2 define_default('DAEMON_UPDATE_LOGIN_LIMIT', 30);
3 define_default('DAEMON_FEED_LIMIT', 500);
4 define_default('DAEMON_SLEEP_INTERVAL', 120);
5 define_default('_MIN_CACHE_IMAGE_SIZE', 1024);
7 function calculate_article_hash($article, $pluginhost) {
10 foreach ($article as $k => $v) {
11 if ($k != "feed" && isset($v)) {
12 $x = strip_tags(is_array($v) ?
implode(",", $v) : $v);
14 //_debug("$k:" . sha1($x) . ":" . htmlspecialchars($x), true);
16 $tmp .= sha1("$k:" . sha1($x));
20 return sha1(implode(",", $pluginhost->get_plugin_names()) . $tmp);
23 function update_feedbrowser_cache() {
25 $result = db_query("SELECT feed_url, site_url, title, COUNT(id) AS subscribers
26 FROM ttrss_feeds WHERE (SELECT COUNT(id) = 0 FROM ttrss_feeds AS tf
27 WHERE tf.feed_url = ttrss_feeds.feed_url
28 AND (private IS true OR auth_login != '' OR auth_pass != '' OR feed_url LIKE '%:%@%/%'))
29 GROUP BY feed_url, site_url, title ORDER BY subscribers DESC LIMIT 1000");
33 db_query("DELETE FROM ttrss_feedbrowser_cache");
37 while ($line = db_fetch_assoc($result)) {
38 $subscribers = db_escape_string($line["subscribers"]);
39 $feed_url = db_escape_string($line["feed_url"]);
40 $title = db_escape_string($line["title"]);
41 $site_url = db_escape_string($line["site_url"]);
43 $tmp_result = db_query("SELECT subscribers FROM
44 ttrss_feedbrowser_cache WHERE feed_url = '$feed_url'");
46 if (db_num_rows($tmp_result) == 0) {
48 db_query("INSERT INTO ttrss_feedbrowser_cache
49 (feed_url, site_url, title, subscribers) VALUES ('$feed_url',
50 '$site_url', '$title', '$subscribers')");
66 * Update a feed batch.
67 * Used by daemons to update n feeds by run.
68 * Only update feed needing a update, and not being processed
71 * @param mixed $link Database link
72 * @param integer $limit Maximum number of feeds in update batch. Default to DAEMON_FEED_LIMIT.
73 * @param boolean $from_http Set to true if you call this function from http to disable cli specific code.
74 * @param boolean $debug Set to false to disable debug output. Default to true.
77 function update_daemon_common($limit = DAEMON_FEED_LIMIT
, $from_http = false, $debug = true) {
78 // Process all other feeds using last_updated and interval parameters
80 $schema_version = get_schema_version();
82 if ($schema_version != SCHEMA_VERSION
) {
83 die("Schema version is wrong, please upgrade the database.\n");
86 define('PREFS_NO_CACHE', true);
88 // Test if the user has loggued in recently. If not, it does not update its feeds.
89 if (!SINGLE_USER_MODE
&& DAEMON_UPDATE_LOGIN_LIMIT
> 0) {
90 if (DB_TYPE
== "pgsql") {
91 $login_thresh_qpart = "AND ttrss_users.last_login >= NOW() - INTERVAL '".DAEMON_UPDATE_LOGIN_LIMIT
." days'";
93 $login_thresh_qpart = "AND ttrss_users.last_login >= DATE_SUB(NOW(), INTERVAL ".DAEMON_UPDATE_LOGIN_LIMIT
." DAY)";
96 $login_thresh_qpart = "";
99 // Test if the feed need a update (update interval exceeded).
100 if (DB_TYPE
== "pgsql") {
101 $update_limit_qpart = "AND ((
102 ttrss_feeds.update_interval = 0
103 AND ttrss_user_prefs.value != '-1'
104 AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_user_prefs.value || ' minutes') AS INTERVAL)
106 ttrss_feeds.update_interval > 0
107 AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_feeds.update_interval || ' minutes') AS INTERVAL)
108 ) OR (ttrss_feeds.last_updated IS NULL
109 AND ttrss_user_prefs.value != '-1')
110 OR (last_updated = '1970-01-01 00:00:00'
111 AND ttrss_user_prefs.value != '-1'))";
113 $update_limit_qpart = "AND ((
114 ttrss_feeds.update_interval = 0
115 AND ttrss_user_prefs.value != '-1'
116 AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL CONVERT(ttrss_user_prefs.value, SIGNED INTEGER) MINUTE)
118 ttrss_feeds.update_interval > 0
119 AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL ttrss_feeds.update_interval MINUTE)
120 ) OR (ttrss_feeds.last_updated IS NULL
121 AND ttrss_user_prefs.value != '-1')
122 OR (last_updated = '1970-01-01 00:00:00'
123 AND ttrss_user_prefs.value != '-1'))";
126 // Test if feed is currently being updated by another process.
127 if (DB_TYPE
== "pgsql") {
128 $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < NOW() - INTERVAL '10 minutes')";
130 $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < DATE_SUB(NOW(), INTERVAL 10 MINUTE))";
133 // Test if there is a limit to number of updated feeds
135 if($limit) $query_limit = sprintf("LIMIT %d", $limit);
137 // Update the least recently updated feeds first
138 $query_order = "ORDER BY last_updated";
139 if (DB_TYPE
== "pgsql") $query_order .= " NULLS FIRST";
141 $query = "SELECT DISTINCT ttrss_feeds.feed_url, ttrss_feeds.last_updated
143 ttrss_feeds, ttrss_users, ttrss_user_prefs
145 ttrss_feeds.owner_uid = ttrss_users.id
146 AND ttrss_user_prefs.profile IS NULL
147 AND ttrss_users.id = ttrss_user_prefs.owner_uid
148 AND ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL'
149 $login_thresh_qpart $update_limit_qpart
150 $updstart_thresh_qpart
151 $query_order $query_limit";
153 // We search for feed needing update.
154 $result = db_query($query);
156 if($debug) _debug(sprintf("Scheduled %d feeds to update...", db_num_rows($result)));
158 // Here is a little cache magic in order to minimize risk of double feed updates.
