-/* $NetBSD: strfile.c,v 1.4 1995/04/24 12:23:09 cgd Exp $ */
+/* $NetBSD: strfile.c,v 1.21 2001/07/22 13:34:00 wiz Exp $ */
/*-
* Copyright (c) 1989, 1993
* SUCH DAMAGE.
*/
+#ifdef __NetBSD__
+#include <sys/cdefs.h>
#ifndef lint
-static char copyright[] =
-"@(#) Copyright (c) 1989, 1993\n\
- The Regents of the University of California. All rights reserved.\n";
+__COPYRIGHT("@(#) Copyright (c) 1989, 1993\n\
+ The Regents of the University of California. All rights reserved.\n");
#endif /* not lint */
#ifndef lint
#if 0
static char sccsid[] = "@(#)strfile.c 8.1 (Berkeley) 5/31/93";
#else
-static char rcsid[] = "$NetBSD: strfile.c,v 1.4 1995/04/24 12:23:09 cgd Exp $";
+__RCSID("$NetBSD: strfile.c,v 1.21 2001/07/22 13:34:00 wiz Exp $");
#endif
#endif /* not lint */
+#endif /* __NetBSD__ */
-# include <machine/endian.h>
+# include <sys/types.h>
# include <sys/param.h>
+# include <ctype.h>
# include <stdio.h>
+# include <stdlib.h>
# include <string.h>
-# include <ctype.h>
+# include <time.h>
+# include <unistd.h>
+
+# ifndef u_int32_t
+# define u_int32_t unsigned int
+# endif
# include "strfile.h"
# ifndef MAXPATHLEN
# define MAXPATHLEN 1024
# endif /* MAXPATHLEN */
+u_int32_t
+h2nl(u_int32_t h)
+{
+ unsigned char c[4];
+ u_int32_t rv;
+
+ c[0] = (h >> 24) & 0xff;
+ c[1] = (h >> 16) & 0xff;
+ c[2] = (h >> 8) & 0xff;
+ c[3] = (h >> 0) & 0xff;
+ memcpy(&rv, c, sizeof rv);
+
+ return (rv);
+}
+
/*
- * This program takes a file composed of strings seperated by
+ * This program takes a file composed of strings separated by
* lines starting with two consecutive delimiting character (default
* character is '%') and creates another file which consists of a table
* describing the file (structure from "strfile.h"), a table of seek
# define STORING_PTRS (Oflag || Rflag)
# define CHUNKSIZE 512
-#ifdef lint
-# define ALWAYS atoi("1")
-#else
-# define ALWAYS 1
-#endif
-# define ALLOC(ptr,sz) if (ALWAYS) { \
+# define ALLOC(ptr,sz) do { \
if (ptr == NULL) \
- ptr = malloc((unsigned int) (CHUNKSIZE * sizeof *ptr)); \
+ ptr = malloc(CHUNKSIZE * sizeof *ptr); \
else if (((sz) + 1) % CHUNKSIZE == 0) \
- ptr = realloc((void *) ptr, ((unsigned int) ((sz) + CHUNKSIZE) * sizeof *ptr)); \
- if (ptr == NULL) { \
- fprintf(stderr, "out of space\n"); \
- exit(1); \
- } \
- } else
-
-#ifdef NO_VOID
-# define void char
-#endif
+ ptr = realloc(ptr, ((sz) + CHUNKSIZE) * sizeof *ptr); \
+ if (ptr == NULL) \
+ die("out of space"); \
+ } while (0)
typedef struct {
char first;
STR *Firstch; /* first chars of each string */
-char *fgets(), *strcpy(), *strcat();
+#ifdef __GNUC__
+#define NORETURN __attribute__((__noreturn__))
+#else
+#define NORETURN
+#endif
+
+void add_offset(FILE *, off_t);
+int cmp_str(const void *, const void *);
+void die(const char *) NORETURN;
+void dieperror(const char *, char *) NORETURN;
+void do_order(void);
+void fwrite_be_offt(off_t, FILE *);
+void getargs(int, char *[]);
+int main(int, char *[]);
+void randomize(void);
+void usage(void) NORETURN;
-void *malloc(), *realloc();
/*
* main:
* CHUNKSIZE blocks; if the latter, we just write each pointer,
* and then seek back to the beginning to write in the table.
