]> git.wh0rd.org - tt-rss.git/blob - classes/rssutils.php
git.wh0rd.org / tt-rss.git / blob
? search:
summary | shortlog | log | commit | commitdiff | tree
blame | history | raw | HEAD
Merge branch 'master' of git.fakecake.org:tt-rss into pdo-experimental
[tt-rss.git] / classes / rssutils.php
1 <?php
2 class RSSUtils {
3 static function calculate_article_hash($article, $pluginhost) {
4 $tmp = "";
5
6 foreach ($article as $k => $v) {
7 if ($k != "feed" && isset($v)) {
8 $x = strip_tags(is_array($v) ? implode(",", $v) : $v);
9
10 //_debug("$k:" . sha1($x) . ":" . htmlspecialchars($x), true);
11
12 $tmp .= sha1("$k:" . sha1($x));
13 }
14 }
15
16 return sha1(implode(",", $pluginhost->get_plugin_names()) . $tmp);
17 }
18
19 static function update_feedbrowser_cache() {
20
21 $pdo = Db::pdo();
22
23 $sth = $pdo->query("SELECT feed_url, site_url, title, COUNT(id) AS subscribers
24 FROM ttrss_feeds WHERE feed_url NOT IN (SELECT feed_url FROM ttrss_feeds
25 WHERE private IS true OR auth_login != '' OR auth_pass != '' OR feed_url LIKE '%:%@%/%')
26 GROUP BY feed_url, site_url, title ORDER BY subscribers DESC LIMIT 1000");
27
28 $pdo->beginTransaction();
29
30 $pdo->query("DELETE FROM ttrss_feedbrowser_cache");
31
32 $count = 0;
33
34 while ($line = $sth->fetch()) {
35
36 $subscribers = $line["subscribers"];
37 $feed_url = $line["feed_url"];
38 $title = $line["title"];
39 $site_url = $line["site_url"];
40
41 $tmph = $pdo->prepare("SELECT subscribers FROM
42 ttrss_feedbrowser_cache WHERE feed_url = ?");
43 $tmph->execute([$feed_url]);
44
45 if (!$tmph->fetch()) {
46
47 $tmph = $pdo->prepare("INSERT INTO ttrss_feedbrowser_cache
48 (feed_url, site_url, title, subscribers)
49 VALUES
50 (?, ?, ?, ?)");
51
52 $tmph->execute([$feed_url, $site_url, $title, $subscribers]);
53
54 ++$count;
55
56 }
57
58 }
59
60 $pdo->commit();
61
62 return $count;
63
64 }
65
66 static function update_daemon_common($limit = DAEMON_FEED_LIMIT, $debug = true) {
67 $schema_version = get_schema_version();
68
69 if ($schema_version != SCHEMA_VERSION) {
70 die("Schema version is wrong, please upgrade the database.\n");
71 }
72
73 $pdo = Db::pdo();
74
75 if (!SINGLE_USER_MODE && DAEMON_UPDATE_LOGIN_LIMIT > 0) {
76 if (DB_TYPE == "pgsql") {
77 $login_thresh_qpart = "AND ttrss_users.last_login >= NOW() - INTERVAL '".DAEMON_UPDATE_LOGIN_LIMIT." days'";
78 } else {
79 $login_thresh_qpart = "AND ttrss_users.last_login >= DATE_SUB(NOW(), INTERVAL ".DAEMON_UPDATE_LOGIN_LIMIT." DAY)";
80 }
81 } else {
82 $login_thresh_qpart = "";
83 }
84
85 if (DB_TYPE == "pgsql") {
86 $update_limit_qpart = "AND ((
87 ttrss_feeds.update_interval = 0
88 AND ttrss_user_prefs.value != '-1'
89 AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_user_prefs.value || ' minutes') AS INTERVAL)
90 ) OR (
91 ttrss_feeds.update_interval > 0
92 AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_feeds.update_interval || ' minutes') AS INTERVAL)
93 ) OR (ttrss_feeds.last_updated IS NULL
94 AND ttrss_user_prefs.value != '-1')
95 OR (last_updated = '1970-01-01 00:00:00'
96 AND ttrss_user_prefs.value != '-1'))";
97 } else {
98 $update_limit_qpart = "AND ((
99 ttrss_feeds.update_interval = 0
100 AND ttrss_user_prefs.value != '-1'
101 AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL CONVERT(ttrss_user_prefs.value, SIGNED INTEGER) MINUTE)
102 ) OR (
103 ttrss_feeds.update_interval > 0
104 AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL ttrss_feeds.update_interval MINUTE)
105 ) OR (ttrss_feeds.last_updated IS NULL
106 AND ttrss_user_prefs.value != '-1')
107 OR (last_updated = '1970-01-01 00:00:00'
108 AND ttrss_user_prefs.value != '-1'))";
109 }
110
111 // Test if feed is currently being updated by another process.
112 if (DB_TYPE == "pgsql") {
113 $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < NOW() - INTERVAL '10 minutes')";
114 } else {
115 $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < DATE_SUB(NOW(), INTERVAL 10 MINUTE))";
116 }
117
118 $query_limit = $limit ? sprintf("LIMIT %d", $limit) : "";
119
120 // Update the least recently updated feeds first
121 $query_order = "ORDER BY last_updated";
122 if (DB_TYPE == "pgsql") $query_order .= " NULLS FIRST";
123
124 $query = "SELECT DISTINCT ttrss_feeds.feed_url, ttrss_feeds.last_updated
125 FROM
126 ttrss_feeds, ttrss_users, ttrss_user_prefs
127 WHERE
128 ttrss_feeds.owner_uid = ttrss_users.id
129 AND ttrss_user_prefs.profile IS NULL
130 AND ttrss_users.id = ttrss_user_prefs.owner_uid
131 AND ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL'
132 $login_thresh_qpart $update_limit_qpart
133 $updstart_thresh_qpart
134 $query_order $query_limit";
135
136 $res = $pdo->query($query);
137
138 $feeds_to_update = array();
139 while ($line = $res->fetch()) {
140 array_push($feeds_to_update, $line['feed_url']);
141 }
142
143 if ($debug) _debug(sprintf("Scheduled %d feeds to update...", count($feeds_to_update)));
144
145 // Update last_update_started before actually starting the batch
146 // in order to minimize collision risk for parallel daemon tasks
147 if (count($feeds_to_update) > 0) {
148 $feeds_qmarks = arr_qmarks($feeds_to_update);
149
150 $tmph = $pdo->prepare("UPDATE ttrss_feeds SET last_update_started = NOW()
151 WHERE feed_url IN ($feeds_qmarks)");
152 $tmph->execute($feeds_to_update);
153 }
154
155 $nf = 0;
156 $bstarted = microtime(true);
157
158 $batch_owners = array();
159
160 // since we have the data cached, we can deal with other feeds with the same url
161 $usth = $pdo->prepare("SELECT DISTINCT ttrss_feeds.id,last_updated,ttrss_feeds.owner_uid
162 FROM ttrss_feeds, ttrss_users, ttrss_user_prefs WHERE
163 ttrss_user_prefs.owner_uid = ttrss_feeds.owner_uid AND
164 ttrss_users.id = ttrss_user_prefs.owner_uid AND
165 ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL' AND
166 ttrss_user_prefs.profile IS NULL AND
167 feed_url = ?
