]> git.wh0rd.org - tt-rss.git/blob - include/rssfuncs.php
git.wh0rd.org / tt-rss.git / blob
? search:


95ac4076f52af6461d18873cf4d84d8e56d2c09d
[tt-rss.git] / include / rssfuncs.php
1 <?php
2 define_default('DAEMON_UPDATE_LOGIN_LIMIT', 30);
3 define_default('DAEMON_FEED_LIMIT', 500);
4 define_default('DAEMON_SLEEP_INTERVAL', 120);
5 define_default('_MIN_CACHE_IMAGE_SIZE', 1024);
6
7 function calculate_article_hash($article, $pluginhost) {
8 $tmp = "";
9
10 foreach ($article as $k => $v) {
11 if ($k != "feed" && isset($v)) {
12 $x = strip_tags(is_array($v) ? implode(",", $v) : $v);
13
14 //_debug("$k:" . sha1($x) . ":" . htmlspecialchars($x), true);
15
16 $tmp .= sha1("$k:" . sha1($x));
17 }
18 }
19
20 return sha1(implode(",", $pluginhost->get_plugin_names()) . $tmp);
21 }
22
23 function update_feedbrowser_cache() {
24
25 $result = db_query("SELECT feed_url, site_url, title, COUNT(id) AS subscribers
26 FROM ttrss_feeds WHERE (SELECT COUNT(id) = 0 FROM ttrss_feeds AS tf
27 WHERE tf.feed_url = ttrss_feeds.feed_url
28 AND (private IS true OR auth_login != '' OR auth_pass != '' OR feed_url LIKE '%:%@%/%'))
29 GROUP BY feed_url, site_url, title ORDER BY subscribers DESC LIMIT 1000");
30
31 db_query("BEGIN");
32
33 db_query("DELETE FROM ttrss_feedbrowser_cache");
34
35 $count = 0;
36
37 while ($line = db_fetch_assoc($result)) {
38 $subscribers = db_escape_string($line["subscribers"]);
39 $feed_url = db_escape_string($line["feed_url"]);
40 $title = db_escape_string($line["title"]);
41 $site_url = db_escape_string($line["site_url"]);
42
43 $tmp_result = db_query("SELECT subscribers FROM
44 ttrss_feedbrowser_cache WHERE feed_url = '$feed_url'");
45
46 if (db_num_rows($tmp_result) == 0) {
47
48 db_query("INSERT INTO ttrss_feedbrowser_cache
49 (feed_url, site_url, title, subscribers) VALUES ('$feed_url',
50 '$site_url', '$title', '$subscribers')");
51
52 ++$count;
53
54 }
55
56 }
57
58 db_query("COMMIT");
59
60 return $count;
61
62 }
63
64
65 /**
66 * Update a feed batch.
67 * Used by daemons to update n feeds by run.
68 * Only update feed needing a update, and not being processed
69 * by another process.
70 *
71 * @param mixed $link Database link
72 * @param integer $limit Maximum number of feeds in update batch. Default to DAEMON_FEED_LIMIT.
73 * @param boolean $from_http Set to true if you call this function from http to disable cli specific code.
74 * @param boolean $debug Set to false to disable debug output. Default to true.
75 * @return void
76 */
77 function update_daemon_common($limit = DAEMON_FEED_LIMIT, $from_http = false, $debug = true) {
78 // Process all other feeds using last_updated and interval parameters
79
80 $schema_version = get_schema_version();
81
82 if ($schema_version != SCHEMA_VERSION) {
83 die("Schema version is wrong, please upgrade the database.\n");
84 }
85
86 define('PREFS_NO_CACHE', true);
87
88 // Test if the user has loggued in recently. If not, it does not update its feeds.
89 if (!SINGLE_USER_MODE && DAEMON_UPDATE_LOGIN_LIMIT > 0) {
90 if (DB_TYPE == "pgsql") {
91 $login_thresh_qpart = "AND ttrss_users.last_login >= NOW() - INTERVAL '".DAEMON_UPDATE_LOGIN_LIMIT." days'";
92 } else {
93 $login_thresh_qpart = "AND ttrss_users.last_login >= DATE_SUB(NOW(), INTERVAL ".DAEMON_UPDATE_LOGIN_LIMIT." DAY)";
94 }
95 } else {
96 $login_thresh_qpart = "";
97 }
98
99 // Test if the feed need a update (update interval exceeded).
100 if (DB_TYPE == "pgsql") {
101 $update_limit_qpart = "AND ((
102 ttrss_feeds.update_interval = 0
103 AND ttrss_user_prefs.value != '-1'
104 AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_user_prefs.value || ' minutes') AS INTERVAL)
105 ) OR (
106 ttrss_feeds.update_interval > 0
107 AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_feeds.update_interval || ' minutes') AS INTERVAL)
108 ) OR (ttrss_feeds.last_updated IS NULL
109 AND ttrss_user_prefs.value != '-1')
110 OR (last_updated = '1970-01-01 00:00:00'
111 AND ttrss_user_prefs.value != '-1'))";
112 } else {
113 $update_limit_qpart = "AND ((
114 ttrss_feeds.update_interval = 0
115 AND ttrss_user_prefs.value != '-1'
116 AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL CONVERT(ttrss_user_prefs.value, SIGNED INTEGER) MINUTE)
117 ) OR (
118 ttrss_feeds.update_interval > 0
119 AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL ttrss_feeds.update_interval MINUTE)
120 ) OR (ttrss_feeds.last_updated IS NULL
121 AND ttrss_user_prefs.value != '-1')
122 OR (last_updated = '1970-01-01 00:00:00'
123 AND ttrss_user_prefs.value != '-1'))";
124 }
125
126 // Test if feed is currently being updated by another process.
127 if (DB_TYPE == "pgsql") {
128 $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < NOW() - INTERVAL '10 minutes')";
129 } else {
130 $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < DATE_SUB(NOW(), INTERVAL 10 MINUTE))";
131 }
132
133 // Test if there is a limit to number of updated feeds
134 $query_limit = "";
135 if($limit) $query_limit = sprintf("LIMIT %d", $limit);
136
137 // Update the least recently updated feeds first
138 $query_order = "ORDER BY last_updated";
139 if (DB_TYPE == "pgsql") $query_order .= " NULLS FIRST";
140
141 $query = "SELECT DISTINCT ttrss_feeds.feed_url, ttrss_feeds.last_updated
142 FROM
143 ttrss_feeds, ttrss_users, ttrss_user_prefs
144 WHERE
145 ttrss_feeds.owner_uid = ttrss_users.id
146 AND ttrss_user_prefs.profile IS NULL
147 AND ttrss_users.id = ttrss_user_prefs.owner_uid
148 AND ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL'
149 $login_thresh_qpart $update_limit_qpart
150 $updstart_thresh_qpart
151 $query_order $query_limit";
152
153 // We search for feed needing update.
154 $result = db_query($query);
155
156 if($debug) _debug(sprintf("Scheduled %d feeds to update...", db_num_rows($result)));
157
158 // Here is a little cache magic in order to minimize risk of double feed updates.
