2 define('DAEMON_UPDATE_LOGIN_LIMIT', 30);
3 define('DAEMON_FEED_LIMIT', 100);
4 define('DAEMON_SLEEP_INTERVAL', 60);
6 function update_feedbrowser_cache($link) {
8 $result = db_query($link, "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");
14 db_query($link, "BEGIN");
16 db_query($link, "DELETE FROM ttrss_feedbrowser_cache");
20 while ($line = db_fetch_assoc($result)) {
21 $subscribers = db_escape_string($link, $line["subscribers"]);
22 $feed_url = db_escape_string($link, $line["feed_url"]);
23 $title = db_escape_string($link, $line["title"]);
24 $site_url = db_escape_string($link, $line["site_url"]);
26 $tmp_result = db_query($link, "SELECT subscribers FROM
27 ttrss_feedbrowser_cache WHERE feed_url = '$feed_url'");
29 if (db_num_rows($tmp_result) == 0) {
31 db_query($link, "INSERT INTO ttrss_feedbrowser_cache
32 (feed_url, site_url, title, subscribers) VALUES ('$feed_url',
33 '$site_url', '$title', '$subscribers')");
41 db_query($link, "COMMIT");
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
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 $debug Set to false to disable debug output. Default to true.
60 function update_daemon_common($link, $limit = DAEMON_FEED_LIMIT
, $from_http = false, $debug = true) {
61 // Process all other feeds using last_updated and interval parameters
63 define('PREFS_NO_CACHE', true);
65 // Test if the user has loggued in recently. If not, it does not update its feeds.
66 if (!SINGLE_USER_MODE
&& DAEMON_UPDATE_LOGIN_LIMIT
> 0) {
67 if (DB_TYPE
== "pgsql") {
68 $login_thresh_qpart = "AND ttrss_users.last_login >= NOW() - INTERVAL '".DAEMON_UPDATE_LOGIN_LIMIT
." days'";
70 $login_thresh_qpart = "AND ttrss_users.last_login >= DATE_SUB(NOW(), INTERVAL ".DAEMON_UPDATE_LOGIN_LIMIT
." DAY)";
73 $login_thresh_qpart = "";
76 // Test if the feed need a update (update interval exceded).
77 if (DB_TYPE
== "pgsql") {
78 $update_limit_qpart = "AND ((
79 ttrss_feeds.update_interval = 0
80 AND ttrss_user_prefs.value != '-1'
81 AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_user_prefs.value || ' minutes') AS INTERVAL)
83 ttrss_feeds.update_interval > 0
84 AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_feeds.update_interval || ' minutes') AS INTERVAL)
85 ) OR ttrss_feeds.last_updated IS NULL
86 OR last_updated = '1970-01-01 00:00:00')";
88 $update_limit_qpart = "AND ((
89 ttrss_feeds.update_interval = 0
90 AND ttrss_user_prefs.value != '-1'
91 AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL CONVERT(ttrss_user_prefs.value, SIGNED INTEGER) MINUTE)
93 ttrss_feeds.update_interval > 0
94 AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL ttrss_feeds.update_interval MINUTE)
95 ) OR ttrss_feeds.last_updated IS NULL
96 OR last_updated = '1970-01-01 00:00:00')";
99 // Test if feed is currently being updated by another process.
100 if (DB_TYPE
== "pgsql") {
101 $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < NOW() - INTERVAL '5 minutes')";
103 $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < DATE_SUB(NOW(), INTERVAL 5 MINUTE))";
106 // Test if there is a limit to number of updated feeds
108 if($limit) $query_limit = sprintf("LIMIT %d", $limit);
110 $random_qpart = sql_random_function();
112 // We search for feed needing update.
113 $result = db_query($link, "SELECT DISTINCT ttrss_feeds.feed_url,$random_qpart
115 ttrss_feeds, ttrss_users, ttrss_user_prefs
117 ttrss_feeds.owner_uid = ttrss_users.id
118 AND ttrss_users.id = ttrss_user_prefs.owner_uid
119 AND ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL'
120 $login_thresh_qpart $update_limit_qpart
121 $updstart_thresh_qpart
122 ORDER BY $random_qpart $query_limit");
124 $user_prefs_cache = array();
126 if($debug) _debug(sprintf("Scheduled %d feeds to update...", db_num_rows($result)));
128 // Here is a little cache magic in order to minimize risk of double feed updates.
129 $feeds_to_update = array();
130 while ($line = db_fetch_assoc($result)) {
131 array_push($feeds_to_update, db_escape_string($link, $line['feed_url']));
134 // We update the feed last update started date before anything else.
135 // There is no lag due to feed contents downloads
136 // It prevent an other process to update the same feed.
138 if(count($feeds_to_update) > 0) {
139 $feeds_quoted = array();
141 foreach ($feeds_to_update as $feed) {
142 array_push($feeds_quoted, "'" . db_escape_string($link, $feed) . "'");
145 db_query($link, sprintf("UPDATE ttrss_feeds SET last_update_started = NOW()
146 WHERE feed_url IN (%s)", implode(',', $feeds_quoted)));
149 expire_cached_files($debug);
150 expire_lock_files($debug);
151 expire_error_log($link, $debug);
155 // For each feed, we call the feed update function.
