Home | History | Annotate | Line # | Download | only in mkdict
mkdict.c revision 1.2 
      1 /*	$NetBSD: mkdict.c,v 1.2 1995/03/21 12:14:49 cgd 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 #ifndef lint
     40 static char copyright[] =
     41 "@(#) 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[] = "@(#)mkdict.c	8.1 (Berkeley) 6/11/93";
     48 #else
     49 static char rcsid[] = "$NetBSD: mkdict.c,v 1.2 1995/03/21 12:14:49 cgd Exp $";
     50 #endif
     51 #endif /* not lint */
     52 
     53 /*
     54  * Filter out words that:
     55  *	1) Are not completely made up of lower case letters
     56  *	2) Contain a 'q' not immediately followed by a 'u'
     57  *	3) Are less that 3 characters long
     58  *	4) Are greater than MAXWORDLEN characters long
     59  */
     60 
     61 #include <ctype.h>
     62 #include <stdio.h>
     63 #include <stdlib.h>
     64 #include <string.h>
     65 
     66 #include "bog.h"
     67 
     68 int
     69 main(argc, argv)
     70 	int argc;
     71 	char *argv[];
     72 {
     73 	register char *p, *q;
     74 	register int ch, common, n, nwords;
     75 	int current, len, prev, qcount;
     76 	char buf[2][MAXWORDLEN + 1];
     77 
     78 	prev = 0;
     79 	current = 1;
     80 	buf[prev][0] = '\0';
     81 	if (argc == 2)
     82 		n = atoi(argv[1]);
     83 
     84 	for (nwords = 1;
     85 	    fgets(buf[current], MAXWORDLEN + 1, stdin) != NULL; ++nwords) {
     86 		if ((p = index(buf[current], '\n')) == NULL) {
     87 			fprintf(stderr,
     88 			    "mkdict: word too long: %s\n", buf[current]);
     89 			while ((ch = getc(stdin)) != EOF && ch != '\n')
     90 				;
     91 			if (ch == EOF)
     92 				break;
     93 			continue;
     94 		}
     95 		len = 0;
     96 		for (p = buf[current]; *p != '\n'; p++) {
     97 			if (!islower(*p))
     98 				break;
     99 			if (*p == 'q') {
    100 				q = p + 1;
    101 				if (*q != 'u')
    102 					break;
    103 				else {
    104 					while (*q = *(q + 1))
    105 						q++;
    106 				}
    107 				len++;
    108 			}
    109 			len++;
    110 		}
    111 		if (*p != '\n' || len < 3 || len > MAXWORDLEN)
    112 			continue;
    113 		if (argc == 2 && nwords % n)
    114 			continue;
    115 
    116 		*p = '\0';
    117 		p = buf[current];
    118 		q = buf[prev];
    119 		qcount = 0;
    120 		while ((ch = *p++) == *q++ && ch != '\0')
    121 			if (ch == 'q')
    122 				qcount++;
    123 		common = p - buf[current] - 1;
    124 		printf("%c%s", common + qcount, p - 1);
    125 		prev = !prev;
    126 		current = !current;
    127 	}
    128 	fprintf(stderr, "%d words\n", nwords);
    129 	exit(0);
    130 }
    131