6e403221 |
1 | /* -*-c-*- |
2 | * |
d9af4a2b |
3 | * $Id: anag.c,v 1.7 2003/11/29 23:38:37 mdw Exp $ |
6e403221 |
4 | * |
5 | * Main driver for anag |
6 | * |
7 | * (c) 2001 Mark Wooding |
8 | */ |
9 | |
10 | /*----- Licensing notice --------------------------------------------------* |
11 | * |
12 | * This file is part of Anag: a simple wordgame helper. |
13 | * |
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. |
18 | * |
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. |
23 | * |
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. |
27 | */ |
28 | |
29 | /*----- Revision history --------------------------------------------------* |
30 | * |
31 | * $Log: anag.c,v $ |
d9af4a2b |
32 | * Revision 1.7 2003/11/29 23:38:37 mdw |
33 | * Debianization. |
34 | * |
fe9969ff |
35 | * Revision 1.6 2003/09/15 02:48:54 mdw |
36 | * Monoalphabetic match filter. |
37 | * |
a10122de |
38 | * Revision 1.5 2002/08/11 12:58:09 mdw |
39 | * Added support for regular expression matching, if supported by the C |
40 | * library. |
41 | * |
1d2d1062 |
42 | * Revision 1.4 2001/02/19 19:18:50 mdw |
43 | * Minor big fixes to parser. |
44 | * |
2668675c |
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. |
48 | * |
60dffc01 |
49 | * Revision 1.2 2001/02/07 09:09:11 mdw |
50 | * Fix spurious error when `-file' is used. |
51 | * |
6e403221 |
52 | * Revision 1.1 2001/02/04 17:14:42 mdw |
53 | * Initial checkin |
54 | * |
55 | */ |
56 | |
57 | /*----- Header files ------------------------------------------------------*/ |
58 | |
59 | #include "anag.h" |
60 | |
61 | /*----- Static variables --------------------------------------------------*/ |
62 | |
63 | static const char *file = DICTIONARY; |
64 | |
65 | /*----- Help text functions -----------------------------------------------*/ |
66 | |
67 | static void usage(FILE *fp) |
68 | { |
69 | pquis(fp, "Usage: $ [-f file] expression\n"); |
70 | } |
71 | |
72 | static void version(FILE *fp) |
73 | { |
74 | pquis(fp, "$, version " VERSION "\n"); |
75 | } |
76 | |
77 | static void help(FILE *fp) |
78 | { |
79 | version(fp); |
80 | fputc('\n', fp); |
81 | usage(fp); |
82 | fputs("\n\ |
83 | Searches a wordlist, printing all of the words which match an expression.\n\ |
2668675c |
84 | \n\ |
85 | Options supported are:\n\ |
86 | \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\ |
91 | \n\ |
6e403221 |
92 | The basic tests in the expression are:\n\ |
93 | \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\ |
fe9969ff |
98 | -mono PATTERN matches words isomorphic to the given PATTERN\n\ |
a10122de |
99 | " |
100 | #ifdef HAVE_REGCOMP |
101 | "\ |
102 | -regexp REGEXP matches with an (extended) regular expression\n\ |
103 | " |
104 | #endif |
d9af4a2b |
105 | #ifdef HAVE_PCRE |
106 | "\ |
107 | -pcre REGEXP matches with a Perl-like regular expression\n\ |
108 | " |
109 | #endif |
a10122de |
110 | "\ |
6e403221 |
111 | \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\ |
d9af4a2b |
115 | ", fp); /*"*/ |
6e403221 |
116 | } |
117 | |
118 | /*----- The options parser ------------------------------------------------*/ |
119 | |
120 | /* --- Options table structure --- */ |
121 | |
122 | struct opt { |
123 | const char *name; |