156 foreach ($feeds_to_update as $feed) {
157 if($debug) _debug("Base feed: $feed");
159 //update_rss_feed($link, $line["id"], true);
161 // since we have the data cached, we can deal with other feeds with the same url
163 $tmp_result = db_query($link, "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($link, $feed)."' AND
169 (ttrss_feeds.update_interval > 0 OR
170 ttrss_user_prefs.value != '-1')
172 ORDER BY ttrss_feeds.id $query_limit");
174 if (db_num_rows($tmp_result) > 0) {
175 while ($tline = db_fetch_assoc($tmp_result)) {
176 if($debug) _debug(" => " . $tline["last_updated"] . ", " . $tline["id"]);
177 update_rss_feed($link, $tline["id"], true);
183 require_once "digest.php";
185 // Send feed digests by email if needed.
186 send_headlines_digests($link, $debug);
190 } // function update_daemon_common
192 // ignore_daemon is not used
193 function update_rss_feed($link, $feed, $ignore_daemon = false, $no_cache = false,
194 $override_url = false) {
196 require_once "lib/simplepie/simplepie.inc";
198 $debug_enabled = defined('DAEMON_EXTENDED_DEBUG') ||
$_REQUEST['xdebug'];
200 if ($debug_enabled) {
201 _debug("update_rss_feed: start");
204 $result = db_query($link, "SELECT id,update_interval,auth_login,
205 feed_url,auth_pass,cache_images,last_updated,
206 mark_unread_on_update, owner_uid,
207 pubsub_state, auth_pass_encrypted
208 FROM ttrss_feeds WHERE id = '$feed'");
210 if (db_num_rows($result) == 0) {
211 if ($debug_enabled) {
212 _debug("update_rss_feed: feed $feed NOT FOUND/SKIPPED");
217 $last_updated = db_fetch_result($result, 0, "last_updated");
218 $owner_uid = db_fetch_result($result, 0, "owner_uid");
219 $mark_unread_on_update = sql_bool_to_bool(db_fetch_result($result,
220 0, "mark_unread_on_update"));
221 $pubsub_state = db_fetch_result($result, 0, "pubsub_state");
222 $auth_pass_encrypted = sql_bool_to_bool(db_fetch_result($result,
223 0, "auth_pass_encrypted"));
225 db_query($link, "UPDATE ttrss_feeds SET last_update_started = NOW()
226 WHERE id = '$feed'");
228 $auth_login = db_fetch_result($result, 0, "auth_login");
229 $auth_pass = db_fetch_result($result, 0, "auth_pass");
231 if ($auth_pass_encrypted) {
232 require_once "crypt.php";
233 $auth_pass = decrypt_string($auth_pass);
236 $cache_images = sql_bool_to_bool(db_fetch_result($result, 0, "cache_images"));
237 $fetch_url = db_fetch_result($result, 0, "feed_url");
239 $feed = db_escape_string($link, $feed);
241 if ($override_url) $fetch_url = $override_url;
243 $date_feed_processed = date('Y-m-d H:i');
245 $cache_filename = CACHE_DIR
. "/simplepie/" . sha1($fetch_url) . ".feed";
247 // Ignore cache if new feed or manual update.
248 $cache_age = ($no_cache ||
is_null($last_updated) ||
$last_updated == '1970-01-01 00:00:00') ?
249 30 : get_feed_update_interval($link, $feed) * 60;
251 if ($debug_enabled) {
252 _debug("update_rss_feed: cache filename: $cache_filename exists: " . file_exists($cache_filename));
253 _debug("update_rss_feed: cache age: $cache_age; no cache: $no_cache");
256 $cached_feed_data_hash = false;
260 $cache_timestamp = file_exists($cache_filename) ?
filemtime($cache_filename) : 0;
261 $last_updated_timestamp = strtotime($last_updated);
263 $force_refetch = isset($_REQUEST["force_refetch"]);
265 if (file_exists($cache_filename) &&
266 is_readable($cache_filename) &&
267 !$auth_login && !$auth_pass &&
268 filemtime($cache_filename) > time() - $cache_age) {
270 if ($debug_enabled) {
271 _debug("update_rss_feed: using local cache.");
274 if ($cache_timestamp > $last_updated_timestamp) {
275 @$rss_data = file_get_contents($cache_filename);
278 $rss_hash = sha1($rss_data);
279 @$rss = unserialize($rss_data);
281 } else if (!$force_refetch) {
282 if ($debug_enabled) {
283 _debug("update_rss_feed: local cache valid and older than last_updated, nothing to do.");
292 if ($debug_enabled) {
293 _debug("update_rss_feed: fetching [$fetch_url] (ts: $cache_timestamp/$last_updated_timestamp)");
296 $feed_data = fetch_file_contents($fetch_url, false,
297 $auth_login, $auth_pass, false,
298 $no_cache ? FEED_FETCH_NO_CACHE_TIMEOUT
: FEED_FETCH_TIMEOUT
,
299 $force_refetch ?
