diff options
author | ArthurHoaro <arthur@hoa.ro> | 2019-07-27 12:30:33 +0200 |
---|---|---|
committer | ArthurHoaro <arthur@hoa.ro> | 2019-07-27 12:30:33 +0200 |
commit | 1b8ed48a0d3964186f4d66d443783f4d250e7147 (patch) | |
tree | 23597f312507ba0c1b461755b9aa086106374a4d /application/bookmark/LinkDB.php | |
parent | 1aa24ed8d2974cda98733f74b36844b02942cc11 (diff) | |
parent | ed3365325d231e044dedc32608fde87b1b39fa34 (diff) | |
download | Shaarli-1b8ed48a0d3964186f4d66d443783f4d250e7147.tar.gz Shaarli-1b8ed48a0d3964186f4d66d443783f4d250e7147.tar.zst Shaarli-1b8ed48a0d3964186f4d66d443783f4d250e7147.zip |
Merge tag 'v0.11.0' into latest
Release v0.11.0
Diffstat (limited to 'application/bookmark/LinkDB.php')
-rw-r--r-- | application/bookmark/LinkDB.php | 575 |
1 files changed, 575 insertions, 0 deletions
diff --git a/application/bookmark/LinkDB.php b/application/bookmark/LinkDB.php new file mode 100644 index 00000000..76ba95f0 --- /dev/null +++ b/application/bookmark/LinkDB.php | |||
@@ -0,0 +1,575 @@ | |||
1 | <?php | ||
2 | |||
3 | namespace Shaarli\Bookmark; | ||
4 | |||
5 | use ArrayAccess; | ||
6 | use Countable; | ||
7 | use DateTime; | ||
8 | use Iterator; | ||
9 | use Shaarli\Bookmark\Exception\LinkNotFoundException; | ||
10 | use Shaarli\Exceptions\IOException; | ||
11 | use Shaarli\FileUtils; | ||
12 | |||
13 | /** | ||
14 | * Data storage for links. | ||
15 | * | ||
16 | * This object behaves like an associative array. | ||
17 | * | ||
18 | * Example: | ||
19 | * $myLinks = new LinkDB(); | ||
20 | * echo $myLinks[350]['title']; | ||
21 | * foreach ($myLinks as $link) | ||
22 | * echo $link['title'].' at url '.$link['url'].'; description:'.$link['description']; | ||
23 | * | ||
24 | * Available keys: | ||
25 | * - id: primary key, incremental integer identifier (persistent) | ||
26 | * - description: description of the entry | ||
27 | * - created: creation date of this entry, DateTime object. | ||
28 | * - updated: last modification date of this entry, DateTime object. | ||
29 | * - private: Is this link private? 0=no, other value=yes | ||
30 | * - tags: tags attached to this entry (separated by spaces) | ||
31 | * - title Title of the link | ||
32 | * - url URL of the link. Used for displayable links. | ||
33 | * Can be absolute or relative in the database but the relative links | ||
34 | * will be converted to absolute ones in templates. | ||
35 | * - real_url Raw URL in stored in the DB (absolute or relative). | ||
36 | * - shorturl Permalink smallhash | ||
37 | * | ||
38 | * Implements 3 interfaces: | ||
39 | * - ArrayAccess: behaves like an associative array; | ||
40 | * - Countable: there is a count() method; | ||
41 | * - Iterator: usable in foreach () loops. | ||
42 | * | ||
43 | * ID mechanism: | ||
44 | * ArrayAccess is implemented in a way that will allow to access a link | ||
45 | * with the unique identifier ID directly with $link[ID]. | ||
46 | * Note that it's not the real key of the link array attribute. | ||
47 | * This mechanism is in place to have persistent link IDs, | ||
48 | * even though the internal array is reordered by date. | ||
49 | * Example: | ||
50 | * - DB: link #1 (2010-01-01) link #2 (2016-01-01) | ||
51 | * - Order: #2 #1 | ||
52 | * - Import links containing: link #3 (2013-01-01) | ||
53 | * - New DB: link #1 (2010-01-01) link #2 (2016-01-01) link #3 (2013-01-01) | ||
54 | * - Real order: #2 #3 #1 | ||
55 | */ | ||
56 | class LinkDB implements Iterator, Countable, ArrayAccess | ||
57 | { | ||
58 | // Links are stored as a PHP serialized string | ||
59 | private $datastore; | ||
60 | |||
61 | // Link date storage format | ||
