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 /*----- Header files ------------------------------------------------------*/
33 /*----- Static variables --------------------------------------------------*/
35 static const char *file
= DICTIONARY
;
37 /*----- Help text functions -----------------------------------------------*/
39 static void usage(FILE *fp
)
41 pquis(fp
, "Usage: $ [-f file] expression\n");
44 static void version(FILE *fp
)
46 pquis(fp
, "$, version " VERSION
"\n");
49 static void help(FILE *fp
)
55 Searches a wordlist, printing all of the words which match an expression.\n\
57 Options supported are:\n\
59 -h, --help display this help text\n\
60 -v, --version display the program's version number\n\
61 -u, --usage display a very brief usage message\n\
62 -f, --file FILE read wordlist from FILE, not `" DICTIONARY
"'\n\
64 The basic tests in the expression are:\n\
66 -anagram WORD matches a full-length anagram\n\
67 -subgram WORD matches words which only use letters in WORD\n\
68 -wildcard PATTERN matches with wildcards `*' and `?'\n\
69 -trackword WORD matches words which can be found in a trackword\n\
70 -mono PATTERN matches words isomorphic to the given PATTERN\n\
74 -regexp REGEXP matches with an (extended) regular expression\n\
79 -pcre REGEXP matches with a Perl-like regular expression\n\
83 -length [+|-]N matches if length is [at least|at most] N\n\
84 -longest output longest matches found here\n\
85 -shortest output shortest matches found here\n\
87 These simple tests can be combined using the operators `-a', `-o' and `-n'\n\
88 (for `and', `or' and `not'; they may also be written `&', `|' and `!' if\n\
89 you like), and grouped using parentheses `(' and `)'.\n\
93 /*----- The options parser ------------------------------------------------*/
95 /* --- Options table structure --- */
105 O_HELP
, O_VERSION
, O_USAGE
,
107 O_AND
, O_OR
, O_NOT
, O_LPAREN
, O_RPAREN
,
108 O_ANAG
, O_SUBG
, O_WILD
, O_TRACK
, O_REGEXP
, O_PCRE
, O_MONO
, O_LENGTH
,
109 O_LONGEST
, O_SHORTEST
,
115 static const struct opt opttab
[] = {
117 /* --- Options -- don't form part of the language --- */
119 { "help", 0, OF_SHORT
, O_HELP
},
120 { "version", 0, OF_SHORT
, O_VERSION
},
121 { "usage", 0, OF_SHORT
, O_USAGE
},
122 { "file", 1, OF_SHORT
, O_FILE
},
124 /* --- Operators -- provide the basic structure of the language --- *
126 * These are also given magical names by the parser.
129 { "and", 0, OF_SHORT
, O_AND
},
130 { "or", 0, OF_SHORT
, O_OR
},
131 { "not", 0, OF_SHORT
, O_NOT
},
133 /* --- Actual matching operations -- do something useful --- */
135 { "anagram", 1, 0, O_ANAG
},
136 { "subgram", 1, 0, O_SUBG
},
137 { "wildcard", 1, 0, O_WILD
},
138 { "trackword", 1, 0, O_TRACK
},
139 { "mono", 1, 0, O_MONO
},
141 { "regexp", 1, 0, O_REGEXP
},
144 { "pcre", 1, 0, O_PCRE
},
146 { "length", 1, 0, O_LENGTH
},
147 { "longest", 0, 0, O_LONGEST
},
148 { "shortest", 0, 0, O_SHORTEST
},
150 /* --- End marker --- */
156 static const char *const *av
;
159 /* --- @nextopt@ --- *
161 * Arguments: @const char ***arg@ = where to store the arg pointer
163 * Returns: The tag of the next option.
165 * Use: Scans the next option off the command line. If the option
166 * doesn't form part of the language, it's processed internally,
167 * and you'll never see it from here. On exit, the @arg@
168 * pointer is set to contain the address of the option scanned,
169 * followed by its arguments if any. You're expected to know
170 * how many arguments there are for your option.
