]> git.wh0rd.org - tt-rss.git/blob - include/rssfuncs.php
git.wh0rd.org / tt-rss.git / blob
? search:
summary | shortlog | log | commit | commitdiff | tree
blame | history | raw | HEAD
experimental: decode numerical utf entities on import in entry title
[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
6 function update_feedbrowser_cache() {
7
8 $result = db_query("SELECT feed_url, site_url, title, COUNT(id) AS subscribers
9 FROM ttrss_feeds WHERE (SELECT COUNT(id) = 0 FROM ttrss_feeds AS tf
10 WHERE tf.feed_url = ttrss_feeds.feed_url
11 AND (private IS true OR auth_login != '' OR auth_pass != '' OR feed_url LIKE '%:%@%/%'))
12 GROUP BY feed_url, site_url, title ORDER BY subscribers DESC LIMIT 1000");
13
14 db_query("BEGIN");
15
16 db_query("DELETE FROM ttrss_feedbrowser_cache");
17
18 $count = 0;
19
20 while ($line = db_fetch_assoc($result)) {
21 $subscribers = db_escape_string($line["subscribers"]);
22 $feed_url = db_escape_string($line["feed_url"]);
23 $title = db_escape_string($line["title"]);
24 $site_url = db_escape_string($line["site_url"]);
25
26 $tmp_result = db_query("SELECT subscribers FROM
27 ttrss_feedbrowser_cache WHERE feed_url = '$feed_url'");
28
29 if (db_num_rows($tmp_result) == 0) {
30
31 db_query("INSERT INTO ttrss_feedbrowser_cache
32 (feed_url, site_url, title, subscribers) VALUES ('$feed_url',
33 '$site_url', '$title', '$subscribers')");
34
35 ++$count;
36
37 }
38
39 }
40
41 db_query("COMMIT");
42
43 return $count;
44
45 }
46
47
48 /**
49 * Update a feed batch.
50 * Used by daemons to update n feeds by run.
51 * Only update feed needing a update, and not being processed
52 * by another process.
53 *
54 * @param mixed $link Database link
55 * @param integer $limit Maximum number of feeds in update batch. Default to DAEMON_FEED_LIMIT.
56 * @param boolean $from_http Set to true if you call this function from http to disable cli specific code.
57 * @param boolean $debug Set to false to disable debug output. Default to true.
58 * @return void
59 */
60 function update_daemon_common($limit = DAEMON_FEED_LIMIT, $from_http = false, $debug = true) {
61 // Process all other feeds using last_updated and interval parameters
62
63 $schema_version = get_schema_version();
64
65 if ($schema_version != SCHEMA_VERSION) {
66 die("Schema version is wrong, please upgrade the database.\n");
67 }
68
69 define('PREFS_NO_CACHE', true);
70
71 // Test if the user has loggued in recently. If not, it does not update its feeds.
72 if (!SINGLE_USER_MODE && DAEMON_UPDATE_LOGIN_LIMIT > 0) {
73 if (DB_TYPE == "pgsql") {
74 $login_thresh_qpart = "AND ttrss_users.last_login >= NOW() - INTERVAL '".DAEMON_UPDATE_LOGIN_LIMIT." days'";
75 } else {
76 $login_thresh_qpart = "AND ttrss_users.last_login >= DATE_SUB(NOW(), INTERVAL ".DAEMON_UPDATE_LOGIN_LIMIT." DAY)";
77 }
78 } else {
79 $login_thresh_qpart = "";
80 }
81
82 // Test if the feed need a update (update interval exceded).
83 if (DB_TYPE == "pgsql") {
84 $update_limit_qpart = "AND ((
85 ttrss_feeds.update_interval = 0
86 AND ttrss_user_prefs.value != '-1'
87 AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_user_prefs.value || ' minutes') AS INTERVAL)
88 ) OR (
89 ttrss_feeds.update_interval > 0
90 AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_feeds.update_interval || ' minutes') AS INTERVAL)
91 ) OR ttrss_feeds.last_updated IS NULL
92 OR last_updated = '1970-01-01 00:00:00')";
93 } else {
94 $update_limit_qpart = "AND ((
95 ttrss_feeds.update_interval = 0
96 AND ttrss_user_prefs.value != '-1'
97 AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL CONVERT(ttrss_user_prefs.value, SIGNED INTEGER) MINUTE)
98 ) OR (
99 ttrss_feeds.update_interval > 0
100 AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL ttrss_feeds.update_interval MINUTE)
101 ) OR ttrss_feeds.last_updated IS NULL
102 OR last_updated = '1970-01-01 00:00:00')";
103 }
104
105 // Test if feed is currently being updated by another process.
106 if (DB_TYPE == "pgsql") {
107 $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < NOW() - INTERVAL '10 minutes')";
108 } else {
109 $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < DATE_SUB(NOW(), INTERVAL 10 MINUTE))";
110 }
111
112 // Test if there is a limit to number of updated feeds
113 $query_limit = "";
114 if($limit) $query_limit = sprintf("LIMIT %d", $limit);
115
116 $query = "SELECT DISTINCT ttrss_feeds.feed_url, ttrss_feeds.last_updated
117 FROM
118 ttrss_feeds, ttrss_users, ttrss_user_prefs
119 WHERE
120 ttrss_feeds.owner_uid = ttrss_users.id
121 AND ttrss_users.id = ttrss_user_prefs.owner_uid
122 AND ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL'
123 $login_thresh_qpart $update_limit_qpart
124 $updstart_thresh_qpart
125 ORDER BY last_updated $query_limit";
126
127 // We search for feed needing update.
128 $result = db_query($query);
129
130 if($debug) _debug(sprintf("Scheduled %d feeds to update...", db_num_rows($result)));
131
132 // Here is a little cache magic in order to minimize risk of double feed updates.
133 $feeds_to_update = array();
134 while ($line = db_fetch_assoc($result)) {
135 array_push($feeds_to_update, db_escape_string($line['feed_url']));
136 }
137
138 // We update the feed last update started date before anything else.
139 // There is no lag due to feed contents downloads
140 // It prevent an other process to update the same feed.
141
142 if(count($feeds_to_update) > 0) {
143 $feeds_quoted = array();
144
145 foreach ($feeds_to_update as $feed) {
146 array_push($feeds_quoted, "'" . db_escape_string($feed) . "'");
147 }
148
149 db_query(sprintf("UPDATE ttrss_feeds SET last_update_started = NOW()
150 WHERE feed_url IN (%s)", implode(',', $feeds_quoted)));
151 }
152
153 $nf = 0;
154
155 // For each feed, we call the feed update function.
