2 define('DAEMON_UPDATE_LOGIN_LIMIT', 30);
3 define('DAEMON_FEED_LIMIT', 100);
4 define('DAEMON_SLEEP_INTERVAL', 60);
6 function update_feedbrowser_cache() {
8 $result = db_query("SELECT feed_url, site_url, title, COUNT(id) AS subscribers
9 FROM ttrss_feeds WHERE (SELECT COUNT(id) = 0 FROM ttrss_feeds AS tf
10 WHERE tf.feed_url = ttrss_feeds.feed_url
11 AND (private IS true OR auth_login != '' OR auth_pass != '' OR feed_url LIKE '%:%@%/%'))
12 GROUP BY feed_url, site_url, title ORDER BY subscribers DESC LIMIT 1000");
16 db_query("DELETE FROM ttrss_feedbrowser_cache");
20 while ($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"]);
26 $tmp_result = db_query("SELECT subscribers FROM
27 ttrss_feedbrowser_cache WHERE feed_url = '$feed_url'");
29 if (db_num_rows($tmp_result) == 0) {
31 db_query("INSERT INTO ttrss_feedbrowser_cache
32 (feed_url, site_url, title, subscribers) VALUES ('$feed_url',
33 '$site_url', '$title', '$subscribers')");
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 $debug Set to false to disable debug output. Default to true.
60 function update_daemon_common($limit = DAEMON_FEED_LIMIT
, $from_http = false, $debug = true) {
61 // Process all other feeds using last_updated and interval parameters
63 $schema_version = get_schema_version();
65 if ($schema_version != SCHEMA_VERSION
) {
66 die("Schema version is wrong, please upgrade the database.\n");
69 define('PREFS_NO_CACHE', true);
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'";
76 $login_thresh_qpart = "AND ttrss_users.last_login >= DATE_SUB(NOW(), INTERVAL ".DAEMON_UPDATE_LOGIN_LIMIT
." DAY)";
79 $login_thresh_qpart = "";
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)
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')";
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)
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')";
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 '5 minutes')";
109 $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < DATE_SUB(NOW(), INTERVAL 5 MINUTE))";
112 // Test if there is a limit to number of updated feeds
114 if($limit) $query_limit = sprintf("LIMIT %d", $limit);
116 $random_qpart = sql_random_function();
118 // We search for feed needing update.
119 $result = db_query("SELECT DISTINCT ttrss_feeds.feed_url,$random_qpart
121 ttrss_feeds, ttrss_users, ttrss_user_prefs
123 ttrss_feeds.owner_uid = ttrss_users.id
124 AND ttrss_users.id = ttrss_user_prefs.owner_uid
125 AND ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL'
126 $login_thresh_qpart $update_limit_qpart
127 $updstart_thresh_qpart
128 ORDER BY $random_qpart $query_limit");
130 if($debug) _debug(sprintf("Scheduled %d feeds to update...", db_num_rows($result)));
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($line['feed_url']));
138 // We update the feed last update started date before anything else.
139 // There is no lag due to feed contents downloads
140 // It prevent an other process to update the same feed.
142 if(count($feeds_to_update) > 0) {
143 $feeds_quoted = array();
145 foreach ($feeds_to_update as $feed) {
146 array_push($feeds_quoted, "'" . db_escape_string($feed) . "'");
149 db_query(sprintf("UPDATE ttrss_feeds SET last_update_started = NOW()
150 WHERE feed_url IN (%s)", implode(',', $feeds_quoted)));
153 expire_cached_files($debug);
154 expire_lock_files($debug);
155 expire_error_log($debug);
159 // For each feed, we call the feed update function.
160 foreach ($feeds_to_update as $feed) {
161 if($debug) _debug("Base feed: $feed");
163 //update_rss_feed($line["id"], true);
165 // since we have the data cached, we can deal with other feeds with the same url
167 $tmp_result = db_query("SELECT DISTINCT ttrss_feeds.id,last_updated
168 FROM ttrss_feeds, ttrss_users, ttrss_user_prefs WHERE
169 ttrss_user_prefs.owner_uid = ttrss_feeds.owner_uid AND
170 ttrss_users.id = ttrss_user_prefs.owner_uid AND
171 ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL' AND
172 feed_url = '".db_escape_string($feed)."' AND
173 (ttrss_feeds.update_interval > 0 OR
174 ttrss_user_prefs.value != '-1')
176 ORDER BY ttrss_feeds.id $query_limit");
178 if (db_num_rows($tmp_result) > 0) {
179 while ($tline = db_fetch_assoc($tmp_result)) {
180 if($debug) _debug(" => " . $tline["last_updated"] . ", " . $tline["id"]);
181 update_rss_feed($tline["id"], true);
187 require_once "digest.php";
189 // Send feed digests by email if needed.