159 $feeds_to_update = array();
160 while ($line = db_fetch_assoc($result)) {
161 array_push($feeds_to_update, db_escape_string($line['feed_url']));
164 // We update the feed last update started date before anything else.
165 // There is no lag due to feed contents downloads
166 // It prevent an other process to update the same feed.
168 if(count($feeds_to_update) > 0) {
169 $feeds_quoted = array();
171 foreach ($feeds_to_update as $feed) {
172 array_push($feeds_quoted, "'" . db_escape_string($feed) . "'");
175 db_query(sprintf("UPDATE ttrss_feeds SET last_update_started = NOW()
176 WHERE feed_url IN (%s)", implode(',', $feeds_quoted)));
180 $bstarted = microtime(true);
182 // For each feed, we call the feed update function.
183 foreach ($feeds_to_update as $feed) {
184 if($debug) _debug("Base feed: $feed");
186 //update_rss_feed($line["id"], true);
188 // since we have the data cached, we can deal with other feeds with the same url
190 $tmp_result = db_query("SELECT DISTINCT ttrss_feeds.id,last_updated,ttrss_feeds.owner_uid
191 FROM ttrss_feeds, ttrss_users, ttrss_user_prefs WHERE
192 ttrss_user_prefs.owner_uid = ttrss_feeds.owner_uid AND
193 ttrss_users.id = ttrss_user_prefs.owner_uid AND
194 ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL' AND
195 ttrss_user_prefs.profile IS NULL AND
196 feed_url = '".db_escape_string($feed)."' AND
197 (ttrss_feeds.update_interval > 0 OR
198 ttrss_user_prefs.value != '-1')
200 ORDER BY ttrss_feeds.id $query_limit");
202 if (db_num_rows($tmp_result) > 0) {
205 while ($tline = db_fetch_assoc($tmp_result)) {
206 if($debug) _debug(" => " . $tline["last_updated"] . ", " . $tline["id"] . " " . $tline["owner_uid"]);
208 $fstarted = microtime(true);
209 $rss = update_rss_feed($tline["id"], true, false);
210 _debug_suppress(false);
212 _debug(sprintf(" %.4f (sec)", microtime(true) - $fstarted));
220 _debug(sprintf("Processed %d feeds in %.4f (sec), %.4f (sec/feed avg)", $nf,
221 microtime(true) - $bstarted, (microtime(true) - $bstarted) / $nf));
224 require_once "digest.php";
226 // Send feed digests by email if needed.
227 send_headlines_digests($debug);
231 } // function update_daemon_common
233 // this is used when subscribing
234 function set_basic_feed_info($feed) {
236 $feed = db_escape_string($feed);
238 $result = db_query("SELECT feed_url,auth_pass,auth_login,auth_pass_encrypted
239 FROM ttrss_feeds WHERE id = '$feed'");
241 $auth_pass_encrypted = sql_bool_to_bool(db_fetch_result($result,
242 0, "auth_pass_encrypted"));
244 $auth_login = db_fetch_result($result, 0, "auth_login");
245 $auth_pass = db_fetch_result($result, 0, "auth_pass");
247 if ($auth_pass_encrypted) {
248 require_once "crypt.php";
249 $auth_pass = decrypt_string($auth_pass);
252 $fetch_url = db_fetch_result($result, 0, "feed_url");
254 $feed_data = fetch_file_contents($fetch_url, false,
255 $auth_login, $auth_pass, false,
256 FEED_FETCH_TIMEOUT_TIMEOUT
,
259 global $fetch_curl_used;
261 if (!$fetch_curl_used) {
262 $tmp = @gzdecode
($feed_data);
264 if ($tmp) $feed_data = $tmp;
267 $feed_data = trim($feed_data);
269 $rss = new FeedParser($feed_data);
272 if (!$rss->error()) {
274 $result = db_query("SELECT title, site_url FROM ttrss_feeds WHERE id = '$feed'");
276 $registered_title = db_fetch_result($result, 0, "title");
277 $orig_site_url = db_fetch_result($result, 0, "site_url");
279 $site_url = db_escape_string(mb_substr(rewrite_relative_url($fetch_url, $rss->get_link()), 0, 245));
280 $feed_title = db_escape_string(mb_substr($rss->get_title(), 0, 199));
282 if ($feed_title && (!$registered_title ||
$registered_title == "[Unknown]")) {
283 db_query("UPDATE ttrss_feeds SET
284 title = '$feed_title' WHERE id = '$feed'");
287 if ($site_url && $orig_site_url != $site_url) {
288 db_query("UPDATE ttrss_feeds SET
289 site_url = '$site_url' WHERE id = '$feed'");
294 // ignore_daemon is not used
295 function update_rss_feed($feed, $ignore_daemon = false, $no_cache = false, $rss = false) {
297 $debug_enabled = defined('DAEMON_EXTENDED_DEBUG') ||
$_REQUEST['xdebug'];
299 _debug_suppress(!$debug_enabled);
300 _debug("start", $debug_enabled);
302 $result = db_query("SELECT title FROM ttrss_feeds
303 WHERE id = '$feed'");
304 $title = db_fetch_result($result, 0, "title");
306 // feed was batch-subscribed or something, we need to get basic info
307 // this is not optimal currently as it fetches stuff separately TODO: optimize
308 if ($title == "[Unknown]") {
309 _debug("setting basic feed info for $feed...");
310 set_basic_feed_info($feed);
313 $result = db_query("SELECT id,update_interval,auth_login,
314 feed_url,auth_pass,cache_images,
315 mark_unread_on_update, owner_uid,
316 pubsub_state, auth_pass_encrypted,
318 (SELECT max(date_entered) FROM
319 ttrss_entries, ttrss_user_entries where ref_id = id AND feed_id = '$feed') AS last_article_timestamp
320 FROM ttrss_feeds WHERE id = '$feed'");
322 if (db_num_rows($result) == 0) {
323 _debug("feed $feed NOT FOUND/SKIPPED", $debug_enabled);
327 $last_article_timestamp = @strtotime
(db_fetch_result($result, 0, "last_article_timestamp"));
329 if (defined('_DISABLE_HTTP_304'))
330 $last_article_timestamp = 0;
332 $owner_uid = db_fetch_result($result, 0, "owner_uid");
333 $mark_unread_on_update = sql_bool_to_bool(db_fetch_result($result,
334 0, "mark_unread_on_update"));
335 $pubsub_state = db_fetch_result($result, 0, "pubsub_state");
336 $auth_pass_encrypted = sql_bool_to_bool(db_fetch_result($result,
337 0, "auth_pass_encrypted"));
339 db_query("UPDATE ttrss_feeds SET last_update_started = NOW()
340 WHERE id = '$feed'");
