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