124 | unsigned nargs; |
125 | unsigned f; |
126 | unsigned tag; |
127 | }; |
128 | |
129 | enum { |
130 | O_HELP, O_VERSION, O_USAGE, |
131 | O_FILE, |
132 | O_AND, O_OR, O_NOT, O_LPAREN, O_RPAREN, |
d9af4a2b |
133 | O_ANAG, O_SUBG, O_WILD, O_TRACK, O_REGEXP, O_PCRE, O_MONO, |
6e403221 |
134 | O_EOF |
135 | }; |
136 | |
137 | #define OF_SHORT 1u |
138 | |
139 | static const struct opt opttab[] = { |
140 | |
141 | /* --- Options -- don't form part of the language --- */ |
142 | |
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 }, |
147 | |
148 | /* --- Operators -- provide the basic structure of the language --- * |
149 | * |
150 | * These are also given magical names by the parser. |
151 | */ |
152 | |
153 | { "and", 0, OF_SHORT, O_AND }, |
154 | { "or", 0, OF_SHORT, O_OR }, |
155 | { "not", 0, OF_SHORT, O_NOT }, |
156 | |
157 | /* --- Actual matching oeprations -- do something useful --- */ |
158 | |
159 | { "anagram", 1, 0, O_ANAG }, |
160 | { "subgram", 1, 0, O_SUBG }, |
161 | { "wildcard", 1, 0, O_WILD }, |
162 | { "trackword", 1, 0, O_TRACK }, |
fe9969ff |
163 | { "mono", 1, 0, O_MONO }, |
a10122de |
164 | #ifdef HAVE_REGCOMP |
165 | { "regexp", 1, 0, O_REGEXP }, |
166 | #endif |
d9af4a2b |
167 | #ifdef HAVE_PCRE |
168 | { "pcre", 1, 0, O_PCRE }, |
169 | #endif |
6e403221 |
170 | |
171 | /* --- End marker --- */ |
172 | |
173 | { 0, 0, 0, 0 } |
174 | }; |
175 | |
176 | static int ac; |
177 | static const char *const *av; |
178 | static int ai; |
179 | |
180 | /* --- @nextopt@ --- * |
181 | * |
182 | * Arguments: @const char ***arg@ = where to store the arg pointer |
183 | * |
184 | * Returns: The tag of the next option. |
185 | * |
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. |
192 | */ |
193 | |
194 | static unsigned nextopt(const char *const **arg) |
195 | { |
196 | for (;;) { |
197 | const struct opt *o, *oo; |
198 | size_t sz; |
199 | const char *p; |
200 | |
201 | /* --- Pick the next option off the front --- */ |
202 | |
203 | *arg = av + ai; |
204 | if (ai >= ac) |
205 | return (O_EOF); |
206 | p = av[ai++]; |
207 | |
208 | /* --- Cope with various forms of magic --- */ |
209 | |
210 | if (p[0] != '-') { |
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); |
217 | } |
218 | goto bad; |
219 | } |
220 | |
221 | /* --- Now cope with other sorts of weirdies --- * |
222 | * |
223 | * By the end of this, a leading `-' or `--' will have been stripped. |
224 | */ |
225 | |
226 | p++; |
227 | if (!*p) |
228 | goto bad; |
229 | if (*p == '-') |
230 | p++; |
231 | if (!*p) { |
232 | if (ai < ac) |
233 | die("syntax error near `--': rubbish at end of line"); |
234 | return (O_EOF); |
235 | } |
236 | |
237 | /* --- Now look the word up in my table --- */ |
238 | |
239 | sz = strlen(p); |
240 | oo = 0; |
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)) { |
244 | oo = o; |
245 | break; |
246 | } |
247 | if (oo) { |
248 | die("ambiguous option name `-%s' (could match `-%s' or `-%s')", |
249 | p, oo->name, o->name); |
250 | } |
251 | oo = o; |
252 | } |
253 | } |
254 | if (!oo) |
255 | die("unrecognized option name `-%s'", p); |
256 | |
257 | /* --- Sort out the arguments --- */ |
258 | |
259 | if (ai + oo->nargs > ac) |
