mirror of
https://git.FreeBSD.org/src.git
synced 2024-12-19 10:53:58 +00:00
85184b8b23
used once on a non-empty pattern space and then again on an empty pattern space, the second usage restores the pattern space length to the length that it had when the first "P" was used. PR: bin/96052 Submitted by: Andrey Zholos <aaz@althenia.net> MFC after: 7 days
719 lines
17 KiB
C
719 lines
17 KiB
C
/*-
|
|
* Copyright (c) 1992 Diomidis Spinellis.
|
|
* Copyright (c) 1992, 1993, 1994
|
|
* The Regents of the University of California. All rights reserved.
|
|
*
|
|
* This code is derived from software contributed to Berkeley by
|
|
* Diomidis Spinellis of Imperial College, University of London.
|
|
*
|
|
* Redistribution and use in source and binary forms, with or without
|
|
* modification, are permitted provided that the following conditions
|
|
* are met:
|
|
* 1. Redistributions of source code must retain the above copyright
|
|
* notice, this list of conditions and the following disclaimer.
|
|
* 2. Redistributions in binary form must reproduce the above copyright
|
|
* notice, this list of conditions and the following disclaimer in the
|
|
* documentation and/or other materials provided with the distribution.
|
|
* 4. Neither the name of the University nor the names of its contributors
|
|
* may be used to endorse or promote products derived from this software
|
|
* without specific prior written permission.
|
|
*
|
|
* THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
|
|
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
|
|
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
|
|
* ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
|
|
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
|
|
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
|
|
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
|
|
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
|
|
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
|
|
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
|
|
* SUCH DAMAGE.
|
|
*/
|
|
|
|
#include <sys/cdefs.h>
|
|
__FBSDID("$FreeBSD$");
|
|
|
|
#ifndef lint
|
|
static const char sccsid[] = "@(#)process.c 8.6 (Berkeley) 4/20/94";
|
|
#endif
|
|
|
|
#include <sys/types.h>
|
|
#include <sys/stat.h>
|
|
#include <sys/ioctl.h>
|
|
#include <sys/uio.h>
|
|
|
|
#include <ctype.h>
|
|
#include <err.h>
|
|
#include <errno.h>
|
|
#include <fcntl.h>
|
|
#include <limits.h>
|
|
#include <regex.h>
|
|
#include <stdio.h>
|
|
#include <stdlib.h>
|
|
#include <string.h>
|
|
#include <unistd.h>
|
|
#include <wchar.h>
|
|
#include <wctype.h>
|
|
|
|
#include "defs.h"
|
|
#include "extern.h"
|
|
|
|
static SPACE HS, PS, SS, YS;
|
|
#define pd PS.deleted
|
|
#define ps PS.space
|
|
#define psl PS.len
|
|
#define hs HS.space
|
|
#define hsl HS.len
|
|
|
|
static __inline int applies(struct s_command *);
|
|
static void do_tr(struct s_tr *);
|
|
static void flush_appends(void);
|
|
static void lputs(char *, size_t);
|
|
static __inline int regexec_e(regex_t *, const char *, int, int, size_t);
|
|
static void regsub(SPACE *, char *, char *);
|
|
static int substitute(struct s_command *);
|
|
|
|
struct s_appends *appends; /* Array of pointers to strings to append. */
|
|
static int appendx; /* Index into appends array. */
|
|
int appendnum; /* Size of appends array. */
|
|
|
|
static int lastaddr; /* Set by applies if last address of a range. */
|
|
static int sdone; /* If any substitutes since last line input. */
|
|
/* Iov structure for 'w' commands. */
|
|
static regex_t *defpreg;
|
|