0 : max($last_updated_timestamp, $cache_timestamp));
301 if ($debug_enabled) {
302 _debug("update_rss_feed: fetch done.");
308 global $fetch_last_error;
309 global $fetch_last_error_code;
311 if ($debug_enabled) {
312 _debug("update_rss_feed: unable to fetch: $fetch_last_error [$fetch_last_error_code]");
318 if ($fetch_last_error_code != 304) {
319 $error_escaped = db_escape_string($link, $fetch_last_error);
321 if ($debug_enabled) {
322 _debug("update_rss_feed: source claims data not modified, nothing to do.");
327 "UPDATE ttrss_feeds SET last_error = '$error_escaped',
328 last_updated = NOW() WHERE id = '$feed'");
334 $pluginhost = new PluginHost($link);
335 $pluginhost->set_debug($debug_enabled);
336 $user_plugins = get_pref($link, "_ENABLED_PLUGINS", $owner_uid);
338 $pluginhost->load(PLUGINS
, $pluginhost::KIND_ALL
);
339 $pluginhost->load($user_plugins, $pluginhost::KIND_USER
, $owner_uid);
340 $pluginhost->load_data();
342 foreach ($pluginhost->get_hooks($pluginhost::HOOK_FEED_FETCHED
) as $plugin) {
343 $feed_data = $plugin->hook_feed_fetched($feed_data);
347 $rss = new SimplePie();
348 $rss->set_sanitize_class("SanitizeDummy");
349 // simplepie ignores the above and creates default sanitizer anyway,
350 // so let's override it...
351 $rss->sanitize
= new SanitizeDummy();
352 $rss->set_output_encoding('UTF-8');
353 $rss->set_raw_data($feed_data);
354 $rss->enable_cache(false);
361 $feed = db_escape_string($link, $feed);
363 if (!$rss->error()) {
365 // cache data for later
366 if (!$auth_pass && !$auth_login && is_writable(CACHE_DIR
. "/simplepie")) {
367 $rss_data = serialize($rss);
368 $new_rss_hash = sha1($rss_data);
370 if ($new_rss_hash != $rss_hash) {
371 if ($debug_enabled) {
372 _debug("update_rss_feed: saving $cache_filename");
374 @file_put_contents
($cache_filename, serialize($rss));
378 // We use local pluginhost here because we need to load different per-user feed plugins
379 $pluginhost->run_hooks($pluginhost::HOOK_FEED_PARSED
, "hook_feed_parsed", $rss);
381 if ($debug_enabled) {
382 _debug("update_rss_feed: processing feed data...");
385 // db_query($link, "BEGIN");
387 if (DB_TYPE
== "pgsql") {
388 $favicon_interval_qpart = "favicon_last_checked < NOW() - INTERVAL '12 hour'";
390 $favicon_interval_qpart = "favicon_last_checked < DATE_SUB(NOW(), INTERVAL 12 HOUR)";
393 $result = db_query($link, "SELECT title,site_url,owner_uid,
394 (favicon_last_checked IS NULL OR $favicon_interval_qpart) AS
396 FROM ttrss_feeds WHERE id = '$feed'");
398 $registered_title = db_fetch_result($result, 0, "title");
399 $orig_site_url = db_fetch_result($result, 0, "site_url");
400 $favicon_needs_check = sql_bool_to_bool(db_fetch_result($result, 0,
401 "favicon_needs_check"));
403 $owner_uid = db_fetch_result($result, 0, "owner_uid");
405 $site_url = db_escape_string($link, mb_substr(rewrite_relative_url($fetch_url, $rss->get_link()), 0, 245));
407 if ($favicon_needs_check ||
$force_refetch) {
408 if ($debug_enabled) {
409 _debug("update_rss_feed: checking favicon...");
412 check_feed_favicon($site_url, $feed, $link);
413 $favicon_file = ICONS_DIR
. "/$feed.ico";
415 if (file_exists($favicon_file)) {
416 require_once "colors.php";
418 $favicon_color = db_escape_string($link,
419 calculate_avg_color($favicon_file));
421 $favicon_colorstring = ",favicon_avg_color = '".$favicon_color."'";
424 db_query($link, "UPDATE ttrss_feeds SET favicon_last_checked = NOW()
426 WHERE id = '$feed'");
429 if (!$registered_title ||
$registered_title == "[Unknown]") {
431 $feed_title = db_escape_string($link, $rss->get_title());
433 if ($debug_enabled) {
434 _debug("update_rss_feed: registering title: $feed_title");
437 db_query($link, "UPDATE ttrss_feeds SET
438 title = '$feed_title' WHERE id = '$feed'");
441 if ($site_url && $orig_site_url != $site_url) {
442 db_query($link, "UPDATE ttrss_feeds SET
443 site_url = '$site_url' WHERE id = '$feed'");
446 if ($debug_enabled) {
447 _debug("update_rss_feed: loading filters & labels...");
450 $filters = load_filters($link, $feed, $owner_uid);
451 $labels = get_all_labels($link, $owner_uid);
453 if ($debug_enabled) {
455 _debug("update_rss_feed: " . count($filters) . " filters loaded.");
458 $items = $rss->get_items();
460 if (!is_array($items)) {
461 if ($debug_enabled) {
462 _debug("update_rss_feed: no articles found.");
465 db_query($link, "UPDATE ttrss_feeds
466 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
468 return; // no articles
471 if ($pubsub_state != 2 && PUBSUBHUBBUB_ENABLED
) {
473 if ($debug_enabled) _debug("update_rss_feed: checking for PUSH hub...");
475 $feed_hub_url = false;
477 $links = $rss->get_links('hub');
479 if ($links && is_array($links)) {
480 foreach ($links as $l) {
486 if ($debug_enabled) _debug("update_rss_feed: feed hub url: $feed_hub_url");
488 if ($feed_hub_url && function_exists('curl_init') &&
489 !ini_get("open_basedir")) {
491 require_once 'lib/pubsubhubbub/subscriber.php';
493 $callback_url = get_self_url_prefix() .
