].rm_eo
- le
);
429 slen
-= (match
[0].rm_eo
- le
);
433 * After a zero-length match, advance one byte,
434 * and at the end of the line, terminate.
436 if (match
[0].rm_so
== match
[0].rm_eo
) {
437 if (*s
== '\0' || *s
== '\n')
442 cspace(&SS
, s
++, 1, APPEND
);
449 } while (n
>= 0 && slen
>= 0 && regexec_e(re
, ps
, 0, 0, le
, psl
));
451 /* Did not find the requested number of matches. */
455 /* Copy the trailing retained string. */
457 cspace(&SS
, s
, slen
, APPEND
);
460 * Swap the substitute space and the pattern space, and make sure
461 * that any leftover pointers into stdio memory get lost.
465 psanl
= tspace
.append_newline
;
469 /* Handle the 'p' flag. */
473 /* Handle the 'w' flag. */
474 if (cp
->u
.s
->wfile
&& !pd
) {
475 if (cp
->u
.s
->wfd
== -1 && (cp
->u
.s
->wfd
= open(cp
->u
.s
->wfile
,
476 O_WRONLY
|O_APPEND
|O_CREAT
|O_TRUNC
, DEFFILEMODE
)) == -1)
477 err(1, "%s", cp
->u
.s
->wfile
);
478 if (write(cp
->u
.s
->wfd
, ps
, psl
) != (ssize_t
)psl
||
479 write(cp
->u
.s
->wfd
, "\n", 1) != 1)
480 err(1, "%s", cp
->u
.s
->wfile
);
487 * Perform translation ('y' command) in the pattern space.
490 do_tr(struct s_tr
*y
)
497 if (MB_CUR_MAX
== 1) {
499 * Single-byte encoding: perform in-place translation
500 * of the pattern space.
502 for (p
= ps
; p
< &ps
[psl
]; p
++)
503 *p
= y
->bytetab
[(u_char
)*p
];
506 * Multi-byte encoding: perform translation into the
507 * translation space, then swap the translation and
510 /* Clean translation space. */
512 for (p
= ps
, left
= psl
; left
> 0; p
+= clen
, left
-= clen
) {
513 if ((c
= y
->bytetab
[(u_char
)*p
]) != '\0') {
514 cspace(&YS
, &c
, 1, APPEND
);
518 for (i
= 0; i
< y
->nmultis
; i
++)
519 if (left
>= y
->multis
[i
].fromlen
&&
520 memcmp(p
, y
->multis
[i
].from
,
521 y
->multis
[i
].fromlen
) == 0)
523 if (i
< y
->nmultis
) {
524 cspace(&YS
, y
->multis
[i
].to
,
525 y
->multis
[i
].tolen
, APPEND
);
526 clen
= y
->multis
[i
].fromlen
;
528 cspace(&YS
, p
, 1, APPEND
);
532 /* Swap the translation space and the pattern space. */
535 psanl
= tmp
.append_newline
;
542 * Flush append requests. Always called before reading a line,
543 * therefore it also resets the substitution done (sdone) flag.
552 for (i
= 0; i
< appendx
; i
++)
553 switch (appends
[i
].type
) {
555 fwrite(appends
[i
].s
, sizeof(char), appends
[i
].len
,
560 * Read files probably shouldn't be cached. Since
561 * it's not an error to read a non-existent file,
562 * it's possible that another program is interacting
563 * with the sed script through the filesystem. It
564 * would be truly bizarre, but possible. It's probably
565 * not that big a performance win, anyhow.