size_t maxnsub;
|
|
regmatch_t *match;
|
|
|
|
#define OUT(s) { fwrite(s, sizeof(u_char), psl, outfile); fputc('\n', outfile); }
|
|
|
|
void
|
|
process(void)
|
|
{
|
|
struct s_command *cp;
|
|
SPACE tspace;
|
|
size_t oldpsl = 0;
|
|
char *p;
|
|
|
|
p = NULL;
|
|
|
|
for (linenum = 0; mf_fgets(&PS, REPLACE);) {
|
|
pd = 0;
|
|
top:
|
|
cp = prog;
|
|
redirect:
|
|
while (cp != NULL) {
|
|
if (!applies(cp)) {
|
|
cp = cp->next;
|
|
continue;
|
|
}
|
|
switch (cp->code) {
|
|
case '{':
|
|
cp = cp->u.c;
|
|
goto redirect;
|
|
case 'a':
|
|
if (appendx >= appendnum)
|
|
if ((appends = realloc(appends,
|
|
sizeof(struct s_appends) *
|
|
(appendnum *= 2))) == NULL)
|
|
err(1, "realloc");
|
|
appends[appendx].type = AP_STRING;
|
|
appends[appendx].s = cp->t;
|
|
appends[appendx].len = strlen(cp->t);
|
|
appendx++;
|
|
break;
|
|
case 'b':
|
|
cp = cp->u.c;
|
|
goto redirect;
|
|
case 'c':
|
|
pd = 1;
|
|
psl = 0;
|
|
if (cp->a2 == NULL || lastaddr)
|
|
(void)fprintf(outfile, "%s", cp->t);
|
|
break;
|
|
case 'd':
|
|
pd = 1;
|
|
goto new;
|
|
case 'D':
|
|
if (pd)
|
|
goto new;
|
|
if (psl == 0 ||
|
|
(p = memchr(ps, '\n', psl)) == NULL) {
|
|
pd = 1;
|
|
goto new;
|
|
} else {
|
|
psl -= (p + 1) - ps;
|
|
memmove(ps, p + 1, psl);
|
|
goto top;
|
|
}
|
|
case 'g':
|
|
cspace(&PS, hs, hsl, REPLACE);
|
|
break;
|
|
case 'G':
|
|
cspace(&PS, "\n", 1, 0);
|
|
cspace(&PS, hs, hsl, 0);
|
|
break;
|
|
case 'h':
|
|
cspace(&HS, ps, psl, REPLACE);
|
|
break;
|
|
case 'H':
|
|
cspace(&HS, "\n", 1, 0);
|
|
cspace(&HS, ps, psl, 0);
|
|
break;
|
|
case 'i':
|
|
(void)fprintf(outfile, "%s", cp->t);
|
|
break;
|
|
case 'l':
|
|
lputs(ps, psl);
|
|
break;
|
|
case 'n':
|
|
if (!nflag && !pd)
|
|
OUT(ps)
|
|
flush_appends();
|
|
if (!mf_fgets(&PS, REPLACE))
|
|
exit(0);
|
|
pd = 0;
|
|
break;
|
|
case 'N':
|
|
flush_appends();
|
|
cspace(&PS, "\n", 1, 0);
|
|
if (!mf_fgets(&PS, 0))
|
|
exit(0);
|
|
break;
|
|
case 'p':
|
|
if (pd)
|
|
break;
|
|
OUT(ps)
|
|
break;
|
|
case 'P':
|
|
if (pd)
|
|
break;
|
|
if ((p = memchr(ps, '\n', psl)) != NULL) {
|
|
oldpsl = psl;
|
|
psl = p - ps;
|
|
}
|
|
OUT(ps)
|
|
if (p != NULL)
|
|
psl = oldpsl;
|
|
break;
|
|
case 'q':
|
|
if (!nflag && !pd)
|
|
OUT(ps)
|
|
flush_appends();
|
|
exit(0);
|
|
case 'r':
|
|
if (appendx >= appendnum)
|
|
if ((appends = realloc(appends,
|
|
sizeof(struct s_appends) *
|
|
(appendnum *= 2))) == NULL)
|
|
err(1, "realloc");
|
|
appends[appendx].type = AP_FILE;
|
|
appends[appendx].s = cp->t;
|
|
appends[appendx].len = strlen(cp->t);
|
|
appendx++;
|
|
break;
|
|
case 's':
|
|
sdone |= substitute(cp);
|
|
break;
|
|
case 't':
|
|
if (sdone) {
|
|
sdone = 0;
|
|
cp = cp->u.c;
|
|
goto redirect;
|
|
}
|
|
break;
|
|
case 'w':
|
|
if (pd)
|
|
break;
|
|
if (cp->u.fd == -1 && (cp->u.fd = open(cp->t,
|
|
O_WRONLY|O_APPEND|O_CREAT|O_TRUNC,
|
|
DEFFILEMODE)) == -1)
|
|
err(1, "%s", cp->t);
|
|
if (write(cp->u.fd, ps, psl) != psl ||
|
|
write(cp->u.fd, "\n", 1) != 1)
|
|
err(1, "%s", cp->t);
|
|
break;
|
|
case 'x':
|
|
if (hs == NULL)
|
|
cspace(&HS, "", 0, REPLACE);
|
|
tspace = PS;
|
|
PS = HS;
|
|
HS = tspace;
|
|
break;
|
|
case 'y':
|
|
if (pd || psl == 0)
|
|
break;
|
|
do_tr(cp->u.y);
|
|
break;
|
|
case ':':
|
|
case '}':
|
|
break;
|
|
case '=':
|
|
(void)fprintf(outfile, "%lu\n", linenum);
|
|
}
|
|
cp = cp->next;
|
|
} /* for all cp */
|
|
|
|
new: if (!nflag && !pd)
|
|
OUT(ps)
|
|
flush_appends();
|
|
} /* for all lines */
|
|
}
|
|
|
|
/*
|
|
* TRUE if the address passed matches the current program state
|
|
* (lastline, linenumber, ps).