156 foreach ($feeds_to_update as $feed) {
157 if($debug) _debug("Base feed: $feed");
158
159 //update_rss_feed($line["id"], true);
160
161 // since we have the data cached, we can deal with other feeds with the same url
162
163 $tmp_result = db_query("SELECT DISTINCT ttrss_feeds.id,last_updated
164 FROM ttrss_feeds, ttrss_users, ttrss_user_prefs WHERE
165 ttrss_user_prefs.owner_uid = ttrss_feeds.owner_uid AND
166 ttrss_users.id = ttrss_user_prefs.owner_uid AND
167 ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL' AND
168 feed_url = '".db_escape_string($feed)."' AND
169 (ttrss_feeds.update_interval > 0 OR
170 ttrss_user_prefs.value != '-1')
171 $login_thresh_qpart
172 ORDER BY ttrss_feeds.id $query_limit");
173
174 if (db_num_rows($tmp_result) > 0) {
175 while ($tline = db_fetch_assoc($tmp_result)) {
176 if($debug) _debug(" => " . $tline["last_updated"] . ", " . $tline["id"]);
177 update_rss_feed($tline["id"], true);
178 ++$nf;
179 }
180 }
181 }
182
183 require_once "digest.php";
184
185 // Send feed digests by email if needed.
186 send_headlines_digests($debug);
187
188 return $nf;
189
190 } // function update_daemon_common
191
192 // ignore_daemon is not used
193 function update_rss_feed($feed, $ignore_daemon = false, $no_cache = false,
194 $override_url = false) {
195
196 require_once "lib/simplepie/simplepie.inc";
197
198 $debug_enabled = defined('DAEMON_EXTENDED_DEBUG') || $_REQUEST['xdebug'];
199
200 _debug("start", $debug_enabled);
201
202 $result = db_query("SELECT id,update_interval,auth_login,
203 feed_url,auth_pass,cache_images,last_updated,
204 mark_unread_on_update, owner_uid,
205 pubsub_state, auth_pass_encrypted,
206 (SELECT max(date_entered) FROM
207 ttrss_entries, ttrss_user_entries where ref_id = id AND feed_id = '$feed') AS last_article_timestamp
208 FROM ttrss_feeds WHERE id = '$feed'");
209
210 if (db_num_rows($result) == 0) {
211 _debug("feed $feed NOT FOUND/SKIPPED", $debug_enabled);
212 return false;
213 }
214
215 $last_updated = db_fetch_result($result, 0, "last_updated");
216 $last_article_timestamp = @strtotime(db_fetch_result($result, 0, "last_article_timestamp"));
217 $owner_uid = db_fetch_result($result, 0, "owner_uid");
218 $mark_unread_on_update = sql_bool_to_bool(db_fetch_result($result,
219 0, "mark_unread_on_update"));
220 $pubsub_state = db_fetch_result($result, 0, "pubsub_state");
221 $auth_pass_encrypted = sql_bool_to_bool(db_fetch_result($result,
222 0, "auth_pass_encrypted"));
223
224 db_query("UPDATE ttrss_feeds SET last_update_started = NOW()
225 WHERE id = '$feed'");
226
227 $auth_login = db_fetch_result($result, 0, "auth_login");
228 $auth_pass = db_fetch_result($result, 0, "auth_pass");
229
230 if ($auth_pass_encrypted) {
231 require_once "crypt.php";
232 $auth_pass = decrypt_string($auth_pass);
233 }
234
235 $cache_images = sql_bool_to_bool(db_fetch_result($result, 0, "cache_images"));
236 $fetch_url = db_fetch_result($result, 0, "feed_url");
237
238 $feed = db_escape_string($feed);
239
240 if ($override_url) $fetch_url = $override_url;
241
242 $date_feed_processed = date('Y-m-d H:i');
243
244 $cache_filename = CACHE_DIR . "/simplepie/" . sha1($fetch_url) . ".feed";
245
246 // Ignore cache if new feed or manual update.
247 $cache_age = ($no_cache || is_null($last_updated) || strpos($last_updated, '1970-01-01') === 0) ? 30 : get_feed_update_interval($feed) * 60;
248
249 _debug("cache filename: $cache_filename exists: " . file_exists($cache_filename), $debug_enabled);
250 _debug("cache age: $cache_age; no cache: $no_cache", $debug_enabled);
251
252 $cached_feed_data_hash = false;
253
254 $rss = false;
255 $rss_hash = false;
256 $cache_timestamp = file_exists($cache_filename) ? filemtime($cache_filename) : 0;
257
258 $force_refetch = isset($_REQUEST["force_refetch"]);
259
260 if (file_exists($cache_filename) &&
261 is_readable($cache_filename) &&
262 !$auth_login && !$auth_pass &&
263 filemtime($cache_filename) > time() - $cache_age) {
264
265 _debug("using local cache.", $debug_enabled);
266
267 if ($cache_timestamp > $last_article_timestamp) {
268 @$rss_data = file_get_contents($cache_filename);
269
270 if ($rss_data) {
271 $rss_hash = sha1($rss_data);
272 @$rss = unserialize($rss_data);
273 }
274 } else if (!$force_refetch) {
275 _debug("local cache valid and older than last_updated, nothing to do.", $debug_enabled);
276 return;
277 }
278 } else {
279 _debug("local cache will not be used for this feed", $debug_enabled);
280 }
281
282 if (!$rss) {
283
284 if (!$feed_data) {
285 _debug("fetching [$fetch_url]...", $debug_enabled);
286 _debug("If-Modified-Since: ".gmdate('D, d M Y H:i:s \G\M\T', $last_article_timestamp), $debug_enabled);
287
288 $feed_data = fetch_file_contents($fetch_url, false,
289 $auth_login, $auth_pass, false,
290 $no_cache ? FEED_FETCH_NO_CACHE_TIMEOUT : FEED_FETCH_TIMEOUT,
291 $force_refetch ? 0 : $last_article_timestamp);
292
293 global $fetch_curl_used;
294
295 if (!$fetch_curl_used) {
296 $tmp = @gzdecode($feed_data);
297
298 if ($tmp) $feed_data = $tmp;
299 }
300
301 $feed_data = trim($feed_data);
302
303 _debug("fetch done.", $debug_enabled);
304
305 if ($feed_data) {
306 $error = verify_feed_xml($feed_data);
307
308 if ($error) {
309 _debug("error verifying XML, code: " . $error->code, $debug_enabled);
310
311 if ($error->code == 26) {
312 _debug("got error 26, trying to decode entities...", $debug_enabled);
313
314 $feed_data = html_entity_decode($feed_data, ENT_COMPAT, 'UTF-8');
315
316 $error = verify_feed_xml($feed_data);
317
318 if ($error) $feed_data = '';
319 }
320 }
321 }
322 }
323
324 if (!$feed_data) {
325 global $fetch_last_error;
326 global $fetch_last_error_code;
327