*/
+int
main(ac, av)
-int ac;
-char **av;
+ int ac;
+ char *av[];
{
- register char *sp, dc;
- register FILE *inf, *outf;
- register off_t last_off, length, pos, *p;
- register int first, cnt;
- register char *nsp;
- register STR *fp;
- static char string[257];
+ char *sp, dc;
+ FILE *inf, *outf;
+ off_t last_off, length, pos, *p;
+ int first, cnt;
+ char *nsp;
+ STR *fp;
+ static char string[257];
+
+ /* sanity test */
+ if (sizeof(u_int32_t) != 4)
+ die("sizeof(unsigned int) != 4");
getargs(ac, av); /* evalute arguments */
dc = Delimch;
- if ((inf = fopen(Infile, "r")) == NULL) {
- perror(Infile);
- exit(1);
- }
+ if ((inf = fopen(Infile, "r")) == NULL)
+ dieperror("open `%s'", Infile);
- if ((outf = fopen(Outfile, "w")) == NULL) {
- perror(Outfile);
- exit(1);
- }
+ if ((outf = fopen(Outfile, "w")) == NULL)
+ dieperror("open `%s'", Outfile);
if (!STORING_PTRS)
- (void) fseek(outf, sizeof Tbl, 0);
+ (void) fseek(outf, sizeof Tbl, SEEK_SET);
/*
* Write the strings onto the file
*/
Tbl.str_longlen = 0;
- Tbl.str_shortlen = (unsigned int) 0xffffffff;
+ Tbl.str_shortlen = (unsigned int) 0x7fffffff;
Tbl.str_delim = dc;
Tbl.str_version = VERSION;
first = Oflag;
last_off = 0;
do {
sp = fgets(string, 256, inf);
- if (sp == NULL || sp[0] == dc && sp[1] == '\n') {
+ if (sp == NULL || (sp[0] == dc && sp[1] == '\n')) {
pos = ftell(inf);
length = pos - last_off - (sp ? strlen(sp) : 0);
last_off = pos;
if (!length)
continue;
add_offset(outf, pos);
- if (Tbl.str_longlen < length)
+ if ((off_t)Tbl.str_longlen < length)
Tbl.str_longlen = length;
- if (Tbl.str_shortlen > length)
+ if ((off_t)Tbl.str_shortlen > length)
Tbl.str_shortlen = length;
first = Oflag;
}
if (Num_pts == 2)
puts("There was 1 string");
else
- printf("There were %d strings\n", Num_pts - 1);
- printf("Longest string: %lu byte%s\n", Tbl.str_longlen,
+ printf("There were %d strings\n", (int)(Num_pts - 1));
+ printf("Longest string: %lu byte%s\n", (unsigned long)Tbl.str_longlen,
Tbl.str_longlen == 1 ? "" : "s");
- printf("Shortest string: %lu byte%s\n", Tbl.str_shortlen,
+ printf("Shortest string: %lu byte%s\n", (unsigned long)Tbl.str_shortlen,
Tbl.str_shortlen == 1 ? "" : "s");
}
- (void) fseek(outf, (off_t) 0, 0);
- Tbl.str_version = htonl(Tbl.str_version);
- Tbl.str_numstr = htonl(Num_pts - 1);
- Tbl.str_longlen = htonl(Tbl.str_longlen);
- Tbl.str_shortlen = htonl(Tbl.str_shortlen);
- Tbl.str_flags = htonl(Tbl.str_flags);
+ (void) fseek(outf, (off_t) 0, SEEK_SET);
+ Tbl.str_version = h2nl(Tbl.str_version);
+ Tbl.str_numstr = h2nl(Num_pts - 1);
+ Tbl.str_longlen = h2nl(Tbl.str_longlen);
+ Tbl.str_shortlen = h2nl(Tbl.str_shortlen);
+ Tbl.str_flags = h2nl(Tbl.str_flags);
(void) fwrite((char *) &Tbl, sizeof Tbl, 1, outf);
if (STORING_PTRS) {
for (p = Seekpts, cnt = Num_pts; cnt--; ++p)
- *p = htonl(*p);
- (void) fwrite((char *) Seekpts, sizeof *Seekpts, (int) Num_pts, outf);
+ fwrite_be_offt(*p, outf);
}
+ fflush(outf);
+ if (ferror(outf))
+ dieperror("fwrite %s", Outfile);
(void) fclose(outf);
exit(0);
}
/*
* This routine evaluates arguments from the command line
*/
+void
getargs(argc, argv)
-int argc;
-char **argv;
+ int argc;
+ char **argv;
{
- extern char *optarg;
- extern int optind;
int ch;
+ extern int optind;
+ extern char *optarg;
- while ((ch = getopt(argc, argv, "c:iorsx")) != EOF)
+ while ((ch = getopt(argc, argv, "c:iorsx")) != -1)
switch(ch) {
case 'c': /* new delimiting char */
Delimch = *optarg;
}
}
+void
usage()
{
(void) fprintf(stderr,
exit(1);
}
+void
+die(str)
+ const char *str;
+{
+ fprintf(stderr, "strfile: %s\n", str);
+ exit(1);
+}
+
+void
+dieperror(fmt, file)
+ const char *fmt;
+ char *file;
+{
+ fprintf(stderr, "strfile: ");
+ fprintf(stderr, fmt, file);
+ fprintf(stderr, ": ");
+ perror(NULL);
+ exit(1);
+}
+
/*
* add_offset:
* Add an offset to the list, or write it out, as appropriate.
