]> git.wh0rd.org - tt-rss.git/blob - include/rssfuncs.php
remove unused old-style image rewritign
[tt-rss.git] / include / rssfuncs.php
1 <?php
2 define_default('DAEMON_UPDATE_LOGIN_LIMIT', 30);
3 define_default('DAEMON_FEED_LIMIT', 500);
4 define_default('DAEMON_SLEEP_INTERVAL', 120);
5
6 function update_feedbrowser_cache() {
7
8 $result = db_query("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("BEGIN");
15
16 db_query("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 $tmp_result = db_query("SELECT subscribers FROM
27 ttrss_feedbrowser_cache WHERE feed_url = '$feed_url'");
28
29 if (db_num_rows($tmp_result) == 0) {
30
31 db_query("INSERT INTO ttrss_feedbrowser_cache
32 (feed_url, site_url, title, subscribers) VALUES ('$feed_url',
33 '$site_url', '$title', '$subscribers')");
34
35 ++$count;
36
37 }
38
39 }
40
41 db_query("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 $debug Set to false to disable debug output. Default to true.
58 * @return void
59 */
60 function update_daemon_common($limit = DAEMON_FEED_LIMIT, $from_http = false, $debug = true) {
61 // Process all other feeds using last_updated and interval parameters
62
63 $schema_version = get_schema_version();
64
65 if ($schema_version != SCHEMA_VERSION) {
66 die("Schema version is wrong, please upgrade the database.\n");
67 }
68
69 define('PREFS_NO_CACHE', true);
70
71 // Test if the user has loggued in recently. If not, it does not update its feeds.
72 if (!SINGLE_USER_MODE && DAEMON_UPDATE_LOGIN_LIMIT > 0) {
73 if (DB_TYPE == "pgsql") {
74 $login_thresh_qpart = "AND ttrss_users.last_login >= NOW() - INTERVAL '".DAEMON_UPDATE_LOGIN_LIMIT." days'";
75 } else {
76 $login_thresh_qpart = "AND ttrss_users.last_login >= DATE_SUB(NOW(), INTERVAL ".DAEMON_UPDATE_LOGIN_LIMIT." DAY)";
77 }
78 } else {
79 $login_thresh_qpart = "";
80 }
81
82 // Test if the feed need a update (update interval exceded).
83 if (DB_TYPE == "pgsql") {
84 $update_limit_qpart = "AND ((
85 ttrss_feeds.update_interval = 0
86 AND ttrss_user_prefs.value != '-1'
87 AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_user_prefs.value || ' minutes') AS INTERVAL)
88 ) OR (
89 ttrss_feeds.update_interval > 0
90 AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_feeds.update_interval || ' minutes') AS INTERVAL)
91 ) OR ttrss_feeds.last_updated IS NULL
92 OR last_updated = '1970-01-01 00:00:00')";
93 } else {
94 $update_limit_qpart = "AND ((
95 ttrss_feeds.update_interval = 0
96 AND ttrss_user_prefs.value != '-1'
97 AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL CONVERT(ttrss_user_prefs.value, SIGNED INTEGER) MINUTE)
98 ) OR (
99 ttrss_feeds.update_interval > 0
100 AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL ttrss_feeds.update_interval MINUTE)
101 ) OR ttrss_feeds.last_updated IS NULL
102 OR last_updated = '1970-01-01 00:00:00')";
103 }
104
105 // Test if feed is currently being updated by another process.
106 if (DB_TYPE == "pgsql") {
107 $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < NOW() - INTERVAL '10 minutes')";
108 } else {
109 $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < DATE_SUB(NOW(), INTERVAL 10 MINUTE))";
110 }
111
112 // Test if there is a limit to number of updated feeds
113 $query_limit = "";
114 if($limit) $query_limit = sprintf("LIMIT %d", $limit);
115
116 $query = "SELECT DISTINCT ttrss_feeds.feed_url, ttrss_feeds.last_updated
117 FROM
118 ttrss_feeds, ttrss_users, ttrss_user_prefs
119 WHERE
120 ttrss_feeds.owner_uid = ttrss_users.id
121 AND ttrss_users.id = ttrss_user_prefs.owner_uid
122 AND ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL'
123 $login_thresh_qpart $update_limit_qpart
124 $updstart_thresh_qpart
125 ORDER BY last_updated $query_limit";
126
127 // We search for feed needing update.