159 $feeds_to_update = array();
160 while ($line = db_fetch_assoc($result)) {
161 array_push($feeds_to_update, db_escape_string($line['feed_url']));
162 }
163
164 // We update the feed last update started date before anything else.
165 // There is no lag due to feed contents downloads
166 // It prevent an other process to update the same feed.
167
168 if(count($feeds_to_update) > 0) {
169 $feeds_quoted = array();
170
171 foreach ($feeds_to_update as $feed) {
172 array_push($feeds_quoted, "'" . db_escape_string($feed) . "'");
173 }
174
175 db_query(sprintf("UPDATE ttrss_feeds SET last_update_started = NOW()
176 WHERE feed_url IN (%s)", implode(',', $feeds_quoted)));
177 }
178
179 $nf = 0;
180 $bstarted = microtime(true);
181
182 // For each feed, we call the feed update function.
183 foreach ($feeds_to_update as $feed) {
184 if($debug) _debug("Base feed: $feed");
185
186 //update_rss_feed($line["id"], true);
187
188 // since we have the data cached, we can deal with other feeds with the same url
189
190 $tmp_result = db_query("SELECT DISTINCT ttrss_feeds.id,last_updated,ttrss_feeds.owner_uid
191 FROM ttrss_feeds, ttrss_users, ttrss_user_prefs WHERE
192 ttrss_user_prefs.owner_uid = ttrss_feeds.owner_uid AND
193 ttrss_users.id = ttrss_user_prefs.owner_uid AND
194 ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL' AND
195 ttrss_user_prefs.profile IS NULL AND
196 feed_url = '".db_escape_string($feed)."' AND
197 (ttrss_feeds.update_interval > 0 OR
198 ttrss_user_prefs.value != '-1')
199 $login_thresh_qpart
200 ORDER BY ttrss_feeds.id $query_limit");
201
202 if (db_num_rows($tmp_result) > 0) {
203 $rss = false;
204
205 while ($tline = db_fetch_assoc($tmp_result)) {
206 if($debug) _debug(" => " . $tline["last_updated"] . ", " . $tline["id"] . " " . $tline["owner_uid"]);
207
208 $fstarted = microtime(true);
209 $rss = update_rss_feed($tline["id"], true, false);
210 _debug_suppress(false);
211
212 _debug(sprintf(" %.4f (sec)", microtime(true) - $fstarted));
213
214 ++$nf;
215 }
216 }
217 }
218
219 if ($nf > 0) {
220 _debug(sprintf("Processed %d feeds in %.4f (sec), %.4f (sec/feed avg)", $nf,
221 microtime(true) - $bstarted, (microtime(true) - $bstarted) / $nf));
222 }
223
224 require_once "digest.php";
225
226 // Send feed digests by email if needed.
227 send_headlines_digests($debug);
228
229 return $nf;
230
231 } // function update_daemon_common
232
233 // this is used when subscribing
234 function set_basic_feed_info($feed) {
235
236 $feed = db_escape_string($feed);
237
238 $result = db_query("SELECT feed_url,auth_pass,auth_login,auth_pass_encrypted
239 FROM ttrss_feeds WHERE id = '$feed'");
240
241 $auth_pass_encrypted = sql_bool_to_bool(db_fetch_result($result,
242 0, "auth_pass_encrypted"));
243
244 $auth_login = db_fetch_result($result, 0, "auth_login");
245 $auth_pass = db_fetch_result($result, 0, "auth_pass");
246
247 if ($auth_pass_encrypted) {
248 require_once "crypt.php";
249 $auth_pass = decrypt_string($auth_pass);
250 }
251
252 $fetch_url = db_fetch_result($result, 0, "feed_url");
253
254 $feed_data = fetch_file_contents($fetch_url, false,
255 $auth_login, $auth_pass, false,
256 FEED_FETCH_TIMEOUT_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
274 $result = db_query("SELECT title, site_url FROM ttrss_feeds WHERE id = '$feed'");
275
276 $registered_title = db_fetch_result($result, 0, "title");
277 $orig_site_url = db_fetch_result($result, 0, "site_url");
278
279 $site_url = db_escape_string(mb_substr(rewrite_relative_url($fetch_url, $rss->get_link()), 0, 245));
280 $feed_title = db_escape_string(mb_substr($rss->get_title(), 0, 199));
281
282 if ($feed_title && (!$registered_title || $registered_title == "[Unknown]")) {
283 db_query("UPDATE ttrss_feeds SET
284 title = '$feed_title' WHERE id = '$feed'");
285 }
286
287 if ($site_url && $orig_site_url != $site_url) {
288 db_query("UPDATE ttrss_feeds SET
289 site_url = '$site_url' WHERE id = '$feed'");
290 }
291 }
292 }
293
294 // ignore_daemon is not used
295 function update_rss_feed($feed, $ignore_daemon = false, $no_cache = false, $rss = false) {
296
297 $debug_enabled = defined('DAEMON_EXTENDED_DEBUG') || $_REQUEST['xdebug'];
298
299 _debug_suppress(!$debug_enabled);
300 _debug("start", $debug_enabled);
301
302 $result = db_query("SELECT title FROM ttrss_feeds
303 WHERE id = '$feed'");
304 $title = db_fetch_result($result, 0, "title");
305
306 // feed was batch-subscribed or something, we need to get basic info
307 // this is not optimal currently as it fetches stuff separately TODO: optimize
308 if ($title == "[Unknown]") {
309 _debug("setting basic feed info for $feed...");
310 set_basic_feed_info($feed);
311 }
312
313 $result = db_query("SELECT id,update_interval,auth_login,
314 feed_url,auth_pass,cache_images,
315 mark_unread_on_update, owner_uid,
316 pubsub_state, auth_pass_encrypted,
317 feed_language,
318 (SELECT max(date_entered) FROM
319 ttrss_entries, ttrss_user_entries where ref_id = id AND feed_id = '$feed') AS last_article_timestamp
320 FROM ttrss_feeds WHERE id = '$feed'");
321
322 if (db_num_rows($result) == 0) {
323 _debug("feed $feed NOT FOUND/SKIPPED", $debug_enabled);
324 return false;
325 }
326
327 $last_article_timestamp = @strtotime(db_fetch_result($result, 0, "last_article_timestamp"));
328
329 if (defined('_DISABLE_HTTP_304'))
330 $last_article_timestamp = 0;
331
332 $owner_uid = db_fetch_result($result, 0, "owner_uid");
333 $mark_unread_on_update = sql_bool_to_bool(db_fetch_result($result,
334 0, "mark_unread_on_update"));
335 $pubsub_state = db_fetch_result($result, 0, "pubsub_state");
336 $auth_pass_encrypted = sql_bool_to_bool(db_fetch_result($result,
337 0, "auth_pass_encrypted"));
338
339 db_query("UPDATE ttrss_feeds SET last_update_started = NOW()
340 WHERE id = '$feed'");
341
342 $auth_login = db_fetch_result($result, 0, "auth_login");
343 $auth_pass = db_fetch_result($result, 0, "auth_pass");
344
345 if ($auth_pass_encrypted) {
346 require_once "crypt.php";
347 $auth_pass = decrypt_string($auth_pass);
348 }
349
350 $cache_images = sql_bool_to_bool(db_fetch_result($result, 0, "cache_images"));
351 $fetch_url = db_fetch_result($result, 0, "feed_url");
352 $feed_language = db_escape_string(mb_strtolower(db_fetch_result($result, 0, "feed_language")));
353 if (!$feed_language) $feed_language = 'english';
354
355 $feed = db_escape_string($feed);
356
357 $date_feed_processed = date('Y-m-d H:i');