190 send_headlines_digests($debug);
194 } // function update_daemon_common
196 // ignore_daemon is not used
197 function update_rss_feed($feed, $ignore_daemon = false, $no_cache = false,
198 $override_url = false) {
200 require_once "lib/simplepie/simplepie.inc";
202 $debug_enabled = defined('DAEMON_EXTENDED_DEBUG') ||
$_REQUEST['xdebug'];
204 if ($debug_enabled) {
205 _debug("update_rss_feed: start");
208 $result = db_query("SELECT id,update_interval,auth_login,
209 feed_url,auth_pass,cache_images,last_updated,
210 mark_unread_on_update, owner_uid,
211 pubsub_state, auth_pass_encrypted
212 FROM ttrss_feeds WHERE id = '$feed'");
214 if (db_num_rows($result) == 0) {
215 if ($debug_enabled) {
216 _debug("update_rss_feed: feed $feed NOT FOUND/SKIPPED");
221 $last_updated = db_fetch_result($result, 0, "last_updated");
222 $owner_uid = db_fetch_result($result, 0, "owner_uid");
223 $mark_unread_on_update = sql_bool_to_bool(db_fetch_result($result,
224 0, "mark_unread_on_update"));
225 $pubsub_state = db_fetch_result($result, 0, "pubsub_state");
226 $auth_pass_encrypted = sql_bool_to_bool(db_fetch_result($result,
227 0, "auth_pass_encrypted"));
229 db_query("UPDATE ttrss_feeds SET last_update_started = NOW()
230 WHERE id = '$feed'");
232 $auth_login = db_fetch_result($result, 0, "auth_login");
233 $auth_pass = db_fetch_result($result, 0, "auth_pass");
235 if ($auth_pass_encrypted) {
236 require_once "crypt.php";
237 $auth_pass = decrypt_string($auth_pass);
240 $cache_images = sql_bool_to_bool(db_fetch_result($result, 0, "cache_images"));
241 $fetch_url = db_fetch_result($result, 0, "feed_url");
243 $feed = db_escape_string($feed);
245 if ($override_url) $fetch_url = $override_url;
247 $date_feed_processed = date('Y-m-d H:i');
249 $cache_filename = CACHE_DIR
. "/simplepie/" . sha1($fetch_url) . ".feed";
251 // Ignore cache if new feed or manual update.
252 $cache_age = ($no_cache ||
is_null($last_updated) ||
$last_updated == '1970-01-01 00:00:00') ?
253 30 : get_feed_update_interval($feed) * 60;
255 if ($debug_enabled) {
256 _debug("update_rss_feed: cache filename: $cache_filename exists: " . file_exists($cache_filename));
257 _debug("update_rss_feed: cache age: $cache_age; no cache: $no_cache");
260 $cached_feed_data_hash = false;
264 $cache_timestamp = file_exists($cache_filename) ?
filemtime($cache_filename) : 0;
265 $last_updated_timestamp = strtotime($last_updated);
267 $force_refetch = isset($_REQUEST["force_refetch"]);
269 if (file_exists($cache_filename) &&
270 is_readable($cache_filename) &&
271 !$auth_login && !$auth_pass &&
272 filemtime($cache_filename) > time() - $cache_age) {
274 if ($debug_enabled) {
275 _debug("update_rss_feed: using local cache.");
278 if ($cache_timestamp > $last_updated_timestamp) {
279 @$rss_data = file_get_contents($cache_filename);
282 $rss_hash = sha1($rss_data);
283 @$rss = unserialize($rss_data);
285 } else if (!$force_refetch) {
286 if ($debug_enabled) {
287 _debug("update_rss_feed: local cache valid and older than last_updated, nothing to do.");
296 if ($debug_enabled) {
297 _debug("update_rss_feed: fetching [$fetch_url] (ts: $cache_timestamp/$last_updated_timestamp)");
300 $feed_data = fetch_file_contents($fetch_url, false,
301 $auth_login, $auth_pass, false,
302 $no_cache ? FEED_FETCH_NO_CACHE_TIMEOUT
: FEED_FETCH_TIMEOUT
,
303 $force_refetch ?
0 : max($last_updated_timestamp, $cache_timestamp));
305 global $fetch_curl_used;
307 if (!$fetch_curl_used) {
308 $tmp = @gzdecode
($feed_data);
310 if ($tmp) $feed_data = $tmp;
313 $feed_data = trim($feed_data);
315 if ($debug_enabled) {
316 _debug("update_rss_feed: fetch done.");
319 $error = verify_feed_xml($feed_data);
322 if ($debug_enabled) {
323 _debug("update_rss_feed: error verifying XML, code: " . $error->code
);
326 if ($error->code
== 26) {
327 if ($debug_enabled) {
328 _debug("update_rss_feed: got error 26, trying to decode entities...");
331 $feed_data = html_entity_decode($feed_data, ENT_COMPAT
, 'UTF-8');
333 $error = verify_feed_xml($feed_data);
335 if ($error) $feed_data = '';
341 global $fetch_last_error;
342 global $fetch_last_error_code;
344 if ($debug_enabled) {
345 _debug("update_rss_feed: unable to fetch: $fetch_last_error [$fetch_last_error_code]");
351 if ($fetch_last_error_code != 304) {
352 $error_escaped = db_escape_string($fetch_last_error);
354 if ($debug_enabled) {
355 _debug("update_rss_feed: source claims data not modified, nothing to do.");
360 "UPDATE ttrss_feeds SET last_error = '$error_escaped',
361 last_updated = NOW() WHERE id = '$feed'");
367 $pluginhost = new PluginHost();
368 $pluginhost->set_debug($debug_enabled);
369 $user_plugins = get_pref("_ENABLED_PLUGINS", $owner_uid);
371 $pluginhost->load(PLUGINS
, PluginHost
::KIND_ALL
);
372 $pluginhost->load($user_plugins, PluginHost
::KIND_USER
, $owner_uid);
373 $pluginhost->load_data();
375 foreach ($pluginhost->get_hooks(PluginHost
::HOOK_FEED_FETCHED
) as $plugin) {
376 $feed_data = $plugin->hook_feed_fetched($feed_data);
380 $rss = new SimplePie();
381 $rss->set_sanitize_class("SanitizeDummy");
382 // simplepie ignores the above and creates default sanitizer anyway,
383 // so let's override it...