128 $result = db_query($query);
129
130 if($debug) _debug(sprintf("Scheduled %d feeds to update...", db_num_rows($result)));
131
132 // Here is a little cache magic in order to minimize risk of double feed updates.
133 $feeds_to_update = array();
134 while ($line = db_fetch_assoc($result)) {
135 array_push($feeds_to_update, db_escape_string($line['feed_url']));
136 }
137
138 // We update the feed last update started date before anything else.
139 // There is no lag due to feed contents downloads
140 // It prevent an other process to update the same feed.
141
142 if(count($feeds_to_update) > 0) {
143 $feeds_quoted = array();
144
145 foreach ($feeds_to_update as $feed) {
146 array_push($feeds_quoted, "'" . db_escape_string($feed) . "'");
147 }
148
149 db_query(sprintf("UPDATE ttrss_feeds SET last_update_started = NOW()
150 WHERE feed_url IN (%s)", implode(',', $feeds_quoted)));
151 }
152
153 $nf = 0;
154
155 // For each feed, we call the feed update function.
156 foreach ($feeds_to_update as $feed) {
157 if($debug) _debug("Base feed: $feed");
158
159 //update_rss_feed($line["id"], true);
160
161 // since we have the data cached, we can deal with other feeds with the same url
162
163 $tmp_result = db_query("SELECT DISTINCT ttrss_feeds.id,last_updated,ttrss_feeds.owner_uid
164 FROM ttrss_feeds, ttrss_users, ttrss_user_prefs WHERE
165 ttrss_user_prefs.owner_uid = ttrss_feeds.owner_uid AND
166 ttrss_users.id = ttrss_user_prefs.owner_uid AND
167 ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL' AND
168 feed_url = '".db_escape_string($feed)."' AND
169 (ttrss_feeds.update_interval > 0 OR
170 ttrss_user_prefs.value != '-1')
171 $login_thresh_qpart
172 ORDER BY ttrss_feeds.id $query_limit");
173
174 if (db_num_rows($tmp_result) > 0) {
175 while ($tline = db_fetch_assoc($tmp_result)) {
176 if($debug) _debug(" => " . $tline["last_updated"] . ", " . $tline["id"] . " " . $tline["owner_uid"]);
177 update_rss_feed($tline["id"], true);
178 ++$nf;
179 }
180 }
181 }
182
183 require_once "digest.php";
184
185 // Send feed digests by email if needed.
