]>
Commit | Line | Data |
---|---|---|
ef416fc2 | 1 | /* |
2 | * "$Id: help-index.c 4863 2005-12-03 04:28:10Z mike $" | |
3 | * | |
4 | * On-line help index routines for the Common UNIX Printing System (CUPS). | |
5 | * | |
6 | * Copyright 1997-2005 by Easy Software Products. | |
7 | * | |
8 | * These coded instructions, statements, and computer programs are the | |
9 | * property of Easy Software Products and are protected by Federal | |
10 | * copyright law. Distribution and use rights are outlined in the file | |
11 | * "LICENSE.txt" which should have been included with this file. If this | |
12 | * file is missing or damaged please contact Easy Software Products | |
13 | * at: | |
14 | * | |
15 | * Attn: CUPS Licensing Information | |
16 | * Easy Software Products | |
17 | * 44141 Airport View Drive, Suite 204 | |
18 | * Hollywood, Maryland 20636 USA | |
19 | * | |
20 | * Voice: (301) 373-9600 | |
21 | * EMail: cups-info@cups.org | |
22 | * WWW: http://www.cups.org | |
23 | * | |
24 | * Contents: | |
25 | * | |
26 | * helpDeleteIndex() - Delete an index, freeing all memory used. | |
27 | * helpFindNode() - Find a node in an index. | |
28 | * helpLoadIndex() - Load a help index from disk. | |
29 | * helpSaveIndex() - Save a help index to disk. | |
30 | * helpSearchIndex() - Search an index. | |
31 | * help_compile_search() - Convert a search string into a regular expression. | |
32 | * help_create_sorted() - Create the sorted node array. | |
33 | * help_delete_node() - Free all memory used by a node. | |
34 | * help_insert_node() - Insert a node in an index. | |
35 | * help_load_directory() - Load a directory of files into an index. | |
36 | * help_load_file() - Load a HTML files into an index. | |
37 | * help_new_node() - Create a new node and add it to an index. | |
38 | * help_sort_nodes_by_name() - Sort nodes by section, filename, and anchor. | |
39 | * help_sort_nodes_by_score() - Sort nodes by score and text. | |
40 | */ | |
41 | ||
42 | /* | |
43 | * Include necessary headers... | |
44 | */ | |
45 | ||
46 | #include "cgi-private.h" | |
47 | #include <cups/dir.h> | |
48 | ||
49 | ||
50 | /* | |
51 | * Local functions... | |
52 | */ | |
53 | ||
54 | static void help_create_sorted(help_index_t *hi); | |
55 | static void help_delete_node(help_node_t *n); | |
56 | static void help_insert_node(help_index_t *hi, help_node_t *n); | |
57 | static int help_load_directory(help_index_t *hi, | |
58 | const char *directory, | |
59 | const char *relative); | |
60 | static int help_load_file(help_index_t *hi, | |
61 | const char *filename, | |
62 | const char *relative, | |
63 | time_t mtime); | |
64 | static help_node_t *help_new_node(const char *filename, const char *anchor, | |
65 | const char *section, const char *text, | |
66 | time_t mtime, off_t offset, | |
67 | size_t length); | |
68 | static int help_sort_by_name(const void *p1, const void *p2); | |
69 | static int help_sort_by_score(const void *p1, const void *p2); | |
70 | ||
71 | ||
72 | /* | |
73 | * 'helpDeleteIndex()' - Delete an index, freeing all memory used. | |
74 | */ | |
75 | ||
76 | void | |
77 | helpDeleteIndex(help_index_t *hi) | |
78 | { | |
79 | int i; /* Looping var */ | |
80 | ||
81 | ||
82 | DEBUG_printf(("helpDeleteIndex(hi=%p)\n", hi)); | |
83 | ||
84 | if (!hi) | |
85 | return; | |
86 | ||
87 | if (!hi->search) | |
88 | { | |
89 | for (i = 0; i < hi->num_nodes; i ++) | |
90 | help_delete_node(hi->nodes[i]); | |
91 | } | |
92 | ||
93 | free(hi->nodes); | |
94 | ||
95 | if (hi->sorted) | |
96 | free(hi->sorted); | |
97 | ||
98 | free(hi); | |
99 | } | |
100 | ||
101 | ||
102 | /* | |
103 | * 'helpFindNode()' - Find a node in an index. | |
104 | */ | |
105 | ||
106 | help_node_t ** /* O - Node pointer or NULL */ | |
107 | helpFindNode(help_index_t *hi, /* I - Index */ | |