494 "/public.php?op=pubsub&id=$feed";
496 $s = new Subscriber($feed_hub_url, $callback_url);
498 $rc = $s->subscribe($fetch_url);
501 _debug("update_rss_feed: feed hub url found, subscribe request sent.");
503 db_query($link, "UPDATE ttrss_feeds SET pubsub_state = 1
504 WHERE id = '$feed'");
508 if ($debug_enabled) {
509 _debug("update_rss_feed: processing articles...");
512 foreach ($items as $item) {
513 if ($_REQUEST['xdebug'] == 3) {
517 $entry_guid = $item->get_id();
518 if (!$entry_guid) $entry_guid = $item->get_link();
519 if (!$entry_guid) $entry_guid = make_guid_from_title($item->get_title());
521 if (!$entry_guid) continue;
523 $entry_guid = "$owner_uid,$entry_guid";
525 $entry_guid_hashed = db_escape_string($link, 'SHA1:' . sha1($entry_guid));
527 if ($debug_enabled) {
528 _debug("update_rss_feed: guid $entry_guid / $entry_guid_hashed");
531 $entry_timestamp = "";
533 $entry_timestamp = strtotime($item->get_date());
535 if ($entry_timestamp == -1 ||
!$entry_timestamp ||
$entry_timestamp > time()) {
536 $entry_timestamp = time();
537 $no_orig_date = 'true';
539 $no_orig_date = 'false';
542 $entry_timestamp_fmt = strftime("%Y/%m/%d %H:%M:%S", $entry_timestamp);
544 if ($debug_enabled) {
545 _debug("update_rss_feed: date $entry_timestamp [$entry_timestamp_fmt]");
548 $entry_title = html_entity_decode($item->get_title());
550 $entry_link = rewrite_relative_url($site_url, $item->get_link());
552 if ($debug_enabled) {
553 _debug("update_rss_feed: title $entry_title");
554 _debug("update_rss_feed: link $entry_link");
557 if (!$entry_title) $entry_title = date("Y-m-d H:i:s", $entry_timestamp);;
559 $entry_content = $item->get_content();
560 if (!$entry_content) $entry_content = $item->get_description();
562 if ($_REQUEST["xdebug"] == 2) {
563 print "update_rss_feed: content: ";
564 print $entry_content;
568 $entry_comments = $item->data
["comments"];
570 if ($item->get_author()) {
571 $entry_author_item = $item->get_author();
572 $entry_author = $entry_author_item->get_name();
573 if (!$entry_author) $entry_author = $entry_author_item->get_email();
575 $entry_author = db_escape_string($link, $entry_author);
578 $entry_guid = db_escape_string($link, mb_substr($entry_guid, 0, 245));
580 $entry_comments = db_escape_string($link, mb_substr($entry_comments, 0, 245));
581 $entry_author = db_escape_string($link, mb_substr($entry_author, 0, 245));
583 $num_comments = $item->get_item_tags('http://purl.org/rss/1.0/modules/slash/', 'comments');
585 if (is_array($num_comments) && is_array($num_comments[0])) {
586 $num_comments = (int) $num_comments[0]["data"];
591 if ($debug_enabled) {
592 _debug("update_rss_feed: num_comments: $num_comments");
593 _debug("update_rss_feed: looking for tags [1]...");
596 // parse <category> entries into tags
598 $additional_tags = array();
600 $additional_tags_src = $item->get_categories();
602 if (is_array($additional_tags_src)) {
603 foreach ($additional_tags_src as $tobj) {
604 array_push($additional_tags, $tobj->get_term());
608 if ($debug_enabled) {
609 _debug("update_rss_feed: category tags:");
610 print_r($additional_tags);
613 if ($debug_enabled) {
614 _debug("update_rss_feed: looking for tags [2]...");
617 $entry_tags = array_unique($additional_tags);
619 for ($i = 0; $i < count($entry_tags); $i++
)
620 $entry_tags[$i] = mb_strtolower($entry_tags[$i], 'utf-8');
622 if ($debug_enabled) {
623 //_debug("update_rss_feed: unfiltered tags found:");
624 //print_r($entry_tags);
627 if ($debug_enabled) {
628 _debug("update_rss_feed: done collecting data.");
631 // TODO: less memory-hungry implementation
633 if ($debug_enabled) {
634 _debug("update_rss_feed: applying plugin filters..");
637 // FIXME not sure if owner_uid is a good idea here, we may have a base entry without user entry (?)