186 send_headlines_digests($debug);
187
188 return $nf;
189
190 } // function update_daemon_common
191
192 // ignore_daemon is not used
193 function update_rss_feed($feed, $ignore_daemon = false, $no_cache = false) {
194
195 $debug_enabled = defined('DAEMON_EXTENDED_DEBUG') || $_REQUEST['xdebug'];
196
197 _debug("start", $debug_enabled);
198
199 $result = db_query("SELECT id,update_interval,auth_login,
200 feed_url,auth_pass,cache_images,last_updated,
201 mark_unread_on_update, owner_uid,
202 pubsub_state, auth_pass_encrypted,
203 (SELECT max(date_entered) FROM
204 ttrss_entries, ttrss_user_entries where ref_id = id AND feed_id = '$feed') AS last_article_timestamp
205 FROM ttrss_feeds WHERE id = '$feed'");
206
207 if (db_num_rows($result) == 0) {
208 _debug("feed $feed NOT FOUND/SKIPPED", $debug_enabled);
209 return false;
210 }
211
212 $last_updated = db_fetch_result($result, 0, "last_updated");
213 $last_article_timestamp = @strtotime(db_fetch_result($result, 0, "last_article_timestamp"));
214
215 if (defined('_DISABLE_HTTP_304'))
216 $last_article_timestamp = 0;
217
218 $owner_uid = db_fetch_result($result, 0, "owner_uid");
219 $mark_unread_on_update = sql_bool_to_bool(db_fetch_result($result,
220 0, "mark_unread_on_update"));
221 $pubsub_state = db_fetch_result($result, 0, "pubsub_state");
222 $auth_pass_encrypted = sql_bool_to_bool(db_fetch_result($result,
223 0, "auth_pass_encrypted"));
224
225 db_query("UPDATE ttrss_feeds SET last_update_started = NOW()
226 WHERE id = '$feed'");
227
228 $auth_login = db_fetch_result($result, 0, "auth_login");
229 $auth_pass = db_fetch_result($result, 0, "auth_pass");
230
231 if ($auth_pass_encrypted) {
232 require_once "crypt.php";
233 $auth_pass = decrypt_string($auth_pass);
234 }
235
236 $cache_images = sql_bool_to_bool(db_fetch_result($result, 0, "cache_images"));
237 $fetch_url = db_fetch_result($result, 0, "feed_url");
238
239 $feed = db_escape_string($feed);
240
241 $date_feed_processed = date('Y-m-d H:i');
242
243 $cache_filename = CACHE_DIR . "/simplepie/" . sha1($fetch_url) . ".xml";
244
245 $pluginhost = new PluginHost();
246 $pluginhost->set_debug($debug_enabled);
247 $user_plugins = get_pref("_ENABLED_PLUGINS", $owner_uid);
248
249 $pluginhost->load(PLUGINS, PluginHost::KIND_ALL);
250 $pluginhost->load($user_plugins, PluginHost::KIND_USER, $owner_uid);
251 $pluginhost->load_data();
252
253 $rss = false;
254 $rss_hash = false;
255
256 $force_refetch = isset($_REQUEST["force_refetch"]);
257
258 if (file_exists($cache_filename) &&
259 is_readable($cache_filename) &&
260 !$auth_login && !$auth_pass &&
261 filemtime($cache_filename) > time() - 30) {
262
263 _debug("using local cache.", $debug_enabled);
264
265 @$feed_data = file_get_contents($cache_filename);
266
267 if ($feed_data) {
268 $rss_hash = sha1($feed_data);
269 }
270
271 } else {
272 _debug("local cache will not be used for this feed", $debug_enabled);
273 }
274
275 if (!$rss) {
276
277 foreach ($pluginhost->get_hooks(PluginHost::HOOK_FETCH_FEED) as $plugin) {
278 $feed_data = $plugin->hook_fetch_feed($feed_data, $fetch_url, $owner_uid, $feed);
279 }
280
281 if (!$feed_data) {
282 _debug("fetching [$fetch_url]...", $debug_enabled);
283 _debug("If-Modified-Since: ".gmdate('D, d M Y H:i:s \G\M\T', $last_article_timestamp), $debug_enabled);
