Home | History | Annotate | Line # | Download | only in mkdict
mkdict.c revision 1.1 
      1 /*-
      2  * Copyright (c) 1993
      3  *	The Regents of the University of California.  All rights reserved.
      4  *
      5  * This code is derived from software contributed to Berkeley by
      6  * Barry Brachman.
      7  *
      8  * Redistribution and use in source and binary forms, with or without
      9  * modification, are permitted provided that the following conditions
     10  * are met:
     11  * 1. Redistributions of source code must retain the above copyright
     12  *    notice, this list of conditions and the following disclaimer.
     13  * 2. Redistributions in binary form must reproduce the above copyright
     14  *    notice, this list of conditions and the following disclaimer in the
     15  *    documentation and/or other materials provided with the distribution.
     16  * 3. All advertising materials mentioning features or use of this software
     17  *    must display the following acknowledgement:
     18  *	This product includes software developed by the University of
     19  *	California, Berkeley and its contributors.
     20  * 4. Neither the name of the University nor the names of its contributors
     21  *    may be used to endorse or promote products derived from this software
     22  *    without specific prior written permission.
     23  *
     24  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
     25  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
     26  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
     27  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
     28  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
     29  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
     30  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
     31  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
     32  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
     33  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
     34  * SUCH DAMAGE.
     35  */
     36 
     37 #ifndef lint
     38 static char copyright[] =
     39 "@(#) Copyright (c) 1993\n\
     40 	The Regents of the University of California.  All rights reserved.\n";
     41 #endif /* not lint */
     42 
     43 #ifndef lint
     44 static char sccsid[] = "@(#)mkdict.c	8.1 (Berkeley) 6/11/93";
     45 #endif /* not lint */
     46 
     47 /*
     48  * Filter out words that:
     49  *	1) Are not completely made up of lower case letters
     50  *	2) Contain a 'q' not immediately followed by a 'u'
     51  *	3) Are less that 3 characters long
     52  *	4) Are greater than MAXWORDLEN characters long
     53  */
     54 
     55 #include <ctype.h>
     56 #include <stdio.h>
     57 #include <stdlib.h>
     58 #include <string.h>
     59 
     60 #include "bog.h"
     61 
     62 int
     63 main(argc, argv)
     64 	int argc;
     65 	char *argv[];
     66 {
     67 	register char *p, *q;
     68 	register int ch, common, n, nwords;
     69 	int current, len, prev, qcount;
     70 	char buf[2][MAXWORDLEN + 1];
     71 
     72 	prev = 0;
     73 	current = 1;
     74 	buf[prev][0] = '\0';
     75 	if (argc == 2)
     76 		n = atoi(argv[1]);
     77 
     78 	for (nwords = 1;
     79 	    fgets(buf[current], MAXWORDLEN + 1, stdin) != NULL; ++nwords) {
     80 		if ((p = index(buf[current], '\n')) == NULL) {
     81 			fprintf(stderr,
     82 			    "mkdict: word too long: %s\n", buf[current]);
     83 			while ((ch = getc(stdin)) != EOF && ch != '\n')
     84 				;
     85 			if (ch == EOF)
     86 				break;
     87 			continue;
     88 		}
     89 		len = 0;
     90 		for (p = buf[current]; *p != '\n'; p++) {
     91 			if (!islower(*p))
     92 				break;
     93 			if (*p == 'q') {
     94 				q = p + 1;
     95 				if (*q != 'u')
     96 					break;
     97 				else {
     98 					while (*q = *(q + 1))
     99 						q++;
    100 				}
    101 				len++;
    102 			}
    103 			len++;
    104 		}
    105 		if (*p != '\n' || len < 3 || len > MAXWORDLEN)
    106 			continue;
    107 		if (argc == 2 && nwords % n)
    108 			continue;
    109 
    110 		*p = '\0';
    111 		p = buf[current];
    112 		q = buf[prev];
    113 		qcount = 0;
    114 		while ((ch = *p++) == *q++ && ch != '\0')
    115 			if (ch == 'q')
    116 				qcount++;
    117 		common = p - buf[current] - 1;
    118 		printf("%c%s", common + qcount, p - 1);
    119 		prev = !prev;
    120 		current = !current;
    121 	}
    122 	fprintf(stderr, "%d words\n", nwords);
    123 	exit(0);
    124 }
    125