*/
+void
add_offset(fp, off)
-FILE *fp;
-off_t off;
+ FILE *fp;
+ off_t off;
{
- off_t net;
if (!STORING_PTRS) {
- net = htonl(off);
- fwrite(&net, 1, sizeof net, fp);
+ fwrite_be_offt(off, fp);
} else {
ALLOC(Seekpts, Num_pts + 1);
Seekpts[Num_pts] = off;
* do_order:
* Order the strings alphabetically (possibly ignoring case).
*/
+void
do_order()
{
- register int i;
- register off_t *lp;
- register STR *fp;
- extern int cmp_str();
+ int i;
+ off_t *lp;
+ STR *fp;
Sort_1 = fopen(Infile, "r");
Sort_2 = fopen(Infile, "r");
Tbl.str_flags |= STR_ORDERED;
}
-/*
- * cmp_str:
- * Compare two strings in the file
- */
-char *
-unctrl(c)
-char c;
+int
+cmp_str(vp1, vp2)
+ const void *vp1, *vp2;
{
- static char buf[3];
-
- if (isprint(c)) {
- buf[0] = c;
- buf[1] = '\0';
- }
- else if (c == 0177) {
- buf[0] = '^';
- buf[1] = '?';
- }
- else {
- buf[0] = '^';
- buf[1] = c + 'A' - 1;
- }
- return buf;
-}
+ const STR *p1, *p2;
+ int c1, c2;
+ int n1, n2;
-cmp_str(p1, p2)
-STR *p1, *p2;
-{
- register int c1, c2;
- register int n1, n2;
+ p1 = (const STR *)vp1;
+ p2 = (const STR *)vp2;
# define SET_N(nf,ch) (nf = (ch == '\n'))
# define IS_END(ch,nf) (ch == Delimch && nf)
if (c1 != c2)
return c1 - c2;
- (void) fseek(Sort_1, p1->pos, 0);
- (void) fseek(Sort_2, p2->pos, 0);
+ (void) fseek(Sort_1, p1->pos, SEEK_SET);
+ (void) fseek(Sort_2, p2->pos, SEEK_SET);
n1 = FALSE;
n2 = FALSE;
* not to randomize across delimiter boundaries. All
* randomization is done within each block.
*/
+void
randomize()
{
- register int cnt, i;
- register off_t tmp;
- register off_t *sp;
- extern time_t time();
+ int cnt, i;
+ off_t tmp;
+ off_t *sp;
srandom((int)(time((time_t *) NULL) + getpid()));
sp[i] = tmp;
}
}
+
+/*
+ * fwrite_be_offt:
+ * Write out the off paramater as a 64 bit big endian number
+ */
+
+void
+fwrite_be_offt(off, f)
+ off_t off;
+ FILE *f;
+{
+ int i;
+ unsigned char c[8];
+
+ for (i = 7; i >= 0; i--) {
+ c[i] = off & 0xff;
+ off >>= 8;
+ }
+ fwrite(c, sizeof(c), 1, f);
+}