]> 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
strip utf8mb4 characters in enclosures on mysql
[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()) . $tmp);
17 }
18
19 // Strips utf8mb4 characters (i.e. emoji) for mysql
20 static function strip_utf8mb4($str) {
21 return preg_replace('/[\x{10000}-\x{10FFFF}]/u', "\xEF\xBF\xBD", $str);
22 }
23
24 static function update_feedbrowser_cache() {
25
26 $pdo = Db::pdo();
27
28 $sth = $pdo->query("SELECT feed_url, site_url, title, COUNT(id) AS subscribers
29 FROM ttrss_feeds WHERE feed_url NOT IN (SELECT feed_url FROM ttrss_feeds
30 WHERE private IS true OR auth_login != '' OR auth_pass != '' OR feed_url LIKE '%:%@%/%')
31 GROUP BY feed_url, site_url, title ORDER BY subscribers DESC LIMIT 1000");
32
33 $pdo->beginTransaction();
34
35 $pdo->query("DELETE FROM ttrss_feedbrowser_cache");
36
37 $count = 0;
38
39 while ($line = $sth->fetch()) {
40
41 $subscribers = $line["subscribers"];
42 $feed_url = $line["feed_url"];
43 $title = $line["title"];
44 $site_url = $line["site_url"];
45
46 $tmph = $pdo->prepare("SELECT subscribers FROM
47 ttrss_feedbrowser_cache WHERE feed_url = ?");
48 $tmph->execute([$feed_url]);
49
50 if (!$tmph->fetch()) {
51
52 $tmph = $pdo->prepare("INSERT INTO ttrss_feedbrowser_cache
53 (feed_url, site_url, title, subscribers)
54 VALUES
55 (?, ?, ?, ?)");
56
57 $tmph->execute([$feed_url, $site_url, $title, $subscribers]);
58
59 ++$count;
60
61 }
62
63 }
64
65 $pdo->commit();
66
67 return $count;
68
69 }
70
71 static function update_daemon_common($limit = DAEMON_FEED_LIMIT, $debug = true) {
72 $schema_version = get_schema_version();
73
74 if ($schema_version != SCHEMA_VERSION) {
75 die("Schema version is wrong, please upgrade the database.\n");
76 }
77
78 $pdo = Db::pdo();
79
80 if (!SINGLE_USER_MODE && DAEMON_UPDATE_LOGIN_LIMIT > 0) {
81 if (DB_TYPE == "pgsql") {
82 $login_thresh_qpart = "AND ttrss_users.last_login >= NOW() - INTERVAL '".DAEMON_UPDATE_LOGIN_LIMIT." days'";
83 } else {
84 $login_thresh_qpart = "AND ttrss_users.last_login >= DATE_SUB(NOW(), INTERVAL ".DAEMON_UPDATE_LOGIN_LIMIT." DAY)";
85 }
86 } else {
87 $login_thresh_qpart = "";
88 }
89
90 if (DB_TYPE == "pgsql") {
91 $update_limit_qpart = "AND ((
92 ttrss_feeds.update_interval = 0
93 AND ttrss_user_prefs.value != '-1'
94 AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_user_prefs.value || ' minutes') AS INTERVAL)
95 ) OR (
96 ttrss_feeds.update_interval > 0
97 AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_feeds.update_interval || ' minutes') AS INTERVAL)
98 ) OR (ttrss_feeds.last_updated IS NULL
99 AND ttrss_user_prefs.value != '-1')
100 OR (last_updated = '1970-01-01 00:00:00'
101 AND ttrss_user_prefs.value != '-1'))";
102 } else {
103 $update_limit_qpart = "AND ((
104 ttrss_feeds.update_interval = 0
105 AND ttrss_user_prefs.value != '-1'
106 AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL CONVERT(ttrss_user_prefs.value, SIGNED INTEGER) MINUTE)
107 ) OR (
108 ttrss_feeds.update_interval > 0
109 AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL ttrss_feeds.update_interval MINUTE)
110 ) OR (ttrss_feeds.last_updated IS NULL
111 AND ttrss_user_prefs.value != '-1')
112 OR (last_updated = '1970-01-01 00:00:00'
113 AND ttrss_user_prefs.value != '-1'))";
114 }
115
116 // Test if feed is currently being updated by another process.
117 if (DB_TYPE == "pgsql") {
118 $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < NOW() - INTERVAL '10 minutes')";
119 } else {
120 $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < DATE_SUB(NOW(), INTERVAL 10 MINUTE))";
121 }
122
123 $query_limit = $limit ? sprintf("LIMIT %d", $limit) : "";
124
125 // Update the least recently updated feeds first
126 $query_order = "ORDER BY last_updated";
127 if (DB_TYPE == "pgsql") $query_order .= " NULLS FIRST";
128
129 $query = "SELECT DISTINCT ttrss_feeds.feed_url, ttrss_feeds.last_updated
130 FROM
131 ttrss_feeds, ttrss_users, ttrss_user_prefs
132 WHERE
133 ttrss_feeds.owner_uid = ttrss_users.id
134 AND ttrss_user_prefs.profile IS NULL
135 AND ttrss_users.id = ttrss_user_prefs.owner_uid
136 AND ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL'
137 $login_thresh_qpart $update_limit_qpart
138 $updstart_thresh_qpart
139 $query_order $query_limit";
140
141 $res = $pdo->query($query);
142
143 $feeds_to_update = array();
144 while ($line = $res->fetch()) {
145 array_push($feeds_to_update, $line['feed_url']);
146 }
147
148 if ($debug) _debug(sprintf("Scheduled %d feeds to update...", count($feeds_to_update)));
149
150 // Update last_update_started before actually starting the batch
151 // in order to minimize collision risk for parallel daemon tasks
152 if (count($feeds_to_update) > 0) {
153 $feeds_qmarks = arr_qmarks($feeds_to_update);
154
155 $tmph = $pdo->prepare("UPDATE ttrss_feeds SET last_update_started = NOW()
156 WHERE feed_url IN ($feeds_qmarks)");
157 $tmph->execute($feeds_to_update);
158 }
159
160 $nf = 0;
161 $bstarted = microtime(true);
162
163 $batch_owners = array();
164
165 // since we have the data cached, we can deal with other feeds with the same url
166 $usth = $pdo->prepare("SELECT DISTINCT ttrss_feeds.id,last_updated,ttrss_feeds.owner_uid
167 FROM ttrss_feeds, ttrss_users, ttrss_user_prefs WHERE
168 ttrss_user_prefs.owner_uid = ttrss_feeds.owner_uid AND
169 ttrss_users.id = ttrss_user_prefs.owner_uid AND
170 ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL' AND
171 ttrss_user_prefs.profile IS NULL AND
172 feed_url = ?
