Home | History | Annotate | Line # | Download | only in mkdict
mkdict.c revision 1.7 
      1 /* $NetBSD: mkdict.c,v 1.7 2000/07/31 11:28:48 simonb 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 const char copyright[] =
     41     "@(#) Copyright (c) 1993\n\
     42 	The Regents of the University of California.  All rights reserved.\n";
     43 #if 0
     44 static char sccsid[] = "@(#)mkdict.c	8.1 (Berkeley) 6/11/93";
     45 #else
     46 static const char rcsid[] =
     47     "$NetBSD: mkdict.c,v 1.7 2000/07/31 11:28:48 simonb Exp $";
     48 #endif
     49 #endif /* not lint */
     50 
     51 /*
     52  * Filter out words that:
     53  *	1) Are not completely made up of lower case letters
     54  *	2) Contain a 'q' not immediately followed by a 'u'
     55  *	3) Are less that 3 characters long
     56  *	4) Are greater than MAXWORDLEN characters long
     57  */
     58 
     59 #include <ctype.h>
     60 #include <stdio.h>
     61 #include <stdlib.h>
     62 #include <string.h>
     63 
     64 #include "bog.h"
     65 
     66 int main(int, char *[]);
     67 
     68 int
     69 main(argc, argv)
     70 	int argc;
     71 	char *argv[];
     72 {
     73 	char *p, *q;
     74 	int ch, common, 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 
     82 	for (nwords = 1;
     83 	    fgets(buf[current], MAXWORDLEN + 1, stdin) != NULL; ++nwords) {
     84 		if ((p = strchr(buf[current], '\n')) == NULL) {
     85 			fprintf(stderr, "word too long: %s", buf[current]);
     86 			while ((ch = getc(stdin)) != EOF && ch != '\n')
     87 				;
     88 			if (ch == EOF)
     89 				break;
     90 			continue;
     91 		}
     92 		len = 0;
     93 		for (p = buf[current]; *p != '\n'; p++) {
     94 			if (!islower(*p))
     95 				break;
     96 			if (*p == 'q') {
     97 				q = p + 1;
     98 				if (*q != 'u')
     99 					break;
    100 				else {
    101 					while ((*q = *(q + 1)))
    102 						q++;
    103 				}
    104 				len++;
    105 			}
    106 			len++;
    107 		}
    108 		if (*p != '\n' || len < 3 || len > MAXWORDLEN)
    109 			continue;
    110 		if (argc == 2 && nwords % atoi(argv[1]))
    111 			continue;
    112 
    113 		*p = '\0';
    114 		p = buf[current];
    115 		q = buf[prev];
    116 		qcount = 0;
    117 		while ((ch = *p++) == *q++ && ch != '\0')
    118 			if (ch == 'q')
    119 				qcount++;
    120 		common = p - buf[current] - 1;
    121 		printf("%c%s", common + qcount, p - 1);
    122 		prev = !prev;
    123 		current = !current;
    124 	}
    125 	fprintf(stderr, "%d words\n", nwords);
    126 	fflush(stdout);
    127 	if (ferror(stdout)) {
    128 		perror("error writing standard output");
    129 		exit(1);
    130 	}
    131 	exit(0);
    132 }
    133