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