]> 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
daemon: show feeds/minute stats
[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_feeds.last_updated < NOW() - CAST((ttrss_user_prefs.value || ' minutes') AS INTERVAL)
81 ) OR (
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')";
86 } else {
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)
90 ) OR (
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')";
95 }
96
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')";
100 } else {
101 $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < DATE_SUB(NOW(), INTERVAL 5 MINUTE))";
102 }
103
104 // Test if there is a limit to number of updated feeds
105 $query_limit = "";
106 if($limit) $query_limit = sprintf("LIMIT %d", $limit);
107
108 $random_qpart = sql_random_function();
109
110 // We search for feed needing update.
111 $result = db_query($link, "SELECT DISTINCT ttrss_feeds.feed_url
112 FROM
113 ttrss_feeds, ttrss_users, ttrss_user_prefs
114 WHERE
115 ttrss_feeds.owner_uid = ttrss_users.id
116 AND ttrss_users.id = ttrss_user_prefs.owner_uid
117 AND ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL'
118 $login_thresh_qpart $update_limit_qpart
119 $updstart_thresh_qpart
120 ORDER BY feed_url $query_limit");
121
122 $user_prefs_cache = array();
123
124 if($debug) _debug(sprintf("Scheduled %d feeds to update...", db_num_rows($result)));
125
126 // Here is a little cache magic in order to minimize risk of double feed updates.
127 $feeds_to_update = array();
128 while ($line = db_fetch_assoc($result)) {
129 array_push($feeds_to_update, db_escape_string($link, $line['feed_url']));
130 }
131
132 // We update the feed last update started date before anything else.
133 // There is no lag due to feed contents downloads
134 // It prevent an other process to update the same feed.
135
136 if(count($feeds_to_update) > 0) {
137 $feeds_quoted = array();
138
139 foreach ($feeds_to_update as $feed) {
140 array_push($feeds_quoted, "'" . db_escape_string($link, $feed) . "'");
141 }
142
143 db_query($link, sprintf("UPDATE ttrss_feeds SET last_update_started = NOW()
144 WHERE feed_url IN (%s)", implode(',', $feeds_quoted)));
145 }
146
147 expire_cached_files($debug);
148 expire_lock_files($debug);
149
150 $nf = 0;
151
152 // For each feed, we call the feed update function.
153 foreach ($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, $tline["id"], true);
172 ++$nf;
173 }
174 }
175 }
176
177 require_once "digest.php";
178
179 // Send feed digests by email if needed.
180 send_headlines_digests($link, $debug);
181
182 return $nf;
183
184 } // function update_daemon_common
185
186 // ignore_daemon is not used
187 function update_rss_feed($link, $feed, $ignore_daemon = false, $no_cache = false,
188 $override_url = false) {
189
190 require_once "lib/simplepie/simplepie.inc";
191
192 $debug_enabled = defined('DAEMON_EXTENDED_DEBUG') || $_REQUEST['xdebug'];
193
194 if ($debug_enabled) {
195 _debug("update_rss_feed: start");
196 }
197
198 $result = db_query($link, "SELECT id,update_interval,auth_login,
199 feed_url,auth_pass,cache_images,last_updated,
200 mark_unread_on_update, owner_uid,
201 pubsub_state
202 FROM ttrss_feeds WHERE id = '$feed'");
203
204 if (db_num_rows($result) == 0) {
205 if ($debug_enabled) {
206 _debug("update_rss_feed: feed $feed NOT FOUND/SKIPPED");
207 }
208 return false;
209 }
210
211 $last_updated = db_fetch_result($result, 0, "last_updated");
212 $owner_uid = db_fetch_result($result, 0, "owner_uid");
213 $mark_unread_on_update = sql_bool_to_bool(db_fetch_result($result,
214 0, "mark_unread_on_update"));
215 $pubsub_state = db_fetch_result($result, 0, "pubsub_state");
216
217 db_query($link, "UPDATE ttrss_feeds SET last_update_started = NOW()
218 WHERE id = '$feed'");
219
220 $auth_login = db_fetch_result($result, 0, "auth_login");
221 $auth_pass = db_fetch_result($result, 0, "auth_pass");
222
223 $cache_images = sql_bool_to_bool(db_fetch_result($result, 0, "cache_images"));
224 $fetch_url = db_fetch_result($result, 0, "feed_url");
225
226 $feed = db_escape_string($link, $feed);
227
228 if ($override_url) $fetch_url = $override_url;
229
230 $date_feed_processed = date('Y-m-d H:i');
231
232 $cache_filename = CACHE_DIR . "/simplepie/" . sha1($fetch_url) . ".feed";
233
234 // Ignore cache if new feed or manual update.
235 $cache_age = ($no_cache || is_null($last_updated) || $last_updated == '1970-01-01 00:00:00') ?