173 $update_limit_qpart
174 $login_thresh_qpart
175 ORDER BY ttrss_feeds.id $query_limit");
176
177 foreach ($feeds_to_update as $feed) {
178 if($debug) _debug("Base feed: $feed");
179
180 $usth->execute([$feed]);
181 //update_rss_feed($line["id"], true);
182
183 if ($tline = $usth->fetch()) {
184 if ($debug) _debug(" => " . $tline["last_updated"] . ", " . $tline["id"] . " " . $tline["owner_uid"]);
185
186 if (array_search($tline["owner_uid"], $batch_owners) === FALSE)
187 array_push($batch_owners, $tline["owner_uid"]);
188
189 $fstarted = microtime(true);
190 RSSUtils::update_rss_feed($tline["id"], true, false);
191 _debug_suppress(false);
192
193 _debug(sprintf(" %.4f (sec)", microtime(true) - $fstarted));
194
195 ++$nf;
196 }
197 }
198
199 if ($nf > 0) {
200 _debug(sprintf("Processed %d feeds in %.4f (sec), %.4f (sec/feed avg)", $nf,
201 microtime(true) - $bstarted, (microtime(true) - $bstarted) / $nf));
202 }
203
204 foreach ($batch_owners as $owner_uid) {
205 _debug("Running housekeeping tasks for user $owner_uid...");
206
207 RSSUtils::housekeeping_user($owner_uid);
208 }
209
210 // Send feed digests by email if needed.
211 Digest::send_headlines_digests($debug);
212
213 return $nf;
214 }
215
216 // this is used when subscribing
217 static function set_basic_feed_info($feed) {
218
219 $pdo = Db::pdo();
220
221 $sth = $pdo->prepare("SELECT owner_uid,feed_url,auth_pass,auth_login,auth_pass_encrypted
222 FROM ttrss_feeds WHERE id = ?");
223 $sth->execute([$feed]);
224
225 if ($row = $sth->fetch()) {
226
227 $owner_uid = $row["owner_uid"];
228
229 $auth_pass_encrypted = $row["auth_pass_encrypted"];
230
231 $auth_login = $row["auth_login"];
232 $auth_pass = $row["auth_pass"];
233
234 if ($auth_pass_encrypted && function_exists("mcrypt_decrypt")) {
235 require_once "crypt.php";
236 $auth_pass = decrypt_string($auth_pass);
237 }
238
239 $fetch_url = $row["feed_url"];
240
241 $pluginhost = new PluginHost();
242 $user_plugins = get_pref("_ENABLED_PLUGINS", $owner_uid);
243
244 $pluginhost->load(PLUGINS, PluginHost::KIND_ALL);
245 $pluginhost->load($user_plugins, PluginHost::KIND_USER, $owner_uid);
246 $pluginhost->load_data();
247
248 $basic_info = array();
249 foreach ($pluginhost->get_hooks(PluginHost::HOOK_FEED_BASIC_INFO) as $plugin) {
250 $basic_info = $plugin->hook_feed_basic_info($basic_info, $fetch_url, $owner_uid, $feed, $auth_login, $auth_pass);
251 }
252
253 if (!$basic_info) {
254 $feed_data = fetch_file_contents($fetch_url, false,
255 $auth_login, $auth_pass, false,
256 FEED_FETCH_TIMEOUT,
257 0);
258
259 global $fetch_curl_used;
260
261 if (!$fetch_curl_used) {
262 $tmp = @gzdecode($feed_data);
263
264 if ($tmp) $feed_data = $tmp;
265 }
266
267 $feed_data = trim($feed_data);
268
269 $rss = new FeedParser($feed_data);
270 $rss->init();
271
272 if (!$rss->error()) {
273 $basic_info = array(
274 'title' => mb_substr($rss->get_title(), 0, 199),
275 'site_url' => mb_substr(rewrite_relative_url($fetch_url, $rss->get_link()), 0, 245)
276 );
277 }
278 }
279
280 if ($basic_info && is_array($basic_info)) {
281 $sth = $pdo->prepare("SELECT title, site_url FROM ttrss_feeds WHERE id = ?");
282 $sth->execute([$feed]);
283
284 if ($row = $sth->fetch()) {
285
286 $registered_title = $row["title"];
287 $orig_site_url = $row["site_url"];
288
289 if ($basic_info['title'] && (!$registered_title || $registered_title == "[Unknown]")) {
290
291 $sth = $pdo->prepare("UPDATE ttrss_feeds SET
292 title = ? WHERE id = ?");
293 $sth->execute([$basic_info['title'], $feed]);
294 }
295
296 if ($basic_info['site_url'] && $orig_site_url != $basic_info['site_url']) {
297 $sth = $pdo->prepare("UPDATE ttrss_feeds SET
298 site_url = ? WHERE id = ?");
299 $sth->execute([$basic_info['site_url'], $feed]);
300 }
301
302 }
303 }
304 }
305 }
306
307 /**
308 * @SuppressWarnings(PHPMD.UnusedFormalParameter)
309 */
310 static function update_rss_feed($feed, $no_cache = false) {
311
312 $debug_enabled = defined('DAEMON_EXTENDED_DEBUG') || clean($_REQUEST['xdebug']);
313
314 _debug_suppress(!$debug_enabled);
315 _debug("start", $debug_enabled);
316
317 $pdo = Db::pdo();
318
319 $sth = $pdo->prepare("SELECT title FROM ttrss_feeds WHERE id = ?");
320 $sth->execute([$feed]);
321
322 if (!$row = $sth->fetch()) {
323 _debug("feed $feed NOT FOUND/SKIPPED", $debug_enabled);
324 user_error("Attempt to update unknown/invalid feed $feed", E_USER_WARNING);
325 return false;
326 }
327
328 $title = $row["title"];
329
330 // feed was batch-subscribed or something, we need to get basic info
331 // this is not optimal currently as it fetches stuff separately TODO: optimize
332 if ($title == "[Unknown]") {
333 _debug("setting basic feed info for $feed...");
334 RSSUtils::set_basic_feed_info($feed);
335 }
336
337 $sth = $pdo->prepare("SELECT id,update_interval,auth_login,
338 feed_url,auth_pass,cache_images,
339 mark_unread_on_update, owner_uid,
340 auth_pass_encrypted, feed_language,
341 last_modified,
342 ".SUBSTRING_FOR_DATE."(last_unconditional, 1, 19) AS last_unconditional
343 FROM ttrss_feeds WHERE id = ?");
344 $sth->execute([$feed]);
345
346 if ($row = $sth->fetch()) {
347
348 $owner_uid = $row["owner_uid"];
349 $mark_unread_on_update = $row["mark_unread_on_update"];
350 $auth_pass_encrypted = $row["auth_pass_encrypted"];
351
352 $sth = $pdo->prepare("UPDATE ttrss_feeds SET last_update_started = NOW()
353 WHERE id = ?");
354 $sth->execute([$feed]);
355
356 $auth_login = $row["auth_login"];
357 $auth_pass = $row["auth_pass"];
358
359 if ($auth_pass_encrypted && function_exists("mcrypt_decrypt")) {
360 require_once "crypt.php";
361 $auth_pass = decrypt_string($auth_pass);
362 }
363
364 $stored_last_modified = $row["last_modified"];
365 $last_unconditional = $row["last_unconditional"];
366 $cache_images = $row["cache_images"];
367 $fetch_url = $row["feed_url"];
368 $feed_language = mb_strtolower($row["feed_language"]);
369 if (!$feed_language) $feed_language = 'english';
370
371 } else {
372 return false;
373 }
374