342 $auth_login = db_fetch_result($result, 0, "auth_login");
343 $auth_pass = db_fetch_result($result, 0, "auth_pass");
345 if ($auth_pass_encrypted) {
346 require_once "crypt.php";
347 $auth_pass = decrypt_string($auth_pass);
350 $cache_images = sql_bool_to_bool(db_fetch_result($result, 0, "cache_images"));
351 $fetch_url = db_fetch_result($result, 0, "feed_url");
352 $feed_language = db_escape_string(mb_strtolower(db_fetch_result($result, 0, "feed_language")));
353 if (!$feed_language) $feed_language = 'english';
355 $feed = db_escape_string($feed);
357 $date_feed_processed = date('Y-m-d H:i');
359 $cache_filename = CACHE_DIR
. "/simplepie/" . sha1($fetch_url) . ".xml";
361 $pluginhost = new PluginHost();
362 $pluginhost->set_debug($debug_enabled);
363 $user_plugins = get_pref("_ENABLED_PLUGINS", $owner_uid);
365 $pluginhost->load(PLUGINS
, PluginHost
::KIND_ALL
);
366 $pluginhost->load($user_plugins, PluginHost
::KIND_USER
, $owner_uid);
367 $pluginhost->load_data();
369 if ($rss && is_object($rss) && get_class($rss) == "FeedParser") {
370 _debug("using previously initialized parser object");
374 $force_refetch = isset($_REQUEST["force_refetch"]);
376 foreach ($pluginhost->get_hooks(PluginHost
::HOOK_FETCH_FEED
) as $plugin) {
377 $feed_data = $plugin->hook_fetch_feed($feed_data, $fetch_url, $owner_uid, $feed, $last_article_timestamp, $auth_login, $auth_pass);
382 file_exists($cache_filename) &&
383 is_readable($cache_filename) &&
384 !$auth_login && !$auth_pass &&
385 filemtime($cache_filename) > time() - 30) {
387 _debug("using local cache [$cache_filename].", $debug_enabled);
389 @$feed_data = file_get_contents($cache_filename);
392 $rss_hash = sha1($feed_data);
396 _debug("local cache will not be used for this feed", $debug_enabled);
399 // fetch feed from source
401 _debug("fetching [$fetch_url]...", $debug_enabled);
402 _debug("If-Modified-Since: ".gmdate('D, d M Y H:i:s \G\M\T', $last_article_timestamp), $debug_enabled);
404 $feed_data = fetch_file_contents($fetch_url, false,
405 $auth_login, $auth_pass, false,
406 $no_cache ? FEED_FETCH_NO_CACHE_TIMEOUT
: FEED_FETCH_TIMEOUT
,
407 $force_refetch ?
0 : $last_article_timestamp);
409 global $fetch_curl_used;
411 if (!$fetch_curl_used) {
412 $tmp = @gzdecode
($feed_data);
414 if ($tmp) $feed_data = $tmp;
417 $feed_data = trim($feed_data);
419 _debug("fetch done.", $debug_enabled);
421 // cache vanilla feed data for re-use
422 if ($feed_data && !$auth_pass && !$auth_login && is_writable(CACHE_DIR
. "/simplepie")) {
423 $new_rss_hash = sha1($feed_data);
425 if ($new_rss_hash != $rss_hash) {
426 _debug("saving $cache_filename", $debug_enabled);
427 @file_put_contents
($cache_filename, $feed_data);
433 global $fetch_last_error;
434 global $fetch_last_error_code;
436 _debug("unable to fetch: $fetch_last_error [$fetch_last_error_code]", $debug_enabled);
441 if ($fetch_last_error_code != 304) {
442 $error_escaped = db_escape_string($fetch_last_error);
444 _debug("source claims data not modified, nothing to do.", $debug_enabled);
448 "UPDATE ttrss_feeds SET last_error = '$error_escaped',
449 last_updated = NOW() WHERE id = '$feed'");
455 foreach ($pluginhost->get_hooks(PluginHost
::HOOK_FEED_FETCHED
) as $plugin) {
456 $feed_data = $plugin->hook_feed_fetched($feed_data, $fetch_url, $owner_uid, $feed);
459 // set last update to now so if anything *simplepie* crashes later we won't be
460 // continuously failing on the same feed
461 //db_query("UPDATE ttrss_feeds SET last_updated = NOW() WHERE id = '$feed'");
464 $rss = new FeedParser($feed_data);
470 $feed = db_escape_string($feed);
472 if (!$rss->error()) {
474 // We use local pluginhost here because we need to load different per-user feed plugins
475 $pluginhost->run_hooks(PluginHost
::HOOK_FEED_PARSED
, "hook_feed_parsed", $rss);
477 _debug("language: $feed_language", $debug_enabled);
478 _debug("processing feed data...", $debug_enabled);
480 // db_query("BEGIN");
482 if (DB_TYPE
== "pgsql") {
483 $favicon_interval_qpart = "favicon_last_checked < NOW() - INTERVAL '12 hour'";
485 $favicon_interval_qpart = "favicon_last_checked < DATE_SUB(NOW(), INTERVAL 12 HOUR)";
488 $result = db_query("SELECT owner_uid,favicon_avg_color,
489 (favicon_last_checked IS NULL OR $favicon_interval_qpart) AS
491 FROM ttrss_feeds WHERE id = '$feed'");
493 $favicon_needs_check = sql_bool_to_bool(db_fetch_result($result, 0,
494 "favicon_needs_check"));
495 $favicon_avg_color = db_fetch_result($result, 0, "favicon_avg_color");
497 $owner_uid = db_fetch_result($result, 0, "owner_uid");
499 $site_url = db_escape_string(mb_substr(rewrite_relative_url($fetch_url, $rss->get_link()), 0, 245));
501 _debug("site_url: $site_url", $debug_enabled);
502 _debug("feed_title: " . $rss->get_title(), $debug_enabled);
504 if ($favicon_needs_check ||
$force_refetch) {
506 /* terrible hack: if we crash on floicon shit here, we won't check
507 * the icon avgcolor again (unless the icon got updated) */
509 $favicon_file = ICONS_DIR
. "/$feed.ico";
510 $favicon_modified = @filemtime
($favicon_file);
512 _debug("checking favicon...", $debug_enabled);
514 check_feed_favicon($site_url, $feed);
515 $favicon_modified_new = @filemtime
($favicon_file);
517 if ($favicon_modified_new > $favicon_modified)
518 $favicon_avg_color = '';
520 if (file_exists($favicon_file) && function_exists("imagecreatefromstring") && $favicon_avg_color == '') {
521 require_once "colors.php";
523 db_query("UPDATE ttrss_feeds SET favicon_avg_color = 'fail' WHERE
526 $favicon_color = db_escape_string(
527 calculate_avg_color($favicon_file));