108 | const char *filename, /* I - Filename */ | |
109 | const char *anchor) /* I - Anchor */ | |
110 | { | |
111 | help_node_t key, /* Search key */ | |
112 | *ptr; /* Pointer to key */ | |
113 | ||
114 | ||
115 | DEBUG_printf(("helpFindNode(hi=%p, filename=\"%s\", anchor=\"%s\")\n", | |
116 | hi, filename ? filename : "(nil)", anchor ? anchor : "(nil)")); | |
117 | ||
118 | /* | |
119 | * Range check input... | |
120 | */ | |
121 | ||
122 | if (!hi || !filename) | |
123 | return (NULL); | |
124 | ||
125 | /* | |
126 | * Initialize the search key... | |
127 | */ | |
128 | ||
129 | key.filename = (char *)filename; | |
130 | key.anchor = (char *)anchor; | |
131 | ptr = &key; | |
132 | ||
133 | /* | |
134 | * Return any match... | |
135 | */ | |
136 | ||
137 | return ((help_node_t **)bsearch(&ptr, hi->nodes, hi->num_nodes, | |
138 | sizeof(help_node_t *), help_sort_by_name)); | |
139 | } | |
140 | ||
141 | ||
142 | /* | |
143 | * 'helpLoadIndex()' - Load a help index from disk. | |
144 | */ | |
145 | ||
146 | help_index_t * /* O - Index pointer or NULL */ | |
147 | helpLoadIndex(const char *hifile, /* I - Index filename */ | |
148 | const char *directory) /* I - Directory that is indexed */ | |
149 | { | |
150 | help_index_t *hi; /* Help index */ | |
151 | cups_file_t *fp; /* Current file */ | |
152 | char line[2048], /* Line from file */ | |
153 | *ptr, /* Pointer into line */ | |
154 | *filename, /* Filename in line */ | |
155 | *anchor, /* Anchor in line */ | |
156 | *sectptr, /* Section pointer in line */ | |
157 | section[1024], /* Section name */ | |
158 | *text; /* Text in line */ | |
159 | time_t mtime; /* Modification time */ | |
160 | off_t offset; /* Offset into file */ | |
161 | size_t length; /* Length in bytes */ | |
162 | int update; /* Update? */ | |
163 | int i; /* Looping var */ | |
164 | help_node_t *node; /* Current node */ | |
165 | ||
166 | ||
167 | DEBUG_printf(("helpLoadIndex(hifile=\"%s\", directory=\"%s\")\n", | |
168 | hifile, directory)); | |
169 | ||
170 | /* | |
171 | * Create a new, empty index. | |
172 | */ | |
173 | ||
174 | hi = (help_index_t *)calloc(1, sizeof(help_index_t)); | |
175 | ||
176 | /* | |
177 | * Try loading the existing index file... | |
178 | */ | |
179 | ||
180 | if ((fp = cupsFileOpen(hifile, "r")) != NULL) | |
181 | { | |
182 | /* | |
183 | * Lock the file and then read the first line... | |
184 | */ | |
185 | ||
186 | cupsFileLock(fp, 1); | |
187 | ||
188 | if (cupsFileGets(fp, line, sizeof(line)) && !strcmp(line, "HELPV1")) | |
189 | { | |
190 | /* | |
191 | * Got a valid header line, now read the data lines... | |
192 | */ | |
193 | ||
194 | while (cupsFileGets(fp, line, sizeof(line))) | |
195 | { | |
196 | /* | |
197 | * Each line looks like one of the following: | |
198 | * | |
199 | * filename mtime offset length "section" "text" | |
200 | * filename#anchor offset length "text" | |
201 | */ | |
202 | ||
203 | filename = line; | |
204 | ||
205 | if ((ptr = strchr(line, ' ')) == NULL) | |
206 | break; | |
207 | ||
208 | while (isspace(*ptr & 255)) | |
209 | *ptr++ = '\0'; | |
210 | ||
211 | if ((anchor = strrchr(filename, '#')) != NULL) | |
212 | { | |
213 | *anchor++ = '\0'; | |
214 | mtime = 0; | |
215 | } | |
216 | else | |
217 | mtime = strtol(ptr, &ptr, 10); | |
218 | ||
219 | offset = strtoll(ptr, &ptr, 10); | |
220 | length = strtoll(ptr, &ptr, 10); | |
221 | ||
222 | while (isspace(*ptr & 255)) | |
223 | ptr ++; | |
224 | ||
225 | if (!anchor) | |
226 | { | |
227 | /* | |
228 | * Get section... | |
229 | */ | |
230 | ||
231 | if (*ptr != '\"') | |
232 | break; | |
233 | ||
234 | ptr ++; | |
235 | sectptr = ptr; | |
236 | ||
237 | while (*ptr && *ptr != '\"') | |
238 | ptr ++; | |
239 | ||
240 | if (*ptr != '\"') | |
241 | break; | |
242 | ||
243 | *ptr++ = '\0'; | |
