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 $subscribers = db_escape_string($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 $tmp_result = db_query($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_feeds.last_updated < NOW() - CAST((ttrss_user_prefs.value || ' minutes') AS INTERVAL)
82 ttrss_feeds.update_interval > 0
83 AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_feeds.update_interval || ' minutes') AS INTERVAL)
84 ) OR ttrss_feeds.last_updated IS NULL
85 OR last_updated = '1970-01-01 00:00:00')";
87 $update_limit_qpart = "AND ((
88 ttrss_feeds.update_interval = 0
89 AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL CONVERT(ttrss_user_prefs.value, SIGNED INTEGER) MINUTE)
91 ttrss_feeds.update_interval > 0
92 AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL ttrss_feeds.update_interval MINUTE)
93 ) OR ttrss_feeds.last_updated IS NULL
94 OR last_updated = '1970-01-01 00:00:00')";
97 // Test if feed is currently being updated by another process.
98 if (DB_TYPE
== "pgsql") {
99 $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < NOW() - INTERVAL '5 minutes')";
101 $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < DATE_SUB(NOW(), INTERVAL 5 MINUTE))";
104 // Test if there is a limit to number of updated feeds
106 if($limit) $query_limit = sprintf("LIMIT %d", $limit);
108 $random_qpart = sql_random_function();
110 // We search for feed needing update.
111 $result = db_query($link, "SELECT ttrss_feeds.feed_url,ttrss_feeds.id, ttrss_feeds.owner_uid,
112 ".SUBSTRING_FOR_DATE
."(ttrss_feeds.last_updated,1,19) AS last_updated,
113 ttrss_feeds.update_interval
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...\n", 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 $feeds_to_update[$line['id']] = $line;
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.
137 $feed_ids = array_keys($feeds_to_update);
139 db_query($link, sprintf("UPDATE ttrss_feeds SET last_update_started = NOW()
140 WHERE id IN (%s)", implode(',', $feed_ids)));
143 expire_cached_files($debug);
145 // For each feed, we call the feed update function.
146 while ($line = array_pop($feeds_to_update)) {
148 if($debug) _debug("Feed: " . $line["feed_url"] . ", " . $line["last_updated"]);
150 update_rss_feed($link, $line["id"], true);
152 sleep(1); // prevent flood (FIXME make this an option?)
155 // Send feed digests by email if needed.
156 send_headlines_digests($link, $debug);
158 } // function update_daemon_common
160 function update_rss_feed($link, $feed, $ignore_daemon = false, $no_cache = false,
161 $override_url = false) {
163 require_once "lib/simplepie/simplepie.inc";
164 require_once "lib/magpierss/rss_fetch.inc";
165 require_once 'lib/magpierss/rss_utils.inc';
167 $debug_enabled = defined('DAEMON_EXTENDED_DEBUG') ||
$_REQUEST['xdebug'];
169 if (!$_REQUEST["daemon"] && !$ignore_daemon) {
173 if ($debug_enabled) {
174 _debug("update_rss_feed: start");
177 if (!$ignore_daemon) {
179 if (DB_TYPE
== "pgsql") {
180 $updstart_thresh_qpart = "(ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < NOW() - INTERVAL '120 seconds')";
182 $updstart_thresh_qpart = "(ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < DATE_SUB(NOW(), INTERVAL 120 SECOND))";
185 $result = db_query($link, "SELECT id,update_interval,auth_login,
186 auth_pass,cache_images,update_method,last_updated
187 FROM ttrss_feeds WHERE id = '$feed' AND $updstart_thresh_qpart");
191 $result = db_query($link, "SELECT id,update_interval,auth_login,
192 feed_url,auth_pass,cache_images,update_method,last_updated,
193 mark_unread_on_update, owner_uid, update_on_checksum_change,
195 FROM ttrss_feeds WHERE id = '$feed'");
199 if (db_num_rows($result) == 0) {
200 if ($debug_enabled) {
201 _debug("update_rss_feed: feed $feed NOT FOUND/SKIPPED");
206 $update_method = db_fetch_result($result, 0, "update_method");
207 $last_updated = db_fetch_result($result, 0, "last_updated");
208 $owner_uid = db_fetch_result($result, 0, "owner_uid");
209 $mark_unread_on_update = sql_bool_to_bool(db_fetch_result($result,
210 0, "mark_unread_on_update"));
211 $update_on_checksum_change = sql_bool_to_bool(db_fetch_result($result,
212 0, "update_on_checksum_change"));
213 $pubsub_state = db_fetch_result($result, 0, "pubsub_state");
215 db_query($link, "UPDATE ttrss_feeds SET last_update_started = NOW()
216 WHERE id = '$feed'");
218 $auth_login = db_fetch_result($result, 0, "auth_login");
219 $auth_pass = db_fetch_result($result, 0, "auth_pass");
221 if ($update_method == 0)
222 $update_method = DEFAULT_UPDATE_METHOD +
1;
228 if ($update_method == 2)
229 $use_simplepie = true;
231 $use_simplepie = false;
233 if ($debug_enabled) {
234 _debug("update method: $update_method (feed setting: $update_method) (use simplepie: $use_simplepie)\n");
237 if ($update_method == 1) {
238 $auth_login = urlencode($auth_login);
239 $auth_pass = urlencode($auth_pass);
242 $cache_images = sql_bool_to_bool(db_fetch_result($result, 0, "cache_images"));
243 $fetch_url = db_fetch_result($result, 0, "feed_url");
245 $feed = db_escape_string($feed);
247 if ($auth_login && $auth_pass ){
248 $url_parts = array();
249 preg_match("/(^[^:]*):\/\/(.*)/", $fetch_url, $url_parts);
251 if ($url_parts[1] && $url_parts[2]) {
252 $fetch_url = $url_parts[1] . "://$auth_login:$auth_pass@" . $url_parts[2];
258 $fetch_url = $override_url;
260 if ($debug_enabled) {
261 _debug("update_rss_feed: fetching [$fetch_url]...");
264 // Ignore cache if new feed or manual update.
