sbase/sort.c

308 lines
5.7 KiB
C
Raw Normal View History

2011-06-02 12:03:34 +00:00
/* See LICENSE file for copyright and license details. */
#include <ctype.h>
2011-06-02 12:03:34 +00:00
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <unistd.h>
2011-06-02 12:03:34 +00:00
#include "text.h"
#include "util.h"
struct keydef {
int start_column;
int end_column;
int start_char;
int end_char;
int flags;
};
enum {
MOD_N = 1 << 1,
MOD_STARTB = 1 << 2,
MOD_ENDB = 1 << 3,
MOD_R = 1 << 4,
};
struct kdlist {
struct keydef keydef;
struct kdlist *next;
};
static struct kdlist *head = NULL;
static struct kdlist *tail = NULL;
static void addkeydef(char *, int);
2014-11-23 19:35:56 +00:00
static void check(FILE *);
2011-06-02 12:03:34 +00:00
static int linecmp(const char **, const char **);
2014-05-15 18:08:17 +00:00
static char *skipblank(char *);
static int parse_flags(char **, int *, int);
static int parse_keydef(struct keydef *, char *, int);
2014-05-15 18:08:17 +00:00
static char *nextcol(char *);
static char *columns(char *, const struct keydef *);
2011-06-02 12:03:34 +00:00
2014-11-23 19:35:56 +00:00
static int Cflag = 0, cflag = 0, uflag = 0;
2014-05-15 18:08:17 +00:00
static char *fieldsep = NULL;
2013-06-14 18:20:47 +00:00
static void
usage(void)
{
2014-11-23 19:35:56 +00:00
enprintf(2, "usage: %s [-Cbcnru] [-t delim] [-k def]... [file...]\n", argv0);
2013-06-14 18:20:47 +00:00
}
2011-06-02 12:03:34 +00:00
int
main(int argc, char *argv[])
{
long i;
FILE *fp;
2014-05-03 16:28:20 +00:00
struct linebuf linebuf = EMPTY_LINEBUF;
int global_flags = 0;
2011-06-02 12:03:34 +00:00
2013-06-14 18:20:47 +00:00
ARGBEGIN {
2014-11-23 19:35:56 +00:00
case 'C':
Cflag = 1;
break;
2014-05-15 18:08:17 +00:00
case 'b':
global_flags |= MOD_STARTB | MOD_ENDB;
break;
2014-11-23 19:35:56 +00:00
case 'c':
cflag = 1;
break;
2014-05-15 18:08:17 +00:00
case 'k':
addkeydef(EARGF(usage()), global_flags);
break;
2013-12-12 13:08:49 +00:00
case 'n':
global_flags |= MOD_N;
2013-12-12 13:08:49 +00:00
break;
2013-06-14 18:20:47 +00:00
case 'r':
global_flags |= MOD_R;
2013-06-14 18:20:47 +00:00
break;
2014-05-15 18:08:17 +00:00
case 't':
fieldsep = EARGF(usage());
if (strlen(fieldsep) != 1)
2014-05-15 18:08:17 +00:00
usage();
break;
2013-06-14 18:20:47 +00:00
case 'u':
uflag = 1;
2013-06-14 18:20:47 +00:00
break;
default:
usage();
} ARGEND;
if (!head && global_flags)
addkeydef("1", global_flags);
addkeydef("1", global_flags & MOD_R);
if (argc == 0) {
2014-11-23 19:35:56 +00:00
if (Cflag || cflag) {
check(stdin);
} else {
getlines(stdin, &linebuf);
}
} else for (; argc > 0; argc--, argv++) {
if (!(fp = fopen(argv[0], "r"))) {
enprintf(2, "fopen %s:", argv[0]);
continue;
}
2014-11-23 19:35:56 +00:00
if (Cflag || cflag) {
check(fp);
} else {
getlines(fp, &linebuf);
}
2011-06-02 12:03:34 +00:00
fclose(fp);
}
2014-11-23 19:35:56 +00:00
if (!Cflag && !cflag) {
qsort(linebuf.lines, linebuf.nlines, sizeof *linebuf.lines,
(int (*)(const void *, const void *))linecmp);
for (i = 0; i < linebuf.nlines; i++) {
if (!uflag || i == 0 || linecmp((const char **)&linebuf.lines[i],
(const char **)&linebuf.lines[i-1])) {
fputs(linebuf.lines[i], stdout);
}
2013-06-14 18:20:47 +00:00
}
}
2014-10-02 22:46:04 +00:00
return 0;
2011-06-02 12:03:34 +00:00
}
static void
addkeydef(char *def, int flags)
{
struct kdlist *node;
node = malloc(sizeof(*node));
if (!node)
enprintf(2, "malloc:");
if (!head)
head = node;
if (parse_keydef(&node->keydef, def, flags))
enprintf(2, "faulty key definition\n");
if (tail)