244 | ||
245 | strlcpy(section, sectptr, sizeof(section)); | |
246 | ||
247 | while (isspace(*ptr & 255)) | |
248 | ptr ++; | |
249 | } | |
250 | ||
251 | if (*ptr != '\"') | |
252 | break; | |
253 | ||
254 | ptr ++; | |
255 | text = ptr; | |
256 | ||
257 | while (*ptr && *ptr != '\"') | |
258 | ptr ++; | |
259 | ||
260 | if (*ptr != '\"') | |
261 | break; | |
262 | ||
263 | *ptr++ = '\0'; | |
264 | ||
265 | if ((node = help_new_node(filename, anchor, section, text, | |
266 | mtime, offset, length)) == NULL) | |
267 | break; | |
268 | ||
269 | help_insert_node(hi, node); | |
270 | ||
271 | node->score = -1; | |
272 | } | |
273 | } | |
274 | ||
275 | cupsFileClose(fp); | |
276 | } | |
277 | ||
278 | /* | |
279 | * Scan for new/updated files... | |
280 | */ | |
281 | ||
282 | update = help_load_directory(hi, directory, NULL); | |
283 | ||
284 | /* | |
285 | * Remove any files that are no longer installed... | |
286 | */ | |
287 | ||
288 | for (i = 0; i < hi->num_nodes;) | |
289 | { | |
290 | if (hi->nodes[i]->score < 0) | |
291 | { | |
292 | /* | |
293 | * Delete this node... | |
294 | */ | |
295 | ||
296 | help_delete_node(hi->nodes[i]); | |
297 | ||
298 | hi->num_nodes --; | |
299 | if (i < hi->num_nodes) | |
300 | memmove(hi->nodes + i, hi->nodes + i + 1, | |
301 | (hi->num_nodes - i) * sizeof(help_node_t *)); | |
302 | ||
303 | update = 1; | |
304 | } | |
305 | else | |
306 | { | |
307 | /* | |
308 | * Keep this node... | |
309 | */ | |
310 | ||
311 | i ++; | |
312 | } | |
313 | } | |
314 | ||
315 | /* | |
316 | * Save the index if we updated it... | |
317 | */ | |
318 | ||
319 | if (update) | |
320 | helpSaveIndex(hi, hifile); | |
321 | ||
322 | /* | |
323 | * Create the sorted array... | |
324 | */ | |
325 | ||
326 | help_create_sorted(hi); | |
327 | ||
328 | /* | |
329 | * Return the index... | |
330 | */ | |
331 | ||
332 | return (hi); | |
333 | } | |
334 | ||
335 | ||
336 | /* | |
337 | * 'helpSaveIndex()' - Save a help index to disk. | |
338 | */ | |
339 | ||
340 | int /* O - 0 on success, -1 on error */ | |
341 | helpSaveIndex(help_index_t *hi, /* I - Index */ | |
342 | const char *hifile) /* I - Index filename */ | |
343 | { | |
344 | cups_file_t *fp; /* Index file */ | |
345 | int i; /* Looping var */ | |
346 | help_node_t *node; /* Current node */ | |
347 | ||
348 | ||
349 | DEBUG_printf(("helpSaveIndex(hi=%p, hifile=\"%s\")\n", hi, hifile)); | |
350 | ||
351 | /* | |
352 | * Try creating a new index file... | |
353 | */ | |
354 | ||
355 | if ((fp = cupsFileOpen(hifile, "w9")) == NULL) | |
356 | return (-1); | |
357 | ||
358 | /* | |
359 | * Lock the file while we write it... | |
360 | */ | |
361 | ||
362 | cupsFileLock(fp, 1); | |
363 | ||
364 | cupsFilePuts(fp, "HELPV1\n"); | |
365 | ||
366 | for (i = 0; i < hi->num_nodes; i ++) | |
367 | { | |
368 | /* | |
369 | * Write the current node with/without the anchor... | |
370 | */ | |
371 | ||
372 | node = hi->nodes[i]; | |
373 | ||
374 | if (node->anchor) | |
375 | { | |
376 | if (cupsFilePrintf(fp, "%s#%s " CUPS_LLFMT " " CUPS_LLFMT " \"%s\"\n", | |
377 | node->filename, node->anchor, | |
378 | CUPS_LLCAST node->offset, CUPS_LLCAST node->length, | |
379 | node->text) < 0) | |
380 | break; | |
381 | } | |
382 | else | |
383 | { | |
384 | if (cupsFilePrintf(fp, "%s %d " CUPS_LLFMT " " CUPS_LLFMT " \"%s\" \"%s\"\n", | |
385 | node->filename, node->mtime, | |
386 | CUPS_LLCAST node->offset, CUPS_LLCAST node->length, | |
387 | node->section ? node->section : "", node->text) < 0) | |
388 | break; | |
389 | } | |
390 | } | |
391 | ||
392 | if (cupsFileClose(fp) < 0) | |
393 | return (-1); | |
394 | else if (i < hi->num_nodes) | |
395 | return (-1); | |
396 | else | |
397 | return (0); | |
398 | } | |
399 | ||
400 | ||
401 | /* | |
402 | * 'helpSearchIndex()' - Search an index. | |