168 $update_limit_qpart
169 $login_thresh_qpart
170 ORDER BY ttrss_feeds.id $query_limit");
171
172 foreach ($feeds_to_update as $feed) {
173 if($debug) _debug("Base feed: $feed");
174
175 $usth->execute([$feed]);
176 //update_rss_feed($line["id"], true);
177
178 if ($tline = $usth->fetch()) {
179 if ($debug) _debug(" => " . $tline["last_updated"] . ", " . $tline["id"] . " " . $tline["owner_uid"]);
180
181 if (array_search($tline["owner_uid"], $batch_owners) === FALSE)
182 array_push($batch_owners, $tline["owner_uid"]);
183
184 $fstarted = microtime(true);
185 RSSUtils::update_rss_feed($tline["id"], true, false);
186 _debug_suppress(false);
187
188 _debug(sprintf(" %.4f (sec)", microtime(true) - $fstarted));
189
190 ++$nf;
191 }
192 }
193
194 if ($nf > 0) {
195 _debug(sprintf("Processed %d feeds in %.4f (sec), %.4f (sec/feed avg)", $nf,
196 microtime(true) - $bstarted, (microtime(true) - $bstarted) / $nf));
197 }
198
199 foreach ($batch_owners as $owner_uid) {
200 _debug("Running housekeeping tasks for user $owner_uid...");
201
202 RSSUtils::housekeeping_user($owner_uid);
203 }
204
205 // Send feed digests by email if needed.
206 Digest::send_headlines_digests($debug);
207
208 return $nf;
209 }
210
211 // this is used when subscribing
212 static function set_basic_feed_info($feed) {
213
214 $pdo = Db::pdo();
215
216 $sth = $pdo->prepare("SELECT owner_uid,feed_url,auth_pass,auth_login,auth_pass_encrypted
217 FROM ttrss_feeds WHERE id = ?");
218 $sth->execute([$feed]);
219
220 if ($row = $sth->fetch()) {
221
222 $owner_uid = $row["owner_uid"];
223
224 $auth_pass_encrypted = sql_bool_to_bool($row["auth_pass_encrypted"]);
225
226 $auth_login = $row["auth_login"];
227 $auth_pass = $row["auth_pass"];
228
229 if ($auth_pass_encrypted && function_exists("mcrypt_decrypt")) {
230 require_once "crypt.php";
231 $auth_pass = decrypt_string($auth_pass);
232 }
233
234 $fetch_url = $row["feed_url"];
235
236 $pluginhost = new PluginHost();
237 $user_plugins = get_pref("_ENABLED_PLUGINS", $owner_uid);
238
239 $pluginhost->load(PLUGINS, PluginHost::KIND_ALL);
240 $pluginhost->load($user_plugins, PluginHost::KIND_USER, $owner_uid);
241 $pluginhost->load_data();
242
243 $basic_info = array();
244 foreach ($pluginhost->get_hooks(PluginHost::HOOK_FEED_BASIC_INFO) as $plugin) {
245 $basic_info = $plugin->hook_feed_basic_info($basic_info, $fetch_url, $owner_uid, $feed, $auth_login, $auth_pass);
246 }
247
248 if (!$basic_info) {
249 $feed_data = fetch_file_contents($fetch_url, false,
250 $auth_login, $auth_pass, false,
251 FEED_FETCH_TIMEOUT,
252 0);
253
254 global $fetch_curl_used;
255
256 if (!$fetch_curl_used) {
257 $tmp = @gzdecode($feed_data);
258
259 if ($tmp) $feed_data = $tmp;
260 }
261
262 $feed_data = trim($feed_data);
263
264 $rss = new FeedParser($feed_data);
265 $rss->init();
266
267 if (!$rss->error()) {
268 $basic_info = array(
269 'title' => mb_substr($rss->get_title(), 0, 199),
270 'site_url' => mb_substr(rewrite_relative_url($fetch_url, $rss->get_link()), 0, 245)
271 );
272 }
273 }
274
275 if ($basic_info && is_array($basic_info)) {
276 $sth = $pdo->prepare("SELECT title, site_url FROM ttrss_feeds WHERE id = ?");
277 $sth->execute([$feed]);
278
279 if ($row = $sth->fetch()) {
280
281 $registered_title = $row["title"];
282 $orig_site_url = $row["site_url"];
283
284 if ($basic_info['title'] && (!$registered_title || $registered_title == "[Unknown]")) {
285
286 $sth = $pdo->prepare("UPDATE ttrss_feeds SET
287 title = ? WHERE id = ?");
288 $sth->execute([$basic_info['title'], $feed]);
289 }
290
291 if ($basic_info['site_url'] && $orig_site_url != $basic_info['site_url']) {
292 $sth = $pdo->prepare("UPDATE ttrss_feeds SET
293 site_url = ? WHERE id = ?");
294 $sth->execute([$basic_info['site_url'], $feed]);
295 }
296
297 }
298 }
299 }
300 }
301
302 /**
303 * @SuppressWarnings(PHPMD.UnusedFormalParameter)
304 */
305 static function update_rss_feed($feed, $no_cache = false) {
306
307 $debug_enabled = defined('DAEMON_EXTENDED_DEBUG') || $_REQUEST['xdebug'];
308
309 _debug_suppress(!$debug_enabled);
310 _debug("start", $debug_enabled);
311
312 $pdo = Db::pdo();
313
314 $sth = $pdo->prepare("SELECT title FROM ttrss_feeds WHERE id = ?");
315 $sth->execute([$feed]);
316
317 if (!$row = $sth->fetch()) {
318 _debug("feed $feed NOT FOUND/SKIPPED", $debug_enabled);
319 user_error("Attempt to update unknown/invalid feed $feed", E_USER_WARNING);
320 return false;
321 }
322
323 $title = $row["title"];
324
325 // feed was batch-subscribed or something, we need to get basic info
326 // this is not optimal currently as it fetches stuff separately TODO: optimize
327 if ($title == "[Unknown]") {
328 _debug("setting basic feed info for $feed...");
329 RSSUtils::set_basic_feed_info($feed);
330 }
331
332 $sth = $pdo->prepare("SELECT id,update_interval,auth_login,
333 feed_url,auth_pass,cache_images,
334 mark_unread_on_update, owner_uid,
335 auth_pass_encrypted, feed_language,
336 last_modified,
337 ".SUBSTRING_FOR_DATE."(last_unconditional, 1, 19) AS last_unconditional
338 FROM ttrss_feeds WHERE id = ?");
339 $sth->execute([$feed]);
340
341 if ($row = $sth->fetch()) {
342
343 $owner_uid = $row["owner_uid"];
344 $mark_unread_on_update = sql_bool_to_bool($row["mark_unread_on_update"]);
345 $auth_pass_encrypted = sql_bool_to_bool($row["auth_pass_encrypted"]);
346
347 $sth = $pdo->prepare("UPDATE ttrss_feeds SET last_update_started = NOW()
348 WHERE id = ?");
349 $sth->execute([$feed]);
350
351 $auth_login = $row["auth_login"];
352 $auth_pass = $row["auth_pass"];
353
354 if ($auth_pass_encrypted && function_exists("mcrypt_decrypt")) {
355 require_once "crypt.php";
356 $auth_pass = decrypt_string($auth_pass);
357 }
358
359 $stored_last_modified = $row["last_modified"];
360 $last_unconditional = $row["last_unconditional"];
361 $cache_images = sql_bool_to_bool($row["cache_images"]);
362 $fetch_url = $row["feed_url"];
363 $feed_language = mb_strtolower($row["feed_language"]);
364 if (!$feed_language) $feed_language = 'english';
365
366 } else {
367 return false;