284
285 $feed_data = fetch_file_contents($fetch_url, false,
286 $auth_login, $auth_pass, false,
287 $no_cache ? FEED_FETCH_NO_CACHE_TIMEOUT : FEED_FETCH_TIMEOUT,
288 $force_refetch ? 0 : $last_article_timestamp);
289
290 global $fetch_curl_used;
291
292 if (!$fetch_curl_used) {
293 $tmp = @gzdecode($feed_data);
294
295 if ($tmp) $feed_data = $tmp;
296 }
297
298 $feed_data = trim($feed_data);
299
300 _debug("fetch done.", $debug_enabled);
301
302 /* if ($feed_data) {
303 $error = verify_feed_xml($feed_data);
304
305 if ($error) {
306 _debug("error verifying XML, code: " . $error->code, $debug_enabled);
307
308 if ($error->code == 26) {
309 _debug("got error 26, trying to decode entities...", $debug_enabled);
310
311 $feed_data = html_entity_decode($feed_data, ENT_COMPAT, 'UTF-8');
312
313 $error = verify_feed_xml($feed_data);
314
315 if ($error) $feed_data = '';
316 }
317 }
318 } */
319 }
320
321 if (!$feed_data) {
322 global $fetch_last_error;
323 global $fetch_last_error_code;
324
325 _debug("unable to fetch: $fetch_last_error [$fetch_last_error_code]", $debug_enabled);
326
327 $error_escaped = '';
328
329 // If-Modified-Since
330 if ($fetch_last_error_code != 304) {
331 $error_escaped = db_escape_string($fetch_last_error);
332 } else {
333 _debug("source claims data not modified, nothing to do.", $debug_enabled);
334 }
335
336 db_query(
337 "UPDATE ttrss_feeds SET last_error = '$error_escaped',
338 last_updated = NOW() WHERE id = '$feed'");
339
340 return;
341 }
342 }
343
344 foreach ($pluginhost->get_hooks(PluginHost::HOOK_FEED_FETCHED) as $plugin) {
345 $feed_data = $plugin->hook_feed_fetched($feed_data, $fetch_url, $owner_uid, $feed);
346 }
347
348 // set last update to now so if anything *simplepie* crashes later we won't be
349 // continuously failing on the same feed
350 //db_query("UPDATE ttrss_feeds SET last_updated = NOW() WHERE id = '$feed'");
351
352 if (!$rss) {
353 $rss = new FeedParser($feed_data);
354 $rss->init();
355 }
356
357 require_once "lib/languagedetect/LanguageDetect.php";
358
359 $lang = new Text_LanguageDetect();
360 $lang->setNameMode(2);
361
362 // print_r($rss);
363
364 $feed = db_escape_string($feed);
365
366 if (!$rss->error()) {
367
368 // cache data for later
369 if (!$auth_pass && !$auth_login && is_writable(CACHE_DIR . "/simplepie")) {
370 $new_rss_hash = sha1($rss_data);
371
372 if ($new_rss_hash != $rss_hash && count($rss->get_items()) > 0 ) {
373 _debug("saving $cache_filename", $debug_enabled);
374 @file_put_contents($cache_filename, $feed_data);
375 }
376 }
377
378 // We use local pluginhost here because we need to load different per-user feed plugins
379 $pluginhost->run_hooks(PluginHost::HOOK_FEED_PARSED, "hook_feed_parsed", $rss);
380
381 _debug("processing feed data...", $debug_enabled);
382
383 // db_query("BEGIN");
384
385 if (DB_TYPE == "pgsql") {
386 $favicon_interval_qpart = "favicon_last_checked < NOW() - INTERVAL '12 hour'";
387 } else {
388 $favicon_interval_qpart = "favicon_last_checked < DATE_SUB(NOW(), INTERVAL 12 HOUR)";
389 }
390
391 $result = db_query("SELECT title,site_url,owner_uid,favicon_avg_color,
392 (favicon_last_checked IS NULL OR $favicon_interval_qpart) AS
393 favicon_needs_check
394 FROM ttrss_feeds WHERE id = '$feed'");
395