403 | */ | |
404 | ||
405 | help_index_t * /* O - Search index */ | |
406 | helpSearchIndex(help_index_t *hi, /* I - Index */ | |
407 | const char *query, /* I - Query string */ | |
408 | const char *section, /* I - Limit search to this section */ | |
409 | const char *filename) /* I - Limit search to this file */ | |
410 | { | |
411 | int i; /* Looping var */ | |
412 | help_index_t *search; /* Search index */ | |
413 | help_node_t **n; /* Current node */ | |
414 | void *sc; /* Search context */ | |
415 | int matches; /* Number of matches */ | |
416 | ||
417 | ||
418 | DEBUG_printf(("helpSearchIndex(hi=%p, query=\"%s\", filename=\"%s\")\n", | |
419 | hi, query ? query : "(nil)", | |
420 | filename ? filename : "(nil)")); | |
421 | ||
422 | /* | |
423 | * Range check... | |
424 | */ | |
425 | ||
426 | if (!hi || !query) | |
427 | return (NULL); | |
428 | ||
429 | for (i = 0, n = hi->nodes; i < hi->num_nodes; i ++, n ++) | |
430 | n[0]->score = 0; | |
431 | ||
432 | if (filename) | |
433 | { | |
434 | n = helpFindNode(hi, filename, NULL); | |
435 | if (!n) | |
436 | return (NULL); | |
437 | } | |
438 | else | |
439 | n = hi->nodes; | |
440 | ||
441 | /* | |
442 | * Convert the query into a regular expression... | |
443 | */ | |
444 | ||
445 | sc = cgiCompileSearch(query); | |
446 | if (!sc) | |
447 | return (NULL); | |
448 | ||
449 | /* | |
450 | * Allocate a search index... | |
451 | */ | |
452 | ||
453 | search = calloc(1, sizeof(help_index_t)); | |
454 | if (!search) | |
455 | { | |
456 | cgiFreeSearch(sc); | |
457 | return (NULL); | |
458 | } | |
459 | ||
460 | search->search = 1; | |
461 | ||
462 | /* | |
463 | * Check each node in the index, adding matching nodes to the | |
464 | * search index... | |
465 | */ | |
466 | ||
467 | for (i = n - hi->nodes; i < hi->num_nodes; i ++, n ++) | |
468 | if (section && strcmp(n[0]->section, section)) | |
469 | continue; | |
470 | else if (filename && strcmp(n[0]->filename, filename)) | |
471 | continue; | |
472 | else if ((matches = cgiDoSearch(sc, n[0]->text)) > 0) | |
473 | { | |
474 | /* | |
475 | * Found a match, add the node to the search index... | |
476 | */ | |
477 | ||
478 | help_insert_node(search, *n); | |
479 | ||
480 | n[0]->score = matches; | |
481 | } | |
482 | ||
483 | /* | |
484 | * Free the search context... | |
485 | */ | |
486 | ||
487 | cgiFreeSearch(sc); | |
488 | ||
489 | /* | |
490 | * Sort the results... | |
491 | */ | |
492 | ||
493 | help_create_sorted(search); | |
494 | ||
495 | /* | |
496 | * Return the results... | |
497 | */ | |
498 | ||
499 | return (search); | |
500 | } | |
501 | ||
502 | ||
503 | /* | |
504 | * 'help_create_sorted()' - Create the sorted node array. | |
505 | */ | |
506 | ||
507 | static void | |
508 | help_create_sorted(help_index_t *hi) /* I - Index */ | |
509 | { | |
510 | DEBUG_printf(("help_create_sorted(hi=%p)\n", hi)); | |
511 | ||
512 | /* | |
513 | * Free any existing sorted array... | |
514 | */ | |
515 | ||
516 | if (hi->sorted) | |
517 | free(hi->sorted); | |
518 | ||
519 | /* | |
520 | * Create a new sorted array... | |
521 | */ | |
522 | ||
523 | hi->sorted = calloc(hi->num_nodes, sizeof(help_node_t *)); | |
524 | ||
525 | if (!hi->sorted) | |
526 | return; | |
527 | ||
528 | /* | |
529 | * Copy the nodes to the new array... | |
530 | */ | |
531 | ||
532 | memcpy(hi->sorted, hi->nodes, hi->num_nodes * sizeof(help_node_t *)); | |
533 | ||
534 | /* | |
535 | * Sort the new array by score and text. | |
536 | */ | |
537 | ||
538 | if (hi->num_nodes > 1) | |
539 | qsort(hi->sorted, hi->num_nodes, sizeof(help_node_t *), | |
540 | help_sort_by_score); | |
541 | } | |
542 | ||
543 | ||
544 | /* | |
545 | * 'help_delete_node()' - Free all memory used by a node. | |
546 | */ | |
547 | ||
548 | static void | |
549 | help_delete_node(help_node_t *n) /* I - Node */ | |