529 $favicon_colorstring = ",favicon_avg_color = '".$favicon_color."'";
530 } else if ($favicon_avg_color == 'fail') {
531 _debug("floicon failed on this file, not trying to recalculate avg color", $debug_enabled);
534 db_query("UPDATE ttrss_feeds SET favicon_last_checked = NOW()
536 WHERE id = '$feed'");
539 _debug("loading filters & labels...", $debug_enabled);
541 $filters = load_filters($feed, $owner_uid);
543 _debug("" . count($filters) . " filters loaded.", $debug_enabled);
545 $items = $rss->get_items();
547 if (!is_array($items)) {
548 _debug("no articles found.", $debug_enabled);
550 db_query("UPDATE ttrss_feeds
551 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
553 return; // no articles
556 if ($pubsub_state != 2 && PUBSUBHUBBUB_ENABLED
) {
558 _debug("checking for PUSH hub...", $debug_enabled);
560 $feed_hub_url = false;
562 $links = $rss->get_links('hub');
564 if ($links && is_array($links)) {
565 foreach ($links as $l) {
571 _debug("feed hub url: $feed_hub_url", $debug_enabled);
573 $feed_self_url = $fetch_url;
575 $links = $rss->get_links('self');
577 if ($links && is_array($links)) {
578 foreach ($links as $l) {
584 _debug("feed self url = $feed_self_url");
586 if ($feed_hub_url && $feed_self_url && function_exists('curl_init') &&
587 !ini_get("open_basedir")) {
589 require_once 'lib/pubsubhubbub/subscriber.php';
591 $callback_url = get_self_url_prefix() .
592 "/public.php?op=pubsub&id=$feed";
594 $s = new Subscriber($feed_hub_url, $callback_url);
596 $rc = $s->subscribe($feed_self_url);
598 _debug("feed hub url found, subscribe request sent. [rc=$rc]", $debug_enabled);
600 db_query("UPDATE ttrss_feeds SET pubsub_state = 1
601 WHERE id = '$feed'");
605 _debug("processing articles...", $debug_enabled);
609 foreach ($items as $item) {
610 if ($_REQUEST['xdebug'] == 3) {
614 if (ini_get("max_execution_time") > 0 && time() - $tstart >= ini_get("max_execution_time") * 0.7) {
615 _debug("looks like there's too many articles to process at once, breaking out", $debug_enabled);
619 $entry_guid = $item->get_id();
620 if (!$entry_guid) $entry_guid = $item->get_link();
621 if (!$entry_guid) $entry_guid = make_guid_from_title($item->get_title());
622 if (!$entry_guid) continue;
624 $entry_guid = "$owner_uid,$entry_guid";
626 $entry_guid_hashed = db_escape_string('SHA1:' . sha1($entry_guid));
628 _debug("guid $entry_guid / $entry_guid_hashed", $debug_enabled);
630 $entry_timestamp = "";
632 $entry_timestamp = $item->get_date();
634 _debug("orig date: " . $item->get_date(), $debug_enabled);
636 if ($entry_timestamp == -1 ||
!$entry_timestamp ||
$entry_timestamp > time()) {
637 $entry_timestamp = time();
640 $entry_timestamp_fmt = strftime("%Y/%m/%d %H:%M:%S", $entry_timestamp);
642 _debug("date $entry_timestamp [$entry_timestamp_fmt]", $debug_enabled);
644 // $entry_title = html_entity_decode($item->get_title(), ENT_COMPAT, 'UTF-8');
645 // $entry_title = decode_numeric_entities($entry_title);
646 $entry_title = $item->get_title();
648 $entry_link = rewrite_relative_url($site_url, $item->get_link());
650 _debug("title $entry_title", $debug_enabled);
651 _debug("link $entry_link", $debug_enabled);
653 if (!$entry_title) $entry_title = date("Y-m-d H:i:s", $entry_timestamp);;
655 $entry_content = $item->get_content();
656 if (!$entry_content) $entry_content = $item->get_description();
658 if ($_REQUEST["xdebug"] == 2) {
660 print $entry_content;
664 $entry_comments = $item->get_comments_url();
665 $entry_author = $item->get_author();
667 $entry_guid = db_escape_string(mb_substr($entry_guid, 0, 245));
669 $entry_comments = db_escape_string(mb_substr(trim($entry_comments), 0, 245));
670 $entry_author = db_escape_string(mb_substr(trim($entry_author), 0, 245));
672 $num_comments = (int) $item->get_comments_count();
674 _debug("author $entry_author", $debug_enabled);
675 _debug("num_comments: $num_comments", $debug_enabled);
676 _debug("looking for tags...", $debug_enabled);
678 // parse <category> entries into tags
680 $additional_tags = array();
682 $additional_tags_src = $item->get_categories();
684 if (is_array($additional_tags_src)) {
685 foreach ($additional_tags_src as $tobj) {
686 array_push($additional_tags, $tobj);
690 $entry_tags = array_unique($additional_tags);
692 for ($i = 0; $i < count($entry_tags); $i++
)
693 $entry_tags[$i] = mb_strtolower($entry_tags[$i], 'utf-8');
695 _debug("tags found: " . join(",", $entry_tags), $debug_enabled);
697 _debug("done collecting data.", $debug_enabled);
699 $result = db_query("SELECT id, content_hash, lang FROM ttrss_entries
700 WHERE guid = '".db_escape_string($entry_guid)."' OR guid = '$entry_guid_hashed'");
702 if (db_num_rows($result) != 0) {
703 $base_entry_id = db_fetch_result($result, 0, "id");
704 $entry_stored_hash = db_fetch_result($result, 0, "content_hash");
705 $article_labels = get_article_labels($base_entry_id, $owner_uid);
706 $entry_language = db_fetch_result($result, 0, "lang");
708 $existing_tags = get_article_tags($base_entry_id, $owner_uid);
709 $entry_tags = array_unique(array_merge($entry_tags, $existing_tags));
712 $base_entry_id = false;
713 $entry_stored_hash = "";
714 $article_labels = array();
715 $entry_language = "";
718 $article = array("owner_uid" => $owner_uid, // read only
719 "guid" => $entry_guid, // read only
720 "guid_hashed" => $entry_guid_hashed, // read only
721 "title" => $entry_title,
722 "content" => $entry_content,
723 "link" => $entry_link,
724 "labels" => $article_labels, // current limitation: can add labels to article, can't remove them