62 | const LINK_DATE_FORMAT = 'Ymd_His'; | ||
63 | |||
64 | // List of links (associative array) | ||
65 | // - key: link date (e.g. "20110823_124546"), | ||
66 | // - value: associative array (keys: title, description...) | ||
67 | private $links; | ||
68 | |||
69 | // List of all recorded URLs (key=url, value=link offset) | ||
70 | // for fast reserve search (url-->link offset) | ||
71 | private $urls; | ||
72 | |||
73 | /** | ||
74 | * @var array List of all links IDS mapped with their array offset. | ||
75 | * Map: id->offset. | ||
76 | */ | ||
77 | protected $ids; | ||
78 | |||
79 | // List of offset keys (for the Iterator interface implementation) | ||
80 | private $keys; | ||
81 | |||
82 | // Position in the $this->keys array (for the Iterator interface) | ||
83 | private $position; | ||
84 | |||
85 | // Is the user logged in? (used to filter private links) | ||
86 | private $loggedIn; | ||
87 | |||
88 | // Hide public links | ||
89 | private $hidePublicLinks; | ||
90 | |||
91 | /** | ||
92 | * Creates a new LinkDB | ||
93 | * | ||
94 | * Checks if the datastore exists; else, attempts to create a dummy one. | ||
95 | * | ||
96 | * @param string $datastore datastore file path. | ||
97 | * @param boolean $isLoggedIn is the user logged in? | ||
98 | * @param boolean $hidePublicLinks if true all links are private. | ||
99 | */ | ||
100 | public function __construct( | ||
101 | $datastore, | ||
102 | $isLoggedIn, | ||
103 | $hidePublicLinks | ||
104 | ) { | ||
105 | |||
106 | $this->datastore = $datastore; | ||
107 | $this->loggedIn = $isLoggedIn; | ||
108 | $this->hidePublicLinks = $hidePublicLinks; | ||
109 | $this->check(); | ||
110 | $this->read(); | ||
111 | } | ||
112 | |||
113 | /** | ||
114 | * Countable - Counts elements of an object | ||
115 | */ | ||
116 | public function count() | ||
117 | { | ||
118 | return count($this->links); | ||
119 | } | ||
120 | |||
121 | /** | ||
122 | * ArrayAccess - Assigns a value to the specified offset | ||
123 | */ | ||
124 | public function offsetSet($offset, $value) | ||
125 | { | ||
126 | // TODO: use exceptions instead of "die" | ||
127 | if (!$this->loggedIn) { | ||
128 | die(t('You are not authorized to add a link.')); | ||
129 | } | ||
130 | if (!isset($value['id']) || empty($value['url'])) { | ||
131 | die(t('Internal Error: A link should always have an id and URL.')); | ||
132 | } | ||
133 | if (($offset !== null && !is_int($offset)) || !is_int($value['id'])) { | ||
134 | die(t('You must specify an integer as a key.')); | ||
135 | } | ||
136 | if ($offset !== null && $offset !== $value['id']) { | ||
137 | die(t('Array offset and link ID must be equal.')); | ||
138 | } | ||
139 | |||
140 | // If the link exists, we reuse the real offset, otherwise new entry | ||
141 | $existing = $this->getLinkOffset($offset); | ||
142 | if ($existing !== null) { | ||
143 | $offset = $existing; | ||
144 | } else { | ||
145 | $offset = count($this->links); | ||
146 | } | ||
147 | $this->links[$offset] = $value; | ||
148 | $this->urls[$value['url']] = $offset; | ||
149 | $this->ids[$value['id']] = $offset; | ||
150 | } | ||
151 | |||
152 | /** | ||
153 | * ArrayAccess - Whether or not an offset exists | ||
154 | */ | ||
155 | public function offsetExists($offset) | ||
156 | { | ||
157 | return array_key_exists($this->getLinkOffset($offset), $this->links); | ||
158 | } | ||
159 | |||
160 | /** | ||
161 | * ArrayAccess - Unsets an offset | ||
162 | */ | ||
163 | public function offsetUnset($offset) | ||
164 | { | ||
165 | if (!$this->loggedIn) { | ||
166 | // TODO: raise an exception | ||
167 | die('You are not authorized to delete a link.'); | ||