550 | { | |
551 | DEBUG_printf(("help_delete_node(n=%p)\n", n)); | |
552 | ||
553 | if (!n) | |
554 | return; | |
555 | ||
556 | if (n->filename) | |
557 | free(n->filename); | |
558 | ||
559 | if (n->anchor) | |
560 | free(n->anchor); | |
561 | ||
562 | if (n->section) | |
563 | free(n->section); | |
564 | ||
565 | if (n->text) | |
566 | free(n->text); | |
567 | ||
568 | free(n); | |
569 | } | |
570 | ||
571 | ||
572 | /* | |
573 | * 'help_insert_node()' - Insert a node in an index. | |
574 | */ | |
575 | ||
576 | static void | |
577 | help_insert_node(help_index_t *hi, /* I - Index */ | |
578 | help_node_t *n) /* I - Node */ | |
579 | { | |
580 | int current, /* Current node */ | |
581 | left, /* Left side */ | |
582 | right, /* Right side */ | |
583 | diff; /* Difference between nodes */ | |
584 | help_node_t **temp; /* Temporary node pointer */ | |
585 | ||
586 | ||
587 | DEBUG_printf(("help_insert_node(hi=%p, n=%p)\n", hi, n)); | |
588 | ||
589 | /* | |
590 | * Allocate memory as needed... | |
591 | */ | |
592 | ||
593 | if (hi->num_nodes >= hi->alloc_nodes) | |
594 | { | |
595 | /* | |
596 | * Expand the array in 128 node increments... | |
597 | */ | |
598 | ||
599 | hi->alloc_nodes += 128; | |
600 | if (hi->alloc_nodes == 128) | |
601 | temp = (help_node_t **)malloc(hi->alloc_nodes * sizeof(help_node_t *)); | |
602 | else | |
603 | temp = (help_node_t **)realloc(hi->nodes, | |
604 | hi->alloc_nodes * sizeof(help_node_t *)); | |
605 | ||
606 | if (!temp) | |
607 | return; | |
608 | ||
609 | hi->nodes = temp; | |
610 | } | |
611 | ||
612 | /* | |
613 | * Find the insertion point... | |
614 | */ | |
615 | ||
616 | if (hi->num_nodes == 0 || | |
617 | help_sort_by_name(&n, hi->nodes + hi->num_nodes - 1) > 0) | |
618 | { | |
619 | /* | |
620 | * Last node... | |
621 | */ | |
622 | ||
623 | hi->nodes[hi->num_nodes] = n; | |
624 | hi->num_nodes ++; | |
625 | return; | |
626 | } | |
627 | else if (help_sort_by_name(&n, hi->nodes) < 0) | |
628 | { | |
629 | /* | |
630 | * First node... | |
631 | */ | |
632 | ||
633 | memmove(hi->nodes + 1, hi->nodes, hi->num_nodes * sizeof(help_node_t *)); | |
634 | hi->nodes[0] = n; | |
635 | hi->num_nodes ++; | |
636 | return; | |
637 | } | |
638 | ||
639 | /* | |
640 | * Otherwise, do a binary insertion... | |
641 | */ | |
642 | ||
643 | left = 0; | |
644 | right = hi->num_nodes - 1; | |
645 | ||
646 | do | |
647 | { | |
648 | current = (left + right) / 2; | |
649 | diff = help_sort_by_name(&n, hi->nodes + current); | |
650 | ||
651 | if (!diff) | |
652 | break; | |
653 | else if (diff < 0) | |
654 | right = current; | |
655 | else | |
656 | left = current; | |
657 | } | |
658 | while ((right - left) > 1); | |
659 | ||
660 | if (diff > 0) | |
661 | current ++; | |
662 | ||
663 | memmove(hi->nodes + current + 1, hi->nodes + current, | |
664 | (hi->num_nodes - current) * sizeof(help_node_t *)); | |
665 | hi->nodes[current] = n; | |
666 | hi->num_nodes ++; | |
667 | } | |
668 | ||
669 | ||
670 | /* | |
671 | * 'help_load_directory()' - Load a directory of files into an index. | |
672 | */ | |
673 | ||
674 | static int /* O - 0 = success, -1 = error, 1 = updated */ | |
675 | help_load_directory( | |
676 | help_index_t *hi, /* I - Index */ | |
677 | const char *directory, /* I - Directory */ | |
678 | const char *relative) /* I - Relative path */ | |
679 | { | |
680 | int i; /* Looping var */ | |
681 | cups_dir_t *dir; /* Directory file */ | |
682 | cups_dentry_t *dent; /* Directory entry */ | |
683 | char *ext, /* Pointer to extension */ | |
684 | filename[1024], /* Full filename */ | |
685 | relname[1024]; /* Relative filename */ | |
686 | int update; /* Updated? */ | |
687 | help_node_t **node; /* Current node */ | |
688 | ||
689 | ||