|
|
*/
|
|
#define MATCH(a) \
|
|
(a)->type == AT_RE ? regexec_e((a)->u.r, ps, 0, 1, psl) : \
|
|
(a)->type == AT_LINE ? linenum == (a)->u.l : lastline()
|
|
|
|
/*
|
|
* Return TRUE if the command applies to the current line. Sets the inrange
|
|
* flag to process ranges. Interprets the non-select (``!'') flag.
|
|
*/
|
|
static __inline int
|
|
applies(struct s_command *cp)
|
|
{
|
|
int r;
|
|
|
|
lastaddr = 0;
|
|
if (cp->a1 == NULL && cp->a2 == NULL)
|
|
r = 1;
|
|
else if (cp->a2)
|
|
if (cp->inrange) {
|
|
if (MATCH(cp->a2)) {
|
|
cp->inrange = 0;
|
|
lastaddr = 1;
|
|
}
|
|
r = 1;
|
|
} else if (MATCH(cp->a1)) {
|
|
/*
|
|
* If the second address is a number less than or
|
|
* equal to the line number first selected, only
|
|
* one line shall be selected.
|
|
* -- POSIX 1003.2
|
|
*/
|
|
if (cp->a2->type == AT_LINE &&
|
|
linenum >= cp->a2->u.l)
|
|
lastaddr = 1;
|
|
else
|
|
cp->inrange = 1;
|
|
r = 1;
|
|
} else
|
|
r = 0;
|
|
else
|
|
r = MATCH(cp->a1);
|
|
return (cp->nonsel ? ! r : r);
|
|
}
|
|
|
|
/*
|
|
* substitute --
|
|
* Do substitutions in the pattern space. Currently, we build a
|
|
* copy of the new pattern space in the substitute space structure
|
|
* and then swap them.
|
|
*/
|
|
static int
|
|
substitute(struct s_command *cp)
|
|
{
|
|
SPACE tspace;
|
|
regex_t *re;
|
|
regoff_t re_off, slen;
|
|
int lastempty, n;
|
|
char *s;
|
|
|
|
s = ps;
|
|
re = cp->u.s->re;
|
|
if (re == NULL) {
|
|
if (defpreg != NULL && cp->u.s->maxbref > defpreg->re_nsub) {
|
|
linenum = cp->u.s->linenum;
|
|
errx(1, "%lu: %s: \\%d not defined in the RE",
|
|
linenum, fname, cp->u.s->maxbref);
|
|
}
|
|
}
|
|
if (!regexec_e(re, s, 0, 0, psl))
|
|
return (0);
|
|
|
|
SS.len = 0; /* Clean substitute space. */
|
|
slen = psl;
|
|
n = cp->u.s->n;
|
|
lastempty = 1;
|
|
|
|
switch (n) {
|
|
case 0: /* Global */
|
|
do {
|
|
if (lastempty || match[0].rm_so != match[0].rm_eo) {
|
|
/* Locate start of replaced string. */
|
|
re_off = match[0].rm_so;
|
|
/* Copy leading retained string. */
|
|
cspace(&SS, s, re_off, APPEND);
|
|
/* Add in regular expression. */
|
|
regsub(&SS, s, cp->u.s->new);
|
|
}
|
|
|
|
/* Move past this match. */
|
|
if (match[0].rm_so != match[0].rm_eo) {
|
|
s += match[0].rm_eo;
|
|
slen -= match[0].rm_eo;
|
|
lastempty = 0;
|
|
} else {
|
|
if (match[0].rm_so < slen)
|
|
cspace(&SS, s + match[0].rm_so, 1,
|
|
APPEND);
|
|
s += match[0].rm_so + 1;
|
|
slen -= match[0].rm_so + 1;
|
|
lastempty = 1;
|
|
}
|
|
} while (slen >= 0 && regexec_e(re, s, REG_NOTBOL, 0, slen));
|
|
/* Copy trailing retained string. */
|
|
if (slen > 0)
|
|
cspace(&SS, s, slen, APPEND);
|
|
break;
|
|
default: /* Nth occurrence */
|
|
while (--n) {
|
|
if (match[0].rm_eo == match[0].rm_so)