168 | } | ||
169 | $realOffset = $this->getLinkOffset($offset); | ||
170 | $url = $this->links[$realOffset]['url']; | ||
171 | unset($this->urls[$url]); | ||
172 | unset($this->ids[$realOffset]); | ||
173 | unset($this->links[$realOffset]); | ||
174 | } | ||
175 | |||
176 | /** | ||
177 | * ArrayAccess - Returns the value at specified offset | ||
178 | */ | ||
179 | public function offsetGet($offset) | ||
180 | { | ||
181 | $realOffset = $this->getLinkOffset($offset); | ||
182 | return isset($this->links[$realOffset]) ? $this->links[$realOffset] : null; | ||
183 | } | ||
184 | |||
185 | /** | ||
186 | * Iterator - Returns the current element | ||
187 | */ | ||
188 | public function current() | ||
189 | { | ||
190 | return $this[$this->keys[$this->position]]; | ||
191 | } | ||
192 | |||
193 | /** | ||
194 | * Iterator - Returns the key of the current element | ||
195 | */ | ||
196 | public function key() | ||
197 | { | ||
198 | return $this->keys[$this->position]; | ||
199 | } | ||
200 | |||
201 | /** | ||
202 | * Iterator - Moves forward to next element | ||
203 | */ | ||
204 | public function next() | ||
205 | { | ||
206 | ++$this->position; | ||
207 | } | ||
208 | |||
209 | /** | ||
210 | * Iterator - Rewinds the Iterator to the first element | ||
211 | * | ||
212 | * Entries are sorted by date (latest first) | ||
213 | */ | ||
214 | public function rewind() | ||
215 | { | ||
216 | $this->keys = array_keys($this->ids); | ||
217 | $this->position = 0; | ||
218 | } | ||
219 | |||
220 | /** | ||
221 | * Iterator - Checks if current position is valid | ||
222 | */ | ||
223 | public function valid() | ||
224 | { | ||
225 | return isset($this->keys[$this->position]); | ||
226 | } | ||
227 | |||
228 | /** | ||
229 | * Checks if the DB directory and file exist | ||
230 | * | ||
231 | * If no DB file is found, creates a dummy DB. | ||
232 | */ | ||
233 | private function check() | ||
234 | { | ||
235 | if (file_exists($this->datastore)) { | ||
236 | return; | ||
237 | } | ||
238 | |||
239 | // Create a dummy database for example | ||
240 | $this->links = array(); | ||
241 | $link = array( | ||
242 | 'id' => 1, | ||
243 | 'title' => t('The personal, minimalist, super-fast, database free, bookmarking service'), | ||
244 | 'url' => 'https://shaarli.readthedocs.io', | ||
245 | 'description' => t( | ||
246 | 'Welcome to Shaarli! This is your first public bookmark. ' | ||
247 | . 'To edit or delete me, you must first login. | ||
248 | |||
249 | To learn how to use Shaarli, consult the link "Documentation" at the bottom of this page. | ||
250 | |||
251 | You use the community supported version of the original Shaarli project, by Sebastien Sauvage.' | ||
252 | ), | ||
253 | 'private' => 0, | ||
254 | 'created' => new DateTime(), | ||
255 | 'tags' => 'opensource software', | ||
256 | 'sticky' => false, | ||
257 | ); | ||
258 | $link['shorturl'] = link_small_hash($link['created'], $link['id']); | ||
259 | $this->links[1] = $link; | ||
260 | |||
261 | $link = array( | ||
262 | 'id' => 0, | ||
263 | 'title' => t('My secret stuff... - Pastebin.com'), | ||
264 | 'url' => 'http://sebsauvage.net/paste/?8434b27936c09649#bR7XsXhoTiLcqCpQbmOpBi3rq2zzQUC5hBI7ZT1O3x8=', | ||
265 | 'description' => t('Shhhh! I\'m a private link only YOU can see. You can delete me too.'), | ||
266 | 'private' => 1, | ||
267 | 'created' => new DateTime('1 minute ago'), | ||
268 | 'tags' => 'secretstuff', | ||
269 | 'sticky' => false, | ||
270 | ); | ||
271 | $link['shorturl'] = link_small_hash($link['created'], $link['id']); | ||
272 | $this->links[0] = $link; | ||
273 | |||