384 $rss->sanitize
= new SanitizeDummy();
385 $rss->set_output_encoding('UTF-8');
386 $rss->set_raw_data($feed_data);
387 $rss->enable_cache(false);
394 $feed = db_escape_string($feed);
396 if (!$rss->error()) {
398 // cache data for later
399 if (!$auth_pass && !$auth_login && is_writable(CACHE_DIR
. "/simplepie")) {
400 $rss_data = serialize($rss);
401 $new_rss_hash = sha1($rss_data);
403 if ($new_rss_hash != $rss_hash) {
404 if ($debug_enabled) {
405 _debug("update_rss_feed: saving $cache_filename");
407 @file_put_contents
($cache_filename, serialize($rss));
411 // We use local pluginhost here because we need to load different per-user feed plugins
412 $pluginhost->run_hooks(PluginHost
::HOOK_FEED_PARSED
, "hook_feed_parsed", $rss);
414 if ($debug_enabled) {
415 _debug("update_rss_feed: processing feed data...");
418 // db_query("BEGIN");
420 if (DB_TYPE
== "pgsql") {
421 $favicon_interval_qpart = "favicon_last_checked < NOW() - INTERVAL '12 hour'";
423 $favicon_interval_qpart = "favicon_last_checked < DATE_SUB(NOW(), INTERVAL 12 HOUR)";
426 $result = db_query("SELECT title,site_url,owner_uid,
427 (favicon_last_checked IS NULL OR $favicon_interval_qpart) AS
429 FROM ttrss_feeds WHERE id = '$feed'");
431 $registered_title = db_fetch_result($result, 0, "title");
432 $orig_site_url = db_fetch_result($result, 0, "site_url");
433 $favicon_needs_check = sql_bool_to_bool(db_fetch_result($result, 0,
434 "favicon_needs_check"));
436 $owner_uid = db_fetch_result($result, 0, "owner_uid");
438 $site_url = db_escape_string(mb_substr(rewrite_relative_url($fetch_url, $rss->get_link()), 0, 245));
440 if ($favicon_needs_check ||
$force_refetch) {
441 if ($debug_enabled) {
442 _debug("update_rss_feed: checking favicon...");
445 check_feed_favicon($site_url, $feed, $link);
446 $favicon_file = ICONS_DIR
. "/$feed.ico";
448 if (file_exists($favicon_file) && function_exists("imagecreatefromstring")) {
449 require_once "colors.php";
451 $favicon_color = db_escape_string(
452 calculate_avg_color($favicon_file));
454 $favicon_colorstring = ",favicon_avg_color = '".$favicon_color."'";
457 db_query("UPDATE ttrss_feeds SET favicon_last_checked = NOW()
459 WHERE id = '$feed'");
462 if (!$registered_title ||
$registered_title == "[Unknown]") {
464 $feed_title = db_escape_string($rss->get_title());
466 if ($debug_enabled) {
467 _debug("update_rss_feed: registering title: $feed_title");
470 db_query("UPDATE ttrss_feeds SET
471 title = '$feed_title' WHERE id = '$feed'");
474 if ($site_url && $orig_site_url != $site_url) {
475 db_query("UPDATE ttrss_feeds SET
476 site_url = '$site_url' WHERE id = '$feed'");
479 if ($debug_enabled) {
480 _debug("update_rss_feed: loading filters & labels...");
483 $filters = load_filters($feed, $owner_uid);
484 $labels = get_all_labels($owner_uid);
486 if ($debug_enabled) {
488 _debug("update_rss_feed: " . count($filters) . " filters loaded.");
491 $items = $rss->get_items();
493 if (!is_array($items)) {
494 if ($debug_enabled) {
495 _debug("update_rss_feed: no articles found.");
498 db_query("UPDATE ttrss_feeds
499 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
501 return; // no articles
504 if ($pubsub_state != 2 && PUBSUBHUBBUB_ENABLED
) {
506 if ($debug_enabled) _debug("update_rss_feed: checking for PUSH hub...");
508 $feed_hub_url = false;
510 $links = $rss->get_links('hub');
512 if ($links && is_array($links)) {
513 foreach ($links as $l) {
519 if ($debug_enabled) _debug("update_rss_feed: feed hub url: $feed_hub_url");
521 if ($feed_hub_url && function_exists('curl_init') &&
522 !ini_get("open_basedir")) {
524 require_once 'lib/pubsubhubbub/subscriber.php';
526 $callback_url = get_self_url_prefix() .