368 }
369
370 $date_feed_processed = date('Y-m-d H:i');
371
372 $cache_filename = CACHE_DIR . "/simplepie/" . sha1($fetch_url) . ".xml";
373
374 $pluginhost = new PluginHost();
375 $pluginhost->set_debug($debug_enabled);
376 $user_plugins = get_pref("_ENABLED_PLUGINS", $owner_uid);
377
378 $pluginhost->load(PLUGINS, PluginHost::KIND_ALL);
379 $pluginhost->load($user_plugins, PluginHost::KIND_USER, $owner_uid);
380 $pluginhost->load_data();
381
382 $rss_hash = false;
383
384 $force_refetch = isset($_REQUEST["force_refetch"]);
385 $feed_data = "";
386
387 foreach ($pluginhost->get_hooks(PluginHost::HOOK_FETCH_FEED) as $plugin) {
388 $feed_data = $plugin->hook_fetch_feed($feed_data, $fetch_url, $owner_uid, $feed, 0, $auth_login, $auth_pass);
389 }
390
391 // try cache
392 if (!$feed_data &&
393 file_exists($cache_filename) &&
394 is_readable($cache_filename) &&
395 !$auth_login && !$auth_pass &&
396 filemtime($cache_filename) > time() - 30) {
397
398 _debug("using local cache [$cache_filename].", $debug_enabled);
399
400 @$feed_data = file_get_contents($cache_filename);
401
402 if ($feed_data) {
403 $rss_hash = sha1($feed_data);
404 }
405
406 } else {
407 _debug("local cache will not be used for this feed", $debug_enabled);
408 }
409
410 global $fetch_last_modified;
411
412 // fetch feed from source
413 if (!$feed_data) {
414 _debug("last unconditional update request: $last_unconditional");
415
416 if (ini_get("open_basedir") && function_exists("curl_init")) {
417 _debug("not using CURL due to open_basedir restrictions");
418 }
419
420 if (time() - strtotime($last_unconditional) > MAX_CONDITIONAL_INTERVAL) {
421 _debug("maximum allowed interval for conditional requests exceeded, forcing refetch");
422
423 $force_refetch = true;
424 } else {
425 _debug("stored last modified for conditional request: $stored_last_modified", $debug_enabled);
426 }
427
428 _debug("fetching [$fetch_url] (force_refetch: $force_refetch)...", $debug_enabled);
429
430 $feed_data = fetch_file_contents([
431 "url" => $fetch_url,
432 "login" => $auth_login,
433 "pass" => $auth_pass,
434 "timeout" => $no_cache ? FEED_FETCH_NO_CACHE_TIMEOUT : FEED_FETCH_TIMEOUT,
435 "last_modified" => $force_refetch ? "" : $stored_last_modified
436 ]);
437
438 global $fetch_curl_used;
439
440 if (!$fetch_curl_used) {
441 $tmp = @gzdecode($feed_data);
442
443 if ($tmp) $feed_data = $tmp;
444 }
445
446 $feed_data = trim($feed_data);
447
448 _debug("fetch done.", $debug_enabled);
449 _debug("source last modified: " . $fetch_last_modified, $debug_enabled);
450
451 if ($feed_data && $fetch_last_modified != $stored_last_modified) {
452 $sth = $pdo->prepare("UPDATE ttrss_feeds SET last_modified = ? WHERE id = ?");
453 $sth->execute([substr($fetch_last_modified, 0, 245), $feed]);
454 }
455
456 // cache vanilla feed data for re-use
457 if ($feed_data && !$auth_pass && !$auth_login && is_writable(CACHE_DIR . "/simplepie")) {
458 $new_rss_hash = sha1($feed_data);
459
460 if ($new_rss_hash != $rss_hash) {
461 _debug("saving $cache_filename", $debug_enabled);
462 @file_put_contents($cache_filename, $feed_data);
463 }
464 }
465 }
466
467 if (!$feed_data) {
468 global $fetch_last_error;
469 global $fetch_last_error_code;
470
471 _debug("unable to fetch: $fetch_last_error [$fetch_last_error_code]", $debug_enabled);
472
473 // If-Modified-Since
474 if ($fetch_last_error_code != 304) {
475 $error_message = $fetch_last_error;
476 } else {
477 _debug("source claims data not modified, nothing to do.", $debug_enabled);
478 $error_message = "";
479 }
480
481 $sth = $pdo->prepare("UPDATE ttrss_feeds SET last_error = ?,
482 last_updated = NOW() WHERE id = ?");
483 $sth->execute([$error_message, $feed]);
484
485 return;
486 }
487
488 foreach ($pluginhost->get_hooks(PluginHost::HOOK_FEED_FETCHED) as $plugin) {
489 $feed_data = $plugin->hook_feed_fetched($feed_data, $fetch_url, $owner_uid, $feed);
490 }
491
492 $rss = new FeedParser($feed_data);
493 $rss->init();
494
495 $feed = $feed;
496
497 if (!$rss->error()) {
498
499 // We use local pluginhost here because we need to load different per-user feed plugins
500 $pluginhost->run_hooks(PluginHost::HOOK_FEED_PARSED, "hook_feed_parsed", $rss);
501
502 _debug("language: $feed_language", $debug_enabled);
503 _debug("processing feed data...", $debug_enabled);
504
505 if (DB_TYPE == "pgsql") {
506 $favicon_interval_qpart = "favicon_last_checked < NOW() - INTERVAL '12 hour'";
507 } else {
508 $favicon_interval_qpart = "favicon_last_checked < DATE_SUB(NOW(), INTERVAL 12 HOUR)";
509 }
510
511 $sth = $pdo->prepare("SELECT owner_uid,favicon_avg_color,
512 (favicon_last_checked IS NULL OR $favicon_interval_qpart) AS
513 favicon_needs_check
514 FROM ttrss_feeds WHERE id = ?");
515 $sth->execute([$feed]);
516
517 if ($row = $sth->fetch()) {
518 $favicon_needs_check = sql_bool_to_bool($row["favicon_needs_check"]);
519 $favicon_avg_color = $row["favicon_avg_color"];
520 $owner_uid = $row["owner_uid"];
521 } else {
522 return false;
523 }
524
525 $site_url = mb_substr(rewrite_relative_url($fetch_url, $rss->get_link()), 0, 245);
526
527 _debug("site_url: $site_url", $debug_enabled);
528 _debug("feed_title: " . $rss->get_title(), $debug_enabled);
529
530 if ($favicon_needs_check || $force_refetch) {
531
532 /* terrible hack: if we crash on floicon shit here, we won't check
533 * the icon avgcolor again (unless the icon got updated) */
534
535 $favicon_file = ICONS_DIR . "/$feed.ico";
536 $favicon_modified = @filemtime($favicon_file);
537
538 _debug("checking favicon...", $debug_enabled);
539
540 RSSUtils::check_feed_favicon($site_url, $feed);
541 $favicon_modified_new = @filemtime($favicon_file);
542
543 if ($favicon_modified_new > $favicon_modified)
544 $favicon_avg_color = '';
545
546 $favicon_colorstring = "";
547 if (file_exists($favicon_file) && function_exists("imagecreatefromstring") && $favicon_avg_color == '') {
548 require_once "colors.php";
549
550 $sth = $pdo->prepare("UPDATE ttrss_feeds SET favicon_avg_color = 'fail' WHERE
551 id = ?");
552 $sth->execute([$feed]);