236 30 : get_feed_update_interval($link, $feed) * 60;
237
238 if ($debug_enabled) {
239 _debug("update_rss_feed: cache filename: $cache_filename exists: " . file_exists($cache_filename));
240 _debug("update_rss_feed: cache age: $cache_age; no cache: $no_cache");
241 }
242
243 $cached_feed_data_hash = false;
244
245 $rss = false;
246 $rss_hash = false;
247 $cache_timestamp = file_exists($cache_filename) ? filemtime($cache_filename) : 0;
248 $last_updated_timestamp = strtotime($last_updated);
249
250 if (file_exists($cache_filename) &&
251 is_readable($cache_filename) &&
252 !$auth_login && !$auth_pass &&
253 filemtime($cache_filename) > time() - $cache_age) {
254
255 if ($debug_enabled) {
256 _debug("update_rss_feed: using local cache.");
257 }
258
259 if ($cache_timestamp > $last_updated_timestamp) {
260 @$rss_data = file_get_contents($cache_filename);
261
262 if ($rss_data) {
263 $rss_hash = sha1($rss_data);
264 @$rss = unserialize($rss_data);
265 }
266 } else {
267 if ($debug_enabled) {
268 _debug("update_rss_feed: local cache valid and older than last_updated, nothing to do.");
269 }
270 return;
271 }
272 }
273
274 if (!$rss) {
275
276 if (!$feed_data) {
277 if ($debug_enabled) {
278 _debug("update_rss_feed: fetching [$fetch_url] (ts: $cache_timestamp/$last_updated_timestamp)");
279 }
280
281 $feed_data = fetch_file_contents($fetch_url, false,
282 $auth_login, $auth_pass, false, $no_cache ? 15 : 45,
283 max($last_updated_timestamp, $cache_timestamp));
284
285 if ($debug_enabled) {
286 _debug("update_rss_feed: fetch done.");
287 }
288
289 }
290
291 if (!$feed_data) {
292 global $fetch_last_error;
293 global $fetch_last_error_code;
294
295 if ($debug_enabled) {
296 _debug("update_rss_feed: unable to fetch: $fetch_last_error [$fetch_last_error_code]");
297 }
298
299 $error_escaped = '';
300
301 // If-Modified-Since
302 if ($fetch_last_error_code != 304) {
303 $error_escaped = db_escape_string($link, $fetch_last_error);
304 } else {
305 if ($debug_enabled) {
306 _debug("update_rss_feed: source claims data not modified, nothing to do.");
307 }
308 }
309
310 db_query($link,
311 "UPDATE ttrss_feeds SET last_error = '$error_escaped',
312 last_updated = NOW() WHERE id = '$feed'");
313
314 return;
315 }
316 }
317
318 $pluginhost = new PluginHost($link);
319 $pluginhost->set_debug($debug_enabled);
320 $user_plugins = get_pref($link, "_ENABLED_PLUGINS", $owner_uid);
321
322 $pluginhost->load(PLUGINS, $pluginhost::KIND_ALL);
323 $pluginhost->load($user_plugins, $pluginhost::KIND_USER, $owner_uid);
324 $pluginhost->load_data();
325
326 foreach ($pluginhost->get_hooks($pluginhost::HOOK_FEED_FETCHED) as $plugin) {
327 $feed_data = $plugin->hook_feed_fetched($feed_data);
328 }
329
330 if (!$rss) {
331 $rss = new SimplePie();
332 $rss->set_sanitize_class("SanitizeDummy");
333 // simplepie ignores the above and creates default sanitizer anyway,
334 // so let's override it...
335 $rss->sanitize = new SanitizeDummy();
336 $rss->set_output_encoding('UTF-8');
337 $rss->set_raw_data($feed_data);
338 $rss->enable_cache(false);
339
340 @$rss->init();
341 }
342
343 // print_r($rss);
344
345 $feed = db_escape_string($link, $feed);
346
347 if (!$rss->error()) {
348
349 // cache data for later
350 if (!$auth_pass && !$auth_login && is_writable(CACHE_DIR . "/simplepie")) {
351 $rss_data = serialize($rss);
352 $new_rss_hash = sha1($rss_data);
353
354 if ($new_rss_hash != $rss_hash) {
355 if ($debug_enabled) {
356 _debug("update_rss_feed: saving $cache_filename");
357 }
358 @file_put_contents($cache_filename, serialize($rss));
359 }
360 }
361
362 // We use local pluginhost here because we need to load different per-user feed plugins
363 $pluginhost->run_hooks($pluginhost::HOOK_FEED_PARSED, "hook_feed_parsed", $rss);
364
365 if ($debug_enabled) {
366 _debug("update_rss_feed: processing feed data...");
367 }
368
369 // db_query($link, "BEGIN");
370
371 if (DB_TYPE == "pgsql") {
372 $favicon_interval_qpart = "favicon_last_checked < NOW() - INTERVAL '12 hour'";
373 } else {
374 $favicon_interval_qpart = "favicon_last_checked < DATE_SUB(NOW(), INTERVAL 12 HOUR)";
375 }
376
377 $result = db_query($link, "SELECT title,site_url,owner_uid,
378 (favicon_last_checked IS NULL OR $favicon_interval_qpart) AS
379 favicon_needs_check
380 FROM ttrss_feeds WHERE id = '$feed'");
381
382 $registered_title = db_fetch_result($result, 0, "title");
383 $orig_site_url = db_fetch_result($result, 0, "site_url");
384 $favicon_needs_check = sql_bool_to_bool(db_fetch_result($result, 0,
385 "favicon_needs_check"));
386
387 $owner_uid = db_fetch_result($result, 0, "owner_uid");
388