527 "/public.php?op=pubsub&id=$feed";
529 $s = new Subscriber($feed_hub_url, $callback_url);
531 $rc = $s->subscribe($fetch_url);
534 _debug("update_rss_feed: feed hub url found, subscribe request sent.");
536 db_query("UPDATE ttrss_feeds SET pubsub_state = 1
537 WHERE id = '$feed'");
541 if ($debug_enabled) {
542 _debug("update_rss_feed: processing articles...");
545 foreach ($items as $item) {
546 if ($_REQUEST['xdebug'] == 3) {
550 $entry_guid = $item->get_id();
551 if (!$entry_guid) $entry_guid = $item->get_link();
552 if (!$entry_guid) $entry_guid = make_guid_from_title($item->get_title());
554 if (!$entry_guid) continue;
556 $entry_guid = "$owner_uid,$entry_guid";
558 $entry_guid_hashed = db_escape_string('SHA1:' . sha1($entry_guid));
560 if ($debug_enabled) {
561 _debug("update_rss_feed: guid $entry_guid / $entry_guid_hashed");
564 $entry_timestamp = "";
566 $entry_timestamp = strtotime($item->get_date());
568 if ($entry_timestamp == -1 ||
!$entry_timestamp ||
$entry_timestamp > time()) {
569 $entry_timestamp = time();
570 $no_orig_date = 'true';
572 $no_orig_date = 'false';
575 $entry_timestamp_fmt = strftime("%Y/%m/%d %H:%M:%S", $entry_timestamp);
577 if ($debug_enabled) {
578 _debug("update_rss_feed: date $entry_timestamp [$entry_timestamp_fmt]");
581 $entry_title = html_entity_decode($item->get_title(), ENT_COMPAT
, 'UTF-8');
583 $entry_link = rewrite_relative_url($site_url, $item->get_link());
585 if ($debug_enabled) {
586 _debug("update_rss_feed: title $entry_title");
587 _debug("update_rss_feed: link $entry_link");
590 if (!$entry_title) $entry_title = date("Y-m-d H:i:s", $entry_timestamp);;
592 $entry_content = $item->get_content();
593 if (!$entry_content) $entry_content = $item->get_description();
595 if ($_REQUEST["xdebug"] == 2) {
596 print "update_rss_feed: content: ";
597 print $entry_content;
601 $entry_comments = $item->data
["comments"];
603 if ($item->get_author()) {
604 $entry_author_item = $item->get_author();
605 $entry_author = $entry_author_item->get_name();
606 if (!$entry_author) $entry_author = $entry_author_item->get_email();
609 $entry_guid = db_escape_string(mb_substr($entry_guid, 0, 245));
611 $entry_comments = db_escape_string(mb_substr(trim($entry_comments), 0, 245));
612 $entry_author = db_escape_string(mb_substr(trim($entry_author), 0, 245));
614 $num_comments = $item->get_item_tags('http://purl.org/rss/1.0/modules/slash/', 'comments');
616 if (is_array($num_comments) && is_array($num_comments[0])) {
617 $num_comments = (int) $num_comments[0]["data"];
622 if ($debug_enabled) {
623 _debug("update_rss_feed: author $entry_author");
624 _debug("update_rss_feed: num_comments: $num_comments");
625 _debug("update_rss_feed: looking for tags [1]...");
628 // parse <category> entries into tags
630 $additional_tags = array();
632 $additional_tags_src = $item->get_categories();
634 if (is_array($additional_tags_src)) {
635 foreach ($additional_tags_src as $tobj) {
636 array_push($additional_tags, $tobj->get_term());
640 if ($debug_enabled) {
641 _debug("update_rss_feed: category tags:");
642 print_r($additional_tags);
645 if ($debug_enabled) {
646 _debug("update_rss_feed: looking for tags [2]...");
649 $entry_tags = array_unique($additional_tags);
651 for ($i = 0; $i < count($entry_tags); $i++
)
652 $entry_tags[$i] = mb_strtolower($entry_tags[$i], 'utf-8');
654 if ($debug_enabled) {
655 //_debug("update_rss_feed: unfiltered tags found:");
656 //print_r($entry_tags);
659 if ($debug_enabled) {
660 _debug("update_rss_feed: done collecting data.");
663 // TODO: less memory-hungry implementation
665 if ($debug_enabled) {
666 _debug("update_rss_feed: applying plugin filters..");
669 // FIXME not sure if owner_uid is a good idea here, we may have a base entry without user entry (?)