173 static unsigned nextopt(const char *const **arg
)
176 const struct opt
*o
, *oo
;
180 /* --- Pick the next option off the front --- */
187 /* --- Cope with various forms of magic --- */
190 if (!p
[1]) switch (*p
) {
191 case '&': return (O_AND
);
192 case '|': return (O_OR
);
193 case '!': return (O_NOT
);
194 case '(': return (O_LPAREN
);
195 case ')': return (O_RPAREN
);
200 /* --- Now cope with other sorts of weirdies --- *
202 * By the end of this, a leading `-' or `--' will have been stripped.
212 die("syntax error near `--': rubbish at end of line");
216 /* --- Now look the word up in my table --- */
220 for (o
= opttab
; o
->name
; o
++) {
221 if (strncmp(p
, o
->name
, sz
) == 0) {
222 if (strlen(o
->name
) == sz
|| ((o
->f
& OF_SHORT
) && sz
== 1)) {
227 die("ambiguous option name `-%s' (could match `-%s' or `-%s')",
228 p
, oo
->name
, o
->name
);
234 die("unrecognized option name `-%s'", p
);
236 /* --- Sort out the arguments --- */
238 if (ai
+ oo
->nargs
> ac
)
239 die("too few arguments for `-%s' (need %u)", oo
->name
, oo
->nargs
);
242 /* --- Now process the option --- */
262 die("syntax error near `%s': unknown token type", av
[ai
- 1]);
266 /*----- Node types for operators ------------------------------------------*/
268 /* --- Node structures --- */
270 typedef struct node_bin
{
276 typedef struct node_un
{
281 /* --- Node functions --- */
283 static int n_or(node
*nn
, const char *p
, size_t sz
)
285 node_bin
*n
= (node_bin
*)nn
;
286 return (n
->left
->func(n
->left
, p
, sz
) || n
->right
->func(n
->right
, p
, sz
));
289 static int n_and(node
*nn
, const char *p
, size_t sz
)
291 node_bin
*n
= (node_bin
*)nn
;
292 return (n
->left
->func(n
->left
, p
, sz
) && n
->right
->func(n
->right
, p
, sz
));
295 static int n_not(node
*nn
, const char *p
, size_t sz
)
297 node_un
*n
= (node_un
*)nn
;
298 return (!n
->arg
->func(n
->arg
, p
, sz
));
301 /*----- Other simple node types -------------------------------------------*/
303 enum { LESS
= -1, EQUAL
= 0, GREATER
= 1 };
305 typedef struct node_numeric
{
311 static void parse_numeric(const char *p
, int *dir
, int *i
)
318 case '-': *dir
= LESS
; pp
++; break;
319 case '+': *dir
= GREATER
; pp
++; break;
320 default: *dir
= EQUAL
; break;
323 l
= strtol(pp
, &q
, 0);
324 if (*q
|| errno
|| l
< INT_MIN
|| l
> INT_MAX
)
325 die("bad numeric parameter `%s'", p
);
329 static node
*make_numeric(const char *const *av
,
330 int (*func
)(struct node
*, const char *, size_t))
332 node_numeric
*n
= xmalloc(sizeof(*n
));
333 parse_numeric(av
[0], &n
->dir
, &n
->i
);
338 static int cmp_numeric(int x
, int dir
, int n
)
341 case LESS
: return (x
<= n
);
342 case EQUAL
: return (x
== n
);
343 case GREATER
: return (x
>= n
);
348 static int n_length(node
*nn
, const char *p
, size_t sz
)
350 node_numeric
*n
= (node_numeric
*)nn
;
351 return (cmp_numeric(sz
, n
->dir
, n
->i
));
354 /*----- Parser for the expression syntax ----------------------------------*/
356 /* --- A parser context --- */
358 typedef struct p_ctx
{
360 const char *const *a
;
363 /* --- Parser structure --- *
365 * This is a simple recursive descent parser. The context retains
366 * information about the current token. Each function is passed the address
367 * of a node pointer to fill in. This simplifies the binary operator code
368 * somewhat, relative to returning pointers to node trees.