638 $result = db_query($link, "SELECT plugin_data,title,content,link,tag_cache,author FROM ttrss_entries, ttrss_user_entries
639 WHERE ref_id = id AND (guid = '".db_escape_string($link, $entry_guid)."' OR guid = '$entry_guid_hashed') AND owner_uid = $owner_uid");
641 if (db_num_rows($result) != 0) {
642 $entry_plugin_data = db_fetch_result($result, 0, "plugin_data");
643 $stored_article = array("title" => db_fetch_result($result, 0, "title"),
644 "content" => db_fetch_result($result, 0, "content"),
645 "link" => db_fetch_result($result, 0, "link"),
646 "tags" => explode(",", db_fetch_result($result, 0, "tag_cache")),
647 "author" => db_fetch_result($result, 0, "author"));
649 $entry_plugin_data = "";
650 $stored_article = array();
653 $article = array("owner_uid" => $owner_uid, // read only
654 "guid" => $entry_guid, // read only
655 "title" => $entry_title,
656 "content" => $entry_content,
657 "link" => $entry_link,
658 "tags" => $entry_tags,
659 "plugin_data" => $entry_plugin_data,
660 "author" => $entry_author,
661 "stored" => $stored_article);
663 foreach ($pluginhost->get_hooks($pluginhost::HOOK_ARTICLE_FILTER
) as $plugin) {
664 $article = $plugin->hook_article_filter($article);
667 $entry_tags = $article["tags"];
668 $entry_guid = db_escape_string($link, $entry_guid);
669 $entry_title = db_escape_string($link, $article["title"]);
670 $entry_author = db_escape_string($link, $article["author"]);
671 $entry_link = db_escape_string($link, $article["link"]);
672 $entry_plugin_data = db_escape_string($link, $article["plugin_data"]);
673 $entry_content = $article["content"]; // escaped below
676 if ($debug_enabled) {
677 _debug("update_rss_feed: plugin data: $entry_plugin_data");
680 if ($cache_images && is_writable(CACHE_DIR
. '/images'))
681 cache_images($entry_content, $site_url, $debug_enabled);
683 $entry_content = db_escape_string($link, $entry_content, false);
685 $content_hash = "SHA1:" . sha1($entry_content);
687 db_query($link, "BEGIN");
689 $result = db_query($link, "SELECT id FROM ttrss_entries
690 WHERE (guid = '$entry_guid' OR guid = '$entry_guid_hashed')");
692 if (db_num_rows($result) == 0) {
694 if ($debug_enabled) {
695 _debug("update_rss_feed: base guid [$entry_guid] not found");
698 // base post entry does not exist, create it
700 $result = db_query($link,
701 "INSERT INTO ttrss_entries
718 '$entry_guid_hashed',
720 '$entry_timestamp_fmt',
726 '$date_feed_processed',
729 '$entry_plugin_data',
732 $article_labels = array();
735 // we keep encountering the entry in feeds, so we need to
736 // update date_updated column so that we don't get horrible
737 // dupes when the entry gets purged and reinserted again e.g.
738 // in the case of SLOW SLOW OMG SLOW updating feeds
740 $base_entry_id = db_fetch_result($result, 0, "id");
742 db_query($link, "UPDATE ttrss_entries SET date_updated = NOW()
743 WHERE id = '$base_entry_id'");
745 $article_labels = get_article_labels($link, $base_entry_id, $owner_uid);
748 // now it should exist, if not - bad luck then
750 $result = db_query($link, "SELECT
751 id,content_hash,no_orig_date,title,plugin_data,guid,
752 ".SUBSTRING_FOR_DATE
."(date_updated,1,19) as date_updated,
753 ".SUBSTRING_FOR_DATE
."(updated,1,19) as updated,
757 WHERE guid = '$entry_guid' OR guid = '$entry_guid_hashed'");
762 if (db_num_rows($result) == 1) {
764 if ($debug_enabled) {
765 _debug("update_rss_feed: base guid found, checking for user record");
768 // this will be used below in update handler
769 $orig_content_hash = db_fetch_result($result, 0, "content_hash");
770 $orig_title = db_fetch_result($result, 0, "title");
771 $orig_num_comments = db_fetch_result($result, 0, "num_comments");
772 $orig_date_updated = strtotime(db_fetch_result($result,
774 $orig_plugin_data = db_fetch_result($result, 0, "plugin_data");
776 $ref_id = db_fetch_result($result, 0, "id");
777 $entry_ref_id = $ref_id;
779 /* $stored_guid = db_fetch_result($result, 0, "guid");
780 if ($stored_guid != $entry_guid_hashed) {
781 if ($debug_enabled) _debug("upgrading compat guid to hashed one");
783 db_query($link, "UPDATE ttrss_entries SET guid = '$entry_guid_hashed' WHERE
787 // check for user post link to main table
789 // do we allow duplicate posts with same GUID in different feeds?