328 _debug("unable to fetch: $fetch_last_error [$fetch_last_error_code]", $debug_enabled);
329
330 $error_escaped = '';
331
332 // If-Modified-Since
333 if ($fetch_last_error_code != 304) {
334 $error_escaped = db_escape_string($fetch_last_error);
335 } else {
336 _debug("source claims data not modified, nothing to do.", $debug_enabled);
337 }
338
339 db_query(
340 "UPDATE ttrss_feeds SET last_error = '$error_escaped',
341 last_updated = NOW() WHERE id = '$feed'");
342
343 return;
344 }
345 }
346
347 $pluginhost = new PluginHost();
348 $pluginhost->set_debug($debug_enabled, $debug_enabled);
349 $user_plugins = get_pref("_ENABLED_PLUGINS", $owner_uid);
350
351 $pluginhost->load(PLUGINS, PluginHost::KIND_ALL);
352 $pluginhost->load($user_plugins, PluginHost::KIND_USER, $owner_uid);
353 $pluginhost->load_data();
354
355 foreach ($pluginhost->get_hooks(PluginHost::HOOK_FEED_FETCHED) as $plugin) {
356 $feed_data = $plugin->hook_feed_fetched($feed_data);
357 }
358
359 if (!$rss) {
360 $rss = new SimplePie();
361 $rss->set_sanitize_class("SanitizeDummy");
362 // simplepie ignores the above and creates default sanitizer anyway,
363 // so let's override it...
364 $rss->sanitize = new SanitizeDummy();
365 $rss->set_output_encoding('UTF-8');
366 $rss->set_raw_data($feed_data);
367 $rss->enable_cache(false);
368
369 @$rss->init();
370 }
371
372 // print_r($rss);
373
374 $feed = db_escape_string($feed);
375
376 if (!$rss->error()) {
377
378 // cache data for later
379 if (!$auth_pass && !$auth_login && is_writable(CACHE_DIR . "/simplepie")) {
380 $rss_data = serialize($rss);
381 $new_rss_hash = sha1($rss_data);
382
383 if ($new_rss_hash != $rss_hash) {
384 _debug("saving $cache_filename", $debug_enabled);
385 @file_put_contents($cache_filename, serialize($rss));
386 }
387 }
388
389 // We use local pluginhost here because we need to load different per-user feed plugins
390 $pluginhost->run_hooks(PluginHost::HOOK_FEED_PARSED, "hook_feed_parsed", $rss);
391
392 _debug("processing feed data...", $debug_enabled);
393
394 // db_query("BEGIN");
395
396 if (DB_TYPE == "pgsql") {
397 $favicon_interval_qpart = "favicon_last_checked < NOW() - INTERVAL '12 hour'";
398 } else {
399 $favicon_interval_qpart = "favicon_last_checked < DATE_SUB(NOW(), INTERVAL 12 HOUR)";
400 }
401
402 $result = db_query("SELECT title,site_url,owner_uid,favicon_avg_color,
403 (favicon_last_checked IS NULL OR $favicon_interval_qpart) AS
404 favicon_needs_check
405 FROM ttrss_feeds WHERE id = '$feed'");
406
407 $registered_title = db_fetch_result($result, 0, "title");
408 $orig_site_url = db_fetch_result($result, 0, "site_url");
409 $favicon_needs_check = sql_bool_to_bool(db_fetch_result($result, 0,
410 "favicon_needs_check"));
411 $favicon_avg_color = db_fetch_result($result, 0, "favicon_avg_color");
412
413 $owner_uid = db_fetch_result($result, 0, "owner_uid");
414
415 $site_url = db_escape_string(mb_substr(rewrite_relative_url($fetch_url, $rss->get_link()), 0, 245));
416
417 if ($favicon_needs_check || $force_refetch) {
418
419 /* terrible hack: if we crash on floicon shit here, we won't check
420 * the icon avgcolor again (unless the icon got updated) */
421
422 $favicon_file = ICONS_DIR . "/$feed.ico";
423 $favicon_modified = @filemtime($favicon_file);
424
425 _debug("checking favicon...", $debug_enabled);
426
427 check_feed_favicon($site_url, $feed, $link);
428 $favicon_modified_new = @filemtime($favicon_file);
429
430 if ($favicon_modified_new > $favicon_modified)
431 $favicon_avg_color = '';
432
433 if (file_exists($favicon_file) && function_exists("imagecreatefromstring") && $favicon_avg_color == '') {
434 require_once "colors.php";
435
436 db_query("UPDATE ttrss_feeds SET favicon_avg_color = 'fail' WHERE
437 id = '$feed'");
438
439 $favicon_color = db_escape_string(
440 calculate_avg_color($favicon_file));
441
442 $favicon_colorstring = ",favicon_avg_color = '".$favicon_color."'";
443 } else if ($favicon_avg_color == 'fail') {
444 _debug("floicon failed on this file, not trying to recalculate avg color", $debug_enabled);
445 }
446
447 db_query("UPDATE ttrss_feeds SET favicon_last_checked = NOW()
448 $favicon_colorstring
449 WHERE id = '$feed'");
450 }
451
452 if (!$registered_title || $registered_title == "[Unknown]") {
453
454 $feed_title = db_escape_string($rss->get_title());
455
456 _debug("registering title: $feed_title", $debug_enabled);
457
458 db_query("UPDATE ttrss_feeds SET
459 title = '$feed_title' WHERE id = '$feed'");
460 }
461
462 if ($site_url && $orig_site_url != $site_url) {
463 db_query("UPDATE ttrss_feeds SET
464 site_url = '$site_url' WHERE id = '$feed'");
465 }
466
467 _debug("loading filters & labels...", $debug_enabled);
468
469 $filters = load_filters($feed, $owner_uid);
470 $labels = get_all_labels($owner_uid);
471
472 _debug("" . count($filters) . " filters loaded.", $debug_enabled);
473
474 $items = $rss->get_items();
475
476 if (!is_array($items)) {
477 _debug("no articles found.", $debug_enabled);
478
479 db_query("UPDATE ttrss_feeds
480 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
481
482 return; // no articles
483 }
484
485 if ($pubsub_state != 2 && PUBSUBHUBBUB_ENABLED) {
486
487 _debug("checking for PUSH hub...", $debug_enabled);
488
489 $feed_hub_url = false;
490
491 $links = $rss->get_links('hub');
492
493 if ($links && is_array($links)) {
494 foreach ($links as $l) {
495 $feed_hub_url = $l;
496 break;
497 }
498 }
499
500 _debug("feed hub url: $feed_hub_url", $debug_enabled);
501
502 if ($feed_hub_url && function_exists('curl_init') &&
503 !ini_get("open_basedir")) {
504
505 require_once 'lib/pubsubhubbub/subscriber.php';
506
507 $callback_url = get_self_url_prefix() .