389 $site_url = db_escape_string($link, mb_substr(rewrite_relative_url($fetch_url, $rss->get_link()), 0, 245));
390
391 if ($debug_enabled) {
392 _debug("update_rss_feed: checking favicon...");
393 }
394
395 if ($favicon_needs_check) {
396 check_feed_favicon($site_url, $feed, $link);
397
398 db_query($link, "UPDATE ttrss_feeds SET favicon_last_checked = NOW()
399 WHERE id = '$feed'");
400 }
401
402 if (!$registered_title || $registered_title == "[Unknown]") {
403
404 $feed_title = db_escape_string($link, $rss->get_title());
405
406 if ($debug_enabled) {
407 _debug("update_rss_feed: registering title: $feed_title");
408 }
409
410 db_query($link, "UPDATE ttrss_feeds SET
411 title = '$feed_title' WHERE id = '$feed'");
412 }
413
414 if ($site_url && $orig_site_url != $site_url) {
415 db_query($link, "UPDATE ttrss_feeds SET
416 site_url = '$site_url' WHERE id = '$feed'");
417 }
418
419 if ($debug_enabled) {
420 _debug("update_rss_feed: loading filters & labels...");
421 }
422
423 $filters = load_filters($link, $feed, $owner_uid);
424 $labels = get_all_labels($link, $owner_uid);
425
426 if ($debug_enabled) {
427 //print_r($filters);
428 _debug("update_rss_feed: " . count($filters) . " filters loaded.");
429 }
430
431 $items = $rss->get_items();
432
433 if (!is_array($items)) {
434 if ($debug_enabled) {
435 _debug("update_rss_feed: no articles found.");
436 }
437
438 db_query($link, "UPDATE ttrss_feeds
439 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
440
441 return; // no articles
442 }
443
444 if ($pubsub_state != 2 && PUBSUBHUBBUB_ENABLED) {
445
446 if ($debug_enabled) _debug("update_rss_feed: checking for PUSH hub...");
447
448 $feed_hub_url = false;
449
450 $links = $rss->get_links('hub');
451
452 if ($links && is_array($links)) {
453 foreach ($links as $l) {
454 $feed_hub_url = $l;
455 break;
456 }
457 }
458
459 if ($debug_enabled) _debug("update_rss_feed: feed hub url: $feed_hub_url");
460
461 if ($feed_hub_url && function_exists('curl_init') &&
462 !ini_get("open_basedir")) {
463
464 require_once 'lib/pubsubhubbub/subscriber.php';
465
466 $callback_url = get_self_url_prefix() .
467 "/public.php?op=pubsub&id=$feed";
468
469 $s = new Subscriber($feed_hub_url, $callback_url);
470
471 $rc = $s->subscribe($fetch_url);
472
473 if ($debug_enabled)
474 _debug("update_rss_feed: feed hub url found, subscribe request sent.");
475
476 db_query($link, "UPDATE ttrss_feeds SET pubsub_state = 1
477 WHERE id = '$feed'");
478 }
479 }
480
481 if ($debug_enabled) {
482 _debug("update_rss_feed: processing articles...");
483 }
484
485 foreach ($items as $item) {
486 if ($_REQUEST['xdebug'] == 3) {
487 print_r($item);
488 }
489
490 $entry_guid = $item->get_id();
491 if (!$entry_guid) $entry_guid = $item->get_link();
492 if (!$entry_guid) $entry_guid = make_guid_from_title($item->get_title());
493
494 if ($debug_enabled) {
495 _debug("update_rss_feed: guid $entry_guid");
496 }
497
498 if (!$entry_guid) continue;
499
500 $entry_guid = "$owner_uid,$entry_guid";
501
502 $entry_timestamp = "";
503
504 $entry_timestamp = strtotime($item->get_date());
505
506 if ($entry_timestamp == -1 || !$entry_timestamp || $entry_timestamp > time()) {
507 $entry_timestamp = time();
508 $no_orig_date = 'true';
509 } else {
510 $no_orig_date = 'false';
511 }
512
513 $entry_timestamp_fmt = strftime("%Y/%m/%d %H:%M:%S", $entry_timestamp);
514
515 if ($debug_enabled) {
516 _debug("update_rss_feed: date $entry_timestamp [$entry_timestamp_fmt]");
517 }
518
519 $entry_title = $item->get_title();
520
521 $entry_link = rewrite_relative_url($site_url, $item->get_link());
522
523 if ($debug_enabled) {
524 _debug("update_rss_feed: title $entry_title");
525 _debug("update_rss_feed: link $entry_link");
526 }
527
528 if (!$entry_title) $entry_title = date("Y-m-d H:i:s", $entry_timestamp);;
529
530 $entry_content = $item->get_content();
531 if (!$entry_content) $entry_content = $item->get_description();
532
533 if ($_REQUEST["xdebug"] == 2) {
534 print "update_rss_feed: content: ";
535 print $entry_content;
536 print "\n";
537 }
538
539 $entry_comments = $item->data["comments"];
540
541 if ($item->get_author()) {
542 $entry_author_item = $item->get_author();
543 $entry_author = $entry_author_item->get_name();
544 if (!$entry_author) $entry_author = $entry_author_item->get_email();
545
546 $entry_author = db_escape_string($link, $entry_author);
547 }
548
549 $entry_guid = db_escape_string($link, mb_substr($entry_guid, 0, 245));
550
551 $entry_comments = db_escape_string($link, mb_substr($entry_comments, 0, 245));
552 $entry_author = db_escape_string($link, mb_substr($entry_author, 0, 245));
553