396 $registered_title = db_fetch_result($result, 0, "title");
397 $orig_site_url = db_fetch_result($result, 0, "site_url");
398 $favicon_needs_check = sql_bool_to_bool(db_fetch_result($result, 0,
399 "favicon_needs_check"));
400 $favicon_avg_color = db_fetch_result($result, 0, "favicon_avg_color");
401
402 $owner_uid = db_fetch_result($result, 0, "owner_uid");
403
404 $site_url = db_escape_string(mb_substr(rewrite_relative_url($fetch_url, $rss->get_link()), 0, 245));
405
406 _debug("site_url: $site_url", $debug_enabled);
407 _debug("feed_title: " . $rss->get_title(), $debug_enabled);
408
409 if ($favicon_needs_check || $force_refetch) {
410
411 /* terrible hack: if we crash on floicon shit here, we won't check
412 * the icon avgcolor again (unless the icon got updated) */
413
414 $favicon_file = ICONS_DIR . "/$feed.ico";
415 $favicon_modified = @filemtime($favicon_file);
416
417 _debug("checking favicon...", $debug_enabled);
418
419 check_feed_favicon($site_url, $feed);
420 $favicon_modified_new = @filemtime($favicon_file);
421
422 if ($favicon_modified_new > $favicon_modified)
423 $favicon_avg_color = '';
424
425 if (file_exists($favicon_file) && function_exists("imagecreatefromstring") && $favicon_avg_color == '') {
426 require_once "colors.php";
427
428 db_query("UPDATE ttrss_feeds SET favicon_avg_color = 'fail' WHERE
429 id = '$feed'");
430
431 $favicon_color = db_escape_string(
432 calculate_avg_color($favicon_file));
433
434 $favicon_colorstring = ",favicon_avg_color = '".$favicon_color."'";
435 } else if ($favicon_avg_color == 'fail') {
436 _debug("floicon failed on this file, not trying to recalculate avg color", $debug_enabled);
437 }
438
439 db_query("UPDATE ttrss_feeds SET favicon_last_checked = NOW()
440 $favicon_colorstring
441 WHERE id = '$feed'");
442 }
443
444 if (!$registered_title || $registered_title == "[Unknown]") {
445
446 $feed_title = db_escape_string($rss->get_title());
447
448 if ($feed_title) {
449 _debug("registering title: $feed_title", $debug_enabled);
450
451 db_query("UPDATE ttrss_feeds SET
452 title = '$feed_title' WHERE id = '$feed'");
453 }
454 }
455
456 if ($site_url && $orig_site_url != $site_url) {
457 db_query("UPDATE ttrss_feeds SET
458 site_url = '$site_url' WHERE id = '$feed'");
459 }
460
461 _debug("loading filters & labels...", $debug_enabled);
462
463 $filters = load_filters($feed, $owner_uid);
464 $labels = get_all_labels($owner_uid);
465
466 _debug("" . count($filters) . " filters loaded.", $debug_enabled);
467
468 $items = $rss->get_items();
469
470 if (!is_array($items)) {
471 _debug("no articles found.", $debug_enabled);
472
473 db_query("UPDATE ttrss_feeds
474 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
475
476 return; // no articles
477 }
478
479 if ($pubsub_state != 2 && PUBSUBHUBBUB_ENABLED) {
480
481 _debug("checking for PUSH hub...", $debug_enabled);
482
483 $feed_hub_url = false;
484
485 $links = $rss->get_links('hub');
486
487 if ($links && is_array($links)) {
488 foreach ($links as $l) {
489 $feed_hub_url = $l;
490 break;
491 }
492 }
493
494 _debug("feed hub url: $feed_hub_url", $debug_enabled);
495
496 if ($feed_hub_url && function_exists('curl_init') &&
497 !ini_get("open_basedir")) {
498
499 require_once 'lib/pubsubhubbub/subscriber.php';
500
501 $callback_url = get_self_url_prefix() .