508 "/public.php?op=pubsub&id=$feed";
509
510 $s = new Subscriber($feed_hub_url, $callback_url);
511
512 $rc = $s->subscribe($fetch_url);
513
514 _debug("feed hub url found, subscribe request sent.", $debug_enabled);
515
516 db_query("UPDATE ttrss_feeds SET pubsub_state = 1
517 WHERE id = '$feed'");
518 }
519 }
520
521 _debug("processing articles...", $debug_enabled);
522
523 foreach ($items as $item) {
524 if ($_REQUEST['xdebug'] == 3) {
525 print_r($item);
526 }
527
528 $entry_guid = $item->get_id();
529 if (!$entry_guid) $entry_guid = $item->get_link();
530 if (!$entry_guid) $entry_guid = make_guid_from_title($item->get_title());
531
532 if (!$entry_guid) continue;
533
534 $entry_guid = "$owner_uid,$entry_guid";
535
536 $entry_guid_hashed = db_escape_string('SHA1:' . sha1($entry_guid));
537
538 _debug("guid $entry_guid / $entry_guid_hashed", $debug_enabled);
539
540 $entry_timestamp = "";
541
542 $entry_timestamp = strtotime($item->get_date());
543
544 if ($entry_timestamp == -1 || !$entry_timestamp || $entry_timestamp > time()) {
545 $entry_timestamp = time();
546 $no_orig_date = 'true';
547 } else {
548 $no_orig_date = 'false';
549 }
550
551 $entry_timestamp_fmt = strftime("%Y/%m/%d %H:%M:%S", $entry_timestamp);
552
553 _debug("date $entry_timestamp [$entry_timestamp_fmt]", $debug_enabled);
554
555 $entry_title = html_entity_decode($item->get_title(), ENT_COMPAT, 'UTF-8');
556 $entry_title = decode_numeric_entities($entry_title);
557
558 $entry_link = rewrite_relative_url($site_url, $item->get_link());
559
560 _debug("title $entry_title", $debug_enabled);
561 _debug("link $entry_link", $debug_enabled);
562
563 if (!$entry_title) $entry_title = date("Y-m-d H:i:s", $entry_timestamp);;
564
565 $entry_content = $item->get_content();
566 if (!$entry_content) $entry_content = $item->get_description();
567
568 if ($_REQUEST["xdebug"] == 2) {
569 print "content: ";
570 print $entry_content;
571 print "\n";
572 }
573
574 $entry_comments = $item->data["comments"];
575
576 if ($item->get_author()) {
577 $entry_author_item = $item->get_author();
578 $entry_author = $entry_author_item->get_name();
579 if (!$entry_author) $entry_author = $entry_author_item->get_email();
580 }
581
582 $entry_guid = db_escape_string(mb_substr($entry_guid, 0, 245));
583
584 $entry_comments = db_escape_string(mb_substr(trim($entry_comments), 0, 245));
585 $entry_author = db_escape_string(mb_substr(trim($entry_author), 0, 245));
586
587 $num_comments = $item->get_item_tags('http://purl.org/rss/1.0/modules/slash/', 'comments');
588
589 if (is_array($num_comments) && is_array($num_comments[0])) {
590 $num_comments = (int) $num_comments[0]["data"];
591 } else {
592 $num_comments = 0;
593 }
594
595 _debug("author $entry_author", $debug_enabled);
596 _debug("num_comments: $num_comments", $debug_enabled);
597 _debug("looking for tags [1]...", $debug_enabled);
598
599 // parse <category> entries into tags
600
601 $additional_tags = array();
602
603 $additional_tags_src = $item->get_categories();
604
605 if (is_array($additional_tags_src)) {
606 foreach ($additional_tags_src as $tobj) {
607 array_push($additional_tags, $tobj->get_term());
608 }
609 }
610
611 _debug("category tags:", $debug_enabled);
612 _debug("looking for tags [2]...", $debug_enabled);
613
614 $entry_tags = array_unique($additional_tags);
615
616 for ($i = 0; $i < count($entry_tags); $i++)
617 $entry_tags[$i] = mb_strtolower($entry_tags[$i], 'utf-8');
618
619 _debug("done collecting data.", $debug_enabled);
620
621 // TODO: less memory-hungry implementation
622
623 _debug("applying plugin filters..", $debug_enabled);
624
625 // FIXME not sure if owner_uid is a good idea here, we may have a base entry without user entry (?)