553
554 $favicon_color = calculate_avg_color($favicon_file);
555
556 $favicon_colorstring = ",favicon_avg_color = " . $pdo->quote($favicon_color);
557
558 } else if ($favicon_avg_color == 'fail') {
559 _debug("floicon failed on this file, not trying to recalculate avg color", $debug_enabled);
560 }
561
562 $sth = $pdo->prepare("UPDATE ttrss_feeds SET favicon_last_checked = NOW()
563 $favicon_colorstring WHERE id = ?");
564 $sth->execute([$feed]);
565 }
566
567 _debug("loading filters & labels...", $debug_enabled);
568
569 $filters = load_filters($feed, $owner_uid);
570
571 if ($debug_enabled) {
572 print_r($filters);
573 }
574
575 _debug("" . count($filters) . " filters loaded.", $debug_enabled);
576
577 $items = $rss->get_items();
578
579 if (!is_array($items)) {
580 _debug("no articles found.", $debug_enabled);
581
582 $sth = $pdo->prepare("UPDATE ttrss_feeds
583 SET last_updated = NOW(), last_unconditional = NOW(), last_error = '' WHERE id = ?");
584 $sth->execute([$feed]);
585
586 return true; // no articles
587 }
588
589 _debug("processing articles...", $debug_enabled);
590
591 $tstart = time();
592
593 foreach ($items as $item) {
594 $pdo->beginTransaction();
595
596 if ($_REQUEST['xdebug'] == 3) {
597 print_r($item);
598 }
599
600 if (ini_get("max_execution_time") > 0 && time() - $tstart >= ini_get("max_execution_time") * 0.7) {
601 _debug("looks like there's too many articles to process at once, breaking out", $debug_enabled);
602 $pdo->commit();
603 break;
604 }
605
606 $entry_guid = strip_tags($item->get_id());
607 if (!$entry_guid) $entry_guid = strip_tags($item->get_link());
608 if (!$entry_guid) $entry_guid = RSSUtils::make_guid_from_title($item->get_title());
609
610 if (!$entry_guid) {
611 $pdo->commit();
612 continue;
613 }
614
615 $entry_guid = "$owner_uid,$entry_guid";
616
617 $entry_guid_hashed = 'SHA1:' . sha1($entry_guid);
618
619 _debug("guid $entry_guid / $entry_guid_hashed", $debug_enabled);
620
621 $entry_timestamp = strip_tags($item->get_date());
622
623 _debug("orig date: " . $item->get_date(), $debug_enabled);
624
625 if ($entry_timestamp == -1 || !$entry_timestamp || $entry_timestamp > time()) {
626 $entry_timestamp = time();
627 }
628
629 $entry_timestamp_fmt = strftime("%Y/%m/%d %H:%M:%S", $entry_timestamp);
630
631 _debug("date $entry_timestamp [$entry_timestamp_fmt]", $debug_enabled);
632
633 $entry_title = strip_tags($item->get_title());
634
635 $entry_link = rewrite_relative_url($site_url, $item->get_link());
636
637 _debug("title $entry_title", $debug_enabled);
638 _debug("link $entry_link", $debug_enabled);
639
640 if (!$entry_title) $entry_title = date("Y-m-d H:i:s", $entry_timestamp);;
641
642 $entry_content = $item->get_content();
643 if (!$entry_content) $entry_content = $item->get_description();
644
645 if ($_REQUEST["xdebug"] == 2) {
646 print "content: ";
647 print htmlspecialchars($entry_content);
648 print "\n";
649 }
650
651 $entry_comments = mb_substr(strip_tags($item->get_comments_url()), 0, 245);
652 $num_comments = (int) $item->get_comments_count();
653
654 $entry_author = strip_tags($item->get_author());
655 $entry_guid = mb_substr($entry_guid, 0, 245);
656
657 _debug("author $entry_author", $debug_enabled);
658 _debug("num_comments: $num_comments", $debug_enabled);
659 _debug("looking for tags...", $debug_enabled);
660
661 // parse <category> entries into tags
662
663 $additional_tags = array();
664
665 $additional_tags_src = $item->get_categories();
666
667 if (is_array($additional_tags_src)) {
668 foreach ($additional_tags_src as $tobj) {
669 array_push($additional_tags, $tobj);
670 }
671 }
672
673 $entry_tags = array_unique($additional_tags);
674
675 for ($i = 0; $i < count($entry_tags); $i++)
676 $entry_tags[$i] = mb_strtolower($entry_tags[$i], 'utf-8');
677
678 _debug("tags found: " . join(",", $entry_tags), $debug_enabled);
679
680 _debug("done collecting data.", $debug_enabled);
681
682 $sth = $pdo->prepare("SELECT id, content_hash, lang FROM ttrss_entries
683 WHERE guid = ? OR guid = ?");
684 $sth->execute([$entry_guid, $entry_guid_hashed]);
685
686 if ($row = $sth->fetch()) {
687 $base_entry_id = $row["id"];
688 $entry_stored_hash = $row["content_hash"];
689 $article_labels = Article::get_article_labels($base_entry_id, $owner_uid);
690 $entry_language = $row["lang"];
691
692 $existing_tags = Article::get_article_tags($base_entry_id, $owner_uid);
693 $entry_tags = array_unique(array_merge($entry_tags, $existing_tags));
694 } else {
695 $base_entry_id = false;
696 $entry_stored_hash = "";
697 $article_labels = array();
698 $entry_language = "";
699 }
700
701 $article = array("owner_uid" => $owner_uid, // read only
702 "guid" => $entry_guid, // read only
703 "guid_hashed" => $entry_guid_hashed, // read only
704 "title" => $entry_title,
705 "content" => $entry_content,
706 "link" => $entry_link,
707 "labels" => $article_labels, // current limitation: can add labels to article, can't remove them
708 "tags" => $entry_tags,
709 "author" => $entry_author,
710 "force_catchup" => false, // ugly hack for the time being
711 "score_modifier" => 0, // no previous value, plugin should recalculate score modifier based on content if needed
712 "language" => $entry_language,
713 "num_comments" => $num_comments, // read only
714 "feed" => array("id" => $feed,
715 "fetch_url" => $fetch_url,
716 "site_url" => $site_url,
717 "cache_images" => $cache_images)
718 );
719
720 $entry_plugin_data = "";
721 $entry_current_hash = RSSUtils::calculate_article_hash($article, $pluginhost);
722
723 _debug("article hash: $entry_current_hash [stored=$entry_stored_hash]", $debug_enabled);
724
725 if ($entry_current_hash == $entry_stored_hash && !isset($_REQUEST["force_rehash"])) {
726 _debug("stored article seems up to date [IID: $base_entry_id], updating timestamp only", $debug_enabled);
727
728 // we keep encountering the entry in feeds, so we need to
729 // update date_updated column so that we don't get horrible
730 // dupes when the entry gets purged and reinserted again e.g.