375 $date_feed_processed = date('Y-m-d H:i');
376
377 $cache_filename = CACHE_DIR . "/feeds/" . sha1($fetch_url) . ".xml";
378
379 $pluginhost = new PluginHost();
380 $pluginhost->set_debug($debug_enabled);
381 $user_plugins = get_pref("_ENABLED_PLUGINS", $owner_uid);
382
383 $pluginhost->load(PLUGINS, PluginHost::KIND_ALL);
384 $pluginhost->load($user_plugins, PluginHost::KIND_USER, $owner_uid);
385 $pluginhost->load_data();
386
387 $rss_hash = false;
388
389 $force_refetch = isset($_REQUEST["force_refetch"]);
390 $feed_data = "";
391
392 foreach ($pluginhost->get_hooks(PluginHost::HOOK_FETCH_FEED) as $plugin) {
393 $feed_data = $plugin->hook_fetch_feed($feed_data, $fetch_url, $owner_uid, $feed, 0, $auth_login, $auth_pass);
394 }
395
396 // try cache
397 if (!$feed_data &&
398 file_exists($cache_filename) &&
399 is_readable($cache_filename) &&
400 !$auth_login && !$auth_pass &&
401 filemtime($cache_filename) > time() - 30) {
402
403 _debug("using local cache [$cache_filename].", $debug_enabled);
404
405 @$feed_data = file_get_contents($cache_filename);
406
407 if ($feed_data) {
408 $rss_hash = sha1($feed_data);
409 }
410
411 } else {
412 _debug("local cache will not be used for this feed", $debug_enabled);
413 }
414
415 global $fetch_last_modified;
416
417 // fetch feed from source
418 if (!$feed_data) {
419 _debug("last unconditional update request: $last_unconditional");
420
421 if (ini_get("open_basedir") && function_exists("curl_init")) {
422 _debug("not using CURL due to open_basedir restrictions");
423 }
424
425 if (time() - strtotime($last_unconditional) > MAX_CONDITIONAL_INTERVAL) {
426 _debug("maximum allowed interval for conditional requests exceeded, forcing refetch");
427
428 $force_refetch = true;
429 } else {
430 _debug("stored last modified for conditional request: $stored_last_modified", $debug_enabled);
431 }
432
433 _debug("fetching [$fetch_url] (force_refetch: $force_refetch)...", $debug_enabled);
434
435 $feed_data = fetch_file_contents([
436 "url" => $fetch_url,
437 "login" => $auth_login,
438 "pass" => $auth_pass,
439 "timeout" => $no_cache ? FEED_FETCH_NO_CACHE_TIMEOUT : FEED_FETCH_TIMEOUT,
440 "last_modified" => $force_refetch ? "" : $stored_last_modified
441 ]);
442
443 global $fetch_curl_used;
444
445 if (!$fetch_curl_used) {
446 $tmp = @gzdecode($feed_data);
447
448 if ($tmp) $feed_data = $tmp;
449 }
450
451 $feed_data = trim($feed_data);
452
453 _debug("fetch done.", $debug_enabled);
454 _debug("source last modified: " . $fetch_last_modified, $debug_enabled);
455
456 if ($feed_data && $fetch_last_modified != $stored_last_modified) {
457 $sth = $pdo->prepare("UPDATE ttrss_feeds SET last_modified = ? WHERE id = ?");
458 $sth->execute([substr($fetch_last_modified, 0, 245), $feed]);
459 }
460
461 // cache vanilla feed data for re-use
462 if ($feed_data && !$auth_pass && !$auth_login && is_writable(CACHE_DIR . "/feeds")) {
463 $new_rss_hash = sha1($feed_data);
464
465 if ($new_rss_hash != $rss_hash) {
466 _debug("saving $cache_filename", $debug_enabled);
467 @file_put_contents($cache_filename, $feed_data);
468 }
469 }
470 }
471
472 if (!$feed_data) {
473 global $fetch_last_error;
474 global $fetch_last_error_code;
475
476 _debug("unable to fetch: $fetch_last_error [$fetch_last_error_code]", $debug_enabled);
477
478 // If-Modified-Since
479 if ($fetch_last_error_code != 304) {
480 $error_message = $fetch_last_error;
481 } else {
482 _debug("source claims data not modified, nothing to do.", $debug_enabled);
483 $error_message = "";
484 }
485
486 $sth = $pdo->prepare("UPDATE ttrss_feeds SET last_error = ?,
487 last_updated = NOW() WHERE id = ?");
488 $sth->execute([$error_message, $feed]);
489
490 return;
491 }
492
493 foreach ($pluginhost->get_hooks(PluginHost::HOOK_FEED_FETCHED) as $plugin) {
494 $feed_data = $plugin->hook_feed_fetched($feed_data, $fetch_url, $owner_uid, $feed);
495 }
496
497 $rss = new FeedParser($feed_data);
498 $rss->init();
499
500 if (!$rss->error()) {
501
502 // We use local pluginhost here because we need to load different per-user feed plugins
503 $pluginhost->run_hooks(PluginHost::HOOK_FEED_PARSED, "hook_feed_parsed", $rss);
504
505 _debug("language: $feed_language", $debug_enabled);
506 _debug("processing feed data...", $debug_enabled);
507
508 if (DB_TYPE == "pgsql") {
509 $favicon_interval_qpart = "favicon_last_checked < NOW() - INTERVAL '12 hour'";
510 } else {
511 $favicon_interval_qpart = "favicon_last_checked < DATE_SUB(NOW(), INTERVAL 12 HOUR)";
512 }
513
514 $sth = $pdo->prepare("SELECT owner_uid,favicon_avg_color,
515 (favicon_last_checked IS NULL OR $favicon_interval_qpart) AS
516 favicon_needs_check
517 FROM ttrss_feeds WHERE id = ?");
518 $sth->execute([$feed]);
519
520 if ($row = $sth->fetch()) {
521 $favicon_needs_check = $row["favicon_needs_check"];
522 $favicon_avg_color = $row["favicon_avg_color"];
523 $owner_uid = $row["owner_uid"];
524 } else {
525 return false;
526 }
527
528 $site_url = mb_substr(rewrite_relative_url($fetch_url, $rss->get_link()), 0, 245);
529
530 _debug("site_url: $site_url", $debug_enabled);
531 _debug("feed_title: " . $rss->get_title(), $debug_enabled);
532
533 if ($favicon_needs_check || $force_refetch) {
534
535 /* terrible hack: if we crash on floicon shit here, we won't check
536 * the icon avgcolor again (unless the icon got updated) */
537
538 $favicon_file = ICONS_DIR . "/$feed.ico";
539 $favicon_modified = @filemtime($favicon_file);
540
541 _debug("checking favicon...", $debug_enabled);
542
543 RSSUtils::check_feed_favicon($site_url, $feed);
544 $favicon_modified_new = @filemtime($favicon_file);
545
546 if ($favicon_modified_new > $favicon_modified)
547 $favicon_avg_color = '';
548
549 $favicon_colorstring = "";
550 if (file_exists($favicon_file) && function_exists("imagecreatefromstring") && $favicon_avg_color == '') {
551 require_once "colors.php";
552
553 $sth = $pdo->prepare("UPDATE ttrss_feeds SET favicon_avg_color = 'fail' WHERE
554 id = ?");
555 $sth->execute([$feed]);
556
557 $favicon_color = calculate_avg_color($favicon_file);