|
|
match[0].rm_eo = match[0].rm_so + 1;
|
|
s += match[0].rm_eo;
|
|
slen -= match[0].rm_eo;
|
|
if (slen < 0)
|
|
return (0);
|
|
if (!regexec_e(re, s, REG_NOTBOL, 0, slen))
|
|
return (0);
|
|
}
|
|
/* FALLTHROUGH */
|
|
case 1: /* 1st occurrence */
|
|
/* Locate start of replaced string. */
|
|
re_off = match[0].rm_so + (s - ps);
|
|
/* Copy leading retained string. */
|
|
cspace(&SS, ps, re_off, APPEND);
|
|
/* Add in regular expression. */
|
|
regsub(&SS, s, cp->u.s->new);
|
|
/* Copy trailing retained string. */
|
|
s += match[0].rm_eo;
|
|
slen -= match[0].rm_eo;
|
|
cspace(&SS, s, slen, APPEND);
|
|
break;
|
|
}
|
|
|
|
/*
|
|
* Swap the substitute space and the pattern space, and make sure
|
|
* that any leftover pointers into stdio memory get lost.
|
|
*/
|
|
tspace = PS;
|
|
PS = SS;
|
|
SS = tspace;
|
|
SS.space = SS.back;
|
|
|
|
/* Handle the 'p' flag. */
|
|
if (cp->u.s->p)
|
|
OUT(ps)
|
|
|
|
/* Handle the 'w' flag. */
|
|
if (cp->u.s->wfile && !pd) {
|
|
if (cp->u.s->wfd == -1 && (cp->u.s->wfd = open(cp->u.s->wfile,
|
|
O_WRONLY|O_APPEND|O_CREAT|O_TRUNC, DEFFILEMODE)) == -1)
|
|
err(1, "%s", cp->u.s->wfile);
|
|
if (write(cp->u.s->wfd, ps, psl) != psl ||
|
|
write(cp->u.s->wfd, "\n", 1) != 1)
|
|
err(1, "%s", cp->u.s->wfile);
|
|
}
|
|
return (1);
|
|
}
|
|
|
|
/*
|
|
* do_tr --
|
|
* Perform translation ('y' command) in the pattern space.
|
|
*/
|
|
static void
|
|
do_tr(struct s_tr *y)
|
|
{
|
|
SPACE tmp;
|
|
char c, *p;
|
|
size_t clen, left;
|
|
int i;
|
|
|
|
if (MB_CUR_MAX == 1) {
|
|
/*
|
|
* Single-byte encoding: perform in-place translation
|
|
* of the pattern space.
|
|
*/
|
|
for (p = ps; p < &ps[psl]; p++)
|
|
*p = y->bytetab[(u_char)*p];
|
|
} else {
|
|
/*
|
|
* Multi-byte encoding: perform translation into the
|
|
* translation space, then swap the translation and
|
|
* pattern spaces.
|
|
*/
|
|
/* Clean translation space. */
|
|
YS.len = 0;
|
|
for (p = ps, left = psl; left > 0; p += clen, left -= clen) {
|
|
if ((c = y->bytetab[(u_char)*p]) != '\0') {
|
|
cspace(&YS, &c, 1, APPEND);
|
|
clen = 1;
|
|
continue;
|
|
}
|
|
for (i = 0; i < y->nmultis; i++)
|
|
if (left >= y->multis[i].fromlen &&
|
|
memcmp(p, y->multis[i].from,
|
|
y->multis[i].fromlen) == 0)
|
|
break;
|
|
if (i < y->nmultis) {
|
|
cspace(&YS, y->multis[i].to,
|
|
y->multis[i].tolen, APPEND);
|
|
clen = y->multis[i].fromlen;
|
|
} else {
|
|
cspace(&YS, p, 1, APPEND);
|
|
clen = 1;
|
|
}
|
|
}
|
|
/* Swap the translation space and the pattern space. */
|
|
tmp = PS;
|
|
PS = YS;
|
|
YS = tmp;
|
|
YS.space = YS.back;
|
|
}
|
|
}
|
|
|
|
/*
|
|
* Flush append requests. Always called before reading a line,
|
|
* therefore it also resets the substitution done (sdone) flag.