725 "tags" => $entry_tags,
726 "author" => $entry_author,
727 "force_catchup" => false, // ugly hack for the time being
728 "score_modifier" => 0, // no previous value, plugin should recalculate score modifier based on content if needed
729 "language" => $entry_language,
730 "feed" => array("id" => $feed,
731 "fetch_url" => $fetch_url,
732 "site_url" => $site_url)
735 $entry_plugin_data = "";
736 $entry_current_hash = calculate_article_hash($article, $pluginhost);
738 _debug("article hash: $entry_current_hash [stored=$entry_stored_hash]", $debug_enabled);
740 if ($entry_current_hash == $entry_stored_hash && !isset($_REQUEST["force_rehash"])) {
741 _debug("stored article seems up to date [IID: $base_entry_id], updating timestamp only", $debug_enabled);
743 // we keep encountering the entry in feeds, so we need to
744 // update date_updated column so that we don't get horrible
745 // dupes when the entry gets purged and reinserted again e.g.
746 // in the case of SLOW SLOW OMG SLOW updating feeds
748 $base_entry_id = db_fetch_result($result, 0, "id");
750 db_query("UPDATE ttrss_entries SET date_updated = NOW()
751 WHERE id = '$base_entry_id'");
753 // if we allow duplicate posts, we have to continue to
754 // create the user entries for this feed
755 if (!get_pref("ALLOW_DUPLICATE_POSTS", $owner_uid, false)) {
760 _debug("hash differs, applying plugin filters:", $debug_enabled);
762 foreach ($pluginhost->get_hooks(PluginHost
::HOOK_ARTICLE_FILTER
) as $plugin) {
763 _debug("... " . get_class($plugin), $debug_enabled);
765 $start = microtime(true);
766 $article = $plugin->hook_article_filter($article);
768 _debug("=== " . sprintf("%.4f (sec)", microtime(true) - $start), $debug_enabled);
770 $entry_plugin_data .= mb_strtolower(get_class($plugin)) . ",";
773 $entry_plugin_data = db_escape_string($entry_plugin_data);
775 _debug("plugin data: $entry_plugin_data", $debug_enabled);
777 // Workaround: 4-byte unicode requires utf8mb4 in MySQL. See https://tt-rss.org/forum/viewtopic.php?f=1&t=3377&p=20077#p20077
778 if (DB_TYPE
== "mysql") {
779 foreach ($article as $k => $v) {
781 // i guess we'll have to take the risk of 4byte unicode labels & tags here
782 if (!is_array($article[$k])) {
783 $article[$k] = preg_replace('/[\x{10000}-\x{10FFFF}]/u', "\xEF\xBF\xBD", $v);
788 /* Collect article tags here so we could filter by them: */
790 $article_filters = get_article_filters($filters, $article["title"],
791 $article["content"], $article["link"], 0, $article["author"],
794 if ($debug_enabled) {
795 _debug("article filters: ", $debug_enabled);
796 if (count($article_filters) != 0) {
797 print_r($article_filters);
801 $plugin_filter_names = find_article_filters($article_filters, "plugin");
802 $plugin_filter_actions = $pluginhost->get_filter_actions();
804 if (count($plugin_filter_names) > 0) {
805 _debug("applying plugin filter actions...", $debug_enabled);
807 foreach ($plugin_filter_names as $pfn) {
808 list($pfclass,$pfaction) = explode(":", $pfn["param"]);
810 if (isset($plugin_filter_actions[$pfclass])) {
811 $plugin = $pluginhost->get_plugin($pfclass);
813 _debug("... $pfclass: $pfaction", $debug_enabled);
816 $start = microtime(true);
817 $article = $plugin->hook_article_filter_action($article, $pfaction);
819 _debug("=== " . sprintf("%.4f (sec)", microtime(true) - $start), $debug_enabled);
821 _debug("??? $pfclass: plugin object not found.");
824 _debug("??? $pfclass: filter plugin not registered.");
829 $entry_tags = $article["tags"];
830 $entry_guid = db_escape_string($entry_guid);
831 $entry_title = db_escape_string($article["title"]);
832 $entry_author = db_escape_string($article["author"]);
833 $entry_link = db_escape_string($article["link"]);
834 $entry_content = $article["content"]; // escaped below
835 $entry_force_catchup = $article["force_catchup"];
836 $article_labels = $article["labels"];
837 $entry_score_modifier = (int) $article["score_modifier"];
838 $entry_language = db_escape_string($article["language"]);
840 if ($debug_enabled) {
841 _debug("article labels:", $debug_enabled);
842 print_r($article_labels);
845 _debug("force catchup: $entry_force_catchup");
847 if ($cache_images && is_writable(CACHE_DIR
. '/images'))
848 cache_images($entry_content, $site_url, $debug_enabled);
850 $entry_content = db_escape_string($entry_content, false);
854 $result = db_query("SELECT id FROM ttrss_entries
855 WHERE (guid = '$entry_guid' OR guid = '$entry_guid_hashed')");
857 if (db_num_rows($result) == 0) {
859 _debug("base guid [$entry_guid] not found", $debug_enabled);
861 // base post entry does not exist, create it
864 "INSERT INTO ttrss_entries
881 '$entry_guid_hashed',
883 '$entry_timestamp_fmt',
885 '$entry_current_hash',
888 '$date_feed_processed',
891 '$entry_plugin_data',
896 $base_entry_id = db_fetch_result($result, 0, "id");
899 // now it should exist, if not - bad luck then
901 $result = db_query("SELECT id FROM ttrss_entries
902 WHERE guid = '$entry_guid' OR guid = '$entry_guid_hashed'");
907 if (db_num_rows($result) == 1) {
909 _debug("base guid found, checking for user record", $debug_enabled);
911 $ref_id = db_fetch_result($result, 0, "id");
912 $entry_ref_id = $ref_id;
914 /* $stored_guid = db_fetch_result($result, 0, "guid");
915 if ($stored_guid != $entry_guid_hashed) {
916 if ($debug_enabled) _debug("upgrading compat guid to hashed one", $debug_enabled);
918 db_query("UPDATE ttrss_entries SET guid = '$entry_guid_hashed' WHERE
922 // check for user post link to main table
924 // do we allow duplicate posts with same GUID in different feeds?