502 "/public.php?op=pubsub&id=$feed";
503
504 $s = new Subscriber($feed_hub_url, $callback_url);
505
506 $rc = $s->subscribe($fetch_url);
507
508 _debug("feed hub url found, subscribe request sent.", $debug_enabled);
509
510 db_query("UPDATE ttrss_feeds SET pubsub_state = 1
511 WHERE id = '$feed'");
512 }
513 }
514
515 _debug("processing articles...", $debug_enabled);
516
517 foreach ($items as $item) {
518 if ($_REQUEST['xdebug'] == 3) {
519 print_r($item);
520 }
521
522 $entry_guid = $item->get_id();
523 if (!$entry_guid) $entry_guid = $item->get_link();
524 if (!$entry_guid) $entry_guid = make_guid_from_title($item->get_title());
525
526 _debug("f_guid $entry_guid", $debug_enabled);
527
528 if (!$entry_guid) continue;
529
530 $entry_guid = "$owner_uid,$entry_guid";
531
532 $entry_guid_hashed = db_escape_string('SHA1:' . sha1($entry_guid));
533
534 _debug("guid $entry_guid / $entry_guid_hashed", $debug_enabled);
535
536 $entry_timestamp = "";
537
538 $entry_timestamp = $item->get_date();
539
540 _debug("orig date: " . $item->get_date(), $debug_enabled);
541
542 if ($entry_timestamp == -1 || !$entry_timestamp || $entry_timestamp > time()) {
543 $entry_timestamp = time();
544 $no_orig_date = 'true';
545 } else {
546 $no_orig_date = 'false';
547 }
548
549 $entry_timestamp_fmt = strftime("%Y/%m/%d %H:%M:%S", $entry_timestamp);
550
551 _debug("date $entry_timestamp [$entry_timestamp_fmt]", $debug_enabled);
552
553 // $entry_title = html_entity_decode($item->get_title(), ENT_COMPAT, 'UTF-8');
554 // $entry_title = decode_numeric_entities($entry_title);
555 $entry_title = $item->get_title();
556
557 $entry_link = rewrite_relative_url($site_url, $item->get_link());
558
559 _debug("title $entry_title", $debug_enabled);
560 _debug("link $entry_link", $debug_enabled);
561
562 if (!$entry_title) $entry_title = date("Y-m-d H:i:s", $entry_timestamp);;
563
564 $entry_content = $item->get_content();
565 if (!$entry_content) $entry_content = $item->get_description();
566
567 if ($_REQUEST["xdebug"] == 2) {
568 print "content: ";
569 print $entry_content;
570 print "\n";
571 }
572
573 $entry_language = $lang->detect($entry_content, 1);
574
575 if (count($entry_language) > 0) {
576 $entry_language = array_keys($entry_language);
577 $entry_language = db_escape_string($entry_language[0]);
578
579 _debug("detected language: $entry_language", $debug_enabled);
580 }
581
582 $entry_comments = $item->get_comments_url();
583 $entry_author = $item->get_author();
584
585 $entry_guid = db_escape_string(mb_substr($entry_guid, 0, 245));
586
587 $entry_comments = db_escape_string(mb_substr(trim($entry_comments), 0, 245));
588 $entry_author = db_escape_string(mb_substr(trim($entry_author), 0, 245));
589
590 $num_comments = (int) $item->get_comments_count();
591
592 _debug("author $entry_author", $debug_enabled);
593 _debug("num_comments: $num_comments", $debug_enabled);
594 _debug("looking for tags...", $debug_enabled);
595
596 // parse <category> entries into tags
597
598 $additional_tags = array();
599
600 $additional_tags_src = $item->get_categories();
601
602 if (is_array($additional_tags_src)) {
603 foreach ($additional_tags_src as $tobj) {
604 array_push($additional_tags, $tobj);
605 }
606 }
607
608 $entry_tags = array_unique($additional_tags);
609
610 for ($i = 0; $i < count($entry_tags); $i++)
611 $entry_tags[$i] = mb_strtolower($entry_tags[$i], 'utf-8');
612
613 _debug("tags found: " . join(",", $entry_tags), $debug_enabled);
614
615 _debug("done collecting data.", $debug_enabled);
616
617 // TODO: less memory-hungry implementation
618
619 _debug("applying plugin filters..", $debug_enabled);
620
621 // FIXME not sure if owner_uid is a good idea here, we may have a base entry without user entry (?)
