sbase/cut.c

216 lines
4.0 KiB
C
Raw Normal View History

2013-10-08 19:39:08 +00:00
/* See LICENSE file for copyright and license details. */
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
2016-03-07 11:34:41 +00:00
#include "text.h"
#include "utf.h"
2013-10-08 19:39:08 +00:00
#include "util.h"
typedef struct Range {
size_t min, max;
struct Range *next;
} Range;
static Range *list = NULL;
static char mode = 0;
static char *delim = "\t";
static size_t delimlen = 1;
static int nflag = 0;
static int sflag = 0;
2013-10-08 19:39:08 +00:00
static void
insert(Range *r)
{
Range *l, *p, *t;
for (p = NULL, l = list; l; p = l, l = l->next) {
if (r->max && r->max + 1 < l->min) {
2013-10-08 19:39:08 +00:00
r->next = l;
break;
} else if (!l->max || r->min < l->max + 2) {
2013-10-08 19:39:08 +00:00
l->min = MIN(r->min, l->min);
for (p = l, t = l->next; t; p = t, t = t->next)
if (r->max && r->max + 1 < t->min)
break;
2013-10-08 19:39:08 +00:00
l->max = (p->max && r->max) ? MAX(p->max, r->max) : 0;
l->next = t;
return;
}
}
if (p)
p->next = r;
else
list = r;
2013-10-08 19:39:08 +00:00
}
static void
parselist(char *str)
{
char *s;
size_t n = 1;
Range *r;
if (!*str)
eprintf("empty list\n");
for (s = str; *s; s++) {
if (*s == ' ')
*s = ',';
if (*s == ',')
n++;
2013-10-08 19:39:08 +00:00
}
r = ereallocarray(NULL, n, sizeof(*r));
for (s = str; n; n--, s++) {
r->min = (*s == '-') ? 1 : strtoul(s, &s, 10);
r->max = (*s == '-') ? strtoul(s + 1, &s, 10) : r->min;
2013-10-08 19:39:08 +00:00
r->next = NULL;
if (!r->min || (r->max && r->max < r->min) || (*s && *s != ','))
eprintf("bad list value\n");
2013-10-08 19:39:08 +00:00
insert(r++);
}
}
static size_t
2016-03-07 11:34:41 +00:00
seek(struct line *s, size_t pos, size_t *prev, size_t count)
2013-10-08 19:39:08 +00:00
{
2016-03-07 11:34:41 +00:00
size_t n = pos - *prev, i, j;
2013-10-08 19:39:08 +00:00
if (mode == 'b') {
2016-03-07 11:34:41 +00:00
if (n >= s->len)
return s->len;
if (nflag)
2016-03-07 11:34:41 +00:00
while (n && !UTF8_POINT(s->data[n]))
n--;
2013-10-08 19:39:08 +00:00
*prev += n;
return n;
} else if (mode == 'c') {
2016-03-07 11:34:41 +00:00
for (n++, i = 0; i < s->len; i++)
if (UTF8_POINT(s->data[i]) && !--n)
break;
2013-10-08 19:39:08 +00:00
} else {
2016-03-07 11:34:41 +00:00
for (i = (count < delimlen + 1) ? 0 : delimlen; n && i < s->len; ) {
if ((s->len - i) >= delimlen &&
!memcmp(s->data + i, delim, delimlen)) {
if (!--n && count)
break;
2016-03-07 11:34:41 +00:00
i += delimlen;
continue;
}
2016-03-07 11:34:41 +00:00
for (j = 1; j + i <= s->len && !fullrune(s->data + i, j); j++);
i += j;
}
2013-10-08 19:39:08 +00:00
}
*prev = pos;
2016-03-07 11:34:41 +00:00
return i;
2013-10-08 19:39:08 +00:00
}
static void
2015-08-04 10:44:33 +00:00
cut(FILE *fp, const char *fname)
2013-10-08 19:39:08 +00:00
{
2016-03-07 11:34:41 +00:00
Range *r;
struct line s;
static struct line line;
static size_t size;
size_t i, n, p;
ssize_t len;
2013-10-08 19:39:08 +00:00
2016-03-07 11:34:41 +00:00
while ((len = getline(&line.data, &size, fp)) > 0) {
line.len = len;
if (line.data[line.len - 1] == '\n')
line.data[--line.len] = '\0';
if (mode == 'f' && !memmem(line.data, line.len, delim, delimlen)) {
if (!sflag) {
fwrite(line.data, 1, line.len, stdout);
fputc('\n', stdout);
}
2013-10-08 19:39:08 +00:00
continue;
}
2016-03-07 11:34:41 +00:00
for (i = 0, p = 1, s = line, r = list; r; r = r->next) {
n = seek(&s, r->min, &p, i);
s.data += n;
s.len -= n;
i += (mode == 'f') ? delimlen : 1;
2016-03-07 11:34:41 +00:00
if (!s.len)
break;
if (!r->max) {
2016-03-07 11:34:41 +00:00
fwrite(s.data, 1, s.len, stdout);
2013-10-08 19:39:08 +00:00
break;
}
2016-03-07 11:34:41 +00:00
n = seek(&s, r->max + 1, &p, i);
i += (mode == 'f') ? delimlen : 1;
2016-03-07 11:34:41 +00:00
if (fwrite(s.data, 1, n, stdout) != n)
eprintf("fwrite <stdout>:");
2016-03-07 11:34:41 +00:00
s.data += n;
s.len -= n;
2013-10-08 19:39:08 +00:00
}
putchar('\n');
}
if (ferror(fp))
eprintf("getline %s:", fname);
2013-10-08 19:39:08 +00:00
}
static void
usage(void)
{
eprintf("usage: %s -b list [-n] [file ...]\n"
" %s -c list [file ...]\n"
" %s -f list [-d delim] [-s] [file ...]\n",
argv0, argv0, argv0);
}
2013-10-08 19:39:08 +00:00
int
main(int argc, char *argv[])
{
FILE *fp;
int ret = 0;
2013-10-08 19:39:08 +00:00
ARGBEGIN {
case 'b':
case 'c':
case 'f':
mode = ARGC();
parselist(EARGF(usage()));
2013-10-08 19:39:08 +00:00
break;
case 'd':
delim = EARGF(usage());
if (!*delim)
eprintf("empty delimiter\n");
delimlen = unescape(delim);
2013-10-08 19:39:08 +00:00
break;
case 'n':
nflag = 1;
2013-10-08 19:39:08 +00:00
break;
case 's':
sflag = 1;
2013-10-08 19:39:08 +00:00
break;
default:
usage();
} ARGEND
2013-10-08 19:39:08 +00:00
if (!mode)
2013-10-08 19:39:08 +00:00
usage();
if (!argc)
cut(stdin, "<stdin>");
else {
for (; *argv; argc--, argv++) {
if (!strcmp(*argv, "-")) {
*argv = "<stdin>";
fp = stdin;
} else if (!(fp = fopen(*argv, "r"))) {
weprintf("fopen %s:", *argv);
ret = 1;
continue;
}
cut(fp, *argv);
if (fp != stdin && fshut(fp, *argv))
ret = 1;
}
2013-10-08 19:39:08 +00:00
}
ret |= fshut(stdin, "<stdin>") | fshut(stdout, "<stdout>");
return ret;
2013-10-08 19:39:08 +00:00
}