554 $num_comments = $item->get_item_tags('http://purl.org/rss/1.0/modules/slash/', 'comments');
555
556 if (is_array($num_comments) && is_array($num_comments[0])) {
557 $num_comments = (int) $num_comments[0]["data"];
558 } else {
559 $num_comments = 0;
560 }
561
562 if ($debug_enabled) {
563 _debug("update_rss_feed: num_comments: $num_comments");
564 _debug("update_rss_feed: looking for tags [1]...");
565 }
566
567 // parse <category> entries into tags
568
569 $additional_tags = array();
570
571 $additional_tags_src = $item->get_categories();
572
573 if (is_array($additional_tags_src)) {
574 foreach ($additional_tags_src as $tobj) {
575 array_push($additional_tags, $tobj->get_term());
576 }
577 }
578
579 if ($debug_enabled) {
580 _debug("update_rss_feed: category tags:");
581 print_r($additional_tags);
582 }
583
584 if ($debug_enabled) {
585 _debug("update_rss_feed: looking for tags [2]...");
586 }
587
588 $entry_tags = array_unique($additional_tags);
589
590 for ($i = 0; $i < count($entry_tags); $i++)
591 $entry_tags[$i] = mb_strtolower($entry_tags[$i], 'utf-8');
592
593 if ($debug_enabled) {
594 //_debug("update_rss_feed: unfiltered tags found:");
595 //print_r($entry_tags);
596 }
597
598 if ($debug_enabled) {
599 _debug("update_rss_feed: done collecting data.");
600 }
601
602 // TODO: less memory-hungry implementation
603
604 if ($debug_enabled) {
605 _debug("update_rss_feed: applying plugin filters..");
606 }
607
608 // FIXME not sure if owner_uid is a good idea here, we may have a base entry without user entry (?)
609 $result = db_query($link, "SELECT plugin_data,title,content,link,tag_cache,author FROM ttrss_entries, ttrss_user_entries
610 WHERE ref_id = id AND guid = '".db_escape_string($link, $entry_guid)."' AND owner_uid = $owner_uid");
611
612 if (db_num_rows($result) != 0) {
613 $entry_plugin_data = db_fetch_result($result, 0, "plugin_data");
614 $stored_article = array("title" => db_fetch_result($result, 0, "title"),
615 "content" => db_fetch_result($result, 0, "content"),
616 "link" => db_fetch_result($result, 0, "link"),
617 "tags" => explode(",", db_fetch_result($result, 0, "tag_cache")),
618 "author" => db_fetch_result($result, 0, "author"));
619 } else {
620 $entry_plugin_data = "";
621 $stored_article = array();
622 }
623
624 $article = array("owner_uid" => $owner_uid, // read only
625 "guid" => $entry_guid, // read only
626 "title" => $entry_title,
627 "content" => $entry_content,
628 "link" => $entry_link,
629 "tags" => $entry_tags,
630 "plugin_data" => $entry_plugin_data,
631 "author" => $entry_author,
632 "stored" => $stored_article);
633
634 foreach ($pluginhost->get_hooks($pluginhost::HOOK_ARTICLE_FILTER) as $plugin) {
635 $article = $plugin->hook_article_filter($article);
636 }
637
638 $entry_tags = $article["tags"];
639 $entry_guid = db_escape_string($link, $entry_guid);
640 $entry_title = db_escape_string($link, $article["title"]);
641 $entry_author = db_escape_string($link, $article["author"]);
642 $entry_link = db_escape_string($link, $article["link"]);
643 $entry_plugin_data = db_escape_string($link, $article["plugin_data"]);
644 $entry_content = $article["content"]; // escaped below
645
646
647 if ($debug_enabled) {
648 _debug("update_rss_feed: plugin data: $entry_plugin_data");
649 }
650
651 if ($cache_images && is_writable(CACHE_DIR . '/images'))
652 cache_images($entry_content, $site_url, $debug_enabled);
653
654 $entry_content = db_escape_string($link, $entry_content, false);
655
656 $content_hash = "SHA1:" . sha1($entry_content);
657
658 db_query($link, "BEGIN");
659
660 $result = db_query($link, "SELECT id FROM ttrss_entries
661 WHERE guid = '$entry_guid'");
662
663 if (db_num_rows($result) == 0) {
664
665 if ($debug_enabled) {
666 _debug("update_rss_feed: base guid [$entry_guid] not found");
667 }
668
669 // base post entry does not exist, create it
670
671 $result = db_query($link,
672 "INSERT INTO ttrss_entries
673 (title,
674 guid,
675 link,
676 updated,
677 content,
678 content_hash,
679 cached_content,
680 no_orig_date,
681 date_updated,
682 date_entered,
683 comments,
684 num_comments,
685 plugin_data,
686 author)
687 VALUES
688 ('$entry_title',
689 '$entry_guid',
690 '$entry_link',
691 '$entry_timestamp_fmt',
692 '$entry_content',
693 '$content_hash',
694 '',
695 $no_orig_date,
696 NOW(),
697 '$date_feed_processed',
698 '$entry_comments',
699 '$num_comments',
700 '$entry_plugin_data',
701 '$entry_author')");
702
703 $article_labels = array();
704
705 } else {
706 // we keep encountering the entry in feeds, so we need to
707 // update date_updated column so that we don't get horrible
708 // dupes when the entry gets purged and reinserted again e.g.