622 $result = db_query("SELECT plugin_data,title,content,link,tag_cache,author FROM ttrss_entries, ttrss_user_entries
623 WHERE ref_id = id AND (guid = '".db_escape_string($entry_guid)."' OR guid = '$entry_guid_hashed') AND owner_uid = $owner_uid");
624
625 if (db_num_rows($result) != 0) {
626 $entry_plugin_data = db_fetch_result($result, 0, "plugin_data");
627 $stored_article = array("title" => db_fetch_result($result, 0, "title"),
628 "content" => db_fetch_result($result, 0, "content"),
629 "link" => db_fetch_result($result, 0, "link"),
630 "tags" => explode(",", db_fetch_result($result, 0, "tag_cache")),
631 "author" => db_fetch_result($result, 0, "author"));
632 } else {
633 $entry_plugin_data = "";
634 $stored_article = array();
635 }
636
637 $article = array("owner_uid" => $owner_uid, // read only
638 "guid" => $entry_guid, // read only
639 "title" => $entry_title,
640 "content" => $entry_content,
641 "link" => $entry_link,
642 "tags" => $entry_tags,
643 "plugin_data" => $entry_plugin_data,
644 "author" => $entry_author,
645 "stored" => $stored_article);
646
647 foreach ($pluginhost->get_hooks(PluginHost::HOOK_ARTICLE_FILTER) as $plugin) {
648 $article = $plugin->hook_article_filter($article);
649 }
650
651 $entry_tags = $article["tags"];
652 $entry_guid = db_escape_string($entry_guid);
653 $entry_title = db_escape_string($article["title"]);
654 $entry_author = db_escape_string($article["author"]);
655 $entry_link = db_escape_string($article["link"]);
656 $entry_plugin_data = db_escape_string($article["plugin_data"]);
657 $entry_content = $article["content"]; // escaped below
658
659
660 _debug("plugin data: $entry_plugin_data", $debug_enabled);
661
662 if ($cache_images && is_writable(CACHE_DIR . '/images'))
663 cache_images($entry_content, $site_url, $debug_enabled);
664
665 $entry_content = db_escape_string($entry_content, false);
666
667 $content_hash = "SHA1:" . sha1($entry_content);
668
669 db_query("BEGIN");
670
671 $result = db_query("SELECT id FROM ttrss_entries
672 WHERE (guid = '$entry_guid' OR guid = '$entry_guid_hashed')");
673
674 if (db_num_rows($result) == 0) {
675
676 _debug("base guid [$entry_guid] not found", $debug_enabled);
677
678 // base post entry does not exist, create it
679
680 $result = db_query(
681 "INSERT INTO ttrss_entries
682 (title,
683 guid,
684 link,
685 updated,
686 content,
687 content_hash,
688 no_orig_date,
689 date_updated,
690 date_entered,
691 comments,
692 num_comments,
693 plugin_data,
694 lang,
695 author)
696 VALUES
697 ('$entry_title',
698 '$entry_guid_hashed',
699 '$entry_link',
700 '$entry_timestamp_fmt',
701 '$entry_content',
702 '$content_hash',
703 $no_orig_date,
704 NOW(),
705 '$date_feed_processed',
706 '$entry_comments',
707 '$num_comments',
708 '$entry_plugin_data',
709 '$entry_language',
710 '$entry_author')");
711
712 $article_labels = array();
713
714 } else {
715 // we keep encountering the entry in feeds, so we need to
716 // update date_updated column so that we don't get horrible
717 // dupes when the entry gets purged and reinserted again e.g.