626 $result = db_query("SELECT plugin_data,title,content,link,tag_cache,author FROM ttrss_entries, ttrss_user_entries
627 WHERE ref_id = id AND (guid = '".db_escape_string($entry_guid)."' OR guid = '$entry_guid_hashed') AND owner_uid = $owner_uid");
628
629 if (db_num_rows($result) != 0) {
630 $entry_plugin_data = db_fetch_result($result, 0, "plugin_data");
631 $stored_article = array("title" => db_fetch_result($result, 0, "title"),
632 "content" => db_fetch_result($result, 0, "content"),
633 "link" => db_fetch_result($result, 0, "link"),
634 "tags" => explode(",", db_fetch_result($result, 0, "tag_cache")),
635 "author" => db_fetch_result($result, 0, "author"));
636 } else {
637 $entry_plugin_data = "";
638 $stored_article = array();
639 }
640
641 $article = array("owner_uid" => $owner_uid, // read only
642 "guid" => $entry_guid, // read only
643 "title" => $entry_title,
644 "content" => $entry_content,
645 "link" => $entry_link,
646 "tags" => $entry_tags,
647 "plugin_data" => $entry_plugin_data,
648 "author" => $entry_author,
649 "stored" => $stored_article);
650
651 foreach ($pluginhost->get_hooks(PluginHost::HOOK_ARTICLE_FILTER) as $plugin) {
652 $article = $plugin->hook_article_filter($article);
653 }
654
655 $entry_tags = $article["tags"];
656 $entry_guid = db_escape_string($entry_guid);
657 $entry_title = db_escape_string($article["title"]);
658 $entry_author = db_escape_string($article["author"]);
659 $entry_link = db_escape_string($article["link"]);
660 $entry_plugin_data = db_escape_string($article["plugin_data"]);
661 $entry_content = $article["content"]; // escaped below
662
663
664 _debug("plugin data: $entry_plugin_data", $debug_enabled);
665
666 if ($cache_images && is_writable(CACHE_DIR . '/images'))
667 cache_images($entry_content, $site_url, $debug_enabled);
668
669 $entry_content = db_escape_string($entry_content, false);
670
671 $content_hash = "SHA1:" . sha1($entry_content);
672
673 db_query("BEGIN");
674
675 $result = db_query("SELECT id FROM ttrss_entries
676 WHERE (guid = '$entry_guid' OR guid = '$entry_guid_hashed')");
677
678 if (db_num_rows($result) == 0) {
679
680 _debug("base guid [$entry_guid] not found", $debug_enabled);
681
682 // base post entry does not exist, create it
683
684 $result = db_query(
685 "INSERT INTO ttrss_entries
686 (title,
687 guid,
688 link,
689 updated,
690 content,
691 content_hash,
692 cached_content,
693 no_orig_date,
694 date_updated,
695 date_entered,
696 comments,
697 num_comments,
698 plugin_data,
699 author)
700 VALUES
701 ('$entry_title',
702 '$entry_guid_hashed',
703 '$entry_link',
704 '$entry_timestamp_fmt',
705 '$entry_content',
706 '$content_hash',
707 '',
708 $no_orig_date,
709 NOW(),
710 '$date_feed_processed',
711 '$entry_comments',
712 '$num_comments',
713 '$entry_plugin_data',
714 '$entry_author')");
715
716 $article_labels = array();
717
718 } else {
719 // we keep encountering the entry in feeds, so we need to
720 // update date_updated column so that we don't get horrible
721 // dupes when the entry gets purged and reinserted again e.g.
722 // in the case of SLOW SLOW OMG SLOW updating feeds
723
724 $base_entry_id = db_fetch_result($result, 0, "id");
725
726 db_query("UPDATE ttrss_entries SET date_updated = NOW()
727 WHERE id = '$base_entry_id'");
728
729 $article_labels = get_article_labels($base_entry_id, $owner_uid);
730 }
731
732 // now it should exist, if not - bad luck then
733
734 $result = db_query("SELECT
735 id,content_hash,no_orig_date,title,plugin_data,guid,
736 ".SUBSTRING_FOR_DATE."(date_updated,1,19) as date_updated,
737 ".SUBSTRING_FOR_DATE."(updated,1,19) as updated,
738 num_comments
739 FROM
740 ttrss_entries
741 WHERE guid = '$entry_guid' OR guid = '$entry_guid_hashed'");
742
743 $entry_ref_id = 0;
744 $entry_int_id = 0;
745
746 if (db_num_rows($result) == 1) {
747
748 _debug("base guid found, checking for user record", $debug_enabled);
749
750 // this will be used below in update handler
751 $orig_content_hash = db_fetch_result($result, 0, "content_hash");
752 $orig_title = db_fetch_result($result, 0, "title");
753 $orig_num_comments = db_fetch_result($result, 0, "num_comments");
754 $orig_date_updated = strtotime(db_fetch_result($result,
755 0, "date_updated"));
756 $orig_plugin_data = db_fetch_result($result, 0, "plugin_data");
757
758 $ref_id = db_fetch_result($result, 0, "id");
759 $entry_ref_id = $ref_id;
760
761 /* $stored_guid = db_fetch_result($result, 0, "guid");
762 if ($stored_guid != $entry_guid_hashed) {
763 if ($debug_enabled) _debug("upgrading compat guid to hashed one", $debug_enabled);
764
765 db_query("UPDATE ttrss_entries SET guid = '$entry_guid_hashed' WHERE
766 id = '$ref_id'");
767 } */
768
769 // check for user post link to main table
770
771 // do we allow duplicate posts with same GUID in different feeds?
772 if (get_pref("ALLOW_DUPLICATE_POSTS", $owner_uid, false)) {
773 $dupcheck_qpart = "AND (feed_id = '$feed' OR feed_id IS NULL)";
774 } else {
775 $dupcheck_qpart = "";
776 }
777
778 /* Collect article tags here so we could filter by them: */
779
780 $article_filters = get_article_filters($filters, $entry_title,
781 $entry_content, $entry_link, $entry_timestamp, $entry_author,
782 $entry_tags);
783
784 if ($debug_enabled) {
785 _debug("article filters: ", $debug_enabled);
786 if (count($article_filters) != 0) {
787 print_r($article_filters);
788 }
789 }
790
791 if (find_article_filter($article_filters, "filter")) {
792 db_query("COMMIT"); // close transaction in progress
793 continue;
794 }
795
796 $score = calculate_article_score($article_filters);
797
798 _debug("initial score: $score", $debug_enabled);
799
800 $query = "SELECT ref_id, int_id FROM ttrss_user_entries WHERE
801 ref_id = '$ref_id' AND owner_uid = '$owner_uid'
802 $dupcheck_qpart";
803
804 // if ($_REQUEST["xdebug"]) print "$query\n";
805
806 $result = db_query($query);
807
808 // okay it doesn't exist - create user entry
809 if (db_num_rows($result) == 0) {
810
811 _debug("user record not found, creating...", $debug_enabled);
812
813 if ($score >= -500 && !find_article_filter($article_filters, 'catchup')) {
814 $unread = 'true';
815 $last_read_qpart = 'NULL';
816 } else {
817 $unread = 'false';
818 $last_read_qpart = 'NOW()';
819 }
820
821 if (find_article_filter($article_filters, 'mark') || $score > 1000) {
822 $marked = 'true';
823 } else {
824 $marked = 'false';
825 }
826
827 if (find_article_filter($article_filters, 'publish')) {
828 $published = 'true';
829 } else {
830 $published = 'false';