790 if (get_pref($link, "ALLOW_DUPLICATE_POSTS", $owner_uid, false)) {
791 $dupcheck_qpart = "AND (feed_id = '$feed' OR feed_id IS NULL)";
793 $dupcheck_qpart = "";
796 /* Collect article tags here so we could filter by them: */
798 $article_filters = get_article_filters($filters, $entry_title,
799 $entry_content, $entry_link, $entry_timestamp, $entry_author,
802 if ($debug_enabled) {
803 _debug("update_rss_feed: article filters: ");
804 if (count($article_filters) != 0) {
805 print_r($article_filters);
809 if (find_article_filter($article_filters, "filter")) {
810 db_query($link, "COMMIT"); // close transaction in progress
814 $score = calculate_article_score($article_filters);
816 if ($debug_enabled) {
817 _debug("update_rss_feed: initial score: $score");
820 $query = "SELECT ref_id, int_id FROM ttrss_user_entries WHERE
821 ref_id = '$ref_id' AND owner_uid = '$owner_uid'
824 // if ($_REQUEST["xdebug"]) print "$query\n";
826 $result = db_query($link, $query);
828 // okay it doesn't exist - create user entry
829 if (db_num_rows($result) == 0) {
831 if ($debug_enabled) {
832 _debug("update_rss_feed: user record not found, creating...");
835 if ($score >= -500 && !find_article_filter($article_filters, 'catchup')) {
837 $last_read_qpart = 'NULL';
840 $last_read_qpart = 'NOW()';
843 if (find_article_filter($article_filters, 'mark') ||
$score > 1000) {
849 if (find_article_filter($article_filters, 'publish')) {
852 $published = 'false';
857 if (DB_TYPE
== "pgsql" and defined('_NGRAM_TITLE_DUPLICATE_THRESHOLD')) {
859 $result = db_query($link, "SELECT COUNT(*) AS similar FROM
860 ttrss_entries,ttrss_user_entries
861 WHERE ref_id = id AND updated >= NOW() - INTERVAL '7 day'
862 AND similarity(title, '$entry_title') >= "._NGRAM_TITLE_DUPLICATE_THRESHOLD
."
863 AND owner_uid = $owner_uid");
865 $ngram_similar = db_fetch_result($result, 0, "similar");
867 if ($debug_enabled) {
868 _debug("update_rss_feed: N-gram similar results: $ngram_similar");
871 if ($ngram_similar > 0) {
876 $last_marked = ($marked == 'true') ?
'NOW()' : 'NULL';
877 $last_published = ($published == 'true') ?
'NOW()' : 'NULL';
879 $result = db_query($link,
880 "INSERT INTO ttrss_user_entries
881 (ref_id, owner_uid, feed_id, unread, last_read, marked,
882 published, score, tag_cache, label_cache, uuid,
883 last_marked, last_published)
884 VALUES ('$ref_id', '$owner_uid', '$feed', $unread,
885 $last_read_qpart, $marked, $published, '$score', '', '',
886 '', $last_marked, $last_published)");
888 if (PUBSUBHUBBUB_HUB
&& $published == 'true') {
889 $rss_link = get_self_url_prefix() .
890 "/public.php?op=rss&id=-2&key=" .
891 get_feed_access_key($link, -2, false, $owner_uid);
893 $p = new Publisher(PUBSUBHUBBUB_HUB
);
895 $pubsub_result = $p->publish_update($rss_link);
898 $result = db_query($link,
899 "SELECT int_id FROM ttrss_user_entries WHERE
900 ref_id = '$ref_id' AND owner_uid = '$owner_uid' AND
901 feed_id = '$feed' LIMIT 1");
903 if (db_num_rows($result) == 1) {
904 $entry_int_id = db_fetch_result($result, 0, "int_id");
907 if ($debug_enabled) {
908 _debug("update_rss_feed: user record FOUND");
911 $entry_ref_id = db_fetch_result($result, 0, "ref_id");
912 $entry_int_id = db_fetch_result($result, 0, "int_id");
915 if ($debug_enabled) {
916 _debug("update_rss_feed: RID: $entry_ref_id, IID: $entry_int_id");
919 $post_needs_update = false;
920 $update_insignificant = false;
922 if ($orig_num_comments != $num_comments) {
923 $post_needs_update = true;
924 $update_insignificant = true;
927 if ($entry_plugin_data != $orig_plugin_data) {
928 $post_needs_update = true;
929 $update_insignificant = true;
932 if ($content_hash != $orig_content_hash) {
933 $post_needs_update = true;
934 $update_insignificant = false;
937 if (db_escape_string($link, $orig_title) != $entry_title) {
938 $post_needs_update = true;
939 $update_insignificant = false;
942 // if post needs update, update it and mark all user entries
943 // linking to this post as updated
944 if ($post_needs_update) {
946 if (defined('DAEMON_EXTENDED_DEBUG')) {
947 _debug("update_rss_feed: post $entry_guid_hashed needs update...");
950 // print "<!-- post $orig_title needs update : $post_needs_update -->";
952 db_query($link, "UPDATE ttrss_entries
953 SET title = '$entry_title', content = '$entry_content',
954 content_hash = '$content_hash',
955 updated = '$entry_timestamp_fmt',
956 num_comments = '$num_comments',
957 plugin_data = '$entry_plugin_data'
958 WHERE id = '$ref_id'");
960 if (!$update_insignificant) {
961 if ($mark_unread_on_update) {
962 db_query($link, "UPDATE ttrss_user_entries
963 SET last_read = null, unread = true WHERE ref_id = '$ref_id'");
969 db_query($link, "COMMIT");
971 if ($debug_enabled) {
972 _debug("update_rss_feed: assigning labels...");
975 assign_article_to_label_filters($link, $entry_ref_id, $article_filters,
976 $owner_uid, $article_labels);
978 if ($debug_enabled) {
979 _debug("update_rss_feed: looking for enclosures...");
984 $enclosures = array();
986 $encs = $item->get_enclosures();