265 $cache_age = (is_null($last_updated) ||
$last_updated == '1970-01-01 00:00:00') ?
266 -1 : get_feed_update_interval($link, $feed) * 60;
268 if ($update_method == 1) {
270 define('MAGPIE_CACHE_AGE', $cache_age);
271 define('MAGPIE_CACHE_ON', !$no_cache);
272 define('MAGPIE_FETCH_TIME_OUT', 60);
273 define('MAGPIE_CACHE_DIR', CACHE_DIR
. "/magpie");
275 $rss = @fetch_rss
($fetch_url);
277 $simplepie_cache_dir = CACHE_DIR
. "/simplepie";
279 if (!is_dir($simplepie_cache_dir)) {
280 mkdir($simplepie_cache_dir);
283 $rss = new SimplePie();
284 $rss->set_useragent(SELF_USER_AGENT
);
285 # $rss->set_timeout(10);
286 $rss->set_feed_url($fetch_url);
287 $rss->set_output_encoding('UTF-8');
288 //$rss->force_feed(true);
290 if ($debug_enabled) {
291 _debug("feed update interval (sec): " .
292 get_feed_update_interval($link, $feed)*60);
295 $rss->enable_cache(!$no_cache);
298 $rss->set_cache_location($simplepie_cache_dir);
299 $rss->set_cache_duration($cache_age);
307 if ($debug_enabled) {
308 _debug("update_rss_feed: fetch done, parsing...");
311 $feed = db_escape_string($feed);
313 if ($update_method == 2) {
314 $fetch_ok = !$rss->error();
321 if ($debug_enabled) {
322 _debug("update_rss_feed: processing feed data...");
325 // db_query($link, "BEGIN");
327 if (DB_TYPE
== "pgsql") {
328 $favicon_interval_qpart = "favicon_last_checked < NOW() - INTERVAL '12 hour'";
330 $favicon_interval_qpart = "favicon_last_checked < DATE_SUB(NOW(), INTERVAL 12 HOUR)";
333 $result = db_query($link, "SELECT title,icon_url,site_url,owner_uid,
334 (favicon_last_checked IS NULL OR $favicon_interval_qpart) AS
336 FROM ttrss_feeds WHERE id = '$feed'");
338 $registered_title = db_fetch_result($result, 0, "title");
339 $orig_icon_url = db_fetch_result($result, 0, "icon_url");
340 $orig_site_url = db_fetch_result($result, 0, "site_url");
341 $favicon_needs_check = sql_bool_to_bool(db_fetch_result($result, 0,
342 "favicon_needs_check"));
344 $owner_uid = db_fetch_result($result, 0, "owner_uid");
346 if ($use_simplepie) {
347 $site_url = db_escape_string(trim($rss->get_link()));
349 $site_url = db_escape_string(trim($rss->channel
["link"]));
352 // weird, weird Magpie
353 if (!$use_simplepie) {
354 if (!$site_url) $site_url = db_escape_string($rss->channel
["link_"]);
357 $site_url = rewrite_relative_url($fetch_url, $site_url);
358 $site_url = substr($site_url, 0, 250);
360 if ($debug_enabled) {
361 _debug("update_rss_feed: checking favicon...");
364 if ($favicon_needs_check) {
365 check_feed_favicon($site_url, $feed, $link);
367 db_query($link, "UPDATE ttrss_feeds SET favicon_last_checked = NOW()
368 WHERE id = '$feed'");
371 if (!$registered_title ||
$registered_title == "[Unknown]") {
373 if ($use_simplepie) {