558
559 $favicon_colorstring = ",favicon_avg_color = " . $pdo->quote($favicon_color);
560
561 } else if ($favicon_avg_color == 'fail') {
562 _debug("floicon failed on this file, not trying to recalculate avg color", $debug_enabled);
563 }
564
565 $sth = $pdo->prepare("UPDATE ttrss_feeds SET favicon_last_checked = NOW()
566 $favicon_colorstring WHERE id = ?");
567 $sth->execute([$feed]);
568 }
569
570 _debug("loading filters & labels...", $debug_enabled);
571
572 $filters = load_filters($feed, $owner_uid);
573
574 if ($debug_enabled) {
575 print_r($filters);
576 }
577
578 _debug("" . count($filters) . " filters loaded.", $debug_enabled);
579
580 $items = $rss->get_items();
581
582 if (!is_array($items)) {
583 _debug("no articles found.", $debug_enabled);
584
585 $sth = $pdo->prepare("UPDATE ttrss_feeds
586 SET last_updated = NOW(), last_unconditional = NOW(), last_error = '' WHERE id = ?");
587 $sth->execute([$feed]);
588
589 return true; // no articles
590 }
591
592 _debug("processing articles...", $debug_enabled);
593
594 $tstart = time();
595
596 foreach ($items as $item) {
597 $pdo->beginTransaction();
598
599 if (clean($_REQUEST['xdebug']) == 3) {
600 print_r($item);
601 }
602
603 if (ini_get("max_execution_time") > 0 && time() - $tstart >= ini_get("max_execution_time") * 0.7) {
604 _debug("looks like there's too many articles to process at once, breaking out", $debug_enabled);
605 $pdo->commit();
606 break;
607 }
608
609 $entry_guid = strip_tags($item->get_id());
610 if (!$entry_guid) $entry_guid = strip_tags($item->get_link());
611 if (!$entry_guid) $entry_guid = RSSUtils::make_guid_from_title($item->get_title());
612
613 if (!$entry_guid) {
614 $pdo->commit();
615 continue;
616 }
617
618 $entry_guid = "$owner_uid,$entry_guid";
619
620 $entry_guid_hashed = 'SHA1:' . sha1($entry_guid);
621
622 _debug("guid $entry_guid / $entry_guid_hashed", $debug_enabled);
623
624 $entry_timestamp = strip_tags($item->get_date());
625
626 _debug("orig date: " . $item->get_date(), $debug_enabled);
627
628 if ($entry_timestamp == -1 || !$entry_timestamp || $entry_timestamp > time()) {
629 $entry_timestamp = time();
630 }
631
632 $entry_timestamp_fmt = strftime("%Y/%m/%d %H:%M:%S", $entry_timestamp);
633
634 _debug("date $entry_timestamp [$entry_timestamp_fmt]", $debug_enabled);
635
636 $entry_title = strip_tags($item->get_title());
637
638 $entry_link = rewrite_relative_url($site_url, $item->get_link());
639
640 _debug("title $entry_title", $debug_enabled);
641 _debug("link $entry_link", $debug_enabled);
642
643 if (!$entry_title) $entry_title = date("Y-m-d H:i:s", $entry_timestamp);;
644
645 $entry_content = $item->get_content();
646 if (!$entry_content) $entry_content = $item->get_description();
647
648 if (clean($_REQUEST["xdebug"]) == 2) {
649 print "content: ";
650 print htmlspecialchars($entry_content);
651 print "\n";
652 }
653
654 $entry_comments = mb_substr(strip_tags($item->get_comments_url()), 0, 245);
655 $num_comments = (int) $item->get_comments_count();
656
657 $entry_author = strip_tags($item->get_author());
658 $entry_guid = mb_substr($entry_guid, 0, 245);
659
660 _debug("author $entry_author", $debug_enabled);
661 _debug("num_comments: $num_comments", $debug_enabled);
662 _debug("looking for tags...", $debug_enabled);
663
664 // parse <category> entries into tags
665
666 $additional_tags = array();
667
668 $additional_tags_src = $item->get_categories();
669
670 if (is_array($additional_tags_src)) {
671 foreach ($additional_tags_src as $tobj) {
672 array_push($additional_tags, $tobj);
673 }
674 }
675
676 $entry_tags = array_unique($additional_tags);
677
678 for ($i = 0; $i < count($entry_tags); $i++)
679 $entry_tags[$i] = mb_strtolower($entry_tags[$i], 'utf-8');
680
681 _debug("tags found: " . join(",", $entry_tags), $debug_enabled);
682
683 _debug("done collecting data.", $debug_enabled);
684
685 $sth = $pdo->prepare("SELECT id, content_hash, lang FROM ttrss_entries
686 WHERE guid = ? OR guid = ?");
687 $sth->execute([$entry_guid, $entry_guid_hashed]);
688
689 if ($row = $sth->fetch()) {
690 $base_entry_id = $row["id"];
691 $entry_stored_hash = $row["content_hash"];
692 $article_labels = Article::get_article_labels($base_entry_id, $owner_uid);
693 $entry_language = $row["lang"];
694
695 $existing_tags = Article::get_article_tags($base_entry_id, $owner_uid);
696 $entry_tags = array_unique(array_merge($entry_tags, $existing_tags));
697 } else {
698 $base_entry_id = false;
699 $entry_stored_hash = "";
700 $article_labels = array();
701 $entry_language = "";
702 }
703
704 $article = array("owner_uid" => $owner_uid, // read only
705 "guid" => $entry_guid, // read only
706 "guid_hashed" => $entry_guid_hashed, // read only
707 "title" => $entry_title,
708 "content" => $entry_content,
709 "link" => $entry_link,
710 "labels" => $article_labels, // current limitation: can add labels to article, can't remove them
711 "tags" => $entry_tags,
712 "author" => $entry_author,
713 "force_catchup" => false, // ugly hack for the time being
714 "score_modifier" => 0, // no previous value, plugin should recalculate score modifier based on content if needed
715 "language" => $entry_language,
716 "num_comments" => $num_comments, // read only
717 "feed" => array("id" => $feed,
718 "fetch_url" => $fetch_url,
719 "site_url" => $site_url,
720 "cache_images" => $cache_images)
721 );
722
723 $entry_plugin_data = "";
724 $entry_current_hash = RSSUtils::calculate_article_hash($article, $pluginhost);
725
726 _debug("article hash: $entry_current_hash [stored=$entry_stored_hash]", $debug_enabled);
727
728 if ($entry_current_hash == $entry_stored_hash && !isset($_REQUEST["force_rehash"])) {
729 _debug("stored article seems up to date [IID: $base_entry_id], updating timestamp only", $debug_enabled);
730
731 // we keep encountering the entry in feeds, so we need to
732 // update date_updated column so that we don't get horrible
733 // dupes when the entry gets purged and reinserted again e.g.
