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