374 $feed_title = db_escape_string($rss->get_title());
376 $feed_title = db_escape_string($rss->channel
["title"]);
379 if ($debug_enabled) {
380 _debug("update_rss_feed: registering title: $feed_title");
383 db_query($link, "UPDATE ttrss_feeds SET
384 title = '$feed_title' WHERE id = '$feed'");
387 if ($site_url && $orig_site_url != $site_url) {
388 db_query($link, "UPDATE ttrss_feeds SET
389 site_url = '$site_url' WHERE id = '$feed'");
392 // print "I: " . $rss->channel["image"]["url"];
394 if (!$use_simplepie) {
395 $icon_url = db_escape_string(trim($rss->image
["url"]));
397 $icon_url = db_escape_string(trim($rss->get_image_url()));
400 $icon_url = rewrite_relative_url($fetch_url, $icon_url);
401 $icon_url = substr($icon_url, 0, 250);
403 if ($icon_url && $orig_icon_url != $icon_url) {
404 db_query($link, "UPDATE ttrss_feeds SET icon_url = '$icon_url' WHERE id = '$feed'");
407 if ($debug_enabled) {
408 _debug("update_rss_feed: loading filters...");
411 $filters = load_filters($link, $feed, $owner_uid);
413 if ($debug_enabled) {
415 _debug("update_rss_feed: " . count($filters) . " filters loaded.");
418 if ($use_simplepie) {
419 $iterator = $rss->get_items();
421 $iterator = $rss->items
;
422 if (!$iterator ||
!is_array($iterator)) $iterator = $rss->entries
;
423 if (!$iterator ||
!is_array($iterator)) $iterator = $rss;
426 if (!is_array($iterator)) {
427 /* db_query($link, "UPDATE ttrss_feeds
428 SET last_error = 'Parse error: can\'t find any articles.'
429 WHERE id = '$feed'"); */
431 // clear any errors and mark feed as updated if fetched okay
432 // even if it's blank
434 if ($debug_enabled) {
435 _debug("update_rss_feed: entry iterator is not an array, no articles?");
438 db_query($link, "UPDATE ttrss_feeds
439 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
441 return; // no articles
444 if ($pubsub_state != 2 && PUBSUBHUBBUB_ENABLED
) {
446 if ($debug_enabled) _debug("update_rss_feed: checking for PUSH hub...");
448 $feed_hub_url = false;
449 if ($use_simplepie) {
450 $links = $rss->get_links('hub');
452 if ($links && is_array($links)) {
453 foreach ($links as $l) {
460 $atom = $rss->channel
['atom'];
463 if ($atom['link@rel'] == 'hub') {
464 $feed_hub_url = $atom['link@href'];
467 if (!$feed_hub_url && $atom['link#'] > 1) {
468 for ($i = 2; $i <= $atom['link#']; $i++
) {
469 if ($atom["link#$i@rel"] == 'hub') {
470 $feed_hub_url = $atom["link#$i@href"];
476 $feed_hub_url = $rss->channel
['link_hub'];
480 if ($debug_enabled) _debug("update_rss_feed: feed hub url: $feed_hub_url");
482 if ($feed_hub_url && function_exists('curl_init') &&
483 !ini_get("open_basedir")) {
485 require_once 'lib/pubsubhubbub/subscriber.php';
487 $callback_url = get_self_url_prefix() .