274 | // Write database to disk | ||
275 | $this->write(); | ||
276 | } | ||
277 | |||
278 | /** | ||
279 | * Reads database from disk to memory | ||
280 | */ | ||
281 | private function read() | ||
282 | { | ||
283 | // Public links are hidden and user not logged in => nothing to show | ||
284 | if ($this->hidePublicLinks && !$this->loggedIn) { | ||
285 | $this->links = array(); | ||
286 | return; | ||
287 | } | ||
288 | |||
289 | $this->urls = []; | ||
290 | $this->ids = []; | ||
291 | $this->links = FileUtils::readFlatDB($this->datastore, []); | ||
292 | |||
293 | $toremove = array(); | ||
294 | foreach ($this->links as $key => &$link) { | ||
295 | if (!$this->loggedIn && $link['private'] != 0) { | ||
296 | // Transition for not upgraded databases. | ||
297 | unset($this->links[$key]); | ||
298 | continue; | ||
299 | } | ||
300 | |||
301 | // Sanitize data fields. | ||
302 | sanitizeLink($link); | ||
303 | |||
304 | // Remove private tags if the user is not logged in. | ||
305 | if (!$this->loggedIn) { | ||
306 | $link['tags'] = preg_replace('/(^|\s+)\.[^($|\s)]+\s*/', ' ', $link['tags']); | ||
307 | } | ||
308 | |||
309 | $link['real_url'] = $link['url']; | ||
310 | |||
311 | $link['sticky'] = isset($link['sticky']) ? $link['sticky'] : false; | ||
312 | |||
313 | // To be able to load links before running the update, and prepare the update | ||
314 | if (!isset($link['created'])) { | ||
315 | $link['id'] = $link['linkdate']; | ||
316 | $link['created'] = DateTime::createFromFormat(self::LINK_DATE_FORMAT, $link['linkdate']); | ||
317 | if (!empty($link['updated'])) { | ||
318 | $link['updated'] = DateTime::createFromFormat(self::LINK_DATE_FORMAT, $link['updated']); | ||
319 | } | ||
320 | $link['shorturl'] = smallHash($link['linkdate']); | ||
321 | } | ||
322 | |||
323 | $this->urls[$link['url']] = $key; | ||
324 | $this->ids[$link['id']] = $key; | ||
325 | } | ||
326 | } | ||
327 | |||
328 | /** | ||
329 | * Saves the database from memory to disk | ||
330 | * | ||
331 | * @throws IOException the datastore is not writable | ||
332 | */ | ||
333 | private function write() | ||
334 | { | ||
335 | $this->reorder(); | ||
336 | FileUtils::writeFlatDB($this->datastore, $this->links); | ||
337 | } | ||
338 | |||
339 | /** | ||
340 | * Saves the database from memory to disk | ||
341 | * | ||
342 | * @param string $pageCacheDir page cache directory | ||
343 | */ | ||
344 | public function save($pageCacheDir) | ||
345 | { | ||
346 | if (!$this->loggedIn) { | ||
347 | // TODO: raise an Exception instead | ||
348 | die('You are not authorized to change the database.'); | ||
349 | } | ||
350 | |||
351 | $this->write(); | ||
352 | |||
353 | invalidateCaches($pageCacheDir); | ||
354 | } | ||
355 | |||
356 | /** | ||
357 | * Returns the link for a given URL, or False if it does not exist. | ||
358 | * | ||
359 | * @param string $url URL to search for | ||
360 | * | ||
361 | * @return mixed the existing link if it exists, else 'false' | ||
362 | */ | ||
363 | public function getLinkFromUrl($url) | ||
364 | { | ||
365 | if (isset($this->urls[$url])) { | ||
366 | return $this->links[$this->urls[$url]]; | ||
367 | } | ||
368 | return false; | ||
369 | } | ||
370 | |||
371 | /** | ||
372 | * Returns the shaare corresponding to a smallHash. | ||
373 | * | ||
374 | * @param string $request QUERY_STRING server parameter. | ||
375 | * | ||
376 | * @return array $filtered array containing permalink data. | ||
377 | * | ||
378 | * @throws LinkNotFoundException if the smallhash is malformed or doesn't match any link. | ||
379 | */ | ||
380 | public function filterHash($request) | ||
381 | { | ||