734 // in the case of SLOW SLOW OMG SLOW updating feeds
735
736 $sth = $pdo->prepare("UPDATE ttrss_entries SET date_updated = NOW()
737 WHERE id = ?");
738 $sth->execute([$base_entry_id]);
739
740 $pdo->commit();
741 continue;
742 }
743
744 _debug("hash differs, applying plugin filters:", $debug_enabled);
745
746 foreach ($pluginhost->get_hooks(PluginHost::HOOK_ARTICLE_FILTER) as $plugin) {
747 _debug("... " . get_class($plugin), $debug_enabled);
748
749 $start = microtime(true);
750 $article = $plugin->hook_article_filter($article);
751
752 _debug("=== " . sprintf("%.4f (sec)", microtime(true) - $start), $debug_enabled);
753
754 $entry_plugin_data .= mb_strtolower(get_class($plugin)) . ",";
755 }
756
757 if (clean($_REQUEST["xdebug"]) == 2) {
758 print "processed content: ";
759 print htmlspecialchars($article["content"]);
760 print "\n";
761 }
762
763 _debug("plugin data: $entry_plugin_data", $debug_enabled);
764
765 // Workaround: 4-byte unicode requires utf8mb4 in MySQL. See https://tt-rss.org/forum/viewtopic.php?f=1&t=3377&p=20077#p20077
766 if (DB_TYPE == "mysql") {
767 foreach ($article as $k => $v) {
768 // i guess we'll have to take the risk of 4byte unicode labels & tags here
769 if (is_string($article[$k])) {
770 $article[$k] = RSSUtils::strip_utf8mb4($v);
771 }
772 }
773 }
774
775 /* Collect article tags here so we could filter by them: */
776
777 $matched_rules = array();
778
779 $article_filters = RSSUtils::get_article_filters($filters, $article["title"],
780 $article["content"], $article["link"], $article["author"],
781 $article["tags"], $matched_rules);
782
783 if ($debug_enabled) {
784 _debug("matched filter rules: ", $debug_enabled);
785
786 if (count($matched_rules) != 0) {
787 print_r($matched_rules);
788 }
789
790 _debug("filter actions: ", $debug_enabled);
791
792 if (count($article_filters) != 0) {
793 print_r($article_filters);
794 }
795 }
796
797 $plugin_filter_names = RSSUtils::find_article_filters($article_filters, "plugin");
798 $plugin_filter_actions = $pluginhost->get_filter_actions();
799
800 if (count($plugin_filter_names) > 0) {
801 _debug("applying plugin filter actions...", $debug_enabled);
802
803 foreach ($plugin_filter_names as $pfn) {
804 list($pfclass,$pfaction) = explode(":", $pfn["param"]);
805
806 if (isset($plugin_filter_actions[$pfclass])) {
807 $plugin = $pluginhost->get_plugin($pfclass);
808
809 _debug("... $pfclass: $pfaction", $debug_enabled);
810
811 if ($plugin) {
812 $start = microtime(true);
813 $article = $plugin->hook_article_filter_action($article, $pfaction);
814
815 _debug("=== " . sprintf("%.4f (sec)", microtime(true) - $start), $debug_enabled);
816 } else {
817 _debug("??? $pfclass: plugin object not found.");
818 }
819 } else {
820 _debug("??? $pfclass: filter plugin not registered.");
821 }
822 }
823 }
824
825 $entry_tags = $article["tags"];
826 $entry_title = strip_tags($article["title"]);
827 $entry_author = mb_substr(strip_tags($article["author"]), 0, 245);
828 $entry_link = strip_tags($article["link"]);
829 $entry_content = $article["content"]; // escaped below
830 $entry_force_catchup = $article["force_catchup"];
831 $article_labels = $article["labels"];
832 $entry_score_modifier = (int) $article["score_modifier"];
833 $entry_language = $article["language"];
834
835 if ($debug_enabled) {
836 _debug("article labels:", $debug_enabled);
837
838 if (count($article_labels) != 0) {
839 print_r($article_labels);
840 }
841 }
842
843 _debug("force catchup: $entry_force_catchup");
844
845 if ($cache_images && is_writable(CACHE_DIR . '/images'))
846 RSSUtils::cache_media($entry_content, $site_url, $debug_enabled);
847
848 $csth = $pdo->prepare("SELECT id FROM ttrss_entries
849 WHERE guid = ? OR guid = ?");
850 $csth->execute([$entry_guid, $entry_guid_hashed]);
851
852 if (!$row = $csth->fetch()) {
853
854 _debug("base guid [$entry_guid or $entry_guid_hashed] not found, creating...", $debug_enabled);
855
856 // base post entry does not exist, create it
857
858 $usth = $pdo->prepare(
859 "INSERT INTO ttrss_entries
860 (title,
861 guid,
862 link,
863 updated,
864 content,
865 content_hash,
866 no_orig_date,
867 date_updated,
868 date_entered,
869 comments,
870 num_comments,
871 plugin_data,
872 lang,
873 author)
874 VALUES
875 (?, ?, ?, ?, ?, ?,
876 false,
877 NOW(),
878 ?, ?, ?, ?, ?, ?)");
879
880 $usth->execute([$entry_title,
881 $entry_guid_hashed,
882 $entry_link,
883 $entry_timestamp_fmt,
884 "$entry_content",
885 $entry_current_hash,
886 $date_feed_processed,
887 $entry_comments,
888 (int)$num_comments,
889 $entry_plugin_data,
890 "$entry_language",
891 "$entry_author"]);
892
893 }
894
895 $csth->execute([$entry_guid, $entry_guid_hashed]);
896
897 $entry_ref_id = 0;
898 $entry_int_id = 0;
899
900 if ($row = $csth->fetch()) {
901
902 _debug("base guid found, checking for user record", $debug_enabled);
903
904 $ref_id = $row['id'];
905 $entry_ref_id = $ref_id;
906
907 if (RSSUtils::find_article_filter($article_filters, "filter")) {
908 $pdo->commit();
909 continue;
910 }
911
912 $score = RSSUtils::calculate_article_score($article_filters) + $entry_score_modifier;
913
914 _debug("initial score: $score [including plugin modifier: $entry_score_modifier]", $debug_enabled);
915
916 // check for user post link to main table
917
918 $sth = $pdo->prepare("SELECT ref_id, int_id FROM ttrss_user_entries WHERE
919 ref_id = ? AND owner_uid = ?");
920 $sth->execute([$ref_id, $owner_uid]);
921
922 // okay it doesn't exist - create user entry
923 if ($row = $sth->fetch()) {
924 $entry_ref_id = $row["ref_id"];
925 $entry_int_id = $row["int_id"];
926
927 _debug("user record FOUND: RID: $entry_ref_id, IID: $entry_int_id", $debug_enabled);
928 } else {
929
930 _debug("user record not found, creating...", $debug_enabled);
931
932 if ($score >= -500 && !RSSUtils::find_article_filter($article_filters, 'catchup') && !$entry_force_catchup) {
933 $unread = 1;
934 $last_read_qpart = null;
935 } else {
936 $unread = 0;
937 $last_read_qpart = date("Y-m-d H:i"); // we can't use NOW() here because it gets quoted
938 }
939
940 if (RSSUtils::find_article_filter($article_filters, 'mark') || $score > 1000) {
941 $marked = 1;
942 } else {
943 $marked = 0;
944 }
945
946 if (RSSUtils::find_article_filter($article_filters, 'publish')) {
947 $published = 1;
948 } else {
949 $published = 0;
950 }
951
952 $last_marked = ($marked == 1) ? 'NOW()' : 'NULL';
953 $last_published = ($published == 1) ? 'NOW()' : 'NULL';
954
955 $sth = $pdo->prepare(
956 "INSERT INTO ttrss_user_entries
957 (ref_id, owner_uid, feed_id, unread, last_read, marked,
958 published, score, tag_cache, label_cache, uuid,
959 last_marked, last_published)
960 VALUES (?, ?, ?, ?, ?, ?, ?, ?, '', '', '', ".$last_marked.", ".$last_published.")");
961
962 $sth->execute([$ref_id, $owner_uid, $feed, $unread, $last_read_qpart, $marked,
963 $published, $score]);
964
965 $sth = $pdo->prepare("SELECT int_id FROM ttrss_user_entries WHERE
966 ref_id = ? AND owner_uid = ? AND
967 feed_id = ? LIMIT 1");
968
969 $sth->execute([$ref_id, $owner_uid, $feed]);
970
971 if ($row = $sth->fetch())
972 $entry_int_id = $row['int_id'];
973 }
974
975 _debug("resulting RID: $entry_ref_id, IID: $entry_int_id", $debug_enabled);
976
977 if (DB_TYPE == "pgsql") {
978 $tsvector_combined = mb_substr($entry_title . ' ' .
