Home | History | Annotate | Line # | Download | only in mkindex
mkindex.c revision 1.6 
      1 /*	$NetBSD: mkindex.c,v 1.6 1999/09/18 20:50:29 jsm Exp $	*/
      2 
      3 /*-
      4  * Copyright (c) 1993
      5  *	The Regents of the University of California.  All rights reserved.
      6  *
      7  * This code is derived from software contributed to Berkeley by
      8  * Barry Brachman.
      9  *
     10  * Redistribution and use in source and binary forms, with or without
     11  * modification, are permitted provided that the following conditions
     12  * are met:
     13  * 1. Redistributions of source code must retain the above copyright
     14  *    notice, this list of conditions and the following disclaimer.
     15  * 2. Redistributions in binary form must reproduce the above copyright
     16  *    notice, this list of conditions and the following disclaimer in the
     17  *    documentation and/or other materials provided with the distribution.
     18  * 3. All advertising materials mentioning features or use of this software
     19  *    must display the following acknowledgement:
     20  *	This product includes software developed by the University of
     21  *	California, Berkeley and its contributors.
     22  * 4. Neither the name of the University nor the names of its contributors
     23  *    may be used to endorse or promote products derived from this software
     24  *    without specific prior written permission.
     25  *
     26  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
     27  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
     28  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
     29  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
     30  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
     31  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
     32  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
     33  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
     34  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
     35  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
     36  * SUCH DAMAGE.
     37  */
     38 
     39 #include <sys/cdefs.h>
     40 #ifndef lint
     41 __COPYRIGHT("@(#) Copyright (c) 1993\n\
     42 	The Regents of the University of California.  All rights reserved.\n");
     43 #endif /* not lint */
     44 
     45 #ifndef lint
     46 #if 0
     47 static char sccsid[] = "@(#)mkindex.c	8.1 (Berkeley) 6/11/93";
     48 #else
     49 __RCSID("$NetBSD: mkindex.c,v 1.6 1999/09/18 20:50:29 jsm Exp $");
     50 #endif
     51 #endif /* not lint */
     52 
     53 #include <err.h>
     54 #include <stdio.h>
     55 #include <stdlib.h>
     56 
     57 #include "bog.h"
     58 
     59 int main __P((void));
     60 char *nextword __P((FILE *, char *, int *, int *));
     61 
     62 int
     63 main(void)
     64 {
     65 	int clen, rlen, prev;
     66 	long off, start;
     67 	char buf[MAXWORDLEN + 1];
     68 
     69 	prev = '\0';
     70 	off = start = 0L;
     71 	while (nextword(stdin, buf, &clen, &rlen) != NULL) {
     72 		if (*buf != prev) {
     73 			if (prev != '\0')
     74 				printf("%c %6ld %6ld\n", prev, start, off - 1);
     75 			prev = *buf;
     76 			start = off;
     77 		}
     78 		off += clen + 1;
     79 	}
     80 	printf("%c %6ld %6ld\n", prev, start, off - 1);
     81 	fflush(stdout);
     82 	if (ferror(stdout))
     83 		err(1, "writing standard output");
     84 	exit(0);
     85 }
     86 
     87 /*
     88  * Return the next word in the compressed dictionary in 'buffer' or
     89  * NULL on end-of-file
     90  * Also set clen to the length of the compressed word (for mkindex) and
     91  * rlen to the strlen() of the real word
     92  */
     93 char *
     94 nextword(fp, buffer, clen, rlen)
     95 	FILE *fp;
     96 	char *buffer;
     97 	int *clen, *rlen;
     98 {
     99 	int ch, pcount;
    100 	char *p, *q;
    101 	static char buf[MAXWORDLEN + 1];
    102 	static int first = 1;
    103 	static int lastch = 0;
    104 
    105    	if (first) {
    106 		if ((pcount = getc(fp)) == EOF)
    107 			return (NULL);
    108 		first = 0;
    109 	}
    110 	else if ((pcount = lastch) == EOF)
    111 		return (NULL);
    112 
    113 	p = buf + (*clen = pcount);
    114 
    115 	while ((ch = getc(fp)) != EOF && ch >= 'a')
    116 			*p++ = ch;
    117 		lastch = ch;
    118 	*p = '\0';
    119 
    120 	*rlen = (int) (p - buf);
    121 	*clen = *rlen - *clen;
    122 
    123 	p = buf;
    124 	q = buffer;
    125 	while ((*q++ = *p) != '\0') {
    126 		if (*p++ == 'q')
    127 			*q++ = 'u';
    128 	}
    129 	return (buffer);
    130 }
    131