690 | DEBUG_printf(("help_load_directory(hi=%p, directory=\"%s\", relative=\"%s\")\n", | |
691 | hi, directory ? directory : "(nil)", relative ? relative : "(nil)")); | |
692 | ||
693 | /* | |
694 | * Open the directory and scan it... | |
695 | */ | |
696 | ||
697 | if ((dir = cupsDirOpen(directory)) == NULL) | |
698 | return (0); | |
699 | ||
700 | update = 0; | |
701 | ||
702 | while ((dent = cupsDirRead(dir)) != NULL) | |
703 | { | |
704 | /* | |
705 | * Get absolute and relative filenames... | |
706 | */ | |
707 | ||
708 | snprintf(filename, sizeof(filename), "%s/%s", directory, dent->filename); | |
709 | if (relative) | |
710 | snprintf(relname, sizeof(relname), "%s/%s", relative, dent->filename); | |
711 | else | |
712 | strlcpy(relname, dent->filename, sizeof(relname)); | |
713 | ||
714 | /* | |
715 | * Check if we have a HTML file... | |
716 | */ | |
717 | ||
718 | if ((ext = strstr(dent->filename, ".html")) != NULL && | |
719 | (!ext[5] || !strcmp(ext + 5, ".gz"))) | |
720 | { | |
721 | /* | |
722 | * HTML file, see if we have already indexed the file... | |
723 | */ | |
724 | ||
725 | if ((node = helpFindNode(hi, relname, NULL)) != NULL) | |
726 | { | |
727 | /* | |
728 | * File already indexed - check dates to confirm that the | |
729 | * index is up-to-date... | |
730 | */ | |
731 | ||
732 | if (node[0]->mtime == dent->fileinfo.st_mtime) | |
733 | { | |
734 | /* | |
735 | * Same modification time, so mark all of the nodes | |
736 | * for this file as up-to-date... | |
737 | */ | |
738 | ||
739 | for (i = node - hi->nodes; i < hi->num_nodes; i ++, node ++) | |
740 | if (!strcmp(node[0]->filename, relname)) | |
741 | node[0]->score = 0; | |
742 | else | |
743 | break; | |
744 | ||
745 | continue; | |
746 | } | |
747 | } | |
748 | ||
749 | update = 1; | |
750 | ||
751 | help_load_file(hi, filename, relname, dent->fileinfo.st_mtime); | |
752 | } | |
753 | else if (S_ISDIR(dent->fileinfo.st_mode)) | |
754 | { | |
755 | /* | |
756 | * Process sub-directory... | |
757 | */ | |
758 | ||
759 | if (help_load_directory(hi, filename, relname) == 1) | |
760 | update = 1; | |
761 | } | |
762 | } | |
763 | ||
764 | cupsDirClose(dir); | |
765 | ||
766 | return (update); | |
767 | } | |
768 | ||
769 | ||
770 | /* | |
771 | * 'help_load_file()' - Load a HTML files into an index. | |
772 | */ | |
773 | ||
774 | static int /* O - 0 = success, -1 = error */ | |
775 | help_load_file( | |
776 | help_index_t *hi, /* I - Index */ | |
777 | const char *filename, /* I - Filename */ | |
778 | const char *relative, /* I - Relative path */ | |
779 | time_t mtime) /* I - Modification time */ | |
780 | { | |
781 | cups_file_t *fp; /* HTML file */ | |
782 | help_node_t *node, /* Current node */ | |
783 | **n; /* Node pointer */ | |
784 | char line[1024], /* Line from file */ | |
785 | section[1024], /* Section */ | |
786 | *ptr, /* Pointer into line */ | |
787 | *anchor, /* Anchor name */ | |
788 | *text; /* Text for anchor */ | |
789 | off_t offset; /* File offset */ | |
790 | char quote; /* Quote character */ | |
791 | ||
792 | ||
793 | DEBUG_printf(("help_load_file(hi=%p, filename=\"%s\", relative=\"%s\", mtime=%ld)\n", | |
794 | hi, filename ? filename : "(nil)", | |
795 | relative ? relative : "(nil)", mtime)); | |
796 | ||
797 | if ((fp = cupsFileOpen(filename, "r")) == NULL) | |
798 | return (-1); | |
799 | ||
800 | node = NULL; | |
801 | offset = 0; | |
802 | ||
803 | strcpy(section, "Other"); | |
804 | ||
805 | while (cupsFileGets(fp, line, sizeof(line))) | |
806 | { | |
807 | /* | |
808 | * Look for "<TITLE>", "<A NAME", or "<!-- SECTION:" prefix... | |
809 | */ | |
810 | ||
811 | if (!strncasecmp(line, "<!-- SECTION:", 13)) | |
812 | { | |
813 | /* | |
814 | * Got section line, copy it! | |
815 | */ | |
816 | ||
817 | for (ptr = line + 13; isspace(*ptr & 255); ptr ++); | |
818 | ||