371 static void p_expr(p_ctx
*p
, node
**/
*nn*/
);
373 static void p_next(p_ctx
*p
)
375 static const char *const eof
[] = { "<end>", 0 };
376 p
->t
= nextopt(&p
->a
);
381 static void p_factor(p_ctx
*p
, node
**nn
)
384 if (p
->t
== O_LPAREN
) {
387 if (p
->t
!= O_RPAREN
)
388 die("syntax error near `%s': missing `)'", *p
->a
);
390 } else if (p
->t
== O_NOT
) {
391 n
= xmalloc(sizeof(node_un
));
395 p_factor(p
, &n
->arg
);
398 case O_ANAG
: *nn
= anagram(p
->a
+ 1); break;
399 case O_SUBG
: *nn
= subgram(p
->a
+ 1); break;
400 case O_WILD
: *nn
= wildcard(p
->a
+ 1); break;
401 case O_TRACK
: *nn
= trackword(p
->a
+ 1); break;
403 case O_REGEXP
: *nn
= regexp(p
->a
+ 1); break;
406 case O_PCRE
: *nn
= pcrenode(p
->a
+ 1); break;
408 case O_MONO
: *nn
= mono(p
->a
+ 1); break;
409 case O_LENGTH
: *nn
= make_numeric(p
->a
+ 1, n_length
); break;
410 case O_LONGEST
: *nn
= longest(p
->a
+ 1); break;
411 case O_SHORTEST
: *nn
= shortest(p
->a
+ 1); break;
412 default: die("syntax error near `%s': unexpected token", *p
->a
);
418 static void p_term(p_ctx
*p
, node
**nn
)
433 n
= xmalloc(sizeof(node_bin
));
441 static void p_expr(p_ctx
*p
, node
**nn
)
449 n
= xmalloc(sizeof(node_bin
));
457 /* --- @p_argv@ --- *
459 * Arguments: @int argc@ = number of command-line arguments
460 * @const char *const argv[]@ = vectoor of arguments
462 * Returns: A compiled node, parsed from the arguments.
464 * Use: Does the donkey-work of parsing a command-line.
467 static node
*p_argv(int argc
, const char *const argv
[])
478 pquis(stderr
, "(Run `$ --help' for more detail.)\n");
483 die("syntax error near `%s': rubbish at end of line (too many `)'s?)",
489 /*----- At-end stuff ------------------------------------------------------*/
491 /* --- @atend_register@ --- *
493 * Arguments: @int (*func)(void *)@ = function to call
494 * @void *p@ = handle to pass to it
498 * Use: Adds a function to the list of things to do at the end of the
499 * program. The function should return nonzero if it produced
503 typedef struct atend
{
505 int (*func
)(void */
*p*/
);
509 static atend
*aa_head
= 0, **aa_tail
= &aa_head
;
511 void atend_register(int (*func
)(void */
*p*/
), void *p
)
513 atend
*a
= xmalloc(sizeof(*a
));
521 /*----- Main code ---------------------------------------------------------*/
525 * Arguments: @int argc@ = number of command-line arguments
526 * @char *argv[]@ = vector of argument words
528 * Returns: Zero on success, nonzero on failure.
530 * Use: Picks entries from a word list which match particular
531 * expressions. This might be of assistance to word-game types.
534 int main(int argc
, char *argv
[])
544 n
= p_argv(argc
, (const char *const *)argv
);
546 if ((fp
= fopen(file
, "r")) == 0)
547 die("error opening `%s': %s", file
, strerror(errno
));
550 if (dstr_putline(&d
, fp
) < 0)
553 for (p
= q
= d
.buf
; p
< l
; p
++) {
554 if (!isalnum((unsigned char)*p
))
556 *q
++ = tolower((unsigned char)*p
);
560 if (n
->func(n
, d
.buf
, d
.len
)) {
561 fwrite(d
.buf
, 1, d
.len
, stdout
);
566 if (ferror(fp
) || fclose(fp
))
567 die("error reading `%s': %s", file
, strerror(errno
));
568 for (a
= aa_head
; a
; a
= a
->next
) {
572 if (fflush(stdout
) || ferror(stdout
) || fclose(stdout
))
573 die("error writing output: %s", strerror(errno
));
574 if (!ok
) pquis(stderr
, "$: no matches found\n");
575 return (ok ? EX_OK
: EX_NONE
);
578 /*----- That's all, folks -------------------------------------------------*/