567 if ((f
= fopen(appends
[i
].s
, "r")) == NULL
)
569 while ((count
= fread(buf
, sizeof(char), sizeof(buf
), f
)))
570 (void)fwrite(buf
, sizeof(char), count
, outfile
);
575 errx(1, "%s: %s", outfname
, strerror(errno
? errno
: EIO
));
580 lputs(char *s
, size_t len
)
582 static const char escapes
[] = "\\\a\b\f\r\t\v";
586 static int termwidth
= -1;
591 if (outfile
!= stdout
)
593 if (termwidth
== -1) {
594 if ((p
= getenv("COLUMNS")) && *p
!= '\0')
596 else if (ioctl(STDOUT_FILENO
, TIOCGWINSZ
, &win
) == 0 &&
598 termwidth
= win
.ws_col
;
605 memset(&mbs
, 0, sizeof(mbs
));
608 clen
= mbrtowc(&wc
, s
, len
, &mbs
);
611 if (clen
== (size_t)-1 || clen
== (size_t)-2) {
612 wc
= (unsigned char)*s
;
614 memset(&mbs
, 0, sizeof(mbs
));
617 if (col
+ 1 >= termwidth
)
618 fprintf(outfile
, "\\\n");
620 fputc('\n', outfile
);
622 } else if (iswprint(wc
)) {
624 if (col
+ width
>= termwidth
) {
625 fprintf(outfile
, "\\\n");
628 fwrite(s
, 1, clen
, outfile
);
630 } else if (wc
!= L
'\0' && (c
= wctob(wc
)) != EOF
&&
631 (p
= strchr(escapes
, c
)) != NULL
) {
632 if (col
+ 2 >= termwidth
) {
633 fprintf(outfile
, "\\\n");
636 fprintf(outfile
, "\\%c", "\\abfrtv"[p
- escapes
]);
639 if (col
+ 4 * clen
>= (unsigned)termwidth
) {
640 fprintf(outfile
, "\\\n");
643 for (i
= 0; i
< clen
; i
++)
644 fprintf(outfile
, "\\%03o",
645 (int)(unsigned char)s
[i
]);
651 if (col
+ 1 >= termwidth
)
652 fprintf(outfile
, "\\\n");
653 (void)fputc('$', outfile
);
654 (void)fputc('\n', outfile
);
656 errx(1, "%s: %s", outfname
, strerror(errno
? errno
: EIO
));
660 regexec_e(regex_t
*preg
, const char *string
, int eflags
, int nomatch
,
661 size_t start
, size_t stop
)
667 errx(1, "first RE may not be empty");
672 match
[0].rm_so
= start
;
673 match
[0].rm_eo
= stop
;
675 eval
= regexec(defpreg
, string
,
676 nomatch
? 0 : maxnsub
+ 1, match
, eflags
| REG_STARTEND
);
683 errx(1, "RE error: %s", strregerror(eval
, defpreg
));
688 * regsub - perform substitutions after a regexp match
689 * Based on a routine by Henry Spencer
692 regsub(SPACE
*sp
, char *string
, char *src
)
697 #define NEEDSP(reqlen) \
698 /* XXX What is the +1 for? */ \
699 if (sp->len + (reqlen) + 1 >= sp->blen) { \
700 sp->blen += (reqlen) + 1024; \
701 if ((sp->space = sp->back = realloc(sp->back, sp->blen)) \
704 dst = sp->space + sp->len; \
707 dst
= sp
->space
+ sp
->len
;
708 while ((c
= *src
++) != '\0') {
711 else if (c
== '\\' && isdigit((unsigned char)*src
))
715 if (no
< 0) { /* Ordinary character. */
716 if (c
== '\\' && (*src
== '\\' || *src
== '&'))
721 } else if (match
[no
].rm_so
!= -1 && match
[no
].rm_eo
!= -1) {
722 len
= match
[no
].rm_eo
- match
[no
].rm_so
;
724 memmove(dst
, string
+ match
[no
].rm_so
, len
);
735 * Concatenate space: append the source space to the destination space,
736 * allocating new space as necessary.
739 cspace(SPACE
*sp
, const char *p
, size_t len
, enum e_spflag spflag
)
743 /* Make sure SPACE has enough memory and ramp up quickly. */
744 tlen
= sp
->len
+ len
+ 1;
745 if (tlen
> sp
->blen
) {
746 sp
->blen
= tlen
+ 1024;
747 if ((sp
->space
= sp
->back
= realloc(sp
->back
, sp
->blen
)) ==
752 if (spflag
== REPLACE
)
755 memmove(sp
->space
+ sp
->len
, p
, len
);
757 sp
->space
[sp
->len
+= len
] = '\0';
761 * Close all cached opened files and report any errors
764 cfclose(struct s_command
*cp
, struct s_command
*end
)
767 for (; cp
!= end
; cp
= cp
->next
)
770 if (cp
->u
.s
->wfd
!= -1 && close(cp
->u
.s
->wfd
))
771 err(1, "%s", cp
->u
.s
->wfile
);
775 if (cp
->u
.fd
!= -1 && close(cp
->u
.fd
))
780 cfclose(cp
->u
.c
, cp
->next
);