925 if (get_pref("ALLOW_DUPLICATE_POSTS", $owner_uid, false)) {
926 $dupcheck_qpart = "AND (feed_id = '$feed' OR feed_id IS NULL)";
928 $dupcheck_qpart = "";
931 if (find_article_filter($article_filters, "filter")) {
932 //db_query("COMMIT"); // close transaction in progress
936 $score = calculate_article_score($article_filters) +
$entry_score_modifier;
938 _debug("initial score: $score [including plugin modifier: $entry_score_modifier]", $debug_enabled);
940 $query = "SELECT ref_id, int_id FROM ttrss_user_entries WHERE
941 ref_id = '$ref_id' AND owner_uid = '$owner_uid'
944 // if ($_REQUEST["xdebug"]) print "$query\n";
946 $result = db_query($query);
948 // okay it doesn't exist - create user entry
949 if (db_num_rows($result) == 0) {
951 _debug("user record not found, creating...", $debug_enabled);
953 if ($score >= -500 && !find_article_filter($article_filters, 'catchup') && !$entry_force_catchup) {
955 $last_read_qpart = 'NULL';
958 $last_read_qpart = 'NOW()';
961 if (find_article_filter($article_filters, 'mark') ||
$score > 1000) {
967 if (find_article_filter($article_filters, 'publish')) {
970 $published = 'false';
975 /* if (DB_TYPE == "pgsql" and defined('_NGRAM_TITLE_DUPLICATE_THRESHOLD')) {
977 $result = db_query("SELECT COUNT(*) AS similar FROM
978 ttrss_entries,ttrss_user_entries
979 WHERE ref_id = id AND updated >= NOW() - INTERVAL '7 day'
980 AND similarity(title, '$entry_title') >= "._NGRAM_TITLE_DUPLICATE_THRESHOLD."
981 AND owner_uid = $owner_uid");
983 $ngram_similar = db_fetch_result($result, 0, "similar");
985 _debug("N-gram similar results: $ngram_similar", $debug_enabled);
987 if ($ngram_similar > 0) {
992 $last_marked = ($marked == 'true') ?
'NOW()' : 'NULL';
993 $last_published = ($published == 'true') ?
'NOW()' : 'NULL';
996 "INSERT INTO ttrss_user_entries
997 (ref_id, owner_uid, feed_id, unread, last_read, marked,
998 published, score, tag_cache, label_cache, uuid,
999 last_marked, last_published)
1000 VALUES ('$ref_id', '$owner_uid', '$feed', $unread,
1001 $last_read_qpart, $marked, $published, '$score', '', '',
1002 '', $last_marked, $last_published)");
1004 if (PUBSUBHUBBUB_HUB
&& $published == 'true') {
1005 $rss_link = get_self_url_prefix() .
1006 "/public.php?op=rss&id=-2&key=" .
1007 get_feed_access_key(-2, false, $owner_uid);
1009 $p = new Publisher(PUBSUBHUBBUB_HUB
);
1011 /* $pubsub_result = */ $p->publish_update($rss_link);
1015 "SELECT int_id FROM ttrss_user_entries WHERE
1016 ref_id = '$ref_id' AND owner_uid = '$owner_uid' AND
1017 feed_id = '$feed' LIMIT 1");
1019 if (db_num_rows($result) == 1) {
1020 $entry_int_id = db_fetch_result($result, 0, "int_id");
1023 _debug("user record FOUND", $debug_enabled);
1025 $entry_ref_id = db_fetch_result($result, 0, "ref_id");
1026 $entry_int_id = db_fetch_result($result, 0, "int_id");
1029 _debug("RID: $entry_ref_id, IID: $entry_int_id", $debug_enabled);
1031 if (DB_TYPE
== "pgsql") {
1032 $tsvector_combined = db_escape_string(mb_substr($entry_title . ' ' . strip_tags(str_replace('<', ' <', $entry_content)),
1035 $tsvector_qpart = "tsvector_combined = to_tsvector('$feed_language', '$tsvector_combined'),";
1038 $tsvector_qpart = "";
1041 db_query("UPDATE ttrss_entries
1042 SET title = '$entry_title',
1043 content = '$entry_content',
1044 content_hash = '$entry_current_hash',
1045 updated = '$entry_timestamp_fmt',
1047 num_comments = '$num_comments',
1048 plugin_data = '$entry_plugin_data',
1049 author = '$entry_author',
1050 lang = '$entry_language'
1051 WHERE id = '$ref_id'");
1054 db_query("UPDATE ttrss_user_entries
1055 SET score = '$score' WHERE ref_id = '$ref_id'");
1057 if ($mark_unread_on_update) {
1058 _debug("article updated, marking unread as requested.", $debug_enabled);
1060 db_query("UPDATE ttrss_user_entries
1061 SET last_read = null, unread = true WHERE ref_id = '$ref_id'");
1065 //db_query("COMMIT");
1067 _debug("assigning labels [other]...", $debug_enabled);
1069 foreach ($article_labels as $label) {
1070 label_add_article($entry_ref_id, $label[1], $owner_uid);
1073 _debug("assigning labels [filters]...", $debug_enabled);
1075 assign_article_to_label_filters($entry_ref_id, $article_filters,
1076 $owner_uid, $article_labels);
1078 _debug("looking for enclosures...", $debug_enabled);
1082 $enclosures = array();
1084 $encs = $item->get_enclosures();
1086 if (is_array($encs)) {
1087 foreach ($encs as $e) {
1089 $e->link
, $e->type
, $e->length
, $e->title
, $e->width
, $e->height
);
1090 array_push($enclosures, $e_item);
1094 if ($debug_enabled) {
1095 _debug("article enclosures:", $debug_enabled);
1096 print_r($enclosures);
1099 //db_query("BEGIN");
1102 // db_query("DELETE FROM ttrss_enclosures WHERE post_id = '$entry_ref_id'");
