3 * $Id: anag.c,v 1.7 2003/11/29 23:38:37 mdw Exp $
7 * (c) 2001 Mark Wooding
10 /*----- Licensing notice --------------------------------------------------*
12 * This file is part of Anag: a simple wordgame helper.
14 * Anag is free software; you can redistribute it and/or modify
15 * it under the terms of the GNU General Public License as published by
16 * the Free Software Foundation; either version 2 of the License, or
17 * (at your option) any later version.
19 * Anag is distributed in the hope that it will be useful,
20 * but WITHOUT ANY WARRANTY; without even the implied warranty of
21 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
22 * GNU General Public License for more details.
24 * You should have received a copy of the GNU General Public License
25 * along with Anag; if not, write to the Free Software Foundation,
26 * Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
29 /*----- Revision history --------------------------------------------------*
32 * Revision 1.7 2003/11/29 23:38:37 mdw
35 * Revision 1.6 2003/09/15 02:48:54 mdw
36 * Monoalphabetic match filter.
38 * Revision 1.5 2002/08/11 12:58:09 mdw
39 * Added support for regular expression matching, if supported by the C
42 * Revision 1.4 2001/02/19 19:18:50 mdw
43 * Minor big fixes to parser.
45 * Revision 1.3 2001/02/16 21:45:19 mdw
46 * Be more helpful. Improve full help message. Special-case error for
47 * empty command strings.
49 * Revision 1.2 2001/02/07 09:09:11 mdw
50 * Fix spurious error when `-file' is used.
52 * Revision 1.1 2001/02/04 17:14:42 mdw
57 /*----- Header files ------------------------------------------------------*/
61 /*----- Static variables --------------------------------------------------*/
63 static const char *file = DICTIONARY;
65 /*----- Help text functions -----------------------------------------------*/
67 static void usage(FILE *fp)
69 pquis(fp, "Usage: $ [-f file] expression\n");
72 static void version(FILE *fp)
74 pquis(fp, "$, version " VERSION "\n");
77 static void help(FILE *fp)
83 Searches a wordlist, printing all of the words which match an expression.\n\
85 Options supported are:\n\
87 -h, --help display this help text\n\
88 -v, --version display the program's version number\n\
89 -u, --usage display a very brief usage message\n\
90 -f, --file FILE read wordlist from FILE, not `" DICTIONARY "'\n\
92 The basic tests in the expression are:\n\
94 -anagram WORD matches a full-length anagram\n\
95 -subgram WORD matches words which only use letters in WORD\n\
96 -wildcard PATTERN matches with wildcards `*' and `?'\n\
97 -trackword WORD matches words which can be found in a trackword\n\
98 -mono PATTERN matches words isomorphic to the given PATTERN\n\
102 -regexp REGEXP matches with an (extended) regular expression\n\
107 -pcre REGEXP matches with a Perl-like regular expression\n\
112 These simple tests can be combined using the operators `-a', `-o' and `-n'\n\
113 (for `and', `or' and `not'; they may also be written `&', `|' and `!' if\n\
114 you like), and grouped using parentheses `(' and `)'.\n\
118 /*----- The options parser ------------------------------------------------*/
120 /* --- Options table structure --- */
130 O_HELP, O_VERSION, O_USAGE,
132 O_AND, O_OR, O_NOT, O_LPAREN, O_RPAREN,
133 O_ANAG, O_SUBG, O_WILD, O_TRACK, O_REGEXP, O_PCRE, O_MONO,
139 static const struct opt opttab[] = {
141 /* --- Options -- don't form part of the language --- */
143 { "help", 0, OF_SHORT, O_HELP },
144 { "version", 0, OF_SHORT, O_VERSION },
145 { "usage", 0, OF_SHORT, O_USAGE },
146 { "file", 1, OF_SHORT, O_FILE },
148 /* --- Operators -- provide the basic structure of the language --- *
150 * These are also given magical names by the parser.
153 { "and", 0, OF_SHORT, O_AND },
154 { "or", 0, OF_SHORT, O_OR },
155 { "not", 0, OF_SHORT, O_NOT },
157 /* --- Actual matching oeprations -- do something useful --- */
159 { "anagram", 1, 0, O_ANAG },
160 { "subgram", 1, 0, O_SUBG },
161 { "wildcard", 1, 0, O_WILD },
162 { "trackword", 1, 0, O_TRACK },
163 { "mono", 1, 0, O_MONO },
165 { "regexp", 1, 0, O_REGEXP },
168 { "pcre", 1, 0, O_PCRE },
171 /* --- End marker --- */
177 static const char *const *av;
180 /* --- @nextopt@ --- *
182 * Arguments: @const char ***arg@ = where to store the arg pointer
184 * Returns: The tag of the next option.
186 * Use: Scans the next option off the command line. If the option
187 * doesn't form part of the language, it's processed internally,
188 * and you'll never see it from here. On exit, the @arg@
189 * pointer is set to contain the address of the option scanned,
190 * followed by its arguments if any. You're expected to know
191 * how many arguments there are for your option.