988 if (is_array($encs)) {
989 foreach ($encs as $e) {
991 $e->link
, $e->type
, $e->length
);
992 array_push($enclosures, $e_item);
996 if ($debug_enabled) {
997 _debug("update_rss_feed: article enclosures:");
998 print_r($enclosures);
1001 db_query($link, "BEGIN");
1003 foreach ($enclosures as $enc) {
1004 $enc_url = db_escape_string($link, $enc[0]);
1005 $enc_type = db_escape_string($link, $enc[1]);
1006 $enc_dur = db_escape_string($link, $enc[2]);
1008 $result = db_query($link, "SELECT id FROM ttrss_enclosures
1009 WHERE content_url = '$enc_url' AND post_id = '$entry_ref_id'");
1011 if (db_num_rows($result) == 0) {
1012 db_query($link, "INSERT INTO ttrss_enclosures
1013 (content_url, content_type, title, duration, post_id) VALUES
1014 ('$enc_url', '$enc_type', '', '$enc_dur', '$entry_ref_id')");
1018 db_query($link, "COMMIT");
1020 // check for manual tags (we have to do it here since they're loaded from filters)
1022 foreach ($article_filters as $f) {
1023 if ($f["type"] == "tag") {
1025 $manual_tags = trim_array(explode(",", $f["param"]));
1027 foreach ($manual_tags as $tag) {
1028 if (tag_is_valid($tag)) {
1029 array_push($entry_tags, $tag);
1037 $boring_tags = trim_array(explode(",", mb_strtolower(get_pref($link,
1038 'BLACKLISTED_TAGS', $owner_uid, ''), 'utf-8')));
1040 $filtered_tags = array();
1041 $tags_to_cache = array();
1043 if ($entry_tags && is_array($entry_tags)) {
1044 foreach ($entry_tags as $tag) {
1045 if (array_search($tag, $boring_tags) === false) {
1046 array_push($filtered_tags, $tag);
1051 $filtered_tags = array_unique($filtered_tags);
1053 if ($debug_enabled) {
1054 _debug("update_rss_feed: filtered article tags:");
1055 print_r($filtered_tags);
1058 // Save article tags in the database
1060 if (count($filtered_tags) > 0) {
1062 db_query($link, "BEGIN");
1064 foreach ($filtered_tags as $tag) {
1066 $tag = sanitize_tag($tag);
1067 $tag = db_escape_string($link, $tag);
1069 if (!tag_is_valid($tag)) continue;
1071 $result = db_query($link, "SELECT id FROM ttrss_tags
1072 WHERE tag_name = '$tag' AND post_int_id = '$entry_int_id' AND
1073 owner_uid = '$owner_uid' LIMIT 1");
1075 if ($result && db_num_rows($result) == 0) {
1077 db_query($link, "INSERT INTO ttrss_tags
1078 (owner_uid,tag_name,post_int_id)
1079 VALUES ('$owner_uid','$tag', '$entry_int_id')");
1082 array_push($tags_to_cache, $tag);
1085 /* update the cache */
1087 $tags_to_cache = array_unique($tags_to_cache);
1089 $tags_str = db_escape_string($link, join(",", $tags_to_cache));
1091 db_query($link, "UPDATE ttrss_user_entries
1092 SET tag_cache = '$tags_str' WHERE ref_id = '$entry_ref_id'
1093 AND owner_uid = $owner_uid");
1095 db_query($link, "COMMIT");
1098 if (get_pref($link, "AUTO_ASSIGN_LABELS", $owner_uid, false)) {
1099 if ($debug_enabled) {
1100 _debug("update_rss_feed: auto-assigning labels...");
1103 foreach ($labels as $label) {
1104 $caption = preg_quote($label["caption"]);
1106 if ($caption && preg_match("/\b$caption\b/i", "$tags_str " . strip_tags($entry_content) . " $entry_title")) {
1107 if (!labels_contains_caption($article_labels, $caption)) {
1108 label_add_article($link, $entry_ref_id, $caption, $owner_uid);
1114 if ($debug_enabled) {
1115 _debug("update_rss_feed: article processed");
1119 if (!$last_updated) {
1120 if ($debug_enabled) {
1121 _debug("update_rss_feed: new feed, catching it up...");
1123 catchup_feed($link, $feed, false, $owner_uid);
1126 if ($debug_enabled) {
1127 _debug("purging feed...");
1130 purge_feed($link, $feed, 0, $debug_enabled);
1132 db_query($link, "UPDATE ttrss_feeds
1133 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
1135 // db_query($link, "COMMIT");
1139 $error_msg = db_escape_string($link, mb_substr($rss->error(), 0, 245));
1141 if ($debug_enabled) {
1142 _debug("update_rss_feed: error fetching feed: $error_msg");
1146 "UPDATE ttrss_feeds SET last_error = '$error_msg',
1147 last_updated = NOW() WHERE id = '$feed'");
1152 if ($debug_enabled) {
1153 _debug("update_rss_feed: done");
1158 function cache_images($html, $site_url, $debug) {
1159 $cache_dir = CACHE_DIR
. "/images";
1161 libxml_use_internal_errors(true);
1163 $charset_hack = '<head>
1164 <meta http-equiv="Content-Type" content="text/html; charset=utf-8"/>
1167 $doc = new DOMDocument();
1168 $doc->loadHTML($charset_hack . $html);
1169 $xpath = new DOMXPath($doc);
1171 $entries = $xpath->query('(//img[@src])');
1173 foreach ($entries as $entry) {
1174 if ($entry->hasAttribute('src')) {
1175 $src = rewrite_relative_url($site_url, $entry->getAttribute('src'));
1177 $local_filename = CACHE_DIR
. "/images/" . sha1($src) . ".png";
1179 if ($debug) _debug("cache_images: downloading: $src to $local_filename");
1181 if (!file_exists($local_filename)) {
1182 $file_content = fetch_file_contents($src);
1184 if ($file_content && strlen($file_content) > 1024) {
1185 file_put_contents($local_filename, $file_content);
1189 if (file_exists($local_filename)) {
1190 $entry->setAttribute('src', SELF_URL_PATH
. '/image.php?url=' .