488 "/public.php?op=pubsub&id=$feed";
490 $s = new Subscriber($feed_hub_url, $callback_url);
492 $rc = $s->subscribe($fetch_url);
495 _debug("update_rss_feed: feed hub url found, subscribe request sent.");
497 db_query($link, "UPDATE ttrss_feeds SET pubsub_state = 1
498 WHERE id = '$feed'");
502 if ($debug_enabled) {
503 _debug("update_rss_feed: processing articles...");
506 foreach ($iterator as $item) {
507 if ($_REQUEST['xdebug'] == 2) {
511 if ($use_simplepie) {
512 $entry_guid = $item->get_id();
513 if (!$entry_guid) $entry_guid = $item->get_link();
514 if (!$entry_guid) $entry_guid = make_guid_from_title($item->get_title());
518 $entry_guid = $item["id"];
520 if (!$entry_guid) $entry_guid = $item["guid"];
521 if (!$entry_guid) $entry_guid = $item["about"];
522 if (!$entry_guid) $entry_guid = $item["link"];
523 if (!$entry_guid) $entry_guid = make_guid_from_title($item["title"]);
526 if ($debug_enabled) {
527 _debug("update_rss_feed: guid $entry_guid");
530 if (!$entry_guid) continue;
532 $entry_timestamp = "";
534 if ($use_simplepie) {
535 $entry_timestamp = strtotime($item->get_date());
537 $rss_2_date = $item['pubdate'];
538 $rss_1_date = $item['dc']['date'];
539 $atom_date = $item['issued'];
540 if (!$atom_date) $atom_date = $item['updated'];
542 if ($atom_date != "") $entry_timestamp = parse_w3cdtf($atom_date);
543 if ($rss_1_date != "") $entry_timestamp = parse_w3cdtf($rss_1_date);
544 if ($rss_2_date != "") $entry_timestamp = strtotime($rss_2_date);
548 if ($entry_timestamp == "" ||
$entry_timestamp == -1 ||
!$entry_timestamp) {
549 $entry_timestamp = time();
550 $no_orig_date = 'true';
552 $no_orig_date = 'false';
555 $entry_timestamp_fmt = strftime("%Y/%m/%d %H:%M:%S", $entry_timestamp);
557 if ($debug_enabled) {
558 _debug("update_rss_feed: date $entry_timestamp [$entry_timestamp_fmt]");
561 if ($use_simplepie) {
562 $entry_title = $item->get_title();
564 $entry_title = trim(strip_tags($item["title"]));
567 if ($use_simplepie) {
568 $entry_link = $item->get_link();
570 // strange Magpie workaround
571 $entry_link = $item["link_"];
572 if (!$entry_link) $entry_link = $item["link"];
575 $entry_link = rewrite_relative_url($site_url, $entry_link);
577 if ($debug_enabled) {
578 _debug("update_rss_feed: title $entry_title");
579 _debug("update_rss_feed: link $entry_link");
582 if (!$entry_title) $entry_title = date("Y-m-d H:i:s", $entry_timestamp);;
584 $entry_link = strip_tags($entry_link);
586 if ($use_simplepie) {
587 $entry_content = $item->get_content();
588 if (!$entry_content) $entry_content = $item->get_description();
590 $entry_content = $item["content:escaped"];
592 if (!$entry_content) $entry_content = $item["content:encoded"];
593 if (!$entry_content && is_array($entry_content)) $entry_content = $item["content"]["encoded"];
594 if (!$entry_content) $entry_content = $item["content"];
596 if (is_array($entry_content)) $entry_content = $entry_content[0];
598 // Magpie bugs are getting ridiculous
599 if (trim($entry_content) == "Array") $entry_content = false;
601 if (!$entry_content) $entry_content = $item["atom_content"];
602 if (!$entry_content) $entry_content = $item["summary"];
604 if (!$entry_content ||
605 strlen($entry_content) < strlen($item["description"])) {
606 $entry_content = $item["description"];
610 if (is_array($entry_content)) {
611 $entry_content = $entry_content["encoded"];
612 if (!$entry_content) $entry_content = $entry_content["escaped"];
616 if ($cache_images && is_writable(CACHE_DIR
. '/images'))
617 $entry_content = cache_images($entry_content, $site_url, $debug_enabled);
619 if ($_REQUEST["xdebug"] == 2) {
620 print "update_rss_feed: content: ";
621 print $entry_content;
625 $entry_content_unescaped = $entry_content;
627 if ($use_simplepie) {
628 $entry_comments = strip_tags($item->data
["comments"]);
629 if ($item->get_author()) {
630 $entry_author_item = $item->get_author();
631 $entry_author = $entry_author_item->get_name();
632 if (!$entry_author) $entry_author = $entry_author_item->get_email();
634 $entry_author = db_escape_string($entry_author);
637 $entry_comments = strip_tags($item["comments"]);
639 $entry_author = db_escape_string(strip_tags($item['dc']['creator']));
641 if ($item['author']) {
643 if (is_array($item['author'])) {
645 if (!$entry_author) {
646 $entry_author = db_escape_string(strip_tags($item['author']['name']));
649 if (!$entry_author) {
650 $entry_author = db_escape_string(strip_tags($item['author']['email']));
654 if (!$entry_author) {