358
359 $cache_filename = CACHE_DIR . "/simplepie/" . sha1($fetch_url) . ".xml";
360
361 $pluginhost = new PluginHost();
362 $pluginhost->set_debug($debug_enabled);
363 $user_plugins = get_pref("_ENABLED_PLUGINS", $owner_uid);
364
365 $pluginhost->load(PLUGINS, PluginHost::KIND_ALL);
366 $pluginhost->load($user_plugins, PluginHost::KIND_USER, $owner_uid);
367 $pluginhost->load_data();
368
369 if ($rss && is_object($rss) && get_class($rss) == "FeedParser") {
370 _debug("using previously initialized parser object");
371 } else {
372 $rss_hash = false;
373
374 $force_refetch = isset($_REQUEST["force_refetch"]);
375
376 foreach ($pluginhost->get_hooks(PluginHost::HOOK_FETCH_FEED) as $plugin) {
377 $feed_data = $plugin->hook_fetch_feed($feed_data, $fetch_url, $owner_uid, $feed, $last_article_timestamp, $auth_login, $auth_pass);
378 }
379
380 // try cache
381 if (!$feed_data &&
382 file_exists($cache_filename) &&
383 is_readable($cache_filename) &&
384 !$auth_login && !$auth_pass &&
385 filemtime($cache_filename) > time() - 30) {
386
387 _debug("using local cache [$cache_filename].", $debug_enabled);
388
389 @$feed_data = file_get_contents($cache_filename);
390
391 if ($feed_data) {
392 $rss_hash = sha1($feed_data);
393 }
394
395 } else {
396 _debug("local cache will not be used for this feed", $debug_enabled);
397 }
398
399 // fetch feed from source
400 if (!$feed_data) {
401 _debug("fetching [$fetch_url]...", $debug_enabled);
402
403 if (ini_get("open_basedir") && function_exists("curl_init")) {
404 _debug("not using CURL due to open_basedir restrictions");
405 }
406
407 _debug("If-Modified-Since: ".gmdate('D, d M Y H:i:s \G\M\T', $last_article_timestamp), $debug_enabled);
408
409 $feed_data = fetch_file_contents($fetch_url, false,
410 $auth_login, $auth_pass, false,
411 $no_cache ? FEED_FETCH_NO_CACHE_TIMEOUT : FEED_FETCH_TIMEOUT,
412 $force_refetch ? 0 : $last_article_timestamp);
413
414 global $fetch_curl_used;
415
416 if (!$fetch_curl_used) {
417 $tmp = @gzdecode($feed_data);
418
419 if ($tmp) $feed_data = $tmp;
420 }
421
422 $feed_data = trim($feed_data);
423
424 _debug("fetch done.", $debug_enabled);
425
426 // cache vanilla feed data for re-use
427 if ($feed_data && !$auth_pass && !$auth_login && is_writable(CACHE_DIR . "/simplepie")) {
428 $new_rss_hash = sha1($feed_data);
429
430 if ($new_rss_hash != $rss_hash) {
431 _debug("saving $cache_filename", $debug_enabled);
432 @file_put_contents($cache_filename, $feed_data);
433 }
434 }
435 }
436
437 if (!$feed_data) {
438 global $fetch_last_error;
439 global $fetch_last_error_code;
440
441 _debug("unable to fetch: $fetch_last_error [$fetch_last_error_code]", $debug_enabled);
442
443 $error_escaped = '';
444
445 // If-Modified-Since
446 if ($fetch_last_error_code != 304) {
447 $error_escaped = db_escape_string($fetch_last_error);
448 } else {
449 _debug("source claims data not modified, nothing to do.", $debug_enabled);
450 }
451
452 db_query(
453 "UPDATE ttrss_feeds SET last_error = '$error_escaped',
454 last_updated = NOW() WHERE id = '$feed'");
455
456 return;
457 }
458 }
459
460 foreach ($pluginhost->get_hooks(PluginHost::HOOK_FEED_FETCHED) as $plugin) {
461 $feed_data = $plugin->hook_feed_fetched($feed_data, $fetch_url, $owner_uid, $feed);
462 }
463
464 // set last update to now so if anything *simplepie* crashes later we won't be
465 // continuously failing on the same feed
466 //db_query("UPDATE ttrss_feeds SET last_updated = NOW() WHERE id = '$feed'");
467
468 if (!$rss) {
469 $rss = new FeedParser($feed_data);
470 $rss->init();
471 }
472
473 // print_r($rss);
474
475 $feed = db_escape_string($feed);
476
477 if (!$rss->error()) {
478
479 // We use local pluginhost here because we need to load different per-user feed plugins
480 $pluginhost->run_hooks(PluginHost::HOOK_FEED_PARSED, "hook_feed_parsed", $rss);
481
482 _debug("language: $feed_language", $debug_enabled);
483 _debug("processing feed data...", $debug_enabled);
484
485 // db_query("BEGIN");
486
487 if (DB_TYPE == "pgsql") {
488 $favicon_interval_qpart = "favicon_last_checked < NOW() - INTERVAL '12 hour'";
489 } else {
490 $favicon_interval_qpart = "favicon_last_checked < DATE_SUB(NOW(), INTERVAL 12 HOUR)";
491 }
492
493 $result = db_query("SELECT owner_uid,favicon_avg_color,
494 (favicon_last_checked IS NULL OR $favicon_interval_qpart) AS
495 favicon_needs_check
496 FROM ttrss_feeds WHERE id = '$feed'");
497
498 $favicon_needs_check = sql_bool_to_bool(db_fetch_result($result, 0,
499 "favicon_needs_check"));
500 $favicon_avg_color = db_fetch_result($result, 0, "favicon_avg_color");
501
502 $owner_uid = db_fetch_result($result, 0, "owner_uid");
503
504 $site_url = db_escape_string(mb_substr(rewrite_relative_url($fetch_url, $rss->get_link()), 0, 245));
505
506 _debug("site_url: $site_url", $debug_enabled);
507 _debug("feed_title: " . $rss->get_title(), $debug_enabled);
508
509 if ($favicon_needs_check || $force_refetch) {
510
511 /* terrible hack: if we crash on floicon shit here, we won't check
512 * the icon avgcolor again (unless the icon got updated) */
513
514 $favicon_file = ICONS_DIR . "/$feed.ico";
515 $favicon_modified = @filemtime($favicon_file);
516
517 _debug("checking favicon...", $debug_enabled);
518
519 check_feed_favicon($site_url, $feed);
520 $favicon_modified_new = @filemtime($favicon_file);
521
522 if ($favicon_modified_new > $favicon_modified)
523 $favicon_avg_color = '';
524
525 if (file_exists($favicon_file) && function_exists("imagecreatefromstring") && $favicon_avg_color == '') {
526 require_once "colors.php";
527
528 db_query("UPDATE ttrss_feeds SET favicon_avg_color = 'fail' WHERE
529 id = '$feed'");
530
531 $favicon_color = db_escape_string(
532 calculate_avg_color($favicon_file));
533
534 $favicon_colorstring = ",favicon_avg_color = '".$favicon_color."'";
535 } else if ($favicon_avg_color == 'fail') {
536 _debug("floicon failed on this file, not trying to recalculate avg color", $debug_enabled);
537 }
538
539 db_query("UPDATE ttrss_feeds SET favicon_last_checked = NOW()
540 $favicon_colorstring
541 WHERE id = '$feed'");
542 }
543
544 _debug("loading filters & labels...", $debug_enabled);
545
546 $filters = load_filters($feed, $owner_uid);
547
548 _debug("" . count($filters) . " filters loaded.", $debug_enabled);
549
550 $items = $rss->get_items();
551
552 if (!is_array($items)) {
553 _debug("no articles found.", $debug_enabled);
554