|
|
*/
|
|
static void
|
|
flush_appends(void)
|
|
{
|
|
FILE *f;
|
|
int count, i;
|
|
char buf[8 * 1024];
|
|
|
|
for (i = 0; i < appendx; i++)
|
|
switch (appends[i].type) {
|
|
case AP_STRING:
|
|
fwrite(appends[i].s, sizeof(char), appends[i].len,
|
|
outfile);
|
|
break;
|
|
case AP_FILE:
|
|
/*
|
|
* Read files probably shouldn't be cached. Since
|
|
* it's not an error to read a non-existent file,
|
|
* it's possible that another program is interacting
|
|
* with the sed script through the filesystem. It
|
|
* would be truly bizarre, but possible. It's probably
|
|
* not that big a performance win, anyhow.
|
|
*/
|
|
if ((f = fopen(appends[i].s, "r")) == NULL)
|
|
break;
|
|
while ((count = fread(buf, sizeof(char), sizeof(buf), f)))
|
|
(void)fwrite(buf, sizeof(char), count, outfile);
|
|
(void)fclose(f);
|
|
break;
|
|
}
|
|
if (ferror(outfile))
|
|
errx(1, "%s: %s", outfname, strerror(errno ? errno : EIO));
|
|
appendx = sdone = 0;
|
|
}
|
|
|
|
static void
|
|
lputs(char *s, size_t len)
|
|
{
|
|
static const char escapes[] = "\\\a\b\f\r\t\v";
|
|
int c, col, width;
|
|
char *p;
|
|
struct winsize win;
|
|
static int termwidth = -1;
|
|
size_t clen, i;
|
|
wchar_t wc;
|
|
mbstate_t mbs;
|
|
|
|
if (outfile != stdout)
|
|
termwidth = 60;
|
|
if (termwidth == -1) {
|
|
if ((p = getenv("COLUMNS")) && *p != '\0')
|
|
termwidth = atoi(p);
|
|
else if (ioctl(STDOUT_FILENO, TIOCGWINSZ, &win) == 0 &&
|
|
win.ws_col > 0)
|
|
termwidth = win.ws_col;
|
|
else
|
|
termwidth = 60;
|
|
}
|
|
|
|
memset(&mbs, 0, sizeof(mbs));
|
|
col = 0;
|
|
while (len != 0) {
|
|
clen = mbrtowc(&wc, s, len, &mbs);
|
|
if (clen == 0)
|
|
clen = 1;
|
|
if (clen == (size_t)-1 || clen == (size_t)-2) {
|
|
wc = (unsigned char)*s;
|
|
clen = 1;
|
|
memset(&mbs, 0, sizeof(mbs));
|
|
}
|
|
if (wc == '\n') {
|
|
if (col + 1 >= termwidth)
|
|
fprintf(outfile, "\\\n");
|
|
fputc('$', outfile);
|
|
fputc('\n', outfile);
|
|
col = 0;
|
|
} else if (iswprint(wc)) {
|
|
width = wcwidth(wc);
|
|
if (col + width >= termwidth) {
|
|
fprintf(outfile, "\\\n");
|
|
col = 0;
|
|
}
|
|
fwrite(s, 1, clen, outfile);
|
|
col += width;
|
|
} else if (wc != L'\0' && (c = wctob(wc)) != EOF &&
|
|
(p = strchr(escapes, c)) != NULL) {
|
|
if (col + 2 >= termwidth) {
|
|
fprintf(outfile, "\\\n");
|
|
col = 0;
|
|
}
|
|
fprintf(outfile, "\\%c", "\\abfrtv"[p - escapes]);
|
|
col += 2;
|
|
} else {
|
|
if (col + 4 * clen >= termwidth) {
|
|
fprintf(outfile, "\\\n");
|
|
col = 0;
|
|
}
|
|
for (i = 0; i < clen; i++)
|
|
fprintf(outfile, "\\%03o",
|
|
(int)(unsigned char)s[i]);
|
|
col += 4 * clen;
|
|
}
|
|
s += clen;
|
|
len -= clen;
|
|
}
|
|
if (col + 1 >= termwidth)
|
|
fprintf(outfile, "\\\n");
|
|
(void)fputc('$', outfile);
|
|
(void)fputc('\n', outfile);
|
|
if (ferror(outfile))
|
|
errx(1, "%s: %s", outfname, strerror(errno ? errno : EIO));
|
|
}
|
|
|
|
static __inline int
|
|