382 | $request = substr($request, 0, 6); | ||
383 | $linkFilter = new LinkFilter($this->links); | ||
384 | return $linkFilter->filter(LinkFilter::$FILTER_HASH, $request); | ||
385 | } | ||
386 | |||
387 | /** | ||
388 | * Returns the list of articles for a given day. | ||
389 | * | ||
390 | * @param string $request day to filter. Format: YYYYMMDD. | ||
391 | * | ||
392 | * @return array list of shaare found. | ||
393 | */ | ||
394 | public function filterDay($request) | ||
395 | { | ||
396 | $linkFilter = new LinkFilter($this->links); | ||
397 | return $linkFilter->filter(LinkFilter::$FILTER_DAY, $request); | ||
398 | } | ||
399 | |||
400 | /** | ||
401 | * Filter links according to search parameters. | ||
402 | * | ||
403 | * @param array $filterRequest Search request content. Supported keys: | ||
404 | * - searchtags: list of tags | ||
405 | * - searchterm: term search | ||
406 | * @param bool $casesensitive Optional: Perform case sensitive filter | ||
407 | * @param string $visibility return only all/private/public links | ||
408 | * @param bool $untaggedonly return only untagged links | ||
409 | * | ||
410 | * @return array filtered links, all links if no suitable filter was provided. | ||
411 | */ | ||
412 | public function filterSearch( | ||
413 | $filterRequest = array(), | ||
414 | $casesensitive = false, | ||
415 | $visibility = 'all', | ||
416 | $untaggedonly = false | ||
417 | ) { | ||
418 | |||
419 | // Filter link database according to parameters. | ||
420 | $searchtags = isset($filterRequest['searchtags']) ? escape($filterRequest['searchtags']) : ''; | ||
421 | $searchterm = isset($filterRequest['searchterm']) ? escape($filterRequest['searchterm']) : ''; | ||
422 | |||
423 | // Search tags + fullsearch - blank string parameter will return all links. | ||
424 | $type = LinkFilter::$FILTER_TAG | LinkFilter::$FILTER_TEXT; // == "vuotext" | ||
425 | $request = [$searchtags, $searchterm]; | ||
426 | |||
427 | $linkFilter = new LinkFilter($this); | ||
428 | return $linkFilter->filter($type, $request, $casesensitive, $visibility, $untaggedonly); | ||
429 | } | ||
430 | |||
431 | /** | ||
432 | * Returns the list tags appearing in the links with the given tags | ||
433 | * | ||
434 | * @param array $filteringTags tags selecting the links to consider | ||
435 | * @param string $visibility process only all/private/public links | ||
436 | * | ||
437 | * @return array tag => linksCount | ||
438 | */ | ||
439 | public function linksCountPerTag($filteringTags = [], $visibility = 'all') | ||
440 | { | ||
441 | $links = $this->filterSearch(['searchtags' => $filteringTags], false, $visibility); | ||
442 | $tags = []; | ||
443 | $caseMapping = []; | ||
444 | foreach ($links as $link) { | ||
445 | foreach (preg_split('/\s+/', $link['tags'], 0, PREG_SPLIT_NO_EMPTY) as $tag) { | ||
446 | if (empty($tag)) { | ||
447 | continue; | ||
448 | } | ||
449 | // The first case found will be displayed. | ||
450 | if (!isset($caseMapping[strtolower($tag)])) { | ||
451 | $caseMapping[strtolower($tag)] = $tag; | ||
452 | $tags[$caseMapping[strtolower($tag)]] = 0; | ||
453 | } | ||
454 | $tags[$caseMapping[strtolower($tag)]]++; | ||
455 | } | ||
456 | } | ||
457 | |||
458 | /* | ||
459 | * Formerly used arsort(), which doesn't define the sort behaviour for equal values. | ||
460 | * Also, this function doesn't produce the same result between PHP 5.6 and 7. | ||
461 | * | ||
462 | * So we now use array_multisort() to sort tags by DESC occurrences, | ||
463 | * then ASC alphabetically for equal values. | ||
464 | * | ||
465 | * @see https://github.com/shaarli/Shaarli/issues/1142 | ||