731 // in the case of SLOW SLOW OMG SLOW updating feeds
732
733 $sth = $pdo->prepare("UPDATE ttrss_entries SET date_updated = NOW()
734 WHERE id = ?");
735 $sth->execute([$base_entry_id]);
736
737 $pdo->commit();
738 continue;
739 }
740
741 _debug("hash differs, applying plugin filters:", $debug_enabled);
742
743 foreach ($pluginhost->get_hooks(PluginHost::HOOK_ARTICLE_FILTER) as $plugin) {
744 _debug("... " . get_class($plugin), $debug_enabled);
745
746 $start = microtime(true);
747 $article = $plugin->hook_article_filter($article);
748
749 _debug("=== " . sprintf("%.4f (sec)", microtime(true) - $start), $debug_enabled);
750
751 $entry_plugin_data .= mb_strtolower(get_class($plugin)) . ",";
752 }
753
754 if ($_REQUEST["xdebug"] == 2) {
755 print "processed content: ";
756 print htmlspecialchars($article["content"]);
757 print "\n";
758 }
759
760 _debug("plugin data: $entry_plugin_data", $debug_enabled);
761
762 // Workaround: 4-byte unicode requires utf8mb4 in MySQL. See https://tt-rss.org/forum/viewtopic.php?f=1&t=3377&p=20077#p20077
763 if (DB_TYPE == "mysql") {
764 foreach ($article as $k => $v) {
765
766 // i guess we'll have to take the risk of 4byte unicode labels & tags here
767 if (is_string($article[$k])) {
768 $article[$k] = preg_replace('/[\x{10000}-\x{10FFFF}]/u', "\xEF\xBF\xBD", $v);
769 }
770 }
771 }
772
773 /* Collect article tags here so we could filter by them: */
774
775 $matched_rules = array();
776
777 $article_filters = RSSUtils::get_article_filters($filters, $article["title"],
778 $article["content"], $article["link"], $article["author"],
779 $article["tags"], $matched_rules);
780
781 if ($debug_enabled) {
782 _debug("matched filter rules: ", $debug_enabled);
783
784 if (count($matched_rules) != 0) {
785 print_r($matched_rules);
786 }
787
788 _debug("filter actions: ", $debug_enabled);
789
790 if (count($article_filters) != 0) {
791 print_r($article_filters);
792 }
793 }
794
795 $plugin_filter_names = RSSUtils::find_article_filters($article_filters, "plugin");
796 $plugin_filter_actions = $pluginhost->get_filter_actions();
797
798 if (count($plugin_filter_names) > 0) {
799 _debug("applying plugin filter actions...", $debug_enabled);
800
801 foreach ($plugin_filter_names as $pfn) {
802 list($pfclass,$pfaction) = explode(":", $pfn["param"]);
803
804 if (isset($plugin_filter_actions[$pfclass])) {
805 $plugin = $pluginhost->get_plugin($pfclass);
806
807 _debug("... $pfclass: $pfaction", $debug_enabled);
808
809 if ($plugin) {
810 $start = microtime(true);
811 $article = $plugin->hook_article_filter_action($article, $pfaction);
812
813 _debug("=== " . sprintf("%.4f (sec)", microtime(true) - $start), $debug_enabled);
814 } else {
815 _debug("??? $pfclass: plugin object not found.");
816 }
817 } else {
818 _debug("??? $pfclass: filter plugin not registered.");
819 }
820 }
821 }
822
823 $entry_tags = $article["tags"];
824 $entry_title = strip_tags($article["title"]);
825 $entry_author = mb_substr(strip_tags($article["author"]), 0, 245);
826 $entry_link = strip_tags($article["link"]);
827 $entry_content = $article["content"]; // escaped below
828 $entry_force_catchup = $article["force_catchup"];
829 $article_labels = $article["labels"];
830 $entry_score_modifier = (int) $article["score_modifier"];
831 $entry_language = $article["language"];
832
833 if ($debug_enabled) {
834 _debug("article labels:", $debug_enabled);
835
836 if (count($article_labels) != 0) {
837 print_r($article_labels);
838 }
839 }
840
841 _debug("force catchup: $entry_force_catchup");
842
843 if ($cache_images && is_writable(CACHE_DIR . '/images'))
844 RSSUtils::cache_media($entry_content, $site_url, $debug_enabled);
845
846 $csth = $pdo->prepare("SELECT id FROM ttrss_entries
847 WHERE guid = ? OR guid = ?");
848 $csth->execute([$entry_guid, $entry_guid_hashed]);
849
850 if (!$row = $csth->fetch()) {
851
852 _debug("base guid [$entry_guid or $entry_guid_hashed] not found, creating...", $debug_enabled);
853
854 // base post entry does not exist, create it
855
856 $usth = $pdo->prepare(
857 "INSERT INTO ttrss_entries
858 (title,
859 guid,
860 link,
861 updated,
862 content,
863 content_hash,
864 no_orig_date,
865 date_updated,
866 date_entered,
867 comments,
868 num_comments,
869 plugin_data,
870 lang,
871 author)
872 VALUES
873 (?, ?, ?, ?, ?, ?,
874 false,
875 NOW(),
876 ?, ?, ?, ?, ?, ?)");
877
878 $usth->execute([$entry_title,
879 $entry_guid_hashed,
880 $entry_link,
881 $entry_timestamp_fmt,
882 $entry_content,
883 $entry_current_hash,
884 $date_feed_processed,
885 $entry_comments,
886 $num_comments,
887 $entry_plugin_data,
888 $entry_language,
889 $entry_author]);
890
891 }
892
893 $csth->execute([$entry_guid, $entry_guid_hashed]);
894
895 $entry_ref_id = 0;
896 $entry_int_id = 0;
897
898 if ($row = $csth->fetch()) {
899
900 _debug("base guid found, checking for user record", $debug_enabled);
901
902 $ref_id = $row['id'];
903 $entry_ref_id = $ref_id;
904
905 if (RSSUtils::find_article_filter($article_filters, "filter")) {
906 $pdo->commit();
907 continue;
908 }
909
910 $score = RSSUtils::calculate_article_score($article_filters) + $entry_score_modifier;
911
912 _debug("initial score: $score [including plugin modifier: $entry_score_modifier]", $debug_enabled);
913
914 // check for user post link to main table
915
916 $sth = $pdo->prepare("SELECT ref_id, int_id FROM ttrss_user_entries WHERE
917 ref_id = ? AND owner_uid = ?");
918 $sth->execute([$ref_id, $owner_uid]);
919
920 // okay it doesn't exist - create user entry
921 if ($row = $sth->fetch()) {
922 $entry_ref_id = $row["ref_id"];
923 $entry_int_id = $row["int_id"];
924
925 _debug("user record FOUND: RID: $entry_ref_id, IID: $entry_int_id", $debug_enabled);
926 } else {
927
928 _debug("user record not found, creating...", $debug_enabled);
929
930 if ($score >= -500 && !RSSUtils::find_article_filter($article_filters, 'catchup') && !$entry_force_catchup) {
931 $unread = 1;
932 $last_read_qpart = null;
933 } else {
934 $unread = 0;
935 $last_read_qpart = 'NOW()';
936 }
937
938 if (RSSUtils::find_article_filter($article_filters, 'mark') || $score > 1000) {
939 $marked = 1;
940 } else {
941 $marked = 0;
942 }
943
944 if (RSSUtils::find_article_filter($article_filters, 'publish')) {
945 $published = 1;
946 } else {
947 $published = 0;
948 }
949
950 $last_marked = ($marked == 'true') ? 'NOW()' : null;
951 $last_published = ($published == 'true') ? 'NOW()' : null;
952
953 $sth = $pdo->prepare(
954 "INSERT INTO ttrss_user_entries
955 (ref_id, owner_uid, feed_id, unread, last_read, marked,
956 published, score, tag_cache, label_cache, uuid,
957 last_marked, last_published)
958 VALUES (?, ?, ?, ?, ?, ?, ?, ?, '', '', '', ?, ?)");
959
960 $sth->execute([$ref_id, $owner_uid, $feed, $unread, $last_read_qpart, $marked,
961 $published, $score, $last_marked, $last_published]);
962
963 $sth = $pdo->prepare("SELECT int_id FROM ttrss_user_entries WHERE
964 ref_id = ? AND owner_uid = ? AND
965 feed_id = ? LIMIT 1");
966
967 $sth->execute([$ref_id, $owner_uid, $feed]);
968
969 if ($row = $sth->fetch())
970 $entry_int_id = $row['int_id'];
971 }
972
973 _debug("resulting RID: $entry_ref_id, IID: $entry_int_id", $debug_enabled);
974
975 if (DB_TYPE == "pgsql") {
976 $tsvector_combined = mb_substr($entry_title . ' ' . strip_tags(str_replace('<', ' <', $entry_content)),
977 0, 1000000);
978
979 $tsvector_qpart = "tsvector_combined = to_tsvector('$feed_language', ".$pdo->quote($tsvector_combined)."),";
980
981 } else {
982 $tsvector_qpart = "";
983 }
984
985 $sth = $pdo->prepare("UPDATE ttrss_entries
986 SET title = ?,
987 content = ?,
988 content_hash = ?,
989 updated = ?,
990 $tsvector_qpart
991 num_comments = ?,
992 plugin_data = ?,
993 author = ?,
994 lang = ?