655 $entry_author = db_escape_string(strip_tags($item['author']));
660 if (preg_match('/^[\t\n\r ]*$/', $entry_author)) $entry_author = '';
662 $entry_guid = db_escape_string(strip_tags($entry_guid));
663 $entry_guid = mb_substr($entry_guid, 0, 250);
665 $result = db_query($link, "SELECT id FROM ttrss_entries
666 WHERE guid = '$entry_guid'");
668 $entry_content = db_escape_string($entry_content, false);
670 $content_hash = "SHA1:" . sha1(strip_tags($entry_content));
672 $entry_title = db_escape_string($entry_title);
673 $entry_link = db_escape_string($entry_link);
674 $entry_comments = mb_substr(db_escape_string($entry_comments), 0, 250);
675 $entry_author = mb_substr($entry_author, 0, 250);
677 if ($use_simplepie) {
678 $num_comments = 0; #FIXME#
680 $num_comments = db_escape_string($item["slash"]["comments"]);
683 if (!$num_comments) $num_comments = 0;
685 if ($debug_enabled) {
686 _debug("update_rss_feed: looking for tags [1]...");
689 // parse <category> entries into tags
691 $additional_tags = array();
693 if ($use_simplepie) {
695 $additional_tags_src = $item->get_categories();
697 if (is_array($additional_tags_src)) {
698 foreach ($additional_tags_src as $tobj) {
699 array_push($additional_tags, $tobj->get_term());
703 if ($debug_enabled) {
704 _debug("update_rss_feed: category tags:");
705 print_r($additional_tags);
710 $t_ctr = $item['category#'];
713 $additional_tags = array();
714 } else if ($t_ctr > 0) {
715 $additional_tags = array($item['category']);
717 if ($item['category@term']) {
718 array_push($additional_tags, $item['category@term']);
721 for ($i = 0; $i <= $t_ctr; $i++
) {
722 if ($item["category#$i"]) {
723 array_push($additional_tags, $item["category#$i"]);
726 if ($item["category#$i@term"]) {
727 array_push($additional_tags, $item["category#$i@term"]);
732 // parse <dc:subject> elements
734 $t_ctr = $item['dc']['subject#'];
737 array_push($additional_tags, $item['dc']['subject']);
739 for ($i = 0; $i <= $t_ctr; $i++
) {
740 if ($item['dc']["subject#$i"]) {
741 array_push($additional_tags, $item['dc']["subject#$i"]);
747 if ($debug_enabled) {
748 _debug("update_rss_feed: looking for tags [2]...");
752 // <a href="..." rel="tag">Xorg</a>, //
756 preg_match_all("/<a.*?rel=['\"]tag['\"].*?\>([^<]+)<\/a>/i",
757 $entry_content_unescaped, $entry_tags);
759 $entry_tags = $entry_tags[1];
761 $entry_tags = array_merge($entry_tags, $additional_tags);
762 $entry_tags = array_unique($entry_tags);
764 for ($i = 0; $i < count($entry_tags); $i++
)
765 $entry_tags[$i] = mb_strtolower($entry_tags[$i], 'utf-8');
767 if ($debug_enabled) {
768 //_debug("update_rss_feed: unfiltered tags found:");
769 //print_r($entry_tags);
774 $entry_content = sanitize_article_content($entry_content);
775 $entry_title = sanitize_article_content($entry_title);
777 if ($debug_enabled) {
778 _debug("update_rss_feed: done collecting data [TITLE:$entry_title]");
781 db_query($link, "BEGIN");
783 if (db_num_rows($result) == 0) {
785 if ($debug_enabled) {
786 _debug("update_rss_feed: base guid not found");
789 // base post entry does not exist, create it
791 $result = db_query($link,
792 "INSERT INTO ttrss_entries
809 '$entry_timestamp_fmt',
819 // we keep encountering the entry in feeds, so we need to
820 // update date_updated column so that we don't get horrible
821 // dupes when the entry gets purged and reinserted again e.g.
822 // in the case of SLOW SLOW OMG SLOW updating feeds
824 $base_entry_id = db_fetch_result($result, 0, "id");
826 db_query($link, "UPDATE ttrss_entries SET date_updated = NOW()
827 WHERE id = '$base_entry_id'");
830 // now it should exist, if not - bad luck then
832 $result = db_query($link, "SELECT
833 id,content_hash,no_orig_date,title,
834 ".SUBSTRING_FOR_DATE
."(date_updated,1,19) as date_updated,
835 ".SUBSTRING_FOR_DATE
."(updated,1,19) as updated,
839 WHERE guid = '$entry_guid'");
844 if (db_num_rows($result) == 1) {
846 if ($debug_enabled) {
847 _debug("update_rss_feed: base guid found, checking for user record");
850 // this will be used below in update handler
851 $orig_content_hash = db_fetch_result($result, 0, "content_hash");
852 $orig_title = db_fetch_result($result, 0, "title");
853 $orig_num_comments = db_fetch_result($result, 0, "num_comments");
854 $orig_date_updated = strtotime(db_fetch_result($result,
857 $ref_id = db_fetch_result($result, 0, "id");
858 $entry_ref_id = $ref_id;
860 // check for user post link to main table
862 // do we allow duplicate posts with same GUID in different feeds?