670 $result = db_query("SELECT plugin_data,title,content,link,tag_cache,author FROM ttrss_entries, ttrss_user_entries
671 WHERE ref_id = id AND (guid = '".db_escape_string($entry_guid)."' OR guid = '$entry_guid_hashed') AND owner_uid = $owner_uid");
673 if (db_num_rows($result) != 0) {
674 $entry_plugin_data = db_fetch_result($result, 0, "plugin_data");
675 $stored_article = array("title" => db_fetch_result($result, 0, "title"),
676 "content" => db_fetch_result($result, 0, "content"),
677 "link" => db_fetch_result($result, 0, "link"),
678 "tags" => explode(",", db_fetch_result($result, 0, "tag_cache")),
679 "author" => db_fetch_result($result, 0, "author"));
681 $entry_plugin_data = "";
682 $stored_article = array();
685 $article = array("owner_uid" => $owner_uid, // read only
686 "guid" => $entry_guid, // read only
687 "title" => $entry_title,
688 "content" => $entry_content,
689 "link" => $entry_link,
690 "tags" => $entry_tags,
691 "plugin_data" => $entry_plugin_data,
692 "author" => $entry_author,
693 "stored" => $stored_article);
695 foreach ($pluginhost->get_hooks(PluginHost
::HOOK_ARTICLE_FILTER
) as $plugin) {
696 $article = $plugin->hook_article_filter($article);
699 $entry_tags = $article["tags"];
700 $entry_guid = db_escape_string($entry_guid);
701 $entry_title = db_escape_string($article["title"]);
702 $entry_author = db_escape_string($article["author"]);
703 $entry_link = db_escape_string($article["link"]);
704 $entry_plugin_data = db_escape_string($article["plugin_data"]);
705 $entry_content = $article["content"]; // escaped below
708 if ($debug_enabled) {
709 _debug("update_rss_feed: plugin data: $entry_plugin_data");
712 if ($cache_images && is_writable(CACHE_DIR
. '/images'))
713 cache_images($entry_content, $site_url, $debug_enabled);
715 $entry_content = db_escape_string($entry_content, false);
717 $content_hash = "SHA1:" . sha1($entry_content);
721 $result = db_query("SELECT id FROM ttrss_entries
722 WHERE (guid = '$entry_guid' OR guid = '$entry_guid_hashed')");
724 if (db_num_rows($result) == 0) {
726 if ($debug_enabled) {
727 _debug("update_rss_feed: base guid [$entry_guid] not found");
730 // base post entry does not exist, create it
733 "INSERT INTO ttrss_entries
750 '$entry_guid_hashed',
752 '$entry_timestamp_fmt',
758 '$date_feed_processed',
761 '$entry_plugin_data',
764 $article_labels = array();
767 // we keep encountering the entry in feeds, so we need to
768 // update date_updated column so that we don't get horrible
769 // dupes when the entry gets purged and reinserted again e.g.
770 // in the case of SLOW SLOW OMG SLOW updating feeds
772 $base_entry_id = db_fetch_result($result, 0, "id");
774 db_query("UPDATE ttrss_entries SET date_updated = NOW()
775 WHERE id = '$base_entry_id'");
777 $article_labels = get_article_labels($base_entry_id, $owner_uid);
780 // now it should exist, if not - bad luck then
782 $result = db_query("SELECT
783 id,content_hash,no_orig_date,title,plugin_data,guid,
784 ".SUBSTRING_FOR_DATE
."(date_updated,1,19) as date_updated,
785 ".SUBSTRING_FOR_DATE
."(updated,1,19) as updated,
789 WHERE guid = '$entry_guid' OR guid = '$entry_guid_hashed'");
794 if (db_num_rows($result) == 1) {
796 if ($debug_enabled) {
797 _debug("update_rss_feed: base guid found, checking for user record");
800 // this will be used below in update handler
801 $orig_content_hash = db_fetch_result($result, 0, "content_hash");
802 $orig_title = db_fetch_result($result, 0, "title");
803 $orig_num_comments = db_fetch_result($result, 0, "num_comments");
804 $orig_date_updated = strtotime(db_fetch_result($result,
806 $orig_plugin_data = db_fetch_result($result, 0, "plugin_data");
808 $ref_id = db_fetch_result($result, 0, "id");
809 $entry_ref_id = $ref_id;
811 /* $stored_guid = db_fetch_result($result, 0, "guid");
812 if ($stored_guid != $entry_guid_hashed) {
813 if ($debug_enabled) _debug("upgrading compat guid to hashed one");
815 db_query("UPDATE ttrss_entries SET guid = '$entry_guid_hashed' WHERE
819 // check for user post link to main table
821 // do we allow duplicate posts with same GUID in different feeds?
822 if (get_pref("ALLOW_DUPLICATE_POSTS", $owner_uid, false)) {
823 $dupcheck_qpart = "AND (feed_id = '$feed' OR feed_id IS NULL)";
825 $dupcheck_qpart = "";
828 /* Collect article tags here so we could filter by them: */
830 $article_filters = get_article_filters($filters, $entry_title,
831 $entry_content, $entry_link, $entry_timestamp, $entry_author,
834 if ($debug_enabled) {
835 _debug("update_rss_feed: article filters: ");
836 if (count($article_filters) != 0) {
837 print_r($article_filters);
841 if (find_article_filter($article_filters, "filter")) {
842 db_query("COMMIT"); // close transaction in progress
846 $score = calculate_article_score($article_filters);
848 if ($debug_enabled) {
849 _debug("update_rss_feed: initial score: $score");
852 $query = "SELECT ref_id, int_id FROM ttrss_user_entries WHERE
853 ref_id = '$ref_id' AND owner_uid = '$owner_uid'
856 // if ($_REQUEST["xdebug"]) print "$query\n";
858 $result = db_query($query);
860 // okay it doesn't exist - create user entry
861 if (db_num_rows($result) == 0) {
863 if ($debug_enabled) {
864 _debug("update_rss_feed: user record not found, creating...");
867 if ($score >= -500 && !find_article_filter($article_filters, 'catchup')) {
869 $last_read_qpart = 'NULL';
872 $last_read_qpart = 'NOW()';
875 if (find_article_filter($article_filters, 'mark') ||
$score > 1000) {
881 if (find_article_filter($article_filters, 'publish')) {
884 $published = 'false';
889 if (DB_TYPE
== "pgsql" and defined('_NGRAM_TITLE_DUPLICATE_THRESHOLD')) {
891 $result = db_query("SELECT COUNT(*) AS similar FROM
892 ttrss_entries,ttrss_user_entries
893 WHERE ref_id = id AND updated >= NOW() - INTERVAL '7 day'
894 AND similarity(title, '$entry_title') >= "._NGRAM_TITLE_DUPLICATE_THRESHOLD
."