709 // in the case of SLOW SLOW OMG SLOW updating feeds
710
711 $base_entry_id = db_fetch_result($result, 0, "id");
712
713 db_query($link, "UPDATE ttrss_entries SET date_updated = NOW()
714 WHERE id = '$base_entry_id'");
715
716 $article_labels = get_article_labels($link, $base_entry_id, $owner_uid);
717 }
718
719 // now it should exist, if not - bad luck then
720
721 $result = db_query($link, "SELECT
722 id,content_hash,no_orig_date,title,plugin_data,
723 ".SUBSTRING_FOR_DATE."(date_updated,1,19) as date_updated,
724 ".SUBSTRING_FOR_DATE."(updated,1,19) as updated,
725 num_comments
726 FROM
727 ttrss_entries
728 WHERE guid = '$entry_guid'");
729
730 $entry_ref_id = 0;
731 $entry_int_id = 0;
732
733 if (db_num_rows($result) == 1) {
734
735 if ($debug_enabled) {
736 _debug("update_rss_feed: base guid [$entry_guid] found, checking for user record");
737 }
738
739 // this will be used below in update handler
740 $orig_content_hash = db_fetch_result($result, 0, "content_hash");
741 $orig_title = db_fetch_result($result, 0, "title");
742 $orig_num_comments = db_fetch_result($result, 0, "num_comments");
743 $orig_date_updated = strtotime(db_fetch_result($result,
744 0, "date_updated"));
745 $orig_plugin_data = db_fetch_result($result, 0, "plugin_data");
746
747 $ref_id = db_fetch_result($result, 0, "id");
748 $entry_ref_id = $ref_id;
749
750 // check for user post link to main table
751
752 // do we allow duplicate posts with same GUID in different feeds?
753 if (get_pref($link, "ALLOW_DUPLICATE_POSTS", $owner_uid, false)) {
754 $dupcheck_qpart = "AND (feed_id = '$feed' OR feed_id IS NULL)";
755 } else {
756 $dupcheck_qpart = "";
757 }
758
759 /* Collect article tags here so we could filter by them: */
760
761 $article_filters = get_article_filters($filters, $entry_title,
762 $entry_content, $entry_link, $entry_timestamp, $entry_author,
763 $entry_tags);
764
765 if ($debug_enabled) {
766 _debug("update_rss_feed: article filters: ");
767 if (count($article_filters) != 0) {
768 print_r($article_filters);
769 }
770 }
771
772 if (find_article_filter($article_filters, "filter")) {
773 db_query($link, "COMMIT"); // close transaction in progress
774 continue;
775 }
776
777 $score = calculate_article_score($article_filters);
778
779 if ($debug_enabled) {
780 _debug("update_rss_feed: initial score: $score");
781 }
782
783 $query = "SELECT ref_id, int_id FROM ttrss_user_entries WHERE
784 ref_id = '$ref_id' AND owner_uid = '$owner_uid'
785 $dupcheck_qpart";
786
787 // if ($_REQUEST["xdebug"]) print "$query\n";
788
789 $result = db_query($link, $query);
790
791 // okay it doesn't exist - create user entry
792 if (db_num_rows($result) == 0) {
793
794 if ($debug_enabled) {
795 _debug("update_rss_feed: user record not found, creating...");
796 }
797
798 if ($score >= -500 && !find_article_filter($article_filters, 'catchup')) {
799 $unread = 'true';
800 $last_read_qpart = 'NULL';
801 } else {
802 $unread = 'false';
803 $last_read_qpart = 'NOW()';
804 }
805
806 if (find_article_filter($article_filters, 'mark') || $score > 1000) {
807 $marked = 'true';
808 } else {
809 $marked = 'false';
810 }
811
812 if (find_article_filter($article_filters, 'publish')) {
813 $published = 'true';
814 } else {
815 $published = 'false';
816 }
817
818 // N-grams
819
820 if (DB_TYPE == "pgsql" and defined('_NGRAM_TITLE_DUPLICATE_THRESHOLD')) {
821
822 $result = db_query($link, "SELECT COUNT(*) AS similar FROM
823 ttrss_entries,ttrss_user_entries
824 WHERE ref_id = id AND updated >= NOW() - INTERVAL '7 day'
825 AND similarity(title, '$entry_title') >= "._NGRAM_TITLE_DUPLICATE_THRESHOLD."
826 AND owner_uid = $owner_uid");
827
828 $ngram_similar = db_fetch_result($result, 0, "similar");
829
830 if ($debug_enabled) {
831 _debug("update_rss_feed: N-gram similar results: $ngram_similar");
832 }
833
834 if ($ngram_similar > 0) {
835 $unread = 'false';
836 }
837 }
838
839 $last_marked = ($marked == 'true') ? 'NOW()' : 'NULL';
840 $last_published = ($published == 'true') ? 'NOW()' : 'NULL';
841
842 $result = db_query($link,
843 "INSERT INTO ttrss_user_entries
844 (ref_id, owner_uid, feed_id, unread, last_read, marked,
845 published, score, tag_cache, label_cache, uuid,
846 last_marked, last_published)
847 VALUES ('$ref_id', '$owner_uid', '$feed', $unread,
848 $last_read_qpart, $marked, $published, '$score', '', '',
849 '', $last_marked, $last_published)");
850
851 if (PUBSUBHUBBUB_HUB && $published == 'true') {
852 $rss_link = get_self_url_prefix() .
853 "/public.php?op=rss&id=-2&key=" .