819 | strlcpy(section, ptr, sizeof(section)); | |
820 | if ((ptr = strstr(section, "-->")) != NULL) | |
821 | { | |
822 | /* | |
823 | * Strip comment stuff from end of line... | |
824 | */ | |
825 | ||
826 | for (*ptr-- = '\0'; ptr > line && isspace(*ptr & 255); *ptr-- = '\0'); | |
827 | ||
828 | if (isspace(*ptr & 255)) | |
829 | *ptr = '\0'; | |
830 | } | |
831 | continue; | |
832 | } | |
833 | ||
834 | for (ptr = line; (ptr = strchr(ptr, '<')) != NULL;) | |
835 | { | |
836 | ptr ++; | |
837 | ||
838 | if (!strncasecmp(ptr, "TITLE>", 6)) | |
839 | { | |
840 | /* | |
841 | * Found the title... | |
842 | */ | |
843 | ||
844 | anchor = NULL; | |
845 | ptr += 6; | |
846 | } | |
847 | else if (!strncasecmp(ptr, "A NAME=", 7)) | |
848 | { | |
849 | /* | |
850 | * Found an anchor... | |
851 | */ | |
852 | ||
853 | ptr += 7; | |
854 | ||
855 | if (*ptr == '\"' || *ptr == '\'') | |
856 | { | |
857 | /* | |
858 | * Get quoted anchor... | |
859 | */ | |
860 | ||
861 | quote = *ptr; | |
862 | anchor = ptr + 1; | |
863 | if ((ptr = strchr(anchor, quote)) != NULL) | |
864 | *ptr++ = '\0'; | |
865 | else | |
866 | break; | |
867 | } | |
868 | else | |
869 | { | |
870 | /* | |
871 | * Get unquoted anchor... | |
872 | */ | |
873 | ||
874 | anchor = ptr + 1; | |
875 | ||
876 | for (ptr = anchor; *ptr && *ptr != '>' && !isspace(*ptr & 255); ptr ++); | |
877 | ||
878 | if (*ptr) | |
879 | *ptr++ = '\0'; | |
880 | else | |
881 | break; | |
882 | } | |
883 | ||
884 | /* | |
885 | * Got the anchor, now lets find the end... | |
886 | */ | |
887 | ||
888 | while (*ptr && *ptr != '>') | |
889 | ptr ++; | |
890 | ||
891 | if (*ptr != '>') | |
892 | break; | |
893 | ||
894 | ptr ++; | |
895 | } | |
896 | else | |
897 | continue; | |
898 | ||
899 | /* | |
900 | * Now collect text for the link... | |
901 | */ | |
902 | ||
903 | text = ptr; | |
904 | while ((ptr = strchr(text, '<')) == NULL) | |
905 | { | |
906 | ptr = text + strlen(text); | |
907 | if (ptr >= (line + sizeof(line) - 2)) | |
908 | break; | |
909 | ||
910 | *ptr++ = ' '; | |
911 | ||
912 | if (!cupsFileGets(fp, ptr, sizeof(line) - (ptr - line) - 1)) | |
913 | break; | |
914 | } | |
915 | ||
916 | *ptr = '\0'; | |
917 | ||
918 | if (node) | |
919 | node->length = offset - node->offset; | |
920 | ||
921 | if (!*text) | |
922 | { | |
923 | node = NULL; | |
924 | break; | |
925 | } | |
926 | ||
927 | if ((n = helpFindNode(hi, relative, anchor)) != NULL) | |
928 | { | |
929 | /* | |
930 | * Node already in the index, so replace the text and other | |
931 | * data... | |
932 | */ | |
933 | ||
934 | node = n[0]; | |
935 | ||
936 | if (node->section) | |
937 | free(node->section); | |
938 | ||
939 | if (node->text) | |
940 | free(node->text); | |
941 | ||
942 | node->section = section[0] ? strdup(section) : NULL; | |
943 | node->text = strdup(text); | |
944 | node->mtime = mtime; | |
945 | node->offset = offset; | |
946 | node->score = 0; | |
947 | } | |
948 | else | |
949 | { | |
950 | /* | |
951 | * New node... | |
952 | */ | |
953 | ||
954 | node = help_new_node(relative, anchor, section, text, mtime, offset, 0); | |
955 | help_insert_node(hi, node); | |
956 | } | |
957 | ||
958 | /* | |
959 | * Go through the text value and replace tabs and newlines with | |
960 | * whitespace and eliminate extra whitespace... | |
961 | */ | |
962 | ||
963 | for (ptr = node->text, text = node->text; *ptr;) | |
964 | if (isspace(*ptr & 255)) | |
965 | { | |
966 | while (isspace(*ptr & 255)) | |
967 | *ptr ++; | |
968 | ||
969 | *text++ = ' '; | |
970 | } | |
971 | else if (text != ptr) | |
972 | *text++ = *ptr++; | |
973 | else | |
974 | { | |
975 | text ++; | |
976 | ptr ++; | |
977 | } | |
978 | ||
979 | *text = '\0'; | |
980 | ||
981 | break; | |
982 | } | |
983 | ||
984 | /* | |
985 | * Get the offset of the next line... | |