831 }
832
833 // N-grams
834
835 if (DB_TYPE == "pgsql" and defined('_NGRAM_TITLE_DUPLICATE_THRESHOLD')) {
836
837 $result = db_query("SELECT COUNT(*) AS similar FROM
838 ttrss_entries,ttrss_user_entries
839 WHERE ref_id = id AND updated >= NOW() - INTERVAL '7 day'
840 AND similarity(title, '$entry_title') >= "._NGRAM_TITLE_DUPLICATE_THRESHOLD."
841 AND owner_uid = $owner_uid");
842
843 $ngram_similar = db_fetch_result($result, 0, "similar");
844
845 _debug("N-gram similar results: $ngram_similar", $debug_enabled);
846
847 if ($ngram_similar > 0) {
848 $unread = 'false';
849 }
850 }
851
852 $last_marked = ($marked == 'true') ? 'NOW()' : 'NULL';
853 $last_published = ($published == 'true') ? 'NOW()' : 'NULL';
854
855 $result = db_query(
856 "INSERT INTO ttrss_user_entries
857 (ref_id, owner_uid, feed_id, unread, last_read, marked,
858 published, score, tag_cache, label_cache, uuid,
859 last_marked, last_published)
860 VALUES ('$ref_id', '$owner_uid', '$feed', $unread,
861 $last_read_qpart, $marked, $published, '$score', '', '',
862 '', $last_marked, $last_published)");
863
864 if (PUBSUBHUBBUB_HUB && $published == 'true') {
865 $rss_link = get_self_url_prefix() .
866 "/public.php?op=rss&id=-2&key=" .
867 get_feed_access_key(-2, false, $owner_uid);
868
869 $p = new Publisher(PUBSUBHUBBUB_HUB);
870
871 $pubsub_result = $p->publish_update($rss_link);
872 }
873
874 $result = db_query(
875 "SELECT int_id FROM ttrss_user_entries WHERE
876 ref_id = '$ref_id' AND owner_uid = '$owner_uid' AND
877 feed_id = '$feed' LIMIT 1");
878
879 if (db_num_rows($result) == 1) {
880 $entry_int_id = db_fetch_result($result, 0, "int_id");
881 }
882 } else {
883 _debug("user record FOUND", $debug_enabled);
884
885 $entry_ref_id = db_fetch_result($result, 0, "ref_id");
886 $entry_int_id = db_fetch_result($result, 0, "int_id");
887 }
888
889 _debug("RID: $entry_ref_id, IID: $entry_int_id", $debug_enabled);
890
891 $post_needs_update = false;
892 $update_insignificant = false;
893
894 if ($orig_num_comments != $num_comments) {
895 $post_needs_update = true;
896 $update_insignificant = true;
897 }
898
899 if ($entry_plugin_data != $orig_plugin_data) {
900 $post_needs_update = true;
901 $update_insignificant = true;
902 }
903
904 if ($content_hash != $orig_content_hash) {
905 $post_needs_update = true;
906 $update_insignificant = false;
907 }
908
909 if (db_escape_string($orig_title) != $entry_title) {
910 $post_needs_update = true;
911 $update_insignificant = false;
912 }
913
914 // if post needs update, update it and mark all user entries
915 // linking to this post as updated
916 if ($post_needs_update) {
917
918 if (defined('DAEMON_EXTENDED_DEBUG')) {
919 _debug("post $entry_guid_hashed needs update...", $debug_enabled);
920 }
921
922 // print "<!-- post $orig_title needs update : $post_needs_update -->";
923
924 db_query("UPDATE ttrss_entries
925 SET title = '$entry_title', content = '$entry_content',
926 content_hash = '$content_hash',
927 updated = '$entry_timestamp_fmt',
928 num_comments = '$num_comments',
929 plugin_data = '$entry_plugin_data'
930 WHERE id = '$ref_id'");
931
932 if (!$update_insignificant) {
933 if ($mark_unread_on_update) {
934 db_query("UPDATE ttrss_user_entries
935 SET last_read = null, unread = true WHERE ref_id = '$ref_id'");
936 }
937 }
938 }
939 }
940
941 db_query("COMMIT");
942
943 _debug("assigning labels...", $debug_enabled);
944
945 assign_article_to_label_filters($entry_ref_id, $article_filters,
946 $owner_uid, $article_labels);
947
948 _debug("looking for enclosures...", $debug_enabled);
949
950 // enclosures
951
952 $enclosures = array();
953
954 $encs = $item->get_enclosures();
955
956 if (is_array($encs)) {
957 foreach ($encs as $e) {
958 $e_item = array(
959 $e->link, $e->type, $e->length);
960 array_push($enclosures, $e_item);
961 }
962 }
963
964 if ($debug_enabled) {
965 _debug("article enclosures:", $debug_enabled);
966 print_r($enclosures);
967 }
968
969 db_query("BEGIN");
970
971 foreach ($enclosures as $enc) {
972 $enc_url = db_escape_string($enc[0]);
973 $enc_type = db_escape_string($enc[1]);
974 $enc_dur = db_escape_string($enc[2]);
975
976 $result = db_query("SELECT id FROM ttrss_enclosures
977 WHERE content_url = '$enc_url' AND post_id = '$entry_ref_id'");
978
979 if (db_num_rows($result) == 0) {
980 db_query("INSERT INTO ttrss_enclosures
981 (content_url, content_type, title, duration, post_id) VALUES
982 ('$enc_url', '$enc_type', '', '$enc_dur', '$entry_ref_id')");