979 preg_replace('/[<\?\:]/', ' ', strip_tags($entry_content)),
980 0, 1000000);
981
982 $tsvector_qpart = "tsvector_combined = to_tsvector(".$pdo->quote($feed_language).", ".$pdo->quote($tsvector_combined)."),";
983
984 } else {
985 $tsvector_qpart = "";
986 }
987
988 //_debug($tsvector_qpart);
989
990 $sth = $pdo->prepare("UPDATE ttrss_entries
991 SET title = :title,
992 $tsvector_qpart
993 content = :content,
994 content_hash = :content_hash,
995 updated = :updated,
996 num_comments = :num_comments,
997 plugin_data = :plugin_data,
998 author = :author,
999 lang = :lang
1000 WHERE id = :id");
1001
1002 $sth->execute([":title" => $entry_title,
1003 ":content" => "$entry_content",
1004 ":content_hash" => $entry_current_hash,
1005 ":updated" => $entry_timestamp_fmt,
1006 ":num_comments" => (int)$num_comments,
1007 ":plugin_data" => $entry_plugin_data,
1008 ":author" => "$entry_author",
1009 ":lang" => $entry_language,
1010 ":id" => $ref_id]);
1011
1012 // update aux data
1013 $sth = $pdo->prepare("UPDATE ttrss_user_entries
1014 SET score = ? WHERE ref_id = ?");
1015 $sth->execute([$score, $ref_id]);
1016
1017 if ($mark_unread_on_update) {
1018 _debug("article updated, marking unread as requested.", $debug_enabled);
1019
1020 $sth = $pdo->prepare("UPDATE ttrss_user_entries
1021 SET last_read = null, unread = true WHERE ref_id = ?");
1022 $sth->execute([$ref_id]);
1023 }
1024 }
1025
1026 _debug("assigning labels [other]...", $debug_enabled);
1027
1028 foreach ($article_labels as $label) {
1029 Labels::add_article($entry_ref_id, $label[1], $owner_uid);
1030 }
1031
1032 _debug("assigning labels [filters]...", $debug_enabled);
1033
1034 RSSUtils::assign_article_to_label_filters($entry_ref_id, $article_filters,
1035 $owner_uid, $article_labels);
1036
1037 _debug("looking for enclosures...", $debug_enabled);
1038
1039 // enclosures
1040
1041 $enclosures = array();
1042
1043 $encs = $item->get_enclosures();
1044
1045 if (is_array($encs)) {
1046 foreach ($encs as $e) {
1047 $e_item = array(
1048 rewrite_relative_url($site_url, $e->link),
1049 $e->type, $e->length, $e->title, $e->width, $e->height);
1050
1051 // Yet another episode of "mysql utf8_general_ci is gimped"
1052 if (DB_TYPE == "mysql") {
1053 for ($i = 0; $i < count($e_item); $i++) {
1054 if (is_string($e_item[$i])) {
1055 $e_item[$i] = RSSUtils::strip_utf8mb4($e_item[$i]);
1056 }
1057 }
1058 }
1059
1060 array_push($enclosures, $e_item);
1061 }
1062 }
1063
1064 if ($cache_images && is_writable(CACHE_DIR . '/images'))
1065 RSSUtils::cache_enclosures($enclosures, $site_url, $debug_enabled);
1066
1067 if ($debug_enabled) {
1068 _debug("article enclosures:", $debug_enabled);
1069 print_r($enclosures);
1070 }
1071
1072 $esth = $pdo->prepare("SELECT id FROM ttrss_enclosures
1073 WHERE content_url = ? AND post_id = ?");
1074
1075 $usth = $pdo->prepare("INSERT INTO ttrss_enclosures
1076 (content_url, content_type, title, duration, post_id, width, height) VALUES
1077 (?, ?, ?, ?, ?, ?, ?)");
1078
1079 foreach ($enclosures as $enc) {
1080 $enc_url = $enc[0];
1081 $enc_type = $enc[1];
1082 $enc_dur = (int)$enc[2];
1083 $enc_title = $enc[3];
1084 $enc_width = intval($enc[4]);
1085 $enc_height = intval($enc[5]);
1086
1087 $esth->execute([$enc_url, $entry_ref_id]);
1088
1089 if (!$esth->fetch()) {
1090 $usth->execute([$enc_url, $enc_type, (string)$enc_title, $enc_dur, $entry_ref_id, $enc_width, $enc_height]);
1091 }
1092 }
1093
1094 // check for manual tags (we have to do it here since they're loaded from filters)
1095
1096 foreach ($article_filters as $f) {
1097 if ($f["type"] == "tag") {
1098
1099 $manual_tags = trim_array(explode(",", $f["param"]));
1100
1101 foreach ($manual_tags as $tag) {
1102 if (tag_is_valid($tag)) {
1103 array_push($entry_tags, $tag);
1104 }
1105 }
1106 }
1107 }
1108
1109 // Skip boring tags
1110
1111 $boring_tags = trim_array(explode(",", mb_strtolower(get_pref(
1112 'BLACKLISTED_TAGS', $owner_uid, ''), 'utf-8')));
1113
1114 $filtered_tags = array();
1115 $tags_to_cache = array();
1116
1117 if ($entry_tags && is_array($entry_tags)) {
1118 foreach ($entry_tags as $tag) {
1119 if (array_search($tag, $boring_tags) === false) {
1120 array_push($filtered_tags, $tag);
1121 }
1122 }
1123 }
1124
1125 $filtered_tags = array_unique($filtered_tags);
1126
1127 if ($debug_enabled) {
1128 _debug("filtered article tags:", $debug_enabled);
1129 print_r($filtered_tags);
1130 }
1131
1132 // Save article tags in the database
1133
1134 if (count($filtered_tags) > 0) {
1135
1136 $tsth = $pdo->prepare("SELECT id FROM ttrss_tags
1137 WHERE tag_name = ? AND post_int_id = ? AND
1138 owner_uid = ? LIMIT 1");
1139
1140 $usth = $pdo->prepare("INSERT INTO ttrss_tags
1141 (owner_uid,tag_name,post_int_id)
1142 VALUES (?, ?, ?)");
1143
1144 foreach ($filtered_tags as $tag) {
1145
1146 $tag = sanitize_tag($tag);
1147
1148 if (!tag_is_valid($tag)) continue;
1149
1150 $tsth->execute([$tag, $entry_int_id, $owner_uid]);
1151
1152 if (!$tsth->fetch()) {
1153 $usth->execute([$owner_uid, $tag, $entry_int_id]);
1154 }
1155
1156 array_push($tags_to_cache, $tag);
1157 }
1158
1159 /* update the cache */
1160
1161 $tags_to_cache = array_unique($tags_to_cache);
1162
1163 $tags_str = join(",", $tags_to_cache);