863 if (get_pref($link, "ALLOW_DUPLICATE_POSTS", $owner_uid, false)) {
864 $dupcheck_qpart = "AND (feed_id = '$feed' OR feed_id IS NULL)";
866 $dupcheck_qpart = "";
869 /* Collect article tags here so we could filter by them: */
871 $article_filters = get_article_filters($filters, $entry_title,
872 $entry_content, $entry_link, $entry_timestamp, $entry_author,
875 if ($debug_enabled) {
876 _debug("update_rss_feed: article filters: ");
877 if (count($article_filters) != 0) {
878 print_r($article_filters);
882 if (find_article_filter($article_filters, "filter")) {
883 db_query($link, "COMMIT"); // close transaction in progress
887 $score = calculate_article_score($article_filters);
889 if ($debug_enabled) {
890 _debug("update_rss_feed: initial score: $score");
893 $query = "SELECT ref_id, int_id FROM ttrss_user_entries WHERE
894 ref_id = '$ref_id' AND owner_uid = '$owner_uid'
897 // if ($_REQUEST["xdebug"]) print "$query\n";
899 $result = db_query($link, $query);
901 // okay it doesn't exist - create user entry
902 if (db_num_rows($result) == 0) {
904 if ($debug_enabled) {
905 _debug("update_rss_feed: user record not found, creating...");
908 if ($score >= -500 && !find_article_filter($article_filters, 'catchup')) {
910 $last_read_qpart = 'NULL';
913 $last_read_qpart = 'NOW()';
916 if (find_article_filter($article_filters, 'mark') ||
$score > 1000) {
922 if (find_article_filter($article_filters, 'publish')) {
925 $published = 'false';
930 if (DB_TYPE
== "pgsql" and defined('_NGRAM_TITLE_DUPLICATE_THRESHOLD')) {
932 $result = db_query($link, "SELECT COUNT(*) AS similar FROM
933 ttrss_entries,ttrss_user_entries
934 WHERE ref_id = id AND updated >= NOW() - INTERVAL '7 day'
935 AND similarity(title, '$entry_title') >= "._NGRAM_TITLE_DUPLICATE_THRESHOLD
."
936 AND owner_uid = $owner_uid");
938 $ngram_similar = db_fetch_result($result, 0, "similar");
940 if ($debug_enabled) {
941 _debug("update_rss_feed: N-gram similar results: $ngram_similar");
944 if ($ngram_similar > 0) {
949 $result = db_query($link,
950 "INSERT INTO ttrss_user_entries
951 (ref_id, owner_uid, feed_id, unread, last_read, marked,
952 published, score, tag_cache, label_cache, uuid)
953 VALUES ('$ref_id', '$owner_uid', '$feed', $unread,
954 $last_read_qpart, $marked, $published, '$score', '', '', '')");
956 if (PUBSUBHUBBUB_HUB
&& $published == 'true') {
957 $rss_link = get_self_url_prefix() .
958 "/public.php?op=rss&id=-2&key=" .
959 get_feed_access_key($link, -2, false, $owner_uid);
961 $p = new Publisher(PUBSUBHUBBUB_HUB
);
963 $pubsub_result = $p->publish_update($rss_link);
966 $result = db_query($link,
967 "SELECT int_id FROM ttrss_user_entries WHERE
968 ref_id = '$ref_id' AND owner_uid = '$owner_uid' AND
969 feed_id = '$feed' LIMIT 1");
971 if (db_num_rows($result) == 1) {
972 $entry_int_id = db_fetch_result($result, 0, "int_id");
975 if ($debug_enabled) {
976 _debug("update_rss_feed: user record FOUND");
979 $entry_ref_id = db_fetch_result($result, 0, "ref_id");
980 $entry_int_id = db_fetch_result($result, 0, "int_id");
983 if ($debug_enabled) {
984 _debug("update_rss_feed: RID: $entry_ref_id, IID: $entry_int_id");
987 $post_needs_update = false;
988 $update_insignificant = false;
990 if ($orig_num_comments != $num_comments) {
991 $post_needs_update = true;
992 $update_insignificant = true;
995 if ($content_hash != $orig_content_hash) {
996 $post_needs_update = true;
997 $update_insignificant = false;
1000 if (db_escape_string($orig_title) != $entry_title) {
1001 $post_needs_update = true;
1002 $update_insignificant = false;
1005 // if post needs update, update it and mark all user entries
1006 // linking to this post as updated
1007 if ($post_needs_update) {
1009 if (defined('DAEMON_EXTENDED_DEBUG')) {
1010 _debug("update_rss_feed: post $entry_guid needs update...");