895 AND owner_uid = $owner_uid");
897 $ngram_similar = db_fetch_result($result, 0, "similar");
899 if ($debug_enabled) {
900 _debug("update_rss_feed: N-gram similar results: $ngram_similar");
903 if ($ngram_similar > 0) {
908 $last_marked = ($marked == 'true') ?
'NOW()' : 'NULL';
909 $last_published = ($published == 'true') ?
'NOW()' : 'NULL';
912 "INSERT INTO ttrss_user_entries
913 (ref_id, owner_uid, feed_id, unread, last_read, marked,
914 published, score, tag_cache, label_cache, uuid,
915 last_marked, last_published)
916 VALUES ('$ref_id', '$owner_uid', '$feed', $unread,
917 $last_read_qpart, $marked, $published, '$score', '', '',
918 '', $last_marked, $last_published)");
920 if (PUBSUBHUBBUB_HUB
&& $published == 'true') {
921 $rss_link = get_self_url_prefix() .
922 "/public.php?op=rss&id=-2&key=" .
923 get_feed_access_key(-2, false, $owner_uid);
925 $p = new Publisher(PUBSUBHUBBUB_HUB
);
927 $pubsub_result = $p->publish_update($rss_link);
931 "SELECT int_id FROM ttrss_user_entries WHERE
932 ref_id = '$ref_id' AND owner_uid = '$owner_uid' AND
933 feed_id = '$feed' LIMIT 1");
935 if (db_num_rows($result) == 1) {
936 $entry_int_id = db_fetch_result($result, 0, "int_id");
939 if ($debug_enabled) {
940 _debug("update_rss_feed: user record FOUND");
943 $entry_ref_id = db_fetch_result($result, 0, "ref_id");
944 $entry_int_id = db_fetch_result($result, 0, "int_id");
947 if ($debug_enabled) {
948 _debug("update_rss_feed: RID: $entry_ref_id, IID: $entry_int_id");
951 $post_needs_update = false;
952 $update_insignificant = false;
954 if ($orig_num_comments != $num_comments) {
955 $post_needs_update = true;
956 $update_insignificant = true;
959 if ($entry_plugin_data != $orig_plugin_data) {
960 $post_needs_update = true;
961 $update_insignificant = true;
964 if ($content_hash != $orig_content_hash) {
965 $post_needs_update = true;
966 $update_insignificant = false;
969 if (db_escape_string($orig_title) != $entry_title) {
970 $post_needs_update = true;
971 $update_insignificant = false;
974 // if post needs update, update it and mark all user entries
975 // linking to this post as updated
976 if ($post_needs_update) {
978 if (defined('DAEMON_EXTENDED_DEBUG')) {
979 _debug("update_rss_feed: post $entry_guid_hashed needs update...");
982 // print "<!-- post $orig_title needs update : $post_needs_update -->";
984 db_query("UPDATE ttrss_entries
985 SET title = '$entry_title', content = '$entry_content',
986 content_hash = '$content_hash',
987 updated = '$entry_timestamp_fmt',
988 num_comments = '$num_comments',
989 plugin_data = '$entry_plugin_data'
990 WHERE id = '$ref_id'");
992 if (!$update_insignificant) {
993 if ($mark_unread_on_update) {
994 db_query("UPDATE ttrss_user_entries
995 SET last_read = null, unread = true WHERE ref_id = '$ref_id'");
1003 if ($debug_enabled) {
1004 _debug("update_rss_feed: assigning labels...");
1007 assign_article_to_label_filters($entry_ref_id, $article_filters,
1008 $owner_uid, $article_labels);
1010 if ($debug_enabled) {
1011 _debug("update_rss_feed: looking for enclosures...");
1016 $enclosures = array();
1018 $encs = $item->get_enclosures();
1020 if (is_array($encs)) {
1021 foreach ($encs as $e) {
1023 $e->link
, $e->type
, $e->length
);
1024 array_push($enclosures, $e_item);
1028 if ($debug_enabled) {
1029 _debug("update_rss_feed: article enclosures:");
1030 print_r($enclosures);
1035 foreach ($enclosures as $enc) {
1036 $enc_url = db_escape_string($enc[0]);
1037 $enc_type = db_escape_string($enc[1]);
1038 $enc_dur = db_escape_string($enc[2]);
1040 $result = db_query("SELECT id FROM ttrss_enclosures
1041 WHERE content_url = '$enc_url' AND post_id = '$entry_ref_id'");
1043 if (db_num_rows($result) == 0) {
1044 db_query("INSERT INTO ttrss_enclosures
1045 (content_url, content_type, title, duration, post_id) VALUES
1046 ('$enc_url', '$enc_type', '', '$enc_dur', '$entry_ref_id')");
1052 // check for manual tags (we have to do it here since they're loaded from filters)