1191 base64_encode($src));
1196 $node = $doc->getElementsByTagName('body')->item(0);
1198 return $doc->saveXML($node);
1201 function expire_error_log($link, $debug) {
1202 if ($debug) _debug("Removing old error log entries...");
1204 if (DB_TYPE
== "pgsql") {
1205 db_query($link, "DELETE FROM ttrss_error_log
1206 WHERE created_at < NOW() - INTERVAL '7 days'");
1208 db_query($link, "DELETE FROM ttrss_error_log
1209 WHERE created_at < DATE_SUB(NOW(), INTERVAL 7 DAY)");
1214 function expire_lock_files($debug) {
1215 if ($debug) _debug("Removing old lock files...");
1219 if (is_writable(LOCK_DIRECTORY
)) {
1220 $files = glob(LOCK_DIRECTORY
. "/*.lock");
1223 foreach ($files as $file) {
1224 if (!file_is_locked($file) && time() - filemtime($file) > 86400*2) {
1232 if ($debug) _debug("Removed $num_deleted files.");
1235 function expire_cached_files($debug) {
1236 foreach (array("simplepie", "images", "export", "upload") as $dir) {
1237 $cache_dir = CACHE_DIR
. "/$dir";
1239 if ($debug) _debug("Expiring $cache_dir");
1243 if (is_writable($cache_dir)) {
1244 $files = glob("$cache_dir/*");
1247 foreach ($files as $file) {
1248 if (time() - filemtime($file) > 86400*7) {
1257 if ($debug) _debug("Removed $num_deleted files.");
1262 * Source: http://www.php.net/manual/en/function.parse-url.php#104527
1263 * Returns the url query as associative array
1265 * @param string query
1266 * @return array params
1268 function convertUrlQuery($query) {
1269 $queryParts = explode('&', $query);
1273 foreach ($queryParts as $param) {
1274 $item = explode('=', $param);
1275 $params[$item[0]] = $item[1];
1281 function get_article_filters($filters, $title, $content, $link, $timestamp, $author, $tags) {
1284 foreach ($filters as $filter) {
1285 $match_any_rule = $filter["match_any_rule"];
1286 $inverse = $filter["inverse"];
1287 $filter_match = false;
1289 foreach ($filter["rules"] as $rule) {
1291 $reg_exp = $rule["reg_exp"];
1292 $rule_inverse = $rule["inverse"];
1297 switch ($rule["type"]) {
1299 $match = @preg_match
("/$reg_exp/i", $title);
1302 // we don't need to deal with multiline regexps
1303 $content = preg_replace("/[\r\n\t]/", "", $content);
1305 $match = @preg_match
("/$reg_exp/i", $content);
1308 // we don't need to deal with multiline regexps
1309 $content = preg_replace("/[\r\n\t]/", "", $content);
1311 $match = (@preg_match
("/$reg_exp/i", $title) || @preg_match
("/$reg_exp/i", $content));
1314 $match = @preg_match
("/$reg_exp/i", $link);
1317 $match = @preg_match
("/$reg_exp/i", $author);
1320 $tag_string = join(",", $tags);
1321 $match = @preg_match
("/$reg_exp/i", $tag_string);
1325 if ($rule_inverse) $match = !$match;
1327 if ($match_any_rule) {
1329 $filter_match = true;
1333 $filter_match = $match;
1340 if ($inverse) $filter_match = !$filter_match;
1342 if ($filter_match) {
1343 foreach ($filter["actions"] AS $action) {
1344 array_push($matches, $action);
1346 // if Stop action encountered, perform no further processing
1347 if ($action["type"] == "stop") return $matches;
1355 function find_article_filter($filters, $filter_name) {
1356 foreach ($filters as $f) {
1357 if ($f["type"] == $filter_name) {
1364 function find_article_filters($filters, $filter_name) {
1367 foreach ($filters as $f) {
1368 if ($f["type"] == $filter_name) {
1369 array_push($results, $f);
1375 function calculate_article_score($filters) {
1378 foreach ($filters as $f) {
1379 if ($f["type"] == "score") {
1380 $score +
= $f["param"];
1386 function labels_contains_caption($labels, $caption) {
1387 foreach ($labels as $label) {
1388 if ($label[1] == $caption) {
1396 function assign_article_to_label_filters($link, $id, $filters, $owner_uid, $article_labels) {
1397 foreach ($filters as $f) {
1398 if ($f["type"] == "label") {
1399 if (!labels_contains_caption($article_labels, $f["param"])) {
1400 label_add_article($link, $id, $f["param"], $owner_uid);
1406 function make_guid_from_title($title) {
1407 return preg_replace("/[ \"\',.:;]/", "-",
1408 mb_strtolower(strip_tags($title), 'utf-8'));