1013 // print "<!-- post $orig_title needs update : $post_needs_update -->";
1015 db_query($link, "UPDATE ttrss_entries
1016 SET title = '$entry_title', content = '$entry_content',
1017 content_hash = '$content_hash',
1018 updated = '$entry_timestamp_fmt',
1019 num_comments = '$num_comments'
1020 WHERE id = '$ref_id'");
1022 if (!$update_insignificant) {
1023 if ($mark_unread_on_update) {
1024 db_query($link, "UPDATE ttrss_user_entries
1025 SET last_read = null, unread = true WHERE ref_id = '$ref_id'");
1026 } else if ($update_on_checksum_change) {
1027 db_query($link, "UPDATE ttrss_user_entries
1028 SET last_read = null WHERE ref_id = '$ref_id'
1029 AND unread = false");
1035 db_query($link, "COMMIT");
1037 if ($debug_enabled) {
1038 _debug("update_rss_feed: assigning labels...");
1041 assign_article_to_labels($link, $entry_ref_id, $article_filters,
1044 if ($debug_enabled) {
1045 _debug("update_rss_feed: looking for enclosures...");
1050 $enclosures = array();
1052 if ($use_simplepie) {
1053 $encs = $item->get_enclosures();
1055 if (is_array($encs)) {
1056 foreach ($encs as $e) {
1058 $e->link
, $e->type
, $e->length
);
1060 array_push($enclosures, $e_item);
1067 $e_ctr = $item['enclosure#'];
1070 $e_item = array($item['enclosure@url'],
1071 $item['enclosure@type'],
1072 $item['enclosure@length']);
1074 array_push($enclosures, $e_item);
1076 for ($i = 0; $i <= $e_ctr; $i++
) {
1078 if ($item["enclosure#$i@url"]) {
1079 $e_item = array($item["enclosure#$i@url"],
1080 $item["enclosure#$i@type"],
1081 $item["enclosure#$i@length"]);
1082 array_push($enclosures, $e_item);
1088 // can there be many of those? yes -fox
1090 $m_ctr = $item['media']['content#'];
1093 $e_item = array($item['media']['content@url'],
1094 $item['media']['content@medium'],
1095 $item['media']['content@length']);
1097 array_push($enclosures, $e_item);
1099 for ($i = 0; $i <= $m_ctr; $i++
) {
1101 if ($item["media"]["content#$i@url"]) {
1102 $e_item = array($item["media"]["content#$i@url"],
1103 $item["media"]["content#$i@medium"],
1104 $item["media"]["content#$i@length"]);
1105 array_push($enclosures, $e_item);
1113 if ($debug_enabled) {
1114 _debug("update_rss_feed: article enclosures:");
1115 print_r($enclosures);
1118 db_query($link, "BEGIN");
1120 foreach ($enclosures as $enc) {
1121 $enc_url = db_escape_string($enc[0]);
1122 $enc_type = db_escape_string($enc[1]);
1123 $enc_dur = db_escape_string($enc[2]);
1125 $result = db_query($link, "SELECT id FROM ttrss_enclosures
1126 WHERE content_url = '$enc_url' AND post_id = '$entry_ref_id'");
1128 if (db_num_rows($result) == 0) {
1129 db_query($link, "INSERT INTO ttrss_enclosures
1130 (content_url, content_type, title, duration, post_id) VALUES
1131 ('$enc_url', '$enc_type', '', '$enc_dur', '$entry_ref_id')");
1135 db_query($link, "COMMIT");
1137 // check for manual tags (we have to do it here since they're loaded from filters)
1139 foreach ($article_filters as $f) {
1140 if ($f["type"] == "tag") {
1142 $manual_tags = trim_array(explode(",", $f["param"]));
1144 foreach ($manual_tags as $tag) {
1145 if (tag_is_valid($tag)) {
1146 array_push($entry_tags, $tag);
1154 $boring_tags = trim_array(explode(",", mb_strtolower(get_pref($link,
1155 'BLACKLISTED_TAGS', $owner_uid, ''), 'utf-8')));
1157 $filtered_tags = array();
1158 $tags_to_cache = array();
1160 if ($entry_tags && is_array($entry_tags)) {
1161 foreach ($entry_tags as $tag) {
1162 if (array_search($tag, $boring_tags) === false) {
1163 array_push($filtered_tags, $tag);
1168 $filtered_tags = array_unique($filtered_tags);
1170 if ($debug_enabled) {
1171 _debug("update_rss_feed: filtered article tags:");
1172 print_r($filtered_tags);
1175 // Save article tags in the database
1177 if (count($filtered_tags) > 0) {