260 | die("too few arguments for `-%s' (need %u)", oo->name, oo->nargs); |
261 | ai += oo->nargs; |
262 | |
263 | /* --- Now process the option --- */ |
264 | |
265 | switch (oo->tag) { |
266 | case O_HELP: |
267 | help(stdout); |
268 | exit(0); |
269 | case O_VERSION: |
270 | version(stdout); |
271 | exit(0); |
272 | case O_USAGE: |
273 | usage(stdout); |
274 | exit(0); |
275 | case O_FILE: |
276 | file = (*arg)[1]; |
277 | break; |
278 | default: |
279 | return (oo->tag); |
280 | } |
60dffc01 |
281 | continue; |
6e403221 |
282 | bad: |
283 | die("syntax error near `%s': unknown token type", av[ai - 1]); |
284 | } |
285 | } |
286 | |
287 | /*----- Node types for operators ------------------------------------------*/ |
288 | |
289 | /* --- Node structures --- */ |
290 | |
291 | typedef struct node_bin { |
292 | node n; |
293 | node *left; |
294 | node *right; |
295 | } node_bin; |
296 | |
297 | typedef struct node_un { |
298 | node n; |
299 | node *arg; |
300 | } node_un; |
301 | |
302 | /* --- Node functions --- */ |
303 | |
304 | static int n_or(node *nn, const char *p, size_t sz) |
305 | { |
306 | node_bin *n = (node_bin *)nn; |
307 | return (n->left->func(n->left, p, sz) || n->right->func(n->right, p, sz)); |
308 | } |
309 | |
310 | static int n_and(node *nn, const char *p, size_t sz) |
311 | { |
312 | node_bin *n = (node_bin *)nn; |
313 | return (n->left->func(n->left, p, sz) && n->right->func(n->right, p, sz)); |
314 | } |
315 | |
316 | static int n_not(node *nn, const char *p, size_t sz) |
317 | { |
318 | node_un *n = (node_un *)nn; |
319 | return (!n->arg->func(n->arg, p, sz)); |
320 | } |
321 | |
322 | /*----- Parser for the expression syntax ----------------------------------*/ |
323 | |
324 | /* --- A parser context --- */ |
325 | |
326 | typedef struct p_ctx { |
327 | unsigned t; |
328 | const char *const *a; |
329 | } p_ctx; |
330 | |
331 | /* --- Parser structure --- * |
332 | * |
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. |
337 | */ |
338 | |
339 | static void p_expr(p_ctx *p, node **/*nn*/); |
340 | |
341 | static void p_next(p_ctx *p) |
342 | { |
343 | static const char *const eof[] = { "<end>", 0 }; |
344 | p->t = nextopt(&p->a); |
345 | if (p->t == O_EOF) |
346 | p->a = eof; |
347 | } |
348 | |
349 | static void p_factor(p_ctx *p, node **nn) |
350 | { |
351 | node_un *n; |
352 | if (p->t == O_LPAREN) { |
353 | p_next(p); |
354 | p_expr(p, nn); |
355 | if (p->t != O_RPAREN) |
1d2d1062 |
356 | die("syntax error near `%s': missing `)'", *p->a); |
6e403221 |
357 | p_next(p); |
358 | } else if (p->t == O_NOT) { |
359 | n = xmalloc(sizeof(node_un)); |
360 | n->n.func = n_not; |
361 | *nn = &n->n; |
362 | p_next(p); |
363 | p_factor(p, &n->arg); |
364 | } else { |
365 | switch (p->t) { |
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; |
a10122de |
370 | #ifdef HAVE_REGCOMP |
371 | case O_REGEXP: *nn = regexp(p->a + 1); break; |
372 | #endif |
d9af4a2b |
373 | #ifdef HAVE_PCRE |
374 | case O_PCRE: *nn = pcrenode(p->a + 1); break; |
375 | #endif |
fe9969ff |
376 | case O_MONO: *nn = mono(p->a + 1); break; |
6e403221 |
377 | default: die("syntax error near `%s': unexpected token", *p->a); |
378 | } |
379 | p_next(p); |
380 | } |
381 | } |