995 WHERE id = ?");
996
997 $sth->execute([$entry_title, $entry_content, $entry_current_hash, $entry_timestamp_fmt,
998 $num_comments, $entry_plugin_data, $entry_author, $entry_language, $ref_id]);
999
1000 // update aux data
1001 $sth = $pdo->prepare("UPDATE ttrss_user_entries
1002 SET score = ? WHERE ref_id = ?");
1003 $sth->execute([$score, $ref_id]);
1004
1005 if ($mark_unread_on_update) {
1006 _debug("article updated, marking unread as requested.", $debug_enabled);
1007
1008 $sth = $pdo->prepare("UPDATE ttrss_user_entries
1009 SET last_read = null, unread = true WHERE ref_id = ?");
1010 $sth->execute([$ref_id]);
1011 }
1012 }
1013
1014 _debug("assigning labels [other]...", $debug_enabled);
1015
1016 foreach ($article_labels as $label) {
1017 Labels::add_article($entry_ref_id, $label[1], $owner_uid);
1018 }
1019
1020 _debug("assigning labels [filters]...", $debug_enabled);
1021
1022 RSSUtils::assign_article_to_label_filters($entry_ref_id, $article_filters,
1023 $owner_uid, $article_labels);
1024
1025 _debug("looking for enclosures...", $debug_enabled);
1026
1027 // enclosures
1028
1029 $enclosures = array();
1030
1031 $encs = $item->get_enclosures();
1032
1033 if (is_array($encs)) {
1034 foreach ($encs as $e) {
1035 $e_item = array(
1036 rewrite_relative_url($site_url, $e->link),
1037 $e->type, $e->length, $e->title, $e->width, $e->height);
1038 array_push($enclosures, $e_item);
1039 }
1040 }
1041
1042 if ($cache_images && is_writable(CACHE_DIR . '/images'))
1043 RSSUtils::cache_enclosures($enclosures, $site_url, $debug_enabled);
1044
1045 if ($debug_enabled) {
1046 _debug("article enclosures:", $debug_enabled);
1047 print_r($enclosures);
1048 }
1049
1050 $esth = $pdo->prepare("SELECT id FROM ttrss_enclosures
1051 WHERE content_url = ? AND post_id = ?");
1052
1053 $usth = $pdo->prepare("INSERT INTO ttrss_enclosures
1054 (content_url, content_type, title, duration, post_id, width, height) VALUES
1055 (?, ?, ?, ?, ?, ?, ?)");
1056
1057 foreach ($enclosures as $enc) {
1058 $enc_url = $enc[0];
1059 $enc_type = $enc[1];
1060 $enc_dur = (int)$enc[2];
1061 $enc_title = $enc[3];
1062 $enc_width = intval($enc[4]);
1063 $enc_height = intval($enc[5]);
1064
1065 $esth->execute([$enc_url, $entry_ref_id]);
1066
1067 if (!$esth->fetch()) {
1068 $usth->execute([$enc_url, $enc_type, (string)$enc_title, $enc_dur, $entry_ref_id, $enc_width, $enc_height]);
1069 }
1070 }
1071
1072 // check for manual tags (we have to do it here since they're loaded from filters)
1073
1074 foreach ($article_filters as $f) {
1075 if ($f["type"] == "tag") {
1076
1077 $manual_tags = trim_array(explode(",", $f["param"]));
1078
1079 foreach ($manual_tags as $tag) {
1080 if (tag_is_valid($tag)) {
1081 array_push($entry_tags, $tag);
1082 }
1083 }
1084 }
1085 }
1086
1087 // Skip boring tags
1088
1089 $boring_tags = trim_array(explode(",", mb_strtolower(get_pref(
1090 'BLACKLISTED_TAGS', $owner_uid, ''), 'utf-8')));
1091
1092 $filtered_tags = array();
1093 $tags_to_cache = array();
1094
1095 if ($entry_tags && is_array($entry_tags)) {
1096 foreach ($entry_tags as $tag) {
1097 if (array_search($tag, $boring_tags) === false) {
1098 array_push($filtered_tags, $tag);
1099 }
1100 }
1101 }
1102
1103 $filtered_tags = array_unique($filtered_tags);
1104
1105 if ($debug_enabled) {
1106 _debug("filtered article tags:", $debug_enabled);
1107 print_r($filtered_tags);
1108 }
1109
1110 // Save article tags in the database
1111
1112 if (count($filtered_tags) > 0) {
1113
1114 $tsth = $pdo->prepare("SELECT id FROM ttrss_tags
1115 WHERE tag_name = ? AND post_int_id = ? AND
1116 owner_uid = ? LIMIT 1");
1117
1118 $usth = $pdo->prepare("INSERT INTO ttrss_tags
1119 (owner_uid,tag_name,post_int_id)
1120 VALUES (?, ?, ?)");
1121
1122 foreach ($filtered_tags as $tag) {
1123
1124 $tag = sanitize_tag($tag);
1125
1126 if (!tag_is_valid($tag)) continue;
1127
1128 $tsth->execute([$tag, $entry_int_id, $owner_uid]);
1129
1130 if (!$tsth->fetch()) {
1131 $usth->execute([$owner_uid, $tag, $entry_int_id]);
1132 }
1133
1134 array_push($tags_to_cache, $tag);
1135 }
1136
1137 /* update the cache */
1138
1139 $tags_to_cache = array_unique($tags_to_cache);
1140
1141 $tags_str = join(",", $tags_to_cache);