854 get_feed_access_key($link, -2, false, $owner_uid);
855
856 $p = new Publisher(PUBSUBHUBBUB_HUB);
857
858 $pubsub_result = $p->publish_update($rss_link);
859 }
860
861 $result = db_query($link,
862 "SELECT int_id FROM ttrss_user_entries WHERE
863 ref_id = '$ref_id' AND owner_uid = '$owner_uid' AND
864 feed_id = '$feed' LIMIT 1");
865
866 if (db_num_rows($result) == 1) {
867 $entry_int_id = db_fetch_result($result, 0, "int_id");
868 }
869 } else {
870 if ($debug_enabled) {
871 _debug("update_rss_feed: user record FOUND");
872 }
873
874 $entry_ref_id = db_fetch_result($result, 0, "ref_id");
875 $entry_int_id = db_fetch_result($result, 0, "int_id");
876 }
877
878 if ($debug_enabled) {
879 _debug("update_rss_feed: RID: $entry_ref_id, IID: $entry_int_id");
880 }
881
882 $post_needs_update = false;
883 $update_insignificant = false;
884
885 if ($orig_num_comments != $num_comments) {
886 $post_needs_update = true;
887 $update_insignificant = true;
888 }
889
890 if ($entry_plugin_data != $orig_plugin_data) {
891 $post_needs_update = true;
892 $update_insignificant = true;
893 }
894
895 if ($content_hash != $orig_content_hash) {
896 $post_needs_update = true;
897 $update_insignificant = false;
898 }
899
900 if (db_escape_string($link, $orig_title) != $entry_title) {
901 $post_needs_update = true;
902 $update_insignificant = false;
903 }
904
905 // if post needs update, update it and mark all user entries
906 // linking to this post as updated
907 if ($post_needs_update) {
908
909 if (defined('DAEMON_EXTENDED_DEBUG')) {
910 _debug("update_rss_feed: post $entry_guid needs update...");
911 }
912
913 // print "<!-- post $orig_title needs update : $post_needs_update -->";
914
915 db_query($link, "UPDATE ttrss_entries
916 SET title = '$entry_title', content = '$entry_content',
917 content_hash = '$content_hash',
918 updated = '$entry_timestamp_fmt',
919 num_comments = '$num_comments',
920 plugin_data = '$entry_plugin_data'
921 WHERE id = '$ref_id'");
922
923 if (!$update_insignificant) {
924 if ($mark_unread_on_update) {
925 db_query($link, "UPDATE ttrss_user_entries
926 SET last_read = null, unread = true WHERE ref_id = '$ref_id'");
927 }
928 }
929 }
930 }
931
932 db_query($link, "COMMIT");
933
934 if ($debug_enabled) {
935 _debug("update_rss_feed: assigning labels...");
936 }
937
938 assign_article_to_label_filters($link, $entry_ref_id, $article_filters,
939 $owner_uid, $article_labels);
940
941 if ($debug_enabled) {
942 _debug("update_rss_feed: looking for enclosures...");
943 }
944
945 // enclosures
946
947 $enclosures = array();
948
949 $encs = $item->get_enclosures();
950
951 if (is_array($encs)) {
952 foreach ($encs as $e) {
953 $e_item = array(
954 $e->link, $e->type, $e->length);
955 array_push($enclosures, $e_item);
956 }
957 }
958
959 if ($debug_enabled) {
960 _debug("update_rss_feed: article enclosures:");
961 print_r($enclosures);
962 }
963
964 db_query($link, "BEGIN");
965
966 foreach ($enclosures as $enc) {
967 $enc_url = db_escape_string($link, $enc[0]);
968 $enc_type = db_escape_string($link, $enc[1]);
969 $enc_dur = db_escape_string($link, $enc[2]);
970
971 $result = db_query($link, "SELECT id FROM ttrss_enclosures
972 WHERE content_url = '$enc_url' AND post_id = '$entry_ref_id'");
973
974 if (db_num_rows($result) == 0) {
975 db_query($link, "INSERT INTO ttrss_enclosures
976 (content_url, content_type, title, duration, post_id) VALUES
977 ('$enc_url', '$enc_type', '', '$enc_dur', '$entry_ref_id')");
978 }
979 }
980
981 db_query($link, "COMMIT");
982
983 // check for manual tags (we have to do it here since they're loaded from filters)
984
985 foreach ($article_filters as $f) {
986 if ($f["type"] == "tag") {
987
988 $manual_tags = trim_array(explode(",", $f["param"]));
989
990 foreach ($manual_tags as $tag) {
991 if (tag_is_valid($tag)) {
992 array_push($entry_tags, $tag);
993 }
994 }
995 }
996 }
997
998 // Skip boring tags
999
1000 $boring_tags = trim_array(explode(",", mb_strtolower(get_pref($link,
1001 'BLACKLISTED_TAGS', $owner_uid, ''), 'utf-8')));
1002
1003 $filtered_tags = array();
1004 $tags_to_cache = array();