382 | |
383 | static void p_term(p_ctx *p, node **nn) |
384 | { |
385 | node_bin *n; |
386 | for (;;) { |
387 | p_factor(p, nn); |
388 | switch (p->t) { |
389 | case O_AND: |
390 | p_next(p); |
391 | default: |
392 | break; |
6e403221 |
393 | case O_RPAREN: |
394 | case O_OR: |
395 | case O_EOF: |
396 | return; |
397 | } |
398 | n = xmalloc(sizeof(node_bin)); |
399 | n->left = *nn; |
400 | n->n.func = n_and; |
401 | *nn = &n->n; |
402 | nn = &n->right; |
403 | } |
404 | } |
405 | |
406 | static void p_expr(p_ctx *p, node **nn) |
407 | { |
408 | node_bin *n; |
409 | for (;;) { |
410 | p_term(p, nn); |
411 | if (p->t != O_OR) |
412 | break; |
413 | p_next(p); |
414 | n = xmalloc(sizeof(node_bin)); |
415 | n->left = *nn; |
416 | n->n.func = n_or; |
417 | *nn = &n->n; |
418 | nn = &n->right; |
419 | } |
420 | } |
421 | |
422 | /* --- @p_argv@ --- * |
423 | * |
424 | * Arguments: @int argc@ = number of command-line arguments |
425 | * @const char *const argv[]@ = vectoor of arguments |
426 | * |
427 | * Returns: A compiled node, parsed from the arguments. |
428 | * |
429 | * Use: Does the donkey-work of parsing a command-line. |
430 | */ |
431 | |
432 | static node *p_argv(int argc, const char *const argv[]) |
433 | { |
434 | p_ctx p; |
435 | node *n; |
436 | |
437 | av = argv; |
438 | ac = argc; |
439 | ai = 1; |
440 | p_next(&p); |
2668675c |
441 | if (p.t == O_EOF) { |
442 | usage(stderr); |
443 | pquis(stderr, "(Run `$ --help' for more detail.)\n"); |
444 | exit(EXIT_FAILURE); |
445 | } |
6e403221 |
446 | p_expr(&p, &n); |
447 | if (p.t != O_EOF) { |
448 | die("syntax error near `%s': rubbish at end of line (too many `)'s?)", |
449 | *p.a); |
450 | } |
451 | return (n); |
452 | } |
453 | |
454 | /*----- Main code ---------------------------------------------------------*/ |
455 | |
456 | /* --- @main@ --- * |
457 | * |
458 | * Arguments: @int argc@ = number of command-line arguments |
459 | * @char *argv[]@ = vector of argument words |
460 | * |
461 | * Returns: Zero on success, nonzero on failure. |
462 | * |
463 | * Use: Picks entries from a word list which match particular |
464 | * expressions. This might be of assistance to word-game types. |
465 | */ |
466 | |
467 | int main(int argc, char *argv[]) |
468 | { |
469 | node *n; |
470 | FILE *fp; |
471 | dstr d = DSTR_INIT; |
472 | char *p, *q, *l; |
473 | |
474 | ego(argv[0]); |
475 | n = p_argv(argc, (const char *const *)argv); |
476 | |
477 | if ((fp = fopen(file, "r")) == 0) |
478 | die("error opening `%s': %s", file, strerror(errno)); |
479 | for (;;) { |
480 | dstr_reset(&d); |
481 | if (dstr_putline(&d, fp) < 0) |
482 | break; |
483 | l = d.buf + d.len; |
484 | for (p = q = d.buf; p < l; p++) { |
485 | if (!isalnum((unsigned char)*p)) |
486 | continue; |
487 | *q++ = tolower((unsigned char)*p); |
488 | } |
489 | *q = 0; |
490 | d.len = q - d.buf; |
491 | if (n->func(n, d.buf, d.len)) { |
492 | fwrite(d.buf, 1, d.len, stdout); |
493 | fputc('\n', stdout); |
494 | } |
495 | } |
496 | if (!feof(fp)) |
497 | die("error reading `%s': %s", file, strerror(errno)); |
498 | fclose(fp); |
499 | return (0); |
500 | } |
501 | |
502 | /*----- That's all, folks -------------------------------------------------*/ |