1164
1165 $tsth = $pdo->prepare("UPDATE ttrss_user_entries
1166 SET tag_cache = ? WHERE ref_id = ?
1167 AND owner_uid = ?");
1168 $tsth->execute([$tags_str, $entry_ref_id, $owner_uid]);
1169 }
1170
1171 _debug("article processed", $debug_enabled);
1172
1173 $pdo->commit();
1174 }
1175
1176 _debug("purging feed...", $debug_enabled);
1177
1178 purge_feed($feed, 0, $debug_enabled);
1179
1180 $sth = $pdo->prepare("UPDATE ttrss_feeds
1181 SET last_updated = NOW(), last_unconditional = NOW(), last_error = '' WHERE id = ?");
1182 $sth->execute([$feed]);
1183
1184 } else {
1185
1186 $error_msg = mb_substr($rss->error(), 0, 245);
1187
1188 _debug("fetch error: $error_msg", $debug_enabled);
1189
1190 if (count($rss->errors()) > 1) {
1191 foreach ($rss->errors() as $error) {
1192 _debug("+ $error");
1193 }
1194 }
1195
1196 $sth = $pdo->prepare("UPDATE ttrss_feeds SET last_error = ?,
1197 last_updated = NOW(), last_unconditional = NOW() WHERE id = ?");
1198 $sth->execute([$error_msg, $feed]);
1199
1200 unset($rss);
1201 return false;
1202 }
1203
1204 _debug("done", $debug_enabled);
1205
1206 return true;
1207 }
1208
1209 static function cache_enclosures($enclosures, $site_url, $debug) {
1210 foreach ($enclosures as $enc) {
1211
1212 if (preg_match("/(image|audio|video)/", $enc[1])) {
1213
1214 $src = rewrite_relative_url($site_url, $enc[0]);
1215
1216 $local_filename = CACHE_DIR . "/images/" . sha1($src);
1217
1218 if ($debug) _debug("cache_enclosures: downloading: $src to $local_filename");
1219
1220 if (!file_exists($local_filename)) {
1221 $file_content = fetch_file_contents($src);
1222
1223 if ($file_content && strlen($file_content) > MIN_CACHE_FILE_SIZE) {
1224 file_put_contents($local_filename, $file_content);
1225 }
1226 } else {
1227 touch($local_filename);
1228 }
1229 }
1230 }
1231 }
1232
1233 static function cache_media($html, $site_url, $debug) {
1234 libxml_use_internal_errors(true);
1235
1236 $charset_hack = '<head>
1237 <meta http-equiv="Content-Type" content="text/html; charset=utf-8"/>
1238 </head>';
1239
1240 $doc = new DOMDocument();
1241 $doc->loadHTML($charset_hack . $html);
1242 $xpath = new DOMXPath($doc);
1243
1244 $entries = $xpath->query('(//img[@src])|(//video/source[@src])|(//audio/source[@src])');
1245
1246 foreach ($entries as $entry) {
1247 if ($entry->hasAttribute('src') && strpos($entry->getAttribute('src'), "data:") !== 0) {
1248 $src = rewrite_relative_url($site_url, $entry->getAttribute('src'));
1249
1250 $local_filename = CACHE_DIR . "/images/" . sha1($src);
1251
1252 if ($debug) _debug("cache_media: downloading: $src to $local_filename");
1253
1254 if (!file_exists($local_filename)) {
1255 $file_content = fetch_file_contents($src);
1256
1257 if ($file_content && strlen($file_content) > MIN_CACHE_FILE_SIZE) {
1258 file_put_contents($local_filename, $file_content);
1259 }
1260 } else {
1261 touch($local_filename);
1262 }
1263 }
1264 }
1265 }
1266
1267 static function expire_error_log($debug) {
1268 if ($debug) _debug("Removing old error log entries...");
1269
1270 $pdo = Db::pdo();
1271
1272 if (DB_TYPE == "pgsql") {
1273 $pdo->query("DELETE FROM ttrss_error_log
1274 WHERE created_at < NOW() - INTERVAL '7 days'");
1275 } else {
1276 $pdo->query("DELETE FROM ttrss_error_log
1277 WHERE created_at < DATE_SUB(NOW(), INTERVAL 7 DAY)");
1278 }
1279 }
1280
1281 static function expire_lock_files($debug) {
1282 //if ($debug) _debug("Removing old lock files...");
1283
1284 $num_deleted = 0;
1285
1286 if (is_writable(LOCK_DIRECTORY)) {
1287 $files = glob(LOCK_DIRECTORY . "/*.lock");
1288
1289 if ($files) {
1290 foreach ($files as $file) {
1291 if (!file_is_locked(basename($file)) && time() - filemtime($file) > 86400*2) {
1292 unlink($file);
1293 ++$num_deleted;
1294 }
1295 }
1296 }
1297 }
1298
1299 if ($debug) _debug("Removed $num_deleted old lock files.");
1300 }
1301
1302 static function expire_cached_files($debug) {
1303 foreach (array("simplepie", "feeds", "images", "export", "upload") as $dir) {
1304 $cache_dir = CACHE_DIR . "/$dir";
1305
1306 // if ($debug) _debug("Expiring $cache_dir");
1307
1308 $num_deleted = 0;
1309
1310 if (is_writable($cache_dir)) {
1311 $files = glob("$cache_dir/*");
1312
1313 if ($files) {
1314 foreach ($files as $file) {
1315 if (time() - filemtime($file) > 86400*CACHE_MAX_DAYS) {
1316 unlink($file);
1317
1318 ++$num_deleted;
1319 }
1320 }
1321 }
1322 }
1323
1324 if ($debug) _debug("$cache_dir: removed $num_deleted files.");
1325 }
1326 }
1327
1328 /**
1329 * Source: http://www.php.net/manual/en/function.parse-url.php#104527
1330 * Returns the url query as associative array
1331 *
1332 * @param string query
1333 * @return array params
1334 */
1335 static function convertUrlQuery($query) {
1336 $queryParts = explode('&', $query);
1337
1338 $params = array();
1339
1340 foreach ($queryParts as $param) {
1341 $item = explode('=', $param);
1342 $params[$item[0]] = $item[1];
1343 }
1344
1345 return $params;
1346 }
1347
1348 static function get_article_filters($filters, $title, $content, $link, $author, $tags, &$matched_rules = false) {
1349 $matches = array();
1350
1351 foreach ($filters as $filter) {
1352 $match_any_rule = $filter["match_any_rule"];
1353 $inverse = $filter["inverse"];
1354 $filter_match = false;
1355
1356 foreach ($filter["rules"] as $rule) {
1357 $match = false;
1358 $reg_exp = str_replace('/', '\/', $rule["reg_exp"]);
1359 $rule_inverse = $rule["inverse"];
1360
1361 if (!$reg_exp)
1362 continue;
1363
1364 switch ($rule["type"]) {
1365 case "title":
1366 $match = @preg_match("/$reg_exp/iu", $title);
1367 break;
1368 case "content":
1369 // we don't need to deal with multiline regexps
1370 $content = preg_replace("/[\r\n\t]/", "", $content);
1371
1372 $match = @preg_match("/$reg_exp/iu", $content);
1373 break;
1374 case "both":
1375 // we don't need to deal with multiline regexps
1376 $content = preg_replace("/[\r\n\t]/", "", $content);
1377
1378 $match = (@preg_match("/$reg_exp/iu", $title) || @preg_match("/$reg_exp/iu", $content));