555 db_query("UPDATE ttrss_feeds
556 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
557
558 return; // no articles
559 }
560
561 if ($pubsub_state != 2 && PUBSUBHUBBUB_ENABLED) {
562
563 _debug("checking for PUSH hub...", $debug_enabled);
564
565 $feed_hub_url = false;
566
567 $links = $rss->get_links('hub');
568
569 if ($links && is_array($links)) {
570 foreach ($links as $l) {
571 $feed_hub_url = $l;
572 break;
573 }
574 }
575
576 _debug("feed hub url: $feed_hub_url", $debug_enabled);
577
578 $feed_self_url = $fetch_url;
579
580 $links = $rss->get_links('self');
581
582 if ($links && is_array($links)) {
583 foreach ($links as $l) {
584 $feed_self_url = $l;
585 break;
586 }
587 }
588
589 _debug("feed self url = $feed_self_url");
590
591 if ($feed_hub_url && $feed_self_url && function_exists('curl_init') &&
592 !ini_get("open_basedir")) {
593
594 require_once 'lib/pubsubhubbub/subscriber.php';
595
596 $callback_url = get_self_url_prefix() .
597 "/public.php?op=pubsub&id=$feed";
598
599 $s = new Subscriber($feed_hub_url, $callback_url);
600
601 $rc = $s->subscribe($feed_self_url);
602
603 _debug("feed hub url found, subscribe request sent. [rc=$rc]", $debug_enabled);
604
605 db_query("UPDATE ttrss_feeds SET pubsub_state = 1
606 WHERE id = '$feed'");
607 }
608 }
609
610 _debug("processing articles...", $debug_enabled);
611
612 $tstart = time();
613
614 foreach ($items as $item) {
615 if ($_REQUEST['xdebug'] == 3) {
616 print_r($item);
617 }
618
619 if (ini_get("max_execution_time") > 0 && time() - $tstart >= ini_get("max_execution_time") * 0.7) {
620 _debug("looks like there's too many articles to process at once, breaking out", $debug_enabled);
621 break;
622 }
623
624 $entry_guid = $item->get_id();
625 if (!$entry_guid) $entry_guid = $item->get_link();
626 if (!$entry_guid) $entry_guid = make_guid_from_title($item->get_title());
627 if (!$entry_guid) continue;
628
629 $entry_guid = "$owner_uid,$entry_guid";
630
631 $entry_guid_hashed = db_escape_string('SHA1:' . sha1($entry_guid));
632
633 _debug("guid $entry_guid / $entry_guid_hashed", $debug_enabled);
634
635 $entry_timestamp = "";
636
637 $entry_timestamp = $item->get_date();
638
639 _debug("orig date: " . $item->get_date(), $debug_enabled);
640
641 if ($entry_timestamp == -1 || !$entry_timestamp || $entry_timestamp > time()) {
642 $entry_timestamp = time();
643 }
644
645 $entry_timestamp_fmt = strftime("%Y/%m/%d %H:%M:%S", $entry_timestamp);
646
647 _debug("date $entry_timestamp [$entry_timestamp_fmt]", $debug_enabled);
648
649 // $entry_title = html_entity_decode($item->get_title(), ENT_COMPAT, 'UTF-8');
650 // $entry_title = decode_numeric_entities($entry_title);
651 $entry_title = $item->get_title();
652
653 $entry_link = rewrite_relative_url($site_url, $item->get_link());
654
655 _debug("title $entry_title", $debug_enabled);
656 _debug("link $entry_link", $debug_enabled);
657
658 if (!$entry_title) $entry_title = date("Y-m-d H:i:s", $entry_timestamp);;
659
660 $entry_content = $item->get_content();
661 if (!$entry_content) $entry_content = $item->get_description();
662
663 if ($_REQUEST["xdebug"] == 2) {
664 print "content: ";
665 print htmlspecialchars($entry_content);
666 print "\n";
667 }
668
669 $entry_comments = $item->get_comments_url();
670 $entry_author = $item->get_author();
671
672 $entry_guid = db_escape_string(mb_substr($entry_guid, 0, 245));
673
674 $entry_comments = db_escape_string(mb_substr(trim($entry_comments), 0, 245));
675 $entry_author = db_escape_string(mb_substr(trim($entry_author), 0, 245));
676
677 $num_comments = (int) $item->get_comments_count();
678
679 _debug("author $entry_author", $debug_enabled);
680 _debug("num_comments: $num_comments", $debug_enabled);
681 _debug("looking for tags...", $debug_enabled);
682
683 // parse <category> entries into tags
684
685 $additional_tags = array();
686
687 $additional_tags_src = $item->get_categories();
688
689 if (is_array($additional_tags_src)) {
690 foreach ($additional_tags_src as $tobj) {
691 array_push($additional_tags, $tobj);
692 }
693 }
694
695 $entry_tags = array_unique($additional_tags);
696
697 for ($i = 0; $i < count($entry_tags); $i++)
698 $entry_tags[$i] = mb_strtolower($entry_tags[$i], 'utf-8');
699
700 _debug("tags found: " . join(",", $entry_tags), $debug_enabled);
701
702 _debug("done collecting data.", $debug_enabled);
703
704 $result = db_query("SELECT id, content_hash, lang FROM ttrss_entries
705 WHERE guid = '".db_escape_string($entry_guid)."' OR guid = '$entry_guid_hashed'");
706
707 if (db_num_rows($result) != 0) {
708 $base_entry_id = db_fetch_result($result, 0, "id");
709 $entry_stored_hash = db_fetch_result($result, 0, "content_hash");
710 $article_labels = get_article_labels($base_entry_id, $owner_uid);
711 $entry_language = db_fetch_result($result, 0, "lang");
712
713 $existing_tags = get_article_tags($base_entry_id, $owner_uid);
714 $entry_tags = array_unique(array_merge($entry_tags, $existing_tags));
715
716 } else {
717 $base_entry_id = false;
718 $entry_stored_hash = "";
719 $article_labels = array();
720 $entry_language = "";
721 }
722
723 $article = array("owner_uid" => $owner_uid, // read only
724 "guid" => $entry_guid, // read only
725 "guid_hashed" => $entry_guid_hashed, // read only
726 "title" => $entry_title,
727 "content" => $entry_content,
728 "link" => $entry_link,
729 "labels" => $article_labels, // current limitation: can add labels to article, can't remove them
730 "tags" => $entry_tags,
731 "author" => $entry_author,
732 "force_catchup" => false, // ugly hack for the time being
733 "score_modifier" => 0, // no previous value, plugin should recalculate score modifier based on content if needed
734 "language" => $entry_language,
735 "feed" => array("id" => $feed,
736 "fetch_url" => $fetch_url,
737 "site_url" => $site_url)
738 );
739
740 $entry_plugin_data = "";
741 $entry_current_hash = calculate_article_hash($article, $pluginhost);
742
743 _debug("article hash: $entry_current_hash [stored=$entry_stored_hash]", $debug_enabled);
744
745 if ($entry_current_hash == $entry_stored_hash && !isset($_REQUEST["force_rehash"])) {
746 _debug("stored article seems up to date [IID: $base_entry_id], updating timestamp only", $debug_enabled);
747
748 // we keep encountering the entry in feeds, so we need to
749 // update date_updated column so that we don't get horrible
750 // dupes when the entry gets purged and reinserted again e.g.