1005
1006 if ($entry_tags && is_array($entry_tags)) {
1007 foreach ($entry_tags as $tag) {
1008 if (array_search($tag, $boring_tags) === false) {
1009 array_push($filtered_tags, $tag);
1010 }
1011 }
1012 }
1013
1014 $filtered_tags = array_unique($filtered_tags);
1015
1016 if ($debug_enabled) {
1017 _debug("update_rss_feed: filtered article tags:");
1018 print_r($filtered_tags);
1019 }
1020
1021 // Save article tags in the database
1022
1023 if (count($filtered_tags) > 0) {
1024
1025 db_query($link, "BEGIN");
1026
1027 foreach ($filtered_tags as $tag) {
1028
1029 $tag = sanitize_tag($tag);
1030 $tag = db_escape_string($link, $tag);
1031
1032 if (!tag_is_valid($tag)) continue;
1033
1034 $result = db_query($link, "SELECT id FROM ttrss_tags
1035 WHERE tag_name = '$tag' AND post_int_id = '$entry_int_id' AND
1036 owner_uid = '$owner_uid' LIMIT 1");
1037
1038 if ($result && db_num_rows($result) == 0) {
1039
1040 db_query($link, "INSERT INTO ttrss_tags
1041 (owner_uid,tag_name,post_int_id)
1042 VALUES ('$owner_uid','$tag', '$entry_int_id')");
1043 }
1044
1045 array_push($tags_to_cache, $tag);
1046 }
1047
1048 /* update the cache */
1049
1050 $tags_to_cache = array_unique($tags_to_cache);
1051
1052 $tags_str = db_escape_string($link, join(",", $tags_to_cache));
1053
1054 db_query($link, "UPDATE ttrss_user_entries
1055 SET tag_cache = '$tags_str' WHERE ref_id = '$entry_ref_id'
1056 AND owner_uid = $owner_uid");
1057
1058 db_query($link, "COMMIT");
1059 }
1060
1061 if (get_pref($link, "AUTO_ASSIGN_LABELS", $owner_uid, false)) {
1062 if ($debug_enabled) {
1063 _debug("update_rss_feed: auto-assigning labels...");
1064 }
1065
1066 foreach ($labels as $label) {
1067 $caption = preg_quote($label["caption"]);
1068
1069 if ($caption && preg_match("/\b$caption\b/i", "$tags_str " . strip_tags($entry_content) . " $entry_title")) {
1070 if (!labels_contains_caption($article_labels, $caption)) {
1071 label_add_article($link, $entry_ref_id, $caption, $owner_uid);
1072 }
1073 }
1074 }
1075 }
1076
1077 if ($debug_enabled) {
1078 _debug("update_rss_feed: article processed");
1079 }
1080 }
1081
1082 if (!$last_updated) {
1083 if ($debug_enabled) {
1084 _debug("update_rss_feed: new feed, catching it up...");
1085 }
1086 catchup_feed($link, $feed, false, $owner_uid);
1087 }
1088
1089 if ($debug_enabled) {
1090 _debug("purging feed...");
1091 }
1092
1093 purge_feed($link, $feed, 0, $debug_enabled);
1094
1095 db_query($link, "UPDATE ttrss_feeds
1096 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
1097
1098 // db_query($link, "COMMIT");
1099
1100 } else {
1101
1102 $error_msg = db_escape_string($link, mb_substr($rss->error(), 0, 245));
1103
1104 if ($debug_enabled) {
1105 _debug("update_rss_feed: error fetching feed: $error_msg");
1106 }
1107
1108 db_query($link,
1109 "UPDATE ttrss_feeds SET last_error = '$error_msg',
1110 last_updated = NOW() WHERE id = '$feed'");
1111 }
1112
1113 unset($rss);
1114
1115 if ($debug_enabled) {
1116 _debug("update_rss_feed: done");
1117 }
1118
1119 }
1120
1121 function cache_images($html, $site_url, $debug) {
1122 $cache_dir = CACHE_DIR . "/images";
1123
1124 libxml_use_internal_errors(true);
1125
1126 $charset_hack = '<head>
1127 <meta http-equiv="Content-Type" content="text/html; charset=utf-8"/>
1128 </head>';
1129
1130 $doc = new DOMDocument();
1131 $doc->loadHTML($charset_hack . $html);
1132 $xpath = new DOMXPath($doc);
1133
1134 $entries = $xpath->query('(//img[@src])');
1135
1136 foreach ($entries as $entry) {
1137 if ($entry->hasAttribute('src')) {
1138 $src = rewrite_relative_url($site_url, $entry->getAttribute('src'));
1139
1140 $local_filename = CACHE_DIR . "/images/" . sha1($src) . ".png";
1141
1142 if ($debug) _debug("cache_images: downloading: $src to $local_filename");
1143
1144 if (!file_exists($local_filename)) {
1145 $file_content = fetch_file_contents($src);
1146
1147 if ($file_content && strlen($file_content) > 1024) {
1148 file_put_contents($local_filename, $file_content);
1149 }
1150 }
1151
1152 if (file_exists($local_filename)) {
1153 $entry->setAttribute('src', SELF_URL_PATH . '/image.php?url=' .