1054 foreach ($article_filters as $f) {
1055 if ($f["type"] == "tag") {
1057 $manual_tags = trim_array(explode(",", $f["param"]));
1059 foreach ($manual_tags as $tag) {
1060 if (tag_is_valid($tag)) {
1061 array_push($entry_tags, $tag);
1069 $boring_tags = trim_array(explode(",", mb_strtolower(get_pref(
1070 'BLACKLISTED_TAGS', $owner_uid, ''), 'utf-8')));
1072 $filtered_tags = array();
1073 $tags_to_cache = array();
1075 if ($entry_tags && is_array($entry_tags)) {
1076 foreach ($entry_tags as $tag) {
1077 if (array_search($tag, $boring_tags) === false) {
1078 array_push($filtered_tags, $tag);
1083 $filtered_tags = array_unique($filtered_tags);
1085 if ($debug_enabled) {
1086 _debug("update_rss_feed: filtered article tags:");
1087 print_r($filtered_tags);
1090 // Save article tags in the database
1092 if (count($filtered_tags) > 0) {
1096 foreach ($filtered_tags as $tag) {
1098 $tag = sanitize_tag($tag);
1099 $tag = db_escape_string($tag);
1101 if (!tag_is_valid($tag)) continue;
1103 $result = db_query("SELECT id FROM ttrss_tags
1104 WHERE tag_name = '$tag' AND post_int_id = '$entry_int_id' AND
1105 owner_uid = '$owner_uid' LIMIT 1");
1107 if ($result && db_num_rows($result) == 0) {
1109 db_query("INSERT INTO ttrss_tags
1110 (owner_uid,tag_name,post_int_id)
1111 VALUES ('$owner_uid','$tag', '$entry_int_id')");
1114 array_push($tags_to_cache, $tag);
1117 /* update the cache */
1119 $tags_to_cache = array_unique($tags_to_cache);
1121 $tags_str = db_escape_string(join(",", $tags_to_cache));
1123 db_query("UPDATE ttrss_user_entries
1124 SET tag_cache = '$tags_str' WHERE ref_id = '$entry_ref_id'
1125 AND owner_uid = $owner_uid");
1130 if (get_pref("AUTO_ASSIGN_LABELS", $owner_uid, false)) {
1131 if ($debug_enabled) {
1132 _debug("update_rss_feed: auto-assigning labels...");
1135 foreach ($labels as $label) {
1136 $caption = preg_quote($label["caption"]);
1138 if ($caption && preg_match("/\b$caption\b/i", "$tags_str " . strip_tags($entry_content) . " $entry_title")) {
1139 if (!labels_contains_caption($article_labels, $caption)) {
1140 label_add_article($entry_ref_id, $caption, $owner_uid);
1146 if ($debug_enabled) {
1147 _debug("update_rss_feed: article processed");
1151 if (!$last_updated) {
1152 if ($debug_enabled) {
1153 _debug("update_rss_feed: new feed, catching it up...");
1155 catchup_feed($feed, false, $owner_uid);
1158 if ($debug_enabled) {
1159 _debug("purging feed...");
1162 purge_feed($feed, 0, $debug_enabled);
1164 db_query("UPDATE ttrss_feeds
1165 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
1167 // db_query("COMMIT");
1171 $error_msg = db_escape_string(mb_substr($rss->error(), 0, 245));
1173 if ($debug_enabled) {
1174 _debug("update_rss_feed: error fetching feed: $error_msg");
1178 "UPDATE ttrss_feeds SET last_error = '$error_msg',
1179 last_updated = NOW() WHERE id = '$feed'");
1184 if ($debug_enabled) {
1185 _debug("update_rss_feed: done");
1190 function cache_images($html, $site_url, $debug) {
1191 $cache_dir = CACHE_DIR
. "/images";
1193 libxml_use_internal_errors(true);
1195 $charset_hack = '<head>
1196 <meta http-equiv="Content-Type" content="text/html; charset=utf-8"/>
1199 $doc = new DOMDocument();
1200 $doc->loadHTML($charset_hack . $html);
1201 $xpath = new DOMXPath($doc);
1203 $entries = $xpath->query('(//img[@src])');
1205 foreach ($entries as $entry) {
1206 if ($entry->hasAttribute('src')) {
1207 $src = rewrite_relative_url($site_url, $entry->getAttribute('src'));
1209 $local_filename = CACHE_DIR
. "/images/" . sha1($src) . ".png";
1211 if ($debug) _debug("cache_images: downloading: $src to $local_filename");
1213 if (!file_exists($local_filename)) {
1214 $file_content = fetch_file_contents($src);
1216 if ($file_content && strlen($file_content) > 1024) {
1217 file_put_contents($local_filename, $file_content);
1221 if (file_exists($local_filename)) {
1222 $entry->setAttribute('src', SELF_URL_PATH
. '/image.php?url=' .