751 // in the case of SLOW SLOW OMG SLOW updating feeds
752
753 $base_entry_id = db_fetch_result($result, 0, "id");
754
755 db_query("UPDATE ttrss_entries SET date_updated = NOW()
756 WHERE id = '$base_entry_id'");
757
758 continue;
759 }
760
761 _debug("hash differs, applying plugin filters:", $debug_enabled);
762
763 foreach ($pluginhost->get_hooks(PluginHost::HOOK_ARTICLE_FILTER) as $plugin) {
764 _debug("... " . get_class($plugin), $debug_enabled);
765
766 $start = microtime(true);
767 $article = $plugin->hook_article_filter($article);
768
769 _debug("=== " . sprintf("%.4f (sec)", microtime(true) - $start), $debug_enabled);
770
771 $entry_plugin_data .= mb_strtolower(get_class($plugin)) . ",";
772 }
773
774 if ($_REQUEST["xdebug"] == 2) {
775 print "processed content: ";
776 print htmlspecialchars($article["content"]);
777 print "\n";
778 }
779
780 $entry_plugin_data = db_escape_string($entry_plugin_data);
781
782 _debug("plugin data: $entry_plugin_data", $debug_enabled);
783
784 // Workaround: 4-byte unicode requires utf8mb4 in MySQL. See https://tt-rss.org/forum/viewtopic.php?f=1&t=3377&p=20077#p20077
785 if (DB_TYPE == "mysql") {
786 foreach ($article as $k => $v) {
787
788 // i guess we'll have to take the risk of 4byte unicode labels & tags here
789 if (!is_array($article[$k])) {
790 $article[$k] = preg_replace('/[\x{10000}-\x{10FFFF}]/u', "\xEF\xBF\xBD", $v);
791 }
792 }
793 }
794
795 /* Collect article tags here so we could filter by them: */
796
797 $article_filters = get_article_filters($filters, $article["title"],
798 $article["content"], $article["link"], 0, $article["author"],
799 $article["tags"]);
800
801 if ($debug_enabled) {
802 _debug("article filters: ", $debug_enabled);
803 if (count($article_filters) != 0) {
804 print_r($article_filters);
805 }
806 }
807
808 $plugin_filter_names = find_article_filters($article_filters, "plugin");
809 $plugin_filter_actions = $pluginhost->get_filter_actions();
810
811 if (count($plugin_filter_names) > 0) {
812 _debug("applying plugin filter actions...", $debug_enabled);
813
814 foreach ($plugin_filter_names as $pfn) {
815 list($pfclass,$pfaction) = explode(":", $pfn["param"]);
816
817 if (isset($plugin_filter_actions[$pfclass])) {
818 $plugin = $pluginhost->get_plugin($pfclass);
819
820 _debug("... $pfclass: $pfaction", $debug_enabled);
821
822 if ($plugin) {
823 $start = microtime(true);
824 $article = $plugin->hook_article_filter_action($article, $pfaction);
825
826 _debug("=== " . sprintf("%.4f (sec)", microtime(true) - $start), $debug_enabled);
827 } else {
828 _debug("??? $pfclass: plugin object not found.");
829 }
830 } else {
831 _debug("??? $pfclass: filter plugin not registered.");
832 }
833 }
834 }
835
836 $entry_tags = $article["tags"];
837 $entry_guid = db_escape_string($entry_guid);
838 $entry_title = db_escape_string($article["title"]);
839 $entry_author = db_escape_string($article["author"]);
840 $entry_link = db_escape_string($article["link"]);
841 $entry_content = $article["content"]; // escaped below
842 $entry_force_catchup = $article["force_catchup"];
843 $article_labels = $article["labels"];
844 $entry_score_modifier = (int) $article["score_modifier"];
845 $entry_language = db_escape_string($article["language"]);
846
847 if ($debug_enabled) {
848 _debug("article labels:", $debug_enabled);
849 print_r($article_labels);
850 }
851
852 _debug("force catchup: $entry_force_catchup");
853
854 if ($cache_images && is_writable(CACHE_DIR . '/images'))
855 cache_images($entry_content, $site_url, $debug_enabled);
856
857 $entry_content = db_escape_string($entry_content, false);
858
859 //db_query("BEGIN");
860
861 $result = db_query("SELECT id FROM ttrss_entries
862 WHERE (guid = '$entry_guid' OR guid = '$entry_guid_hashed')");
863
864 if (db_num_rows($result) == 0) {
865
866 _debug("base guid [$entry_guid] not found", $debug_enabled);
867
868 // base post entry does not exist, create it
869
870 $result = db_query(
871 "INSERT INTO ttrss_entries
872 (title,
873 guid,
874 link,
875 updated,
876 content,
877 content_hash,
878 no_orig_date,
879 date_updated,
880 date_entered,
881 comments,
882 num_comments,
883 plugin_data,
884 lang,
885 author)
886 VALUES
887 ('$entry_title',
888 '$entry_guid_hashed',
889 '$entry_link',
890 '$entry_timestamp_fmt',
891 '$entry_content',
892 '$entry_current_hash',
893 false,
894 NOW(),
895 '$date_feed_processed',
896 '$entry_comments',
897 '$num_comments',
898 '$entry_plugin_data',
899 '$entry_language',
900 '$entry_author')");
901
902 } else {
903 $base_entry_id = db_fetch_result($result, 0, "id");
904 }
905
906 // now it should exist, if not - bad luck then
907
908 $result = db_query("SELECT id FROM ttrss_entries
909 WHERE guid = '$entry_guid' OR guid = '$entry_guid_hashed'");
910
911 $entry_ref_id = 0;
912 $entry_int_id = 0;
913
914 if (db_num_rows($result) == 1) {
915
916 _debug("base guid found, checking for user record", $debug_enabled);
917
918 $ref_id = db_fetch_result($result, 0, "id");
919 $entry_ref_id = $ref_id;
920
921 /* $stored_guid = db_fetch_result($result, 0, "guid");
922 if ($stored_guid != $entry_guid_hashed) {
923 if ($debug_enabled) _debug("upgrading compat guid to hashed one", $debug_enabled);
924
925 db_query("UPDATE ttrss_entries SET guid = '$entry_guid_hashed' WHERE
926 id = '$ref_id'");
927 } */
928
929 if (find_article_filter($article_filters, "filter")) {
930 //db_query("COMMIT"); // close transaction in progress
931 continue;
932 }
933
934 $score = calculate_article_score($article_filters) + $entry_score_modifier;