1154 base64_encode($src));
1155 }
1156 }
1157 }
1158
1159 $node = $doc->getElementsByTagName('body')->item(0);
1160
1161 return $doc->saveXML($node);
1162 }
1163
1164 function expire_lock_files($debug) {
1165 if ($debug) _debug("Removing old lock files...");
1166
1167 $num_deleted = 0;
1168
1169 if (is_writable(LOCK_DIRECTORY)) {
1170 $files = glob(LOCK_DIRECTORY . "/*.lock");
1171
1172 if ($files) {
1173 foreach ($files as $file) {
1174 if (!file_is_locked($file) && time() - filemtime($file) > 86400*2) {
1175 unlink($file);
1176 ++$num_deleted;
1177 }
1178 }
1179 }
1180 }
1181
1182 if ($debug) _debug("Removed $num_deleted files.");
1183 }
1184
1185 function expire_cached_files($debug) {
1186 foreach (array("simplepie", "images", "export") as $dir) {
1187 $cache_dir = CACHE_DIR . "/$dir";
1188
1189 if ($debug) _debug("Expiring $cache_dir");
1190
1191 $num_deleted = 0;
1192
1193 if (is_writable($cache_dir)) {
1194 $files = glob("$cache_dir/*");
1195
1196 if ($files) {
1197 foreach ($files as $file) {
1198 if (time() - filemtime($file) > 86400*7) {
1199 unlink($file);
1200
1201 ++$num_deleted;
1202 }
1203 }
1204 }
1205 }
1206
1207 if ($debug) _debug("Removed $num_deleted files.");
1208 }
1209 }
1210
1211 /**
1212 * Source: http://www.php.net/manual/en/function.parse-url.php#104527
1213 * Returns the url query as associative array
1214 *
1215 * @param string query
1216 * @return array params
1217 */
1218 function convertUrlQuery($query) {
1219 $queryParts = explode('&', $query);
1220
1221 $params = array();
1222
1223 foreach ($queryParts as $param) {
1224 $item = explode('=', $param);
1225 $params[$item[0]] = $item[1];
1226 }
1227
1228 return $params;
1229 }
1230
1231 function get_article_filters($filters, $title, $content, $link, $timestamp, $author, $tags) {
1232 $matches = array();
1233
1234 foreach ($filters as $filter) {
1235 $match_any_rule = $filter["match_any_rule"];
1236 $inverse = $filter["inverse"];
1237 $filter_match = false;
1238
1239 foreach ($filter["rules"] as $rule) {
1240 $match = false;
1241 $reg_exp = $rule["reg_exp"];
1242 $rule_inverse = $rule["inverse"];
1243
1244 if (!$reg_exp)
1245 continue;
1246
1247 switch ($rule["type"]) {
1248 case "title":
1249 $match = @preg_match("/$reg_exp/i", $title);
1250 break;
1251 case "content":
1252 // we don't need to deal with multiline regexps
1253 $content = preg_replace("/[\r\n\t]/", "", $content);
1254
1255 $match = @preg_match("/$reg_exp/i", $content);
1256 break;
1257 case "both":
1258 // we don't need to deal with multiline regexps
1259 $content = preg_replace("/[\r\n\t]/", "", $content);
1260
1261 $match = (@preg_match("/$reg_exp/i", $title) || @preg_match("/$reg_exp/i", $content));
1262 break;
1263 case "link":
1264 $match = @preg_match("/$reg_exp/i", $link);
1265 break;
1266 case "author":
1267 $match = @preg_match("/$reg_exp/i", $author);
1268 break;
1269 case "tag":
1270 $tag_string = join(",", $tags);
1271 $match = @preg_match("/$reg_exp/i", $tag_string);
1272 break;
1273 }
1274
1275 if ($rule_inverse) $match = !$match;
1276
1277 if ($match_any_rule) {
1278 if ($match) {
1279 $filter_match = true;
1280 break;
1281 }
1282 } else {
1283 $filter_match = $match;
1284 if (!$match) {
1285 break;
1286 }
1287 }
1288 }
1289
1290 if ($inverse) $filter_match = !$filter_match;
1291
1292 if ($filter_match) {
1293 foreach ($filter["actions"] AS $action) {
1294 array_push($matches, $action);
1295 }
1296 }
1297 }
1298
1299 return $matches;
1300 }
1301
1302 function find_article_filter($filters, $filter_name) {
1303 foreach ($filters as $f) {
1304 if ($f["type"] == $filter_name) {
1305 return $f;
1306 };
1307 }
1308 return false;
1309 }
1310
1311 function find_article_filters($filters, $filter_name) {
1312 $results = array();
1313
1314 foreach ($filters as $f) {
1315 if ($f["type"] == $filter_name) {
1316 array_push($results, $f);
1317 };
1318 }
1319 return $results;
1320 }
1321
1322 function calculate_article_score($filters) {
1323 $score = 0;
1324
1325 foreach ($filters as $f) {
1326 if ($f["type"] == "score") {
1327 $score += $f["param"];
1328 };
1329 }
1330 return $score;
1331 }
1332
1333 function labels_contains_caption($labels, $caption) {
1334 foreach ($labels as $label) {
1335 if ($label[1] == $caption) {
1336 return true;
1337 }
1338 }
1339
1340 return false;
1341 }
1342
1343 function assign_article_to_label_filters($link, $id, $filters, $owner_uid, $article_labels) {
1344 foreach ($filters as $f) {
1345 if ($f["type"] == "label") {
1346 if (!labels_contains_caption($article_labels, $f["param"])) {
1347 label_add_article($link, $id, $f["param"], $owner_uid);
1348 }
1349 }
1350 }
1351 }
1352
1353 function make_guid_from_title($title) {
1354 return preg_replace("/[ \"\',.:;]/", "-",
1355 mb_strtolower(strip_tags($title), 'utf-8'));
1356 }
1357
1358
1359 ?>
web-based news feed (RSS/Atom) reader and aggregator; see https://tt-rss.org/