1379 break;
1380 case "link":
1381 $match = @preg_match("/$reg_exp/iu", $link);
1382 break;
1383 case "author":
1384 $match = @preg_match("/$reg_exp/iu", $author);
1385 break;
1386 case "tag":
1387 foreach ($tags as $tag) {
1388 if (@preg_match("/$reg_exp/iu", $tag)) {
1389 $match = true;
1390 break;
1391 }
1392 }
1393 break;
1394 }
1395
1396 if ($rule_inverse) $match = !$match;
1397
1398 if ($match_any_rule) {
1399 if ($match) {
1400 $filter_match = true;
1401 break;
1402 }
1403 } else {
1404 $filter_match = $match;
1405 if (!$match) {
1406 break;
1407 }
1408 }
1409 }
1410
1411 if ($inverse) $filter_match = !$filter_match;
1412
1413 if ($filter_match) {
1414 if (is_array($matched_rules)) array_push($matched_rules, $rule);
1415
1416 foreach ($filter["actions"] AS $action) {
1417 array_push($matches, $action);
1418
1419 // if Stop action encountered, perform no further processing
1420 if (isset($action["type"]) && $action["type"] == "stop") return $matches;
1421 }
1422 }
1423 }
1424
1425 return $matches;
1426 }
1427
1428 static function find_article_filter($filters, $filter_name) {
1429 foreach ($filters as $f) {
1430 if ($f["type"] == $filter_name) {
1431 return $f;
1432 };
1433 }
1434 return false;
1435 }
1436
1437 static function find_article_filters($filters, $filter_name) {
1438 $results = array();
1439
1440 foreach ($filters as $f) {
1441 if ($f["type"] == $filter_name) {
1442 array_push($results, $f);
1443 };
1444 }
1445 return $results;
1446 }
1447
1448 static function calculate_article_score($filters) {
1449 $score = 0;
1450
1451 foreach ($filters as $f) {
1452 if ($f["type"] == "score") {
1453 $score += $f["param"];
1454 };
1455 }
1456 return $score;
1457 }
1458
1459 static function labels_contains_caption($labels, $caption) {
1460 foreach ($labels as $label) {
1461 if ($label[1] == $caption) {
1462 return true;
1463 }
1464 }
1465
1466 return false;
1467 }
1468
1469 static function assign_article_to_label_filters($id, $filters, $owner_uid, $article_labels) {
1470 foreach ($filters as $f) {
1471 if ($f["type"] == "label") {
1472 if (!RSSUtils::labels_contains_caption($article_labels, $f["param"])) {
1473 Labels::add_article($id, $f["param"], $owner_uid);
1474 }
1475 }
1476 }
1477 }
1478
1479 static function make_guid_from_title($title) {
1480 return preg_replace("/[ \"\',.:;]/", "-",
1481 mb_strtolower(strip_tags($title), 'utf-8'));
1482 }
1483
1484 static function cleanup_counters_cache($debug) {
1485 $pdo = Db::pdo();
1486
1487 $res = $pdo->query("DELETE FROM ttrss_counters_cache
1488 WHERE feed_id > 0 AND
1489 (SELECT COUNT(id) FROM ttrss_feeds WHERE
1490 id = feed_id AND
1491 ttrss_counters_cache.owner_uid = ttrss_feeds.owner_uid) = 0");
1492
1493 $frows = $res->rowCount();
1494
1495 $res = $pdo->query("DELETE FROM ttrss_cat_counters_cache
1496 WHERE feed_id > 0 AND
1497 (SELECT COUNT(id) FROM ttrss_feed_categories WHERE
1498 id = feed_id AND
1499 ttrss_cat_counters_cache.owner_uid = ttrss_feed_categories.owner_uid) = 0");
1500
1501 $crows = $res->rowCount();
1502
1503 if ($debug) _debug("Removed $frows (feeds) $crows (cats) orphaned counter cache entries.");
1504 }
1505
1506 static function housekeeping_user($owner_uid) {
1507 $tmph = new PluginHost();
1508
1509 load_user_plugins($owner_uid, $tmph);
1510
1511 $tmph->run_hooks(PluginHost::HOOK_HOUSE_KEEPING, "hook_house_keeping", "");
1512 }
1513
1514 static function housekeeping_common($debug) {
1515 RSSUtils::expire_cached_files($debug);
1516 RSSUtils::expire_lock_files($debug);
1517 RSSUtils::expire_error_log($debug);
1518
1519 $count = RSSUtils::update_feedbrowser_cache();
1520 _debug("Feedbrowser updated, $count feeds processed.");
1521
1522 Article::purge_orphans( true);
1523 RSSUtils::cleanup_counters_cache($debug);
1524
1525 //$rc = cleanup_tags( 14, 50000);
1526 //_debug("Cleaned $rc cached tags.");
1527
1528 PluginHost::getInstance()->run_hooks(PluginHost::HOOK_HOUSE_KEEPING, "hook_house_keeping", "");
1529 }
1530
1531 static function check_feed_favicon($site_url, $feed) {
1532 # print "FAVICON [$site_url]: $favicon_url\n";
1533
1534 $icon_file = ICONS_DIR . "/$feed.ico";
1535
1536 if (!file_exists($icon_file)) {
1537 $favicon_url = get_favicon_url($site_url);
1538
1539 if ($favicon_url) {
1540 // Limiting to "image" type misses those served with text/plain
1541 $contents = fetch_file_contents($favicon_url); // , "image");
1542
1543 if ($contents) {
1544 // Crude image type matching.
1545 // Patterns gleaned from the file(1) source code.
1546 if (preg_match('/^\x00\x00\x01\x00/', $contents)) {
1547 // 0 string \000\000\001\000 MS Windows icon resource
1548 //error_log("check_feed_favicon: favicon_url=$favicon_url isa MS Windows icon resource");
1549 }
1550 elseif (preg_match('/^GIF8/', $contents)) {
1551 // 0 string GIF8 GIF image data
1552 //error_log("check_feed_favicon: favicon_url=$favicon_url isa GIF image");
1553 }
1554 elseif (preg_match('/^\x89PNG\x0d\x0a\x1a\x0a/', $contents)) {
1555 // 0 string \x89PNG\x0d\x0a\x1a\x0a PNG image data
1556 //error_log("check_feed_favicon: favicon_url=$favicon_url isa PNG image");
1557 }
1558 elseif (preg_match('/^\xff\xd8/', $contents)) {
1559 // 0 beshort 0xffd8 JPEG image data
1560 //error_log("check_feed_favicon: favicon_url=$favicon_url isa JPG image");
1561 }
1562 elseif (preg_match('/^BM/', $contents)) {
1563 // 0 string BM PC bitmap (OS2, Windows BMP files)
1564 //error_log("check_feed_favicon, favicon_url=$favicon_url isa BMP image");
1565 }
1566 else {
1567 //error_log("check_feed_favicon: favicon_url=$favicon_url isa UNKNOWN type");
1568 $contents = "";
1569 }
1570 }
1571
1572 if ($contents) {
1573 $fp = @fopen($icon_file, "w");
1574
1575 if ($fp) {
1576 fwrite($fp, $contents);
1577 fclose($fp);
1578 chmod($icon_file, 0644);
1579 }
1580 }
1581 }
1582 return $icon_file;
1583 }
1584 }
1585
1586
1587
1588 }
web-based news feed (RSS/Atom) reader and aggregator; see https://tt-rss.org/