1104 foreach ($enclosures as $enc) {
1105 $enc_url = db_escape_string($enc[0]);
1106 $enc_type = db_escape_string($enc[1]);
1107 $enc_dur = db_escape_string($enc[2]);
1108 $enc_title = db_escape_string($enc[3]);
1109 $enc_width = intval($enc[4]);
1110 $enc_height = intval($enc[5]);
1112 $result = db_query("SELECT id FROM ttrss_enclosures
1113 WHERE content_url = '$enc_url' AND post_id = '$entry_ref_id'");
1115 if (db_num_rows($result) == 0) {
1116 db_query("INSERT INTO ttrss_enclosures
1117 (content_url, content_type, title, duration, post_id, width, height) VALUES
1118 ('$enc_url', '$enc_type', '$enc_title', '$enc_dur', '$entry_ref_id', $enc_width, $enc_height)");
1122 //db_query("COMMIT");
1124 // check for manual tags (we have to do it here since they're loaded from filters)
1126 foreach ($article_filters as $f) {
1127 if ($f["type"] == "tag") {
1129 $manual_tags = trim_array(explode(",", $f["param"]));
1131 foreach ($manual_tags as $tag) {
1132 if (tag_is_valid($tag)) {
1133 array_push($entry_tags, $tag);
1141 $boring_tags = trim_array(explode(",", mb_strtolower(get_pref(
1142 'BLACKLISTED_TAGS', $owner_uid, ''), 'utf-8')));
1144 $filtered_tags = array();
1145 $tags_to_cache = array();
1147 if ($entry_tags && is_array($entry_tags)) {
1148 foreach ($entry_tags as $tag) {
1149 if (array_search($tag, $boring_tags) === false) {
1150 array_push($filtered_tags, $tag);
1155 $filtered_tags = array_unique($filtered_tags);
1157 if ($debug_enabled) {
1158 _debug("filtered article tags:", $debug_enabled);
1159 print_r($filtered_tags);
1162 // Save article tags in the database
1164 if (count($filtered_tags) > 0) {
1166 //db_query("BEGIN");
1168 foreach ($filtered_tags as $tag) {
1170 $tag = sanitize_tag($tag);
1171 $tag = db_escape_string($tag);
1173 if (!tag_is_valid($tag)) continue;
1175 $result = db_query("SELECT id FROM ttrss_tags
1176 WHERE tag_name = '$tag' AND post_int_id = '$entry_int_id' AND
1177 owner_uid = '$owner_uid' LIMIT 1");
1179 if ($result && db_num_rows($result) == 0) {
1181 db_query("INSERT INTO ttrss_tags
1182 (owner_uid,tag_name,post_int_id)
1183 VALUES ('$owner_uid','$tag', '$entry_int_id')");
1186 array_push($tags_to_cache, $tag);
1189 /* update the cache */
1191 $tags_to_cache = array_unique($tags_to_cache);
1193 $tags_str = db_escape_string(join(",", $tags_to_cache));
1195 db_query("UPDATE ttrss_user_entries
1196 SET tag_cache = '$tags_str' WHERE ref_id = '$entry_ref_id'
1197 AND owner_uid = $owner_uid");
1199 //db_query("COMMIT");
1202 _debug("article processed", $debug_enabled);
1205 _debug("purging feed...", $debug_enabled);
1207 purge_feed($feed, 0, $debug_enabled);
1209 db_query("UPDATE ttrss_feeds
1210 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
1212 // db_query("COMMIT");
1216 $error_msg = db_escape_string(mb_substr($rss->error(), 0, 245));
1218 _debug("fetch error: $error_msg", $debug_enabled);
1220 if (count($rss->errors()) > 1) {
1221 foreach ($rss->errors() as $error) {
1227 "UPDATE ttrss_feeds SET last_error = '$error_msg',
1228 last_updated = NOW() WHERE id = '$feed'");
1233 _debug("done", $debug_enabled);
1238 function cache_images($html, $site_url, $debug) {
1239 libxml_use_internal_errors(true);
1241 $charset_hack = '<head>
1242 <meta http-equiv="Content-Type" content="text/html; charset=utf-8"/>
1245 $doc = new DOMDocument();
1246 $doc->loadHTML($charset_hack . $html);
1247 $xpath = new DOMXPath($doc);
1249 $entries = $xpath->query('(//img[@src])');
1251 foreach ($entries as $entry) {
1252 if ($entry->hasAttribute('src')) {
1253 $src = rewrite_relative_url($site_url, $entry->getAttribute('src'));
1255 $local_filename = CACHE_DIR
. "/images/" . sha1($src) . ".png";
1257 if ($debug) _debug("cache_images: downloading: $src to $local_filename");
1259 if (!file_exists($local_filename)) {
1260 $file_content = fetch_file_contents($src);
1262 if ($file_content && strlen($file_content) > _MIN_CACHE_IMAGE_SIZE
) {
1263 file_put_contents($local_filename, $file_content);
1266 touch($local_filename);
1272 function expire_error_log($debug) {
1273 if ($debug) _debug("Removing old error log entries...");
1275 if (DB_TYPE
== "pgsql") {
1276 db_query("DELETE FROM ttrss_error_log
1277 WHERE created_at < NOW() - INTERVAL '7 days'");
1279 db_query("DELETE FROM ttrss_error_log
1280 WHERE created_at < DATE_SUB(NOW(), INTERVAL 7 DAY)");
1285 function expire_lock_files($debug) {
1286 //if ($debug) _debug("Removing old lock files...");
1290 if (is_writable(LOCK_DIRECTORY
)) {
1291 $files = glob(LOCK_DIRECTORY
. "/*.lock");
1294 foreach ($files as $file) {
1295 if (!file_is_locked(basename($file)) && time() - filemtime($file) > 86400*2) {