regexec_e(regex_t *preg, const char *string, int eflags, int nomatch,
|
|
size_t slen)
|
|
{
|
|
int eval;
|
|
|
|
if (preg == NULL) {
|
|
if (defpreg == NULL)
|
|
errx(1, "first RE may not be empty");
|
|
} else
|
|
defpreg = preg;
|
|
|
|
/* Set anchors */
|
|
match[0].rm_so = 0;
|
|
match[0].rm_eo = slen;
|
|
|
|
eval = regexec(defpreg, string,
|
|
nomatch ? 0 : maxnsub + 1, match, eflags | REG_STARTEND);
|
|
switch(eval) {
|
|
case 0:
|
|
return (1);
|
|
case REG_NOMATCH:
|
|
return (0);
|
|
}
|
|
errx(1, "RE error: %s", strregerror(eval, defpreg));
|
|
/* NOTREACHED */
|
|
}
|
|
|
|
/*
|
|
* regsub - perform substitutions after a regexp match
|
|
* Based on a routine by Henry Spencer
|
|
*/
|
|
static void
|
|
regsub(SPACE *sp, char *string, char *src)
|
|
{
|
|
int len, no;
|
|
char c, *dst;
|
|
|
|
#define NEEDSP(reqlen) \
|
|
/* XXX What is the +1 for? */ \
|
|
if (sp->len + (reqlen) + 1 >= sp->blen) { \
|
|
sp->blen += (reqlen) + 1024; \
|
|
if ((sp->space = sp->back = realloc(sp->back, sp->blen)) \
|
|
== NULL) \
|
|
err(1, "realloc"); \
|
|
dst = sp->space + sp->len; \
|
|
}
|
|
|
|
dst = sp->space + sp->len;
|
|
while ((c = *src++) != '\0') {
|
|
if (c == '&')
|
|
no = 0;
|
|
else if (c == '\\' && isdigit((unsigned char)*src))
|
|
no = *src++ - '0';
|
|
else
|
|
no = -1;
|
|
if (no < 0) { /* Ordinary character. */
|
|
if (c == '\\' && (*src == '\\' || *src == '&'))
|
|
c = *src++;
|
|
NEEDSP(1);
|
|
*dst++ = c;
|
|
++sp->len;
|
|
} else if (match[no].rm_so != -1 && match[no].rm_eo != -1) {
|
|
len = match[no].rm_eo - match[no].rm_so;
|
|
NEEDSP(len);
|
|
memmove(dst, string + match[no].rm_so, len);
|
|
dst += len;
|
|
sp->len += len;
|
|
}
|
|
}
|
|
NEEDSP(1);
|
|
*dst = '\0';
|
|
}
|
|
|
|
/*
|
|
* aspace --
|
|
* Append the source space to the destination space, allocating new
|
|
* space as necessary.
|
|
*/
|
|
void
|
|
cspace(SPACE *sp, const char *p, size_t len, enum e_spflag spflag)
|
|
{
|
|
size_t tlen;
|
|
|
|
/* Make sure SPACE has enough memory and ramp up quickly. */
|
|
tlen = sp->len + len + 1;
|
|
if (tlen > sp->blen) {
|
|
sp->blen = tlen + 1024;
|
|
if ((sp->space = sp->back = realloc(sp->back, sp->blen)) ==
|
|
NULL)
|
|
err(1, "realloc");
|
|
}
|
|
|
|
if (spflag == REPLACE)
|
|
sp->len = 0;
|
|
|
|
memmove(sp->space + sp->len, p, len);
|
|
|
|
sp->space[sp->len += len] = '\0';
|
|
}
|
|
|
|
/*
|
|
* Close all cached opened files and report any errors
|
|
*/
|
|
void
|
|
cfclose(struct s_command *cp, struct s_command *end)
|
|
{
|
|
|
|
for (; cp != end; cp = cp->next)
|
|
switch(cp->code) {
|
|
case 's':
|
|
if (cp->u.s->wfd != -1 && close(cp->u.s->wfd))
|
|
err(1, "%s", cp->u.s->wfile);
|
|
cp->u.s->wfd = -1;
|
|
break;
|
|
case 'w':
|
|
if (cp->u.fd != -1 && close(cp->u.fd))
|
|
err(1, "%s", cp->t);
|
|
cp->u.fd = -1;
|
|
break;
|
|
case '{':
|
|
cfclose(cp->u.c, cp->next);
|
|
break;
|
|
}
|
|
}
|