1142
1143 $tsth = $pdo->prepare("UPDATE ttrss_user_entries
1144 SET tag_cache = ? WHERE ref_id = ?
1145 AND owner_uid = ?");
1146 $tsth->execute([$tags_str, $entry_ref_id, $owner_uid]);
1147 }
1148
1149 _debug("article processed", $debug_enabled);
1150
1151 $pdo->commit();
1152 }
1153
1154 _debug("purging feed...", $debug_enabled);
1155
1156 purge_feed($feed, 0, $debug_enabled);
1157
1158 $sth = $pdo->prepare("UPDATE ttrss_feeds
1159 SET last_updated = NOW(), last_unconditional = NOW(), last_error = '' WHERE id = ?");
1160 $sth->execute([$feed]);
1161
1162 } else {
1163
1164 $error_msg = mb_substr($rss->error(), 0, 245);
1165
1166 _debug("fetch error: $error_msg", $debug_enabled);
1167
1168 if (count($rss->errors()) > 1) {
1169 foreach ($rss->errors() as $error) {
1170 _debug("+ $error");
1171 }
1172 }
1173
1174 $sth = $pdo->prepare("UPDATE ttrss_feeds SET last_error = ?,
1175 last_updated = NOW(), last_unconditional = NOW() WHERE id = ?");
1176 $sth->execute([$error_msg, $feed]);
1177
1178 unset($rss);
1179 return false;
1180 }
1181
1182 _debug("done", $debug_enabled);
1183
1184 return true;
1185 }
1186
1187 static function cache_enclosures($enclosures, $site_url, $debug) {
1188 foreach ($enclosures as $enc) {
1189
1190 if (preg_match("/(image|audio|video)/", $enc[1])) {
1191
1192 $src = rewrite_relative_url($site_url, $enc[0]);
1193
1194 $local_filename = CACHE_DIR . "/images/" . sha1($src);
1195
1196 if ($debug) _debug("cache_enclosures: downloading: $src to $local_filename");
1197
1198 if (!file_exists($local_filename)) {
1199 $file_content = fetch_file_contents($src);
1200
1201 if ($file_content && strlen($file_content) > MIN_CACHE_FILE_SIZE) {
1202 file_put_contents($local_filename, $file_content);
1203 }
1204 } else {
1205 touch($local_filename);
1206 }
1207 }
1208 }
1209 }
1210
1211 static function cache_media($html, $site_url, $debug) {
1212 libxml_use_internal_errors(true);
1213
1214 $charset_hack = '<head>
1215 <meta http-equiv="Content-Type" content="text/html; charset=utf-8"/>
1216 </head>';
1217
1218 $doc = new DOMDocument();
1219 $doc->loadHTML($charset_hack . $html);
1220 $xpath = new DOMXPath($doc);
1221
1222 $entries = $xpath->query('(//img[@src])|(//video/source[@src])|(//audio/source[@src])');
1223
1224 foreach ($entries as $entry) {
1225 if ($entry->hasAttribute('src') && strpos($entry->getAttribute('src'), "data:") !== 0) {
1226 $src = rewrite_relative_url($site_url, $entry->getAttribute('src'));
1227
1228 $local_filename = CACHE_DIR . "/images/" . sha1($src);
1229
1230 if ($debug) _debug("cache_media: downloading: $src to $local_filename");
1231
1232 if (!file_exists($local_filename)) {
1233 $file_content = fetch_file_contents($src);
1234
1235 if ($file_content && strlen($file_content) > MIN_CACHE_FILE_SIZE) {
1236 file_put_contents($local_filename, $file_content);
1237 }
1238 } else {
1239 touch($local_filename);
1240 }
1241 }
1242 }
1243 }
1244
1245 static function expire_error_log($debug) {
1246 if ($debug) _debug("Removing old error log entries...");
1247
1248 $pdo = Db::pdo();
1249
1250 if (DB_TYPE == "pgsql") {
1251 $pdo->query("DELETE FROM ttrss_error_log
1252 WHERE created_at < NOW() - INTERVAL '7 days'");
1253 } else {
1254 $pdo->query("DELETE FROM ttrss_error_log
1255 WHERE created_at < DATE_SUB(NOW(), INTERVAL 7 DAY)");
1256 }
1257 }
1258
1259 static function expire_lock_files($debug) {
1260 //if ($debug) _debug("Removing old lock files...");
1261
1262 $num_deleted = 0;
1263
1264 if (is_writable(LOCK_DIRECTORY)) {
1265 $files = glob(LOCK_DIRECTORY . "/*.lock");
1266
1267 if ($files) {
1268 foreach ($files as $file) {
1269 if (!file_is_locked(basename($file)) && time() - filemtime($file) > 86400*2) {
1270 unlink($file);
1271 ++$num_deleted;
1272 }
1273 }
1274 }
1275 }
1276
1277 if ($debug) _debug("Removed $num_deleted old lock files.");
1278 }
1279
1280 static function expire_cached_files($debug) {
1281 foreach (array("simplepie", "images", "export", "upload") as $dir) {
1282 $cache_dir = CACHE_DIR . "/$dir";
1283
1284 // if ($debug) _debug("Expiring $cache_dir");
1285
1286 $num_deleted = 0;
1287
1288 if (is_writable($cache_dir)) {
1289 $files = glob("$cache_dir/*");
1290
1291 if ($files) {
1292 foreach ($files as $file) {
1293 if (time() - filemtime($file) > 86400*CACHE_MAX_DAYS) {
1294 unlink($file);
1295
1296 ++$num_deleted;
1297 }
1298 }
1299 }
1300 }
1301
1302 if ($debug) _debug("$cache_dir: removed $num_deleted files.");
1303 }
1304 }
1305
1306 /**
1307 * Source: http://www.php.net/manual/en/function.parse-url.php#104527
1308 * Returns the url query as associative array
1309 *
1310 * @param string query
1311 * @return array params
1312 */
1313 static function convertUrlQuery($query) {
1314 $queryParts = explode('&', $query);
1315
1316 $params = array();
1317
1318 foreach ($queryParts as $param) {
1319 $item = explode('=', $param);
1320 $params[$item[0]] = $item[1];
1321 }
1322
1323 return $params;
1324 }
1325
1326 static function get_article_filters($filters, $title, $content, $link, $author, $tags, &$matched_rules = false) {
1327 $matches = array();
1328
1329 foreach ($filters as $filter) {
1330 $match_any_rule = $filter["match_any_rule"];
1331 $inverse = $filter["inverse"];
1332 $filter_match = false;
1333
1334 foreach ($filter["rules"] as $rule) {
1335 $match = false;
1336 $reg_exp = str_replace('/', '\/', $rule["reg_exp"]);
1337 $rule_inverse = $rule["inverse"];
1338
1339 if (!$reg_exp)
1340 continue;
1341
1342 switch ($rule["type"]) {
1343 case "title":
1344 $match = @preg_match("/$reg_exp/iu", $title);
1345 break;
1346 case "content":
1347 // we don't need to deal with multiline regexps
1348 $content = preg_replace("/[\r\n\t]/", "", $content);
1349
1350 $match = @preg_match("/$reg_exp/iu", $content);
1351 break;
1352 case "both":
1353 // we don't need to deal with multiline regexps
1354 $content = preg_replace("/[\r\n\t]/", "", $content);
1355
1356 $match = (@preg_match("/$reg_exp/iu", $title) || @preg_match("/$reg_exp/iu", $content));