1303 if ($debug) _debug("Removed $num_deleted old lock files.");
1306 function expire_cached_files($debug) {
1307 foreach (array("simplepie", "images", "export", "upload") as $dir) {
1308 $cache_dir = CACHE_DIR
. "/$dir";
1310 // if ($debug) _debug("Expiring $cache_dir");
1314 if (is_writable($cache_dir)) {
1315 $files = glob("$cache_dir/*");
1318 foreach ($files as $file) {
1319 if (time() - filemtime($file) > 86400*7) {
1328 if ($debug) _debug("$cache_dir: removed $num_deleted files.");
1333 * Source: http://www.php.net/manual/en/function.parse-url.php#104527
1334 * Returns the url query as associative array
1336 * @param string query
1337 * @return array params
1339 function convertUrlQuery($query) {
1340 $queryParts = explode('&', $query);
1344 foreach ($queryParts as $param) {
1345 $item = explode('=', $param);
1346 $params[$item[0]] = $item[1];
1352 function get_article_filters($filters, $title, $content, $link, $timestamp, $author, $tags) {
1355 foreach ($filters as $filter) {
1356 $match_any_rule = $filter["match_any_rule"];
1357 $inverse = $filter["inverse"];
1358 $filter_match = false;
1360 foreach ($filter["rules"] as $rule) {
1362 $reg_exp = str_replace('/', '\/', $rule["reg_exp"]);
1363 $rule_inverse = $rule["inverse"];
1368 switch ($rule["type"]) {
1370 $match = @preg_match
("/$reg_exp/i", $title);
1373 // we don't need to deal with multiline regexps
1374 $content = preg_replace("/[\r\n\t]/", "", $content);
1376 $match = @preg_match
("/$reg_exp/i", $content);
1379 // we don't need to deal with multiline regexps
1380 $content = preg_replace("/[\r\n\t]/", "", $content);
1382 $match = (@preg_match
("/$reg_exp/i", $title) || @preg_match
("/$reg_exp/i", $content));
1385 $match = @preg_match
("/$reg_exp/i", $link);
1388 $match = @preg_match
("/$reg_exp/i", $author);
1391 foreach ($tags as $tag) {
1392 if (@preg_match
("/$reg_exp/i", $tag)) {
1400 if ($rule_inverse) $match = !$match;
1402 if ($match_any_rule) {
1404 $filter_match = true;
1408 $filter_match = $match;
1415 if ($inverse) $filter_match = !$filter_match;
1417 if ($filter_match) {
1418 foreach ($filter["actions"] AS $action) {
1419 array_push($matches, $action);
1421 // if Stop action encountered, perform no further processing
1422 if (isset($action["type"]) && $action["type"] == "stop") return $matches;
1430 function find_article_filter($filters, $filter_name) {
1431 foreach ($filters as $f) {
1432 if ($f["type"] == $filter_name) {
1439 function find_article_filters($filters, $filter_name) {
1442 foreach ($filters as $f) {
1443 if ($f["type"] == $filter_name) {
1444 array_push($results, $f);
1450 function calculate_article_score($filters) {
1453 foreach ($filters as $f) {
1454 if ($f["type"] == "score") {
1455 $score +
= $f["param"];
1461 function labels_contains_caption($labels, $caption) {
1462 foreach ($labels as $label) {
1463 if ($label[1] == $caption) {
1471 function assign_article_to_label_filters($id, $filters, $owner_uid, $article_labels) {
1472 foreach ($filters as $f) {
1473 if ($f["type"] == "label") {
1474 if (!labels_contains_caption($article_labels, $f["param"])) {
1475 label_add_article($id, $f["param"], $owner_uid);
1481 function make_guid_from_title($title) {
1482 return preg_replace("/[ \"\',.:;]/", "-",
1483 mb_strtolower(strip_tags($title), 'utf-8'));
1486 /* function verify_feed_xml($feed_data) {
1487 libxml_use_internal_errors(true);
1488 $doc = new DOMDocument();
1489 $doc->loadXML($feed_data);
1490 $error = libxml_get_last_error();
1491 libxml_clear_errors();
1495 function cleanup_counters_cache($debug) {
1496 $result = db_query("DELETE FROM ttrss_counters_cache
1497 WHERE feed_id > 0 AND
1498 (SELECT COUNT(id) FROM ttrss_feeds WHERE
1500 ttrss_counters_cache.owner_uid = ttrss_feeds.owner_uid) = 0");
1501 $frows = db_affected_rows($result);
1503 $result = db_query("DELETE FROM ttrss_cat_counters_cache
1504 WHERE feed_id > 0 AND
1505 (SELECT COUNT(id) FROM ttrss_feed_categories WHERE
1507 ttrss_cat_counters_cache.owner_uid = ttrss_feed_categories.owner_uid) = 0");
1508 $crows = db_affected_rows($result);
1510 _debug("Removed $frows (feeds) $crows (cats) orphaned counter cache entries.");
1513 function housekeeping_common($debug) {
1514 expire_cached_files($debug);
1515 expire_lock_files($debug);
1516 expire_error_log($debug);
1518 $count = update_feedbrowser_cache();
1519 _debug("Feedbrowser updated, $count feeds processed.");
1521 purge_orphans( true);
1522 cleanup_counters_cache($debug);
1524 //$rc = cleanup_tags( 14, 50000);
1525 //_debug("Cleaned $rc cached tags.");
1527 PluginHost
::getInstance()->run_hooks(PluginHost
::HOOK_HOUSE_KEEPING
, "hook_house_keeping", "");