986 | */ | |
987 | ||
988 | offset = cupsFileTell(fp); | |
989 | } | |
990 | ||
991 | cupsFileClose(fp); | |
992 | ||
993 | if (node) | |
994 | node->length = offset - node->offset; | |
995 | ||
996 | return (0); | |
997 | } | |
998 | ||
999 | ||
1000 | /* | |
1001 | * 'help_new_node()' - Create a new node and add it to an index. | |
1002 | */ | |
1003 | ||
1004 | static help_node_t * /* O - Node pointer or NULL on error */ | |
1005 | help_new_node(const char *filename, /* I - Filename */ | |
1006 | const char *anchor, /* I - Anchor */ | |
1007 | const char *section, /* I - Section */ | |
1008 | const char *text, /* I - Text */ | |
1009 | time_t mtime, /* I - Modification time */ | |
1010 | off_t offset, /* I - Offset in file */ | |
1011 | size_t length) /* I - Length in bytes */ | |
1012 | { | |
1013 | help_node_t *n; /* Node */ | |
1014 | ||
1015 | ||
1016 | DEBUG_printf(("help_new_node(filename=\"%s\", anchor=\"%s\", text=\"%s\", mtime=%ld, offset=%ld, length=%ld)\n", | |
1017 | filename ? filename : "(nil)", anchor ? anchor : "(nil)", | |
1018 | text ? text : "(nil)", mtime, offset, length)); | |
1019 | ||
1020 | n = (help_node_t *)calloc(1, sizeof(help_node_t)); | |
1021 | if (!n) | |
1022 | return (NULL); | |
1023 | ||
1024 | n->filename = strdup(filename); | |
1025 | n->anchor = anchor ? strdup(anchor) : NULL; | |
1026 | n->section = (section && *section) ? strdup(section) : NULL; | |
1027 | n->text = strdup(text); | |
1028 | n->mtime = mtime; | |
1029 | n->offset = offset; | |
1030 | n->length = length; | |
1031 | ||
1032 | return (n); | |
1033 | } | |
1034 | ||
1035 | ||
1036 | /* | |
1037 | * 'help_sort_nodes_by_name()' - Sort nodes by section, filename, and anchor. | |
1038 | */ | |
1039 | ||
1040 | static int /* O - Difference */ | |
1041 | help_sort_by_name(const void *p1, /* I - First node */ | |
1042 | const void *p2) /* I - Second node */ | |
1043 | { | |
1044 | help_node_t **n1, /* First node */ | |
1045 | **n2; /* Second node */ | |
1046 | int diff; /* Difference */ | |
1047 | ||
1048 | ||
1049 | DEBUG_printf(("help_sort_by_name(p1=%p, p2=%p)\n", p1, p2)); | |
1050 | ||
1051 | n1 = (help_node_t **)p1; | |
1052 | n2 = (help_node_t **)p2; | |
1053 | ||
1054 | if ((diff = strcmp(n1[0]->filename, n2[0]->filename)) != 0) | |
1055 | return (diff); | |
1056 | ||
1057 | if (!n1[0]->anchor && !n2[0]->anchor) | |
1058 | return (0); | |
1059 | else if (!n1[0]->anchor) | |
1060 | return (-1); | |
1061 | else if (!n2[0]->anchor) | |
1062 | return (1); | |
1063 | else | |
1064 | return (strcmp(n1[0]->anchor, n2[0]->anchor)); | |
1065 | } | |
1066 | ||
1067 | ||
1068 | /* | |
1069 | * 'help_sort_nodes_by_score()' - Sort nodes by score and text. | |
1070 | */ | |
1071 | ||
1072 | static int /* O - Difference */ | |
1073 | help_sort_by_score(const void *p1, /* I - First node */ | |
1074 | const void *p2) /* I - Second node */ | |
1075 | { | |
1076 | help_node_t **n1, /* First node */ | |
1077 | **n2; /* Second node */ | |
1078 | int diff; /* Difference */ | |
1079 | ||
1080 | ||
1081 | DEBUG_printf(("help_sort_by_score(p1=%p, p2=%p)\n", p1, p2)); | |
1082 | ||
1083 | n1 = (help_node_t **)p1; | |
1084 | n2 = (help_node_t **)p2; | |
1085 | ||
1086 | if (n1[0]->score != n2[0]->score) | |
1087 | return (n1[0]->score - n2[0]->score); | |
1088 | ||
1089 | if (n1[0]->section && !n2[0]->section) | |
1090 | return (1); | |
1091 | else if (!n1[0]->section && n2[0]->section) | |
1092 | return (-1); | |
1093 | else if (n1[0]->section && n2[0]->section && | |
1094 | (diff = strcmp(n1[0]->section, n2[0]->section)) != 0) | |
1095 | return (diff); | |
1096 | ||
1097 | return (strcasecmp(n1[0]->text, n2[0]->text)); | |
1098 | } | |
1099 | ||
1100 | ||
1101 | /* | |
1102 | * End of "$Id: help-index.c 4863 2005-12-03 04:28:10Z mike $". | |
1103 | */ |