tail->next = node;
node->next = NULL;
tail = node;
}
2014-11-23 19:35:56 +00:00
static void
check(FILE *fp)
{
static struct { char *buf; size_t size; } prev, cur, tmp;
if (!prev.buf)
getline(&prev.buf, &prev.size, fp);
while (getline(&cur.buf, &cur.size, fp) != -1) {
if (uflag > linecmp((const char **) &cur.buf, (const char **) &prev.buf)) {
if (!Cflag)
weprintf("disorder: %s", cur.buf);
exit(1);
}
tmp = cur;
cur = prev;
prev = tmp;
}
}
static int
2011-06-02 12:03:34 +00:00
linecmp(const char **a, const char **b)
{
char *s1, *s2;
int res = 0;
struct kdlist *node;
for (node = head; node && res == 0; node = node->next) {
s1 = columns((char *)*a, &node->keydef);
s2 = columns((char *)*b, &node->keydef);
/* if -u is given, don't use default key definition
* unless it is the only one */
if (uflag && node == tail && head != tail)
res = 0;
else if (node->keydef.flags & MOD_N)
res = strtol(s1, 0, 10) - strtol(s2, 0, 10);
2013-12-12 13:08:49 +00:00
else
res = strcmp(s1, s2);
if (node->keydef.flags & MOD_R)
res = -res;
free(s1);
free(s2);
}
return res;
}
static int
parse_flags(char **s, int *flags, int bflag)
{
while (isalpha((int)**s))
switch (*((*s)++)) {
case 'b':
*flags |= bflag;
break;
case 'n':
*flags |= MOD_N;
break;
case 'r':
*flags |= MOD_R;
break;
default:
return -1;
}
return 0;
}
static int
parse_keydef(struct keydef *kd, char *s, int flags)
{
char *rest = s;
kd->start_column = 1;
kd->start_char = 1;
/* 0 means end of line */
kd->end_column = 0;
kd->end_char = 0;
kd->flags = flags;
kd->start_column = strtol(rest, &rest, 10);
if (kd->start_column < 1)
return -1;
if (*rest == '.')
kd->start_char = strtol(rest+1, &rest, 10);
if (kd->start_char < 1)
return -1;
2014-11-19 19:59:37 +00:00
if (parse_flags(&rest, &kd->flags, MOD_STARTB) < 0)
return -1;
if (*rest == ',') {
kd->end_column = strtol(rest+1, &rest, 10);
if (kd->end_column && kd->end_column < kd->start_column)
return -1;
if (*rest == '.') {
kd->end_char = strtol(rest+1, &rest, 10);
if (kd->end_char < 1)
return -1;
}
2014-11-19 19:59:37 +00:00
if (parse_flags(&rest, &kd->flags, MOD_ENDB) < 0)
return -1;
2013-12-12 13:08:49 +00:00
}
if (*rest != '\0')
return -1;
return 0;
2011-06-02 12:03:34 +00:00
}
2013-06-14 18:20:47 +00:00
static char *
2014-05-15 18:08:17 +00:00
skipblank(char *s)
{
while(*s && isblank(*s))
s++;
return s;
}
static char *
2014-05-15 18:08:17 +00:00
nextcol(char *s)
{
if (!fieldsep) {
2014-05-15 18:08:17 +00:00
s = skipblank(s);
while(*s && !isblank(*s))
s++;
} else {
if (!strchr(s, *fieldsep))
2014-05-15 18:08:17 +00:00
s = strchr(s, '\0');
else
s = strchr(s, *fieldsep) + 1;
}
return s;
}
static char *
columns(char *line, const struct keydef *kd)
{
char *start, *end;
char *res;
2014-05-15 18:08:17 +00:00
int i;
2014-04-18 16:21:31 +00:00
for (i = 1, start = line; i < kd->start_column; i++)
2014-05-15 18:08:17 +00:00
start = nextcol(start);
if (kd->flags & MOD_STARTB)
2014-05-15 18:08:17 +00:00
start = skipblank(start);
start += MIN(kd->start_char, nextcol(start) - start) - 1;
if (kd->end_column) {
for (i = 1, end = line; i < kd->end_column; i++)
2014-05-15 18:08:17 +00:00
end = nextcol(end);
if (kd->flags & MOD_ENDB)
2014-05-15 18:08:17 +00:00
end = skipblank(end);
if (kd->end_char)
2014-05-15 18:08:17 +00:00
end += MIN(kd->end_char, nextcol(end) - end);
2014-04-18 16:21:31 +00:00
else
2014-05-15 18:08:17 +00:00
end = nextcol(end);
} else {
if (!(end = strchr(line, '\n')))
end = strchr(line, '\0');
}
if (!(res = strndup(start, end - start)))
enprintf(2, "strndup:");
return res;
}