935
936 _debug("initial score: $score [including plugin modifier: $entry_score_modifier]", $debug_enabled);
937
938 // check for user post link to main table
939
940 $query = "SELECT ref_id, int_id FROM ttrss_user_entries WHERE
941 ref_id = '$ref_id' AND owner_uid = '$owner_uid'";
942
943 // if ($_REQUEST["xdebug"]) print "$query\n";
944
945 $result = db_query($query);
946
947 // okay it doesn't exist - create user entry
948 if (db_num_rows($result) == 0) {
949
950 _debug("user record not found, creating...", $debug_enabled);
951
952 if ($score >= -500 && !find_article_filter($article_filters, 'catchup') && !$entry_force_catchup) {
953 $unread = 'true';
954 $last_read_qpart = 'NULL';
955 } else {
956 $unread = 'false';
957 $last_read_qpart = 'NOW()';
958 }
959
960 if (find_article_filter($article_filters, 'mark') || $score > 1000) {
961 $marked = 'true';
962 } else {
963 $marked = 'false';
964 }
965
966 if (find_article_filter($article_filters, 'publish')) {
967 $published = 'true';
968 } else {
969 $published = 'false';
970 }
971
972 // N-grams
973
974 /* if (DB_TYPE == "pgsql" and defined('_NGRAM_TITLE_DUPLICATE_THRESHOLD')) {
975
976 $result = db_query("SELECT COUNT(*) AS similar FROM
977 ttrss_entries,ttrss_user_entries
978 WHERE ref_id = id AND updated >= NOW() - INTERVAL '7 day'
979 AND similarity(title, '$entry_title') >= "._NGRAM_TITLE_DUPLICATE_THRESHOLD."
980 AND owner_uid = $owner_uid");
981
982 $ngram_similar = db_fetch_result($result, 0, "similar");
983
984 _debug("N-gram similar results: $ngram_similar", $debug_enabled);
985
986 if ($ngram_similar > 0) {
987 $unread = 'false';
988 }
989 } */
990
991 $last_marked = ($marked == 'true') ? 'NOW()' : 'NULL';
992 $last_published = ($published == 'true') ? 'NOW()' : 'NULL';
993
994 $result = db_query(
995 "INSERT INTO ttrss_user_entries
996 (ref_id, owner_uid, feed_id, unread, last_read, marked,
997 published, score, tag_cache, label_cache, uuid,
998 last_marked, last_published)
999 VALUES ('$ref_id', '$owner_uid', '$feed', $unread,
1000 $last_read_qpart, $marked, $published, '$score', '', '',
1001 '', $last_marked, $last_published)");
1002
1003 if (PUBSUBHUBBUB_HUB && $published == 'true') {
1004 $rss_link = get_self_url_prefix() .
1005 "/public.php?op=rss&id=-2&key=" .
1006 get_feed_access_key(-2, false, $owner_uid);
1007
1008 $p = new Publisher(PUBSUBHUBBUB_HUB);
1009
1010 /* $pubsub_result = */ $p->publish_update($rss_link);
1011 }
1012
1013 $result = db_query(
1014 "SELECT int_id FROM ttrss_user_entries WHERE
1015 ref_id = '$ref_id' AND owner_uid = '$owner_uid' AND
1016 feed_id = '$feed' LIMIT 1");
1017
1018 if (db_num_rows($result) == 1) {
1019 $entry_int_id = db_fetch_result($result, 0, "int_id");
1020 }
1021 } else {
1022 _debug("user record FOUND", $debug_enabled);
1023
1024 $entry_ref_id = db_fetch_result($result, 0, "ref_id");
1025 $entry_int_id = db_fetch_result($result, 0, "int_id");
1026 }
1027
1028 _debug("RID: $entry_ref_id, IID: $entry_int_id", $debug_enabled);
1029
1030 if (DB_TYPE == "pgsql") {
1031 $tsvector_combined = db_escape_string(mb_substr($entry_title . ' ' . strip_tags(str_replace('<', ' <', $entry_content)),
1032 0, 1000000));
1033
1034 $tsvector_qpart = "tsvector_combined = to_tsvector('$feed_language', '$tsvector_combined'),";
1035
1036 } else {
1037 $tsvector_qpart = "";
1038 }
1039
1040 db_query("UPDATE ttrss_entries
1041 SET title = '$entry_title',
1042 content = '$entry_content',
1043 content_hash = '$entry_current_hash',
1044 updated = '$entry_timestamp_fmt',
1045 $tsvector_qpart
1046 num_comments = '$num_comments',
1047 plugin_data = '$entry_plugin_data',
1048 author = '$entry_author',
1049 lang = '$entry_language'
1050 WHERE id = '$ref_id'");
1051
1052 // update aux data
1053 db_query("UPDATE ttrss_user_entries
1054 SET score = '$score' WHERE ref_id = '$ref_id'");
1055
1056 if ($mark_unread_on_update) {
1057 _debug("article updated, marking unread as requested.", $debug_enabled);
1058
1059 db_query("UPDATE ttrss_user_entries
1060 SET last_read = null, unread = true WHERE ref_id = '$ref_id'");
1061 }
1062 }
1063
1064 //db_query("COMMIT");
1065
1066 _debug("assigning labels [other]...", $debug_enabled);
1067
1068 foreach ($article_labels as $label) {
1069 label_add_article($entry_ref_id, $label[1], $owner_uid);
1070 }
1071
1072 _debug("assigning labels [filters]...", $debug_enabled);
1073
1074 assign_article_to_label_filters($entry_ref_id, $article_filters,
1075 $owner_uid, $article_labels);
1076
1077 _debug("looking for enclosures...", $debug_enabled);
1078
1079 // enclosures
1080
1081 $enclosures = array();
1082
1083 $encs = $item->get_enclosures();
1084
1085 if (is_array($encs)) {
1086 foreach ($encs as $e) {
1087 $e_item = array(
1088 $e->link, $e->type, $e->length, $e->title, $e->width, $e->height);
1089 array_push($enclosures, $e_item);
1090 }
1091 }
1092
1093 if ($debug_enabled) {
1094 _debug("article enclosures:", $debug_enabled);
1095 print_r($enclosures);
1096 }
1097
1098 //db_query("BEGIN");
1099
1100 // debugging
1101 // db_query("DELETE FROM ttrss_enclosures WHERE post_id = '$entry_ref_id'");
1102
1103 foreach ($enclosures as $enc) {
1104 $enc_url = db_escape_string($enc[0]);
1105 $enc_type = db_escape_string($enc[1]);
1106 $enc_dur = db_escape_string($enc[2]);
1107 $enc_title = db_escape_string($enc[3]);
1108 $enc_width = intval($enc[4]);
1109 $enc_height = intval($enc[5]);
1110
1111 $result = db_query("SELECT id FROM ttrss_enclosures