983 }
984 }
985
986 db_query("COMMIT");
987
988 // check for manual tags (we have to do it here since they're loaded from filters)
989
990 foreach ($article_filters as $f) {
991 if ($f["type"] == "tag") {
992
993 $manual_tags = trim_array(explode(",", $f["param"]));
994
995 foreach ($manual_tags as $tag) {
996 if (tag_is_valid($tag)) {
997 array_push($entry_tags, $tag);
998 }
999 }
1000 }
1001 }
1002
1003 // Skip boring tags
1004
1005 $boring_tags = trim_array(explode(",", mb_strtolower(get_pref(
1006 'BLACKLISTED_TAGS', $owner_uid, ''), 'utf-8')));
1007
1008 $filtered_tags = array();
1009 $tags_to_cache = array();
1010
1011 if ($entry_tags && is_array($entry_tags)) {
1012 foreach ($entry_tags as $tag) {
1013 if (array_search($tag, $boring_tags) === false) {
1014 array_push($filtered_tags, $tag);
1015 }
1016 }
1017 }
1018
1019 $filtered_tags = array_unique($filtered_tags);
1020
1021 if ($debug_enabled) {
1022 _debug("filtered article tags:", $debug_enabled);
1023 print_r($filtered_tags);
1024 }
1025
1026 // Save article tags in the database
1027
1028 if (count($filtered_tags) > 0) {
1029
1030 db_query("BEGIN");
1031
1032 foreach ($filtered_tags as $tag) {
1033
1034 $tag = sanitize_tag($tag);
1035 $tag = db_escape_string($tag);
1036
1037 if (!tag_is_valid($tag)) continue;
1038
1039 $result = db_query("SELECT id FROM ttrss_tags
1040 WHERE tag_name = '$tag' AND post_int_id = '$entry_int_id' AND
1041 owner_uid = '$owner_uid' LIMIT 1");
1042
1043 if ($result && db_num_rows($result) == 0) {
1044
1045 db_query("INSERT INTO ttrss_tags
1046 (owner_uid,tag_name,post_int_id)
1047 VALUES ('$owner_uid','$tag', '$entry_int_id')");
1048 }
1049
1050 array_push($tags_to_cache, $tag);
1051 }
1052
1053 /* update the cache */
1054
1055 $tags_to_cache = array_unique($tags_to_cache);
1056
1057 $tags_str = db_escape_string(join(",", $tags_to_cache));
1058
1059 db_query("UPDATE ttrss_user_entries
1060 SET tag_cache = '$tags_str' WHERE ref_id = '$entry_ref_id'
1061 AND owner_uid = $owner_uid");
1062
1063 db_query("COMMIT");
1064 }
1065
1066 if (get_pref("AUTO_ASSIGN_LABELS", $owner_uid, false)) {
1067 _debug("auto-assigning labels...", $debug_enabled);
1068
1069 foreach ($labels as $label) {
1070 $caption = preg_quote($label["caption"]);
1071
1072 if ($caption && preg_match("/\b$caption\b/i", "$tags_str " . strip_tags($entry_content) . " $entry_title")) {
1073 if (!labels_contains_caption($article_labels, $caption)) {
1074 label_add_article($entry_ref_id, $caption, $owner_uid);
1075 }
1076 }
1077 }
1078 }
1079
1080 _debug("article processed", $debug_enabled);
1081 }
1082
1083 if (!$last_updated) {
1084 _debug("new feed, catching it up...", $debug_enabled);
1085 catchup_feed($feed, false, $owner_uid);
1086 }
1087
1088 _debug("purging feed...", $debug_enabled);
1089
1090 purge_feed($feed, 0, $debug_enabled);
1091
1092 db_query("UPDATE ttrss_feeds
1093 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
1094
1095 // db_query("COMMIT");
1096
1097 } else {
1098
1099 $error_msg = db_escape_string(mb_substr($rss->error(), 0, 245));
1100
1101 _debug("error fetching feed: $error_msg", $debug_enabled);
1102
1103 db_query(
1104 "UPDATE ttrss_feeds SET last_error = '$error_msg',
1105 last_updated = NOW() WHERE id = '$feed'");
1106 }
1107
1108 unset($rss);
1109
1110 _debug("done", $debug_enabled);
1111 }
1112
1113 function cache_images($html, $site_url, $debug) {
1114 $cache_dir = CACHE_DIR . "/images";
1115
1116 libxml_use_internal_errors(true);
1117
1118 $charset_hack = '<head>
1119 <meta http-equiv="Content-Type" content="text/html; charset=utf-8"/>
1120 </head>';
1121
1122 $doc = new DOMDocument();
1123 $doc->loadHTML($charset_hack . $html);
1124 $xpath = new DOMXPath($doc);
1125
1126 $entries = $xpath->query('(//img[@src])');
1127
1128 foreach ($entries as $entry) {
1129 if ($entry->hasAttribute('src')) {
1130 $src = rewrite_relative_url($site_url, $entry->getAttribute('src'));
1131
1132 $local_filename = CACHE_DIR . "/images/" . sha1($src) . ".png";
1133
1134 if ($debug) _debug("cache_images: downloading: $src to $local_filename");
1135
1136 if (!file_exists($local_filename)) {
1137 $file_content = fetch_file_contents($src);
1138
1139 if ($file_content && strlen($file_content) > 1024) {
1140 file_put_contents($local_filename, $file_content);
1141 }
1142 }
1143
1144 if (file_exists($local_filename)) {
1145 $entry->setAttribute('src', SELF_URL_PATH . '/image.php?url=' .
