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