1179 db_query($link, "BEGIN");
1181 foreach ($filtered_tags as $tag) {
1183 $tag = sanitize_tag($tag);
1184 $tag = db_escape_string($tag);
1186 if (!tag_is_valid($tag)) continue;
1188 $result = db_query($link, "SELECT id FROM ttrss_tags
1189 WHERE tag_name = '$tag' AND post_int_id = '$entry_int_id' AND
1190 owner_uid = '$owner_uid' LIMIT 1");
1192 if ($result && db_num_rows($result) == 0) {
1194 db_query($link, "INSERT INTO ttrss_tags
1195 (owner_uid,tag_name,post_int_id)
1196 VALUES ('$owner_uid','$tag', '$entry_int_id')");
1199 array_push($tags_to_cache, $tag);
1202 /* update the cache */
1204 $tags_to_cache = array_unique($tags_to_cache);
1206 $tags_str = db_escape_string(join(",", $tags_to_cache));
1208 db_query($link, "UPDATE ttrss_user_entries
1209 SET tag_cache = '$tags_str' WHERE ref_id = '$entry_ref_id'
1210 AND owner_uid = $owner_uid");
1212 db_query($link, "COMMIT");
1215 if ($debug_enabled) {
1216 _debug("update_rss_feed: article processed");
1220 if (!$last_updated) {
1221 if ($debug_enabled) {
1222 _debug("update_rss_feed: new feed, catching it up...");
1224 catchup_feed($link, $feed, false, $owner_uid);
1227 if ($debug_enabled) {
1228 _debug("purging feed...");
1231 purge_feed($link, $feed, 0, $debug_enabled);
1233 db_query($link, "UPDATE ttrss_feeds
1234 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
1236 // db_query($link, "COMMIT");
1240 if ($use_simplepie) {
1241 $error_msg = mb_substr($rss->error(), 0, 250);
1243 $error_msg = mb_substr(magpie_error(), 0, 250);
1246 if ($debug_enabled) {
1247 _debug("update_rss_feed: error fetching feed: $error_msg");
1250 $error_msg = db_escape_string($error_msg);
1253 "UPDATE ttrss_feeds SET last_error = '$error_msg',
1254 last_updated = NOW() WHERE id = '$feed'");
1257 if ($use_simplepie) {
1261 if ($debug_enabled) {
1262 _debug("update_rss_feed: done");
1267 function cache_images($html, $site_url, $debug) {
1268 $cache_dir = CACHE_DIR
. "/images";
1270 libxml_use_internal_errors(true);
1272 $charset_hack = '<head>
1273 <meta http-equiv="Content-Type" content="text/html; charset=utf-8"/>
1276 $doc = new DOMDocument();
1277 $doc->loadHTML($charset_hack . $html);
1278 $xpath = new DOMXPath($doc);
1280 $entries = $xpath->query('(//img[@src])');
1282 foreach ($entries as $entry) {
1283 if ($entry->hasAttribute('src')) {
1284 $src = rewrite_relative_url($site_url, $entry->getAttribute('src'));
1286 $local_filename = CACHE_DIR
. "/images/" . sha1($src) . ".png";
1288 if ($debug) _debug("cache_images: downloading: $src to $local_filename");
1290 if (!file_exists($local_filename)) {
1291 $file_content = fetch_file_contents($src);
1293 if ($file_content && strlen($file_content) > 1024) {
1294 file_put_contents($local_filename, $file_content);
1298 if (file_exists($local_filename)) {
1299 $entry->setAttribute('src', SELF_URL_PATH
. '/image.php?url=' .
1300 base64_encode($src));
1305 $node = $doc->getElementsByTagName('body')->item(0);
1307 return $doc->saveXML($node, LIBXML_NOEMPTYTAG
);
1310 function expire_cached_files($debug) {
1311 foreach (array("magpie", "simplepie", "images", "export") as $dir) {
1312 $cache_dir = CACHE_DIR
. "/$dir";
1314 if ($debug) _debug("Expiring $cache_dir");
1318 if (is_writable($cache_dir)) {
1319 $files = glob("$cache_dir/*");
1322 foreach ($files as $file) {
1323 if (time() - filemtime($file) > 86400*7) {
1331 if ($debug) _debug("Removed $num_deleted files.");
1336 * Source: http://www.php.net/manual/en/function.parse-url.php#104527
1337 * Returns the url query as associative array
1339 * @param string query
1340 * @return array params
1342 function convertUrlQuery($query) {
1343 $queryParts = explode('&', $query);
1347 foreach ($queryParts as $param) {
1348 $item = explode('=', $param);
1349 $params[$item[0]] = $item[1];