1112 WHERE content_url = '$enc_url' AND post_id = '$entry_ref_id'");
1113
1114 if (db_num_rows($result) == 0) {
1115 db_query("INSERT INTO ttrss_enclosures
1116 (content_url, content_type, title, duration, post_id, width, height) VALUES
1117 ('$enc_url', '$enc_type', '$enc_title', '$enc_dur', '$entry_ref_id', $enc_width, $enc_height)");
1118 }
1119 }
1120
1121 //db_query("COMMIT");
1122
1123 // check for manual tags (we have to do it here since they're loaded from filters)
1124
1125 foreach ($article_filters as $f) {
1126 if ($f["type"] == "tag") {
1127
1128 $manual_tags = trim_array(explode(",", $f["param"]));
1129
1130 foreach ($manual_tags as $tag) {
1131 if (tag_is_valid($tag)) {
1132 array_push($entry_tags, $tag);
1133 }
1134 }
1135 }
1136 }
1137
1138 // Skip boring tags
1139
1140 $boring_tags = trim_array(explode(",", mb_strtolower(get_pref(
1141 'BLACKLISTED_TAGS', $owner_uid, ''), 'utf-8')));
1142
1143 $filtered_tags = array();
1144 $tags_to_cache = array();
1145
1146 if ($entry_tags && is_array($entry_tags)) {
1147 foreach ($entry_tags as $tag) {
1148 if (array_search($tag, $boring_tags) === false) {
1149 array_push($filtered_tags, $tag);
1150 }
1151 }
1152 }
1153
1154 $filtered_tags = array_unique($filtered_tags);
1155
1156 if ($debug_enabled) {
1157 _debug("filtered article tags:", $debug_enabled);
1158 print_r($filtered_tags);
1159 }
1160
1161 // Save article tags in the database
1162
1163 if (count($filtered_tags) > 0) {
1164
1165 //db_query("BEGIN");
1166
1167 foreach ($filtered_tags as $tag) {
1168
1169 $tag = sanitize_tag($tag);
1170 $tag = db_escape_string($tag);
1171
1172 if (!tag_is_valid($tag)) continue;
1173
1174 $result = db_query("SELECT id FROM ttrss_tags
1175 WHERE tag_name = '$tag' AND post_int_id = '$entry_int_id' AND
1176 owner_uid = '$owner_uid' LIMIT 1");
1177
1178 if ($result && db_num_rows($result) == 0) {
1179
1180 db_query("INSERT INTO ttrss_tags
1181 (owner_uid,tag_name,post_int_id)
1182 VALUES ('$owner_uid','$tag', '$entry_int_id')");
1183 }
1184
1185 array_push($tags_to_cache, $tag);
1186 }
1187
1188 /* update the cache */
1189
1190 $tags_to_cache = array_unique($tags_to_cache);
1191
1192 $tags_str = db_escape_string(join(",", $tags_to_cache));
1193
1194 db_query("UPDATE ttrss_user_entries
1195 SET tag_cache = '$tags_str' WHERE ref_id = '$entry_ref_id'
1196 AND owner_uid = $owner_uid");
1197
1198 //db_query("COMMIT");
1199 }
1200
1201 _debug("article processed", $debug_enabled);
1202 }
1203
1204 _debug("purging feed...", $debug_enabled);
1205
1206 purge_feed($feed, 0, $debug_enabled);
1207
1208 db_query("UPDATE ttrss_feeds
1209 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
1210
1211 // db_query("COMMIT");
1212
1213 } else {
1214
1215 $error_msg = db_escape_string(mb_substr($rss->error(), 0, 245));
1216
1217 _debug("fetch error: $error_msg", $debug_enabled);
1218
1219 if (count($rss->errors()) > 1) {
1220 foreach ($rss->errors() as $error) {
1221 _debug("+ $error");
1222 }
1223 }
1224
1225 db_query(
1226 "UPDATE ttrss_feeds SET last_error = '$error_msg',
1227 last_updated = NOW() WHERE id = '$feed'");
1228
1229 unset($rss);
1230 }
1231
1232 _debug("done", $debug_enabled);
1233
1234 return $rss;
1235 }
1236
1237 function cache_images($html, $site_url, $debug) {
1238 libxml_use_internal_errors(true);
1239
1240 $charset_hack = '<head>
1241 <meta http-equiv="Content-Type" content="text/html; charset=utf-8"/>
1242 </head>';
1243
1244 $doc = new DOMDocument();
1245 $doc->loadHTML($charset_hack . $html);
1246 $xpath = new DOMXPath($doc);
1247
1248 $entries = $xpath->query('(//img[@src])');
1249
1250 foreach ($entries as $entry) {
1251 if ($entry->hasAttribute('src')) {
1252 $src = rewrite_relative_url($site_url, $entry->getAttribute('src'));
1253
1254 $local_filename = CACHE_DIR . "/images/" . sha1($src) . ".png";
1255
1256 if ($debug) _debug("cache_images: downloading: $src to $local_filename");
1257
1258 if (!file_exists($local_filename)) {
1259 $file_content = fetch_file_contents($src);
1260
1261 if ($file_content && strlen($file_content) > _MIN_CACHE_IMAGE_SIZE) {
1262 file_put_contents($local_filename, $file_content);
1263 }
1264 } else {
1265 touch($local_filename);
1266 }
1267 }
1268 }
1269 }
1270
1271 function expire_error_log($debug) {
1272 if ($debug) _debug("Removing old error log entries...");
1273
1274 if (DB_TYPE == "pgsql") {
1275 db_query("DELETE FROM ttrss_error_log
1276 WHERE created_at < NOW() - INTERVAL '7 days'");
1277 } else {
1278 db_query("DELETE FROM ttrss_error_log
1279 WHERE created_at < DATE_SUB(NOW(), INTERVAL 7 DAY)");
1280 }
1281
1282 }
1283
1284 function expire_lock_files($debug) {
1285 //if ($debug) _debug("Removing old lock files...");
1286
1287 $num_deleted = 0;
1288
1289 if (is_writable(LOCK_DIRECTORY)) {
1290 $files = glob(LOCK_DIRECTORY . "/*.lock");
1291
1292 if ($files) {
1293 foreach ($files as $file) {
1294 if (!file_is_locked(basename($file)) && time() - filemtime($file) > 86400*2) {
1295 unlink($file);
1296 ++$num_deleted;
1297 }
1298 }
1299 }
1300 }
1301
1302 if ($debug) _debug("Removed $num_deleted old lock files.");
1303 }
1304
1305 function expire_cached_files($debug) {
1306 foreach (array("simplepie", "images", "export", "upload") as $dir) {
1307 $cache_dir = CACHE_DIR . "/$dir";
1308
1309 // if ($debug) _debug("Expiring $cache_dir");
1310
1311 $num_deleted = 0;
1312
1313 if (is_writable($cache_dir)) {
1314 $files = glob("$cache_dir/*");
1315
1316 if ($files) {
1317 foreach ($files as $file) {
1318 if (time() - filemtime($file) > 86400*7) {