194 static unsigned nextopt(const char *const **arg)
197 const struct opt *o, *oo;
201 /* --- Pick the next option off the front --- */
208 /* --- Cope with various forms of magic --- */
211 if (!p[1]) switch (*p) {
212 case '&': return (O_AND);
213 case '|': return (O_OR);
214 case '!': return (O_NOT);
215 case '(': return (O_LPAREN);
216 case ')': return (O_RPAREN);
221 /* --- Now cope with other sorts of weirdies --- *
223 * By the end of this, a leading `-' or `--' will have been stripped.
233 die("syntax error near `--': rubbish at end of line");
237 /* --- Now look the word up in my table --- */
241 for (o = opttab; o->name; o++) {
242 if (strncmp(p, o->name, sz) == 0) {
243 if (strlen(o->name) == sz || ((o->f & OF_SHORT) && sz == 1)) {
248 die("ambiguous option name `-%s' (could match `-%s' or `-%s')",
249 p, oo->name, o->name);
255 die("unrecognized option name `-%s'", p);
257 /* --- Sort out the arguments --- */
259 if (ai + oo->nargs > ac)
260 die("too few arguments for `-%s' (need %u)", oo->name, oo->nargs);
263 /* --- Now process the option --- */
283 die("syntax error near `%s': unknown token type", av[ai - 1]);
287 /*----- Node types for operators ------------------------------------------*/
289 /* --- Node structures --- */
291 typedef struct node_bin {
297 typedef struct node_un {
302 /* --- Node functions --- */
304 static int n_or(node *nn, const char *p, size_t sz)
306 node_bin *n = (node_bin *)nn;
307 return (n->left->func(n->left, p, sz) || n->right->func(n->right, p, sz));
310 static int n_and(node *nn, const char *p, size_t sz)
312 node_bin *n = (node_bin *)nn;
313 return (n->left->func(n->left, p, sz) && n->right->func(n->right, p, sz));
316 static int n_not(node *nn, const char *p, size_t sz)
318 node_un *n = (node_un *)nn;
319 return (!n->arg->func(n->arg, p, sz));
322 /*----- Parser for the expression syntax ----------------------------------*/
324 /* --- A parser context --- */
326 typedef struct p_ctx {
328 const char *const *a;
331 /* --- Parser structure --- *
333 * This is a simple recursive descent parser. The context retains
334 * information about the current token. Each function is passed the address
335 * of a node pointer to fill in. This simplifies the binary operator code
336 * somewhat, relative to returning pointers to node trees.
339 static void p_expr(p_ctx *p, node **/*nn*/);
341 static void p_next(p_ctx *p)
343 static const char *const eof[] = { "<end>", 0 };
344 p->t = nextopt(&p->a);
349 static void p_factor(p_ctx *p, node **nn)
352 if (p->t == O_LPAREN) {
355 if (p->t != O_RPAREN)
356 die("syntax error near `%s': missing `)'", *p->a);
358 } else if (p->t == O_NOT) {
359 n = xmalloc(sizeof(node_un));
363 p_factor(p, &n->arg);
366 case O_ANAG: *nn = anagram(p->a + 1); break;
367 case O_SUBG: *nn = subgram(p->a + 1); break;
368 case O_WILD: *nn = wildcard(p->a + 1); break;
369 case O_TRACK: *nn = trackword(p->a + 1); break;
371 case O_REGEXP: *nn = regexp(p->a + 1); break;
374 case O_PCRE: *nn = pcrenode(p->a + 1); break;
376 case O_MONO: *nn = mono(p->a + 1); break;
377 default: die("syntax error near `%s': unexpected token", *p->a);
383 static void p_term(p_ctx *p, node **nn)
398 n = xmalloc(sizeof(node_bin));
406 static void p_expr(p_ctx *p, node **nn)
414 n = xmalloc(sizeof(node_bin));
422 /* --- @p_argv@ --- *
424 * Arguments: @int argc@ = number of command-line arguments
425 * @const char *const argv[]@ = vectoor of arguments
427 * Returns: A compiled node, parsed from the arguments.
429 * Use: Does the donkey-work of parsing a command-line.
432 static node *p_argv(int argc, const char *const argv[])
443 pquis(stderr, "(Run `$ --help' for more detail.)\n");
448 die("syntax error near `%s': rubbish at end of line (too many `)'s?)",
454 /*----- Main code ---------------------------------------------------------*/
458 * Arguments: @int argc@ = number of command-line arguments
459 * @char *argv[]@ = vector of argument words
461 * Returns: Zero on success, nonzero on failure.
463 * Use: Picks entries from a word list which match particular
464 * expressions. This might be of assistance to word-game types.
467 int main(int argc, char *argv[])
475 n = p_argv(argc, (const char *const *)argv);
477 if ((fp = fopen(file, "r")) == 0)
478 die("error opening `%s': %s", file, strerror(errno));
481 if (dstr_putline(&d, fp) < 0)
484 for (p = q = d.buf; p < l; p++) {
485 if (!isalnum((unsigned char)*p))
487 *q++ = tolower((unsigned char)*p);
491 if (n->func(n, d.buf, d.len)) {
492 fwrite(d.buf, 1, d.len, stdout);
497 die("error reading `%s': %s", file, strerror(errno));
502 /*----- That's all, folks -------------------------------------------------*/