466 | */ | ||
467 | $keys = array_keys($tags); | ||
468 | $tmpTags = array_combine($keys, $keys); | ||
469 | array_multisort($tags, SORT_DESC, $tmpTags, SORT_ASC, $tags); | ||
470 | return $tags; | ||
471 | } | ||
472 | |||
473 | /** | ||
474 | * Rename or delete a tag across all links. | ||
475 | * | ||
476 | * @param string $from Tag to rename | ||
477 | * @param string $to New tag. If none is provided, the from tag will be deleted | ||
478 | * | ||
479 | * @return array|bool List of altered links or false on error | ||
480 | */ | ||
481 | public function renameTag($from, $to) | ||
482 | { | ||
483 | if (empty($from)) { | ||
484 | return false; | ||
485 | } | ||
486 | $delete = empty($to); | ||
487 | // True for case-sensitive tag search. | ||
488 | $linksToAlter = $this->filterSearch(['searchtags' => $from], true); | ||
489 | foreach ($linksToAlter as $key => &$value) { | ||
490 | $tags = preg_split('/\s+/', trim($value['tags'])); | ||
491 | if (($pos = array_search($from, $tags)) !== false) { | ||
492 | if ($delete) { | ||
493 | unset($tags[$pos]); // Remove tag. | ||
494 | } else { | ||
495 | $tags[$pos] = trim($to); | ||
496 | } | ||
497 | $value['tags'] = trim(implode(' ', array_unique($tags))); | ||
498 | $this[$value['id']] = $value; | ||
499 | } | ||
500 | } | ||
501 | |||
502 | return $linksToAlter; | ||
503 | } | ||
504 | |||
505 | /** | ||
506 | * Returns the list of days containing articles (oldest first) | ||
507 | * Output: An array containing days (in format YYYYMMDD). | ||
508 | */ | ||
509 | public function days() | ||
510 | { | ||
511 | $linkDays = array(); | ||
512 | foreach ($this->links as $link) { | ||
513 | $linkDays[$link['created']->format('Ymd')] = 0; | ||
514 | } | ||
515 | $linkDays = array_keys($linkDays); | ||
516 | sort($linkDays); | ||
517 | |||
518 | return $linkDays; | ||
519 | } | ||
520 | |||
521 | /** | ||
522 | * Reorder links by creation date (newest first). | ||
523 | * | ||
524 | * Also update the urls and ids mapping arrays. | ||
525 | * | ||
526 | * @param string $order ASC|DESC | ||
527 | */ | ||
528 | public function reorder($order = 'DESC') | ||
529 | { | ||
530 | $order = $order === 'ASC' ? -1 : 1; | ||
531 | // Reorder array by dates. | ||
532 | usort($this->links, function ($a, $b) use ($order) { | ||
533 | if (isset($a['sticky']) && isset($b['sticky']) && $a['sticky'] !== $b['sticky']) { | ||
534 | return $a['sticky'] ? -1 : 1; | ||
535 | } | ||
536 | return $a['created'] < $b['created'] ? 1 * $order : -1 * $order; | ||
537 | }); | ||
538 | |||
539 | $this->urls = []; | ||
540 | $this->ids = []; | ||
541 | foreach ($this->links as $key => $link) { | ||
542 | $this->urls[$link['url']] = $key; | ||
543 | $this->ids[$link['id']] = $key; | ||
544 | } | ||
545 | } | ||
546 | |||
547 | /** | ||
548 | * Return the next key for link creation. | ||
549 | * E.g. If the last ID is 597, the next will be 598. | ||
550 | * | ||
551 | * @return int next ID. | ||
552 | */ | ||
553 | public function getNextId() | ||
554 | { | ||
555 | if (!empty($this->ids)) { | ||
556 | return max(array_keys($this->ids)) + 1; | ||
557 | } | ||
558 | return 0; | ||
559 | } | ||
560 | |||
561 | /** | ||
562 | * Returns a link offset in links array from its unique ID. | ||
563 | * | ||
564 | * @param int $id Persistent ID of a link. | ||
565 | * | ||
566 | * @return int Real offset in local array, or null if doesn't exist. | ||
567 | */ | ||
568 | protected function getLinkOffset($id) | ||
569 | { | ||
570 | if (isset($this->ids[$id])) { | ||
571 | return $this->ids[$id]; | ||
572 | } | ||
573 | return null; | ||
574 | } | ||
575 | } | ||