1319 unlink($file);
1320
1321 ++$num_deleted;
1322 }
1323 }
1324 }
1325 }
1326
1327 if ($debug) _debug("$cache_dir: removed $num_deleted files.");
1328 }
1329 }
1330
1331 /**
1332 * Source: http://www.php.net/manual/en/function.parse-url.php#104527
1333 * Returns the url query as associative array
1334 *
1335 * @param string query
1336 * @return array params
1337 */
1338 function convertUrlQuery($query) {
1339 $queryParts = explode('&', $query);
1340
1341 $params = array();
1342
1343 foreach ($queryParts as $param) {
1344 $item = explode('=', $param);
1345 $params[$item[0]] = $item[1];
1346 }
1347
1348 return $params;
1349 }
1350
1351 function get_article_filters($filters, $title, $content, $link, $timestamp, $author, $tags) {
1352 $matches = array();
1353
1354 foreach ($filters as $filter) {
1355 $match_any_rule = $filter["match_any_rule"];
1356 $inverse = $filter["inverse"];
1357 $filter_match = false;
1358
1359 foreach ($filter["rules"] as $rule) {
1360 $match = false;
1361 $reg_exp = str_replace('/', '\/', $rule["reg_exp"]);
1362 $rule_inverse = $rule["inverse"];
1363
1364 if (!$reg_exp)
1365 continue;
1366
1367 switch ($rule["type"]) {
1368 case "title":
1369 $match = @preg_match("/$reg_exp/i", $title);
1370 break;
1371 case "content":
1372 // we don't need to deal with multiline regexps
1373 $content = preg_replace("/[\r\n\t]/", "", $content);
1374
1375 $match = @preg_match("/$reg_exp/i", $content);
1376 break;
1377 case "both":
1378 // we don't need to deal with multiline regexps
1379 $content = preg_replace("/[\r\n\t]/", "", $content);
1380
1381 $match = (@preg_match("/$reg_exp/i", $title) || @preg_match("/$reg_exp/i", $content));
1382 break;
1383 case "link":
1384 $match = @preg_match("/$reg_exp/i", $link);
1385 break;
1386 case "author":
1387 $match = @preg_match("/$reg_exp/i", $author);
1388 break;
1389 case "tag":
1390 foreach ($tags as $tag) {
1391 if (@preg_match("/$reg_exp/i", $tag)) {
1392 $match = true;
1393 break;
1394 }
1395 }
1396 break;
1397 }
1398
1399 if ($rule_inverse) $match = !$match;
1400
1401 if ($match_any_rule) {
1402 if ($match) {
1403 $filter_match = true;
1404 break;
1405 }
1406 } else {
1407 $filter_match = $match;
1408 if (!$match) {
1409 break;
1410 }
1411 }
1412 }
1413
1414 if ($inverse) $filter_match = !$filter_match;
1415
1416 if ($filter_match) {
1417 foreach ($filter["actions"] AS $action) {
1418 array_push($matches, $action);
1419
1420 // if Stop action encountered, perform no further processing
1421 if (isset($action["type"]) && $action["type"] == "stop") return $matches;
1422 }
1423 }
1424 }
1425
1426 return $matches;
1427 }
1428
1429 function find_article_filter($filters, $filter_name) {
1430 foreach ($filters as $f) {
1431 if ($f["type"] == $filter_name) {
1432 return $f;
1433 };
1434 }
1435 return false;
1436 }
1437
1438 function find_article_filters($filters, $filter_name) {
1439 $results = array();
1440
1441 foreach ($filters as $f) {
1442 if ($f["type"] == $filter_name) {
1443 array_push($results, $f);
1444 };
1445 }
1446 return $results;
1447 }
1448
1449 function calculate_article_score($filters) {
1450 $score = 0;
1451
1452 foreach ($filters as $f) {
1453 if ($f["type"] == "score") {
1454 $score += $f["param"];
1455 };
1456 }
1457 return $score;
1458 }
1459
1460 function labels_contains_caption($labels, $caption) {
1461 foreach ($labels as $label) {
1462 if ($label[1] == $caption) {
1463 return true;
1464 }
1465 }
1466
1467 return false;
1468 }
1469
1470 function assign_article_to_label_filters($id, $filters, $owner_uid, $article_labels) {
1471 foreach ($filters as $f) {
1472 if ($f["type"] == "label") {
1473 if (!labels_contains_caption($article_labels, $f["param"])) {
1474 label_add_article($id, $f["param"], $owner_uid);
1475 }
1476 }
1477 }
1478 }
1479
1480 function make_guid_from_title($title) {
1481 return preg_replace("/[ \"\',.:;]/", "-",
1482 mb_strtolower(strip_tags($title), 'utf-8'));
1483 }
1484
1485 /* function verify_feed_xml($feed_data) {
1486 libxml_use_internal_errors(true);
1487 $doc = new DOMDocument();
1488 $doc->loadXML($feed_data);
1489 $error = libxml_get_last_error();
1490 libxml_clear_errors();
1491 return $error;
1492 } */
1493
1494 function cleanup_counters_cache($debug) {
1495 $result = db_query("DELETE FROM ttrss_counters_cache
1496 WHERE feed_id > 0 AND
1497 (SELECT COUNT(id) FROM ttrss_feeds WHERE
1498 id = feed_id AND
1499 ttrss_counters_cache.owner_uid = ttrss_feeds.owner_uid) = 0");
1500 $frows = db_affected_rows($result);
1501
1502 $result = db_query("DELETE FROM ttrss_cat_counters_cache
1503 WHERE feed_id > 0 AND
1504 (SELECT COUNT(id) FROM ttrss_feed_categories WHERE
1505 id = feed_id AND
1506 ttrss_cat_counters_cache.owner_uid = ttrss_feed_categories.owner_uid) = 0");
1507 $crows = db_affected_rows($result);
1508
1509 _debug("Removed $frows (feeds) $crows (cats) orphaned counter cache entries.");
1510 }
1511
1512 function housekeeping_common($debug) {
1513 expire_cached_files($debug);
1514 expire_lock_files($debug);
1515 expire_error_log($debug);
1516
1517 $count = update_feedbrowser_cache();
1518 _debug("Feedbrowser updated, $count feeds processed.");
1519
1520 purge_orphans( true);
1521 cleanup_counters_cache($debug);
1522
1523 //$rc = cleanup_tags( 14, 50000);
1524 //_debug("Cleaned $rc cached tags.");
1525
1526 PluginHost::getInstance()->run_hooks(PluginHost::HOOK_HOUSE_KEEPING, "hook_house_keeping", "");
1527
1528 }
1529 ?>
web-based news feed (RSS/Atom) reader and aggregator; see https://tt-rss.org/