1146 base64_encode($src));
1147 }
1148 }
1149 }
1150
1151 $node = $doc->getElementsByTagName('body')->item(0);
1152
1153 return $doc->saveXML($node);
1154 }
1155
1156 function expire_error_log($debug) {
1157 if ($debug) _debug("Removing old error log entries...");
1158
1159 if (DB_TYPE == "pgsql") {
1160 db_query("DELETE FROM ttrss_error_log
1161 WHERE created_at < NOW() - INTERVAL '7 days'");
1162 } else {
1163 db_query("DELETE FROM ttrss_error_log
1164 WHERE created_at < DATE_SUB(NOW(), INTERVAL 7 DAY)");
1165 }
1166
1167 }
1168
1169 function expire_lock_files($debug) {
1170 //if ($debug) _debug("Removing old lock files...");
1171
1172 $num_deleted = 0;
1173
1174 if (is_writable(LOCK_DIRECTORY)) {
1175 $files = glob(LOCK_DIRECTORY . "/*.lock");
1176
1177 if ($files) {
1178 foreach ($files as $file) {
1179 if (!file_is_locked(basename($file)) && time() - filemtime($file) > 86400*2) {
1180 unlink($file);
1181 ++$num_deleted;
1182 }
1183 }
1184 }
1185 }
1186
1187 if ($debug) _debug("Removed $num_deleted old lock files.");
1188 }
1189
1190 function expire_cached_files($debug) {
1191 foreach (array("simplepie", "images", "export", "upload") as $dir) {
1192 $cache_dir = CACHE_DIR . "/$dir";
1193
1194 // if ($debug) _debug("Expiring $cache_dir");
1195
1196 $num_deleted = 0;
1197
1198 if (is_writable($cache_dir)) {
1199 $files = glob("$cache_dir/*");
1200
1201 if ($files) {
1202 foreach ($files as $file) {
1203 if (time() - filemtime($file) > 86400*7) {
1204 unlink($file);
1205
1206 ++$num_deleted;
1207 }
1208 }
1209 }
1210 }
1211
1212 if ($debug) _debug("$cache_dir: removed $num_deleted files.");
1213 }
1214 }
1215
1216 /**
1217 * Source: http://www.php.net/manual/en/function.parse-url.php#104527
1218 * Returns the url query as associative array
1219 *
1220 * @param string query
1221 * @return array params
1222 */
1223 function convertUrlQuery($query) {
1224 $queryParts = explode('&', $query);
1225
1226 $params = array();
1227
1228 foreach ($queryParts as $param) {
1229 $item = explode('=', $param);
1230 $params[$item[0]] = $item[1];
1231 }
1232
1233 return $params;
1234 }
1235
1236 function get_article_filters($filters, $title, $content, $link, $timestamp, $author, $tags) {
1237 $matches = array();
1238
1239 foreach ($filters as $filter) {
1240 $match_any_rule = $filter["match_any_rule"];
1241 $inverse = $filter["inverse"];
1242 $filter_match = false;
1243
1244 foreach ($filter["rules"] as $rule) {
1245 $match = false;
1246 $reg_exp = str_replace('/', '\/', $rule["reg_exp"]);
1247 $rule_inverse = $rule["inverse"];
1248
1249 if (!$reg_exp)
1250 continue;
1251
1252 switch ($rule["type"]) {
1253 case "title":
1254 $match = @preg_match("/$reg_exp/i", $title);
1255 break;
1256 case "content":
1257 // we don't need to deal with multiline regexps
1258 $content = preg_replace("/[\r\n\t]/", "", $content);
1259
1260 $match = @preg_match("/$reg_exp/i", $content);
1261 break;
1262 case "both":
1263 // we don't need to deal with multiline regexps
1264 $content = preg_replace("/[\r\n\t]/", "", $content);
1265
1266 $match = (@preg_match("/$reg_exp/i", $title) || @preg_match("/$reg_exp/i", $content));
1267 break;
1268 case "link":
1269 $match = @preg_match("/$reg_exp/i", $link);
1270 break;
1271 case "author":
1272 $match = @preg_match("/$reg_exp/i", $author);
1273 break;
1274 case "tag":
1275 foreach ($tags as $tag) {
1276 if (@preg_match("/$reg_exp/i", $tag)) {
1277 $match = true;
1278 break;
1279 }
1280 }
1281 break;
1282 }
1283
1284 if ($rule_inverse) $match = !$match;
1285
1286 if ($match_any_rule) {
1287 if ($match) {
1288 $filter_match = true;
1289 break;
1290 }
1291 } else {
1292 $filter_match = $match;
1293 if (!$match) {
1294 break;
1295 }
1296 }
1297 }
1298
1299 if ($inverse) $filter_match = !$filter_match;
1300
1301 if ($filter_match) {
1302 foreach ($filter["actions"] AS $action) {
1303 array_push($matches, $action);
1304
1305 // if Stop action encountered, perform no further processing
1306 if ($action["type"] == "stop") return $matches;
1307 }
1308 }
1309 }
1310
1311 return $matches;
1312 }
1313
1314 function find_article_filter($filters, $filter_name) {
1315 foreach ($filters as $f) {
1316 if ($f["type"] == $filter_name) {
1317 return $f;
1318 };
1319 }
1320 return false;
1321 }
1322
1323 function find_article_filters($filters, $filter_name) {
1324 $results = array();
1325
1326 foreach ($filters as $f) {
1327 if ($f["type"] == $filter_name) {
1328 array_push($results, $f);
1329 };
1330 }
1331 return $results;
1332 }
1333
1334 function calculate_article_score($filters) {
1335 $score = 0;
1336
1337 foreach ($filters as $f) {
1338 if ($f["type"] == "score") {
1339 $score += $f["param"];
1340 };
1341 }
1342 return $score;
1343 }
1344
1345 function labels_contains_caption($labels, $caption) {
1346 foreach ($labels as $label) {
1347 if ($label[1] == $caption) {
1348 return true;
1349 }
1350 }
1351
1352 return false;
1353 }
1354
1355 function assign_article_to_label_filters($id, $filters, $owner_uid, $article_labels) {
1356 foreach ($filters as $f) {
1357 if ($f["type"] == "label") {
1358 if (!labels_contains_caption($article_labels, $f["param"])) {
1359 label_add_article($id, $f["param"], $owner_uid);
1360 }
1361 }
1362 }
1363 }
1364
1365 function make_guid_from_title($title) {
1366 return preg_replace("/[ \"\',.:;]/", "-",
1367 mb_strtolower(strip_tags($title), 'utf-8'));
1368 }
1369
1370 function verify_feed_xml($feed_data) {
1371 libxml_use_internal_errors(true);
1372 $doc = new DOMDocument();
1373 $doc->loadXML($feed_data);
1374 $error = libxml_get_last_error();
1375 libxml_clear_errors();
1376 return $error;
1377 }
1378
1379 function housekeeping_common($debug) {
1380 expire_cached_files($debug);
1381 expire_lock_files($debug);
1382 expire_error_log($debug);
1383
1384 $count = update_feedbrowser_cache();
1385 _debug("Feedbrowser updated, $count feeds processed.");
1386
1387 purge_orphans( true);
1388 $rc = cleanup_tags( 14, 50000);
1389
1390 _debug("Cleaned $rc cached tags.");
1391 }
1392
1393 function utf8_entity_decode($entity){
1394 $convmap = array(0x0, 0x10000, 0, 0xfffff);
1395 return mb_decode_numericentity($entity, $convmap, 'UTF-8');
1396 }
1397
1398 function decode_numeric_entities($body) {
1399 $body = preg_replace('/&#\d{2,5};/ue', "utf8_entity_decode('$0')", $body );
1400 $body = preg_replace('/&#x([a-fA-F0-7]{2,8});/ue', "utf8_entity_decode('&#'.hexdec('$1').';')", $body );
1401 return $body;
1402 }
1403 ?>
web-based news feed (RSS/Atom) reader and aggregator; see https://tt-rss.org/