1223 base64_encode($src));
1228 $node = $doc->getElementsByTagName('body')->item(0);
1230 return $doc->saveXML($node);
1233 function expire_error_log($debug) {
1234 if ($debug) _debug("Removing old error log entries...");
1236 if (DB_TYPE
== "pgsql") {
1237 db_query("DELETE FROM ttrss_error_log
1238 WHERE created_at < NOW() - INTERVAL '7 days'");
1240 db_query("DELETE FROM ttrss_error_log
1241 WHERE created_at < DATE_SUB(NOW(), INTERVAL 7 DAY)");
1246 function expire_lock_files($debug) {
1247 if ($debug) _debug("Removing old lock files...");
1251 if (is_writable(LOCK_DIRECTORY
)) {
1252 $files = glob(LOCK_DIRECTORY
. "/*.lock");
1255 foreach ($files as $file) {
1256 if (!file_is_locked($file) && time() - filemtime($file) > 86400*2) {
1264 if ($debug) _debug("Removed $num_deleted files.");
1267 function expire_cached_files($debug) {
1268 foreach (array("simplepie", "images", "export", "upload") as $dir) {
1269 $cache_dir = CACHE_DIR
. "/$dir";
1271 if ($debug) _debug("Expiring $cache_dir");
1275 if (is_writable($cache_dir)) {
1276 $files = glob("$cache_dir/*");
1279 foreach ($files as $file) {
1280 if (time() - filemtime($file) > 86400*7) {
1289 if ($debug) _debug("Removed $num_deleted files.");
1294 * Source: http://www.php.net/manual/en/function.parse-url.php#104527
1295 * Returns the url query as associative array
1297 * @param string query
1298 * @return array params
1300 function convertUrlQuery($query) {
1301 $queryParts = explode('&', $query);
1305 foreach ($queryParts as $param) {
1306 $item = explode('=', $param);
1307 $params[$item[0]] = $item[1];
1313 function get_article_filters($filters, $title, $content, $link, $timestamp, $author, $tags) {
1316 foreach ($filters as $filter) {
1317 $match_any_rule = $filter["match_any_rule"];
1318 $inverse = $filter["inverse"];
1319 $filter_match = false;
1321 foreach ($filter["rules"] as $rule) {
1323 $reg_exp = $rule["reg_exp"];
1324 $rule_inverse = $rule["inverse"];
1329 switch ($rule["type"]) {
1331 $match = @preg_match
("/$reg_exp/i", $title);
1334 // we don't need to deal with multiline regexps
1335 $content = preg_replace("/[\r\n\t]/", "", $content);
1337 $match = @preg_match
("/$reg_exp/i", $content);
1340 // we don't need to deal with multiline regexps
1341 $content = preg_replace("/[\r\n\t]/", "", $content);
1343 $match = (@preg_match
("/$reg_exp/i", $title) || @preg_match
("/$reg_exp/i", $content));
1346 $match = @preg_match
("/$reg_exp/i", $link);
1349 $match = @preg_match
("/$reg_exp/i", $author);
1352 $tag_string = join(",", $tags);
1353 $match = @preg_match
("/$reg_exp/i", $tag_string);
1357 if ($rule_inverse) $match = !$match;
1359 if ($match_any_rule) {
1361 $filter_match = true;
1365 $filter_match = $match;
1372 if ($inverse) $filter_match = !$filter_match;
1374 if ($filter_match) {
1375 foreach ($filter["actions"] AS $action) {
1376 array_push($matches, $action);
1378 // if Stop action encountered, perform no further processing
1379 if ($action["type"] == "stop") return $matches;
1387 function find_article_filter($filters, $filter_name) {
1388 foreach ($filters as $f) {
1389 if ($f["type"] == $filter_name) {
1396 function find_article_filters($filters, $filter_name) {
1399 foreach ($filters as $f) {
1400 if ($f["type"] == $filter_name) {
1401 array_push($results, $f);
1407 function calculate_article_score($filters) {
1410 foreach ($filters as $f) {
1411 if ($f["type"] == "score") {
1412 $score +
= $f["param"];
1418 function labels_contains_caption($labels, $caption) {
1419 foreach ($labels as $label) {
1420 if ($label[1] == $caption) {
1428 function assign_article_to_label_filters($id, $filters, $owner_uid, $article_labels) {
1429 foreach ($filters as $f) {
1430 if ($f["type"] == "label") {
1431 if (!labels_contains_caption($article_labels, $f["param"])) {
1432 label_add_article($id, $f["param"], $owner_uid);
1438 function make_guid_from_title($title) {
1439 return preg_replace("/[ \"\',.:;]/", "-",
1440 mb_strtolower(strip_tags($title), 'utf-8'));
1443 function verify_feed_xml($feed_data) {
1444 libxml_use_internal_errors(true);
1445 $doc = new DOMDocument();
1446 $doc->loadXML($feed_data);
1447 $error = libxml_get_last_error();
1448 libxml_clear_errors();