1357 break;
1358 case "link":
1359 $match = @preg_match("/$reg_exp/iu", $link);
1360 break;
1361 case "author":
1362 $match = @preg_match("/$reg_exp/iu", $author);
1363 break;
1364 case "tag":
1365 foreach ($tags as $tag) {
1366 if (@preg_match("/$reg_exp/iu", $tag)) {
1367 $match = true;
1368 break;
1369 }
1370 }
1371 break;
1372 }
1373
1374 if ($rule_inverse) $match = !$match;
1375
1376 if ($match_any_rule) {
1377 if ($match) {
1378 $filter_match = true;
1379 break;
1380 }
1381 } else {
1382 $filter_match = $match;
1383 if (!$match) {
1384 break;
1385 }
1386 }
1387 }
1388
1389 if ($inverse) $filter_match = !$filter_match;
1390
1391 if ($filter_match) {
1392 if (is_array($matched_rules)) array_push($matched_rules, $rule);
1393
1394 foreach ($filter["actions"] AS $action) {
1395 array_push($matches, $action);
1396
1397 // if Stop action encountered, perform no further processing
1398 if (isset($action["type"]) && $action["type"] == "stop") return $matches;
1399 }
1400 }
1401 }
1402
1403 return $matches;
1404 }
1405
1406 static function find_article_filter($filters, $filter_name) {
1407 foreach ($filters as $f) {
1408 if ($f["type"] == $filter_name) {
1409 return $f;
1410 };
1411 }
1412 return false;
1413 }
1414
1415 static function find_article_filters($filters, $filter_name) {
1416 $results = array();
1417
1418 foreach ($filters as $f) {
1419 if ($f["type"] == $filter_name) {
1420 array_push($results, $f);
1421 };
1422 }
1423 return $results;
1424 }
1425
1426 static function calculate_article_score($filters) {
1427 $score = 0;
1428
1429 foreach ($filters as $f) {
1430 if ($f["type"] == "score") {
1431 $score += $f["param"];
1432 };
1433 }
1434 return $score;
1435 }
1436
1437 static function labels_contains_caption($labels, $caption) {
1438 foreach ($labels as $label) {
1439 if ($label[1] == $caption) {
1440 return true;
1441 }
1442 }
1443
1444 return false;
1445 }
1446
1447 static function assign_article_to_label_filters($id, $filters, $owner_uid, $article_labels) {
1448 foreach ($filters as $f) {
1449 if ($f["type"] == "label") {
1450 if (!RSSUtils::labels_contains_caption($article_labels, $f["param"])) {
1451 Labels::add_article($id, $f["param"], $owner_uid);
1452 }
1453 }
1454 }
1455 }
1456
1457 static function make_guid_from_title($title) {
1458 return preg_replace("/[ \"\',.:;]/", "-",
1459 mb_strtolower(strip_tags($title), 'utf-8'));
1460 }
1461
1462 static function cleanup_counters_cache($debug) {
1463 $pdo = Db::pdo();
1464
1465 $res = $pdo->query("DELETE FROM ttrss_counters_cache
1466 WHERE feed_id > 0 AND
1467 (SELECT COUNT(id) FROM ttrss_feeds WHERE
1468 id = feed_id AND
1469 ttrss_counters_cache.owner_uid = ttrss_feeds.owner_uid) = 0");
1470
1471 $frows = $res->rowCount();
1472
1473 $res = $pdo->query("DELETE FROM ttrss_cat_counters_cache
1474 WHERE feed_id > 0 AND
1475 (SELECT COUNT(id) FROM ttrss_feed_categories WHERE
1476 id = feed_id AND
1477 ttrss_cat_counters_cache.owner_uid = ttrss_feed_categories.owner_uid) = 0");
1478
1479 $crows = $res->rowCount();
1480
1481 if ($debug) _debug("Removed $frows (feeds) $crows (cats) orphaned counter cache entries.");
1482 }
1483
1484 static function housekeeping_user($owner_uid) {
1485 $tmph = new PluginHost();
1486
1487 load_user_plugins($owner_uid, $tmph);
1488
1489 $tmph->run_hooks(PluginHost::HOOK_HOUSE_KEEPING, "hook_house_keeping", "");
1490 }
1491
1492 static function housekeeping_common($debug) {
1493 RSSUtils::expire_cached_files($debug);
1494 RSSUtils::expire_lock_files($debug);
1495 RSSUtils::expire_error_log($debug);
1496
1497 $count = RSSUtils::update_feedbrowser_cache();
1498 _debug("Feedbrowser updated, $count feeds processed.");
1499
1500 Article::purge_orphans( true);
1501 RSSUtils::cleanup_counters_cache($debug);
1502
1503 //$rc = cleanup_tags( 14, 50000);
1504 //_debug("Cleaned $rc cached tags.");
1505
1506 PluginHost::getInstance()->run_hooks(PluginHost::HOOK_HOUSE_KEEPING, "hook_house_keeping", "");
1507 }
1508
1509 static function check_feed_favicon($site_url, $feed) {
1510 # print "FAVICON [$site_url]: $favicon_url\n";
1511
1512 $icon_file = ICONS_DIR . "/$feed.ico";
1513
1514 if (!file_exists($icon_file)) {
1515 $favicon_url = get_favicon_url($site_url);
1516
1517 if ($favicon_url) {
1518 // Limiting to "image" type misses those served with text/plain
1519 $contents = fetch_file_contents($favicon_url); // , "image");
1520
1521 if ($contents) {
1522 // Crude image type matching.
1523 // Patterns gleaned from the file(1) source code.
1524 if (preg_match('/^\x00\x00\x01\x00/', $contents)) {
1525 // 0 string \000\000\001\000 MS Windows icon resource
1526 //error_log("check_feed_favicon: favicon_url=$favicon_url isa MS Windows icon resource");
1527 }
1528 elseif (preg_match('/^GIF8/', $contents)) {
1529 // 0 string GIF8 GIF image data
1530 //error_log("check_feed_favicon: favicon_url=$favicon_url isa GIF image");
1531 }
1532 elseif (preg_match('/^\x89PNG\x0d\x0a\x1a\x0a/', $contents)) {
1533 // 0 string \x89PNG\x0d\x0a\x1a\x0a PNG image data
1534 //error_log("check_feed_favicon: favicon_url=$favicon_url isa PNG image");
1535 }
1536 elseif (preg_match('/^\xff\xd8/', $contents)) {
1537 // 0 beshort 0xffd8 JPEG image data
1538 //error_log("check_feed_favicon: favicon_url=$favicon_url isa JPG image");
1539 }
1540 elseif (preg_match('/^BM/', $contents)) {
1541 // 0 string BM PC bitmap (OS2, Windows BMP files)
1542 //error_log("check_feed_favicon, favicon_url=$favicon_url isa BMP image");
1543 }
1544 else {
1545 //error_log("check_feed_favicon: favicon_url=$favicon_url isa UNKNOWN type");
1546 $contents = "";
1547 }
1548 }
1549
1550 if ($contents) {
1551 $fp = @fopen($icon_file, "w");
1552
1553 if ($fp) {
1554 fwrite($fp, $contents);
1555 fclose($fp);
1556 chmod($icon_file, 0644);
1557 }
1558 }
1559 }
1560 return $icon_file;
1561 }
1562 }
1563
1564
1565
1566 }
web-based news feed (RSS/Atom) reader and aggregator; see https://tt-rss.org/