718 // in the case of SLOW SLOW OMG SLOW updating feeds
719
720 $base_entry_id = db_fetch_result($result, 0, "id");
721
722 db_query("UPDATE ttrss_entries SET date_updated = NOW()
723 WHERE id = '$base_entry_id'");
724
725 $article_labels = get_article_labels($base_entry_id, $owner_uid);
726 }
727
728 // now it should exist, if not - bad luck then
729
730 $result = db_query("SELECT
731 id,content_hash,no_orig_date,title,plugin_data,guid,
732 ".SUBSTRING_FOR_DATE."(date_updated,1,19) as date_updated,
733 ".SUBSTRING_FOR_DATE."(updated,1,19) as updated,
734 num_comments
735 FROM
736 ttrss_entries
737 WHERE guid = '$entry_guid' OR guid = '$entry_guid_hashed'");
738
739 $entry_ref_id = 0;
740 $entry_int_id = 0;
741
742 if (db_num_rows($result) == 1) {
743
744 _debug("base guid found, checking for user record", $debug_enabled);
745
746 // this will be used below in update handler
747 $orig_content_hash = db_fetch_result($result, 0, "content_hash");
748 $orig_title = db_fetch_result($result, 0, "title");
749 $orig_num_comments = db_fetch_result($result, 0, "num_comments");
750 $orig_date_updated = strtotime(db_fetch_result($result,
751 0, "date_updated"));
752 $orig_plugin_data = db_fetch_result($result, 0, "plugin_data");
753
754 $ref_id = db_fetch_result($result, 0, "id");
755 $entry_ref_id = $ref_id;
756
757 /* $stored_guid = db_fetch_result($result, 0, "guid");
758 if ($stored_guid != $entry_guid_hashed) {
759 if ($debug_enabled) _debug("upgrading compat guid to hashed one", $debug_enabled);
760
761 db_query("UPDATE ttrss_entries SET guid = '$entry_guid_hashed' WHERE
762 id = '$ref_id'");
763 } */
764
765 // check for user post link to main table
766
767 // do we allow duplicate posts with same GUID in different feeds?
768 if (get_pref("ALLOW_DUPLICATE_POSTS", $owner_uid, false)) {
769 $dupcheck_qpart = "AND (feed_id = '$feed' OR feed_id IS NULL)";
770 } else {
771 $dupcheck_qpart = "";
772 }
773
774 /* Collect article tags here so we could filter by them: */
775
776 $article_filters = get_article_filters($filters, $entry_title,
777 $entry_content, $entry_link, $entry_timestamp, $entry_author,
778 $entry_tags);
779
780 if ($debug_enabled) {
781 _debug("article filters: ", $debug_enabled);
782 if (count($article_filters) != 0) {
783 print_r($article_filters);
784 }
785 }
786
787 if (find_article_filter($article_filters, "filter")) {
788 db_query("COMMIT"); // close transaction in progress
789 continue;
790 }
791
792 $score = calculate_article_score($article_filters);
793
794 _debug("initial score: $score", $debug_enabled);
795
796 $query = "SELECT ref_id, int_id FROM ttrss_user_entries WHERE
797 ref_id = '$ref_id' AND owner_uid = '$owner_uid'
798 $dupcheck_qpart";
799
800 // if ($_REQUEST["xdebug"]) print "$query\n";
801
802 $result = db_query($query);
803
804 // okay it doesn't exist - create user entry
805 if (db_num_rows($result) == 0) {
806
807 _debug("user record not found, creating...", $debug_enabled);
808
809 if ($score >= -500 && !find_article_filter($article_filters, 'catchup')) {
810 $unread = 'true';
811 $last_read_qpart = 'NULL';
812 } else {
813 $unread = 'false';
814 $last_read_qpart = 'NOW()';
815 }
816
817 if (find_article_filter($article_filters, 'mark') || $score > 1000) {
818 $marked = 'true';
819 } else {
820 $marked = 'false';
821 }
822
823 if (find_article_filter($article_filters, 'publish')) {
824 $published = 'true';
825 } else {
826 $published = 'false';
827 }
828
829 // N-grams
830
831 if (DB_TYPE == "pgsql" and defined('_NGRAM_TITLE_DUPLICATE_THRESHOLD')) {
832
833 $result = db_query("SELECT COUNT(*) AS similar FROM
834 ttrss_entries,ttrss_user_entries
835 WHERE ref_id = id AND updated >= NOW() - INTERVAL '7 day'
836 AND similarity(title, '$entry_title') >= "._NGRAM_TITLE_DUPLICATE_THRESHOLD."
837 AND owner_uid = $owner_uid");
838
839 $ngram_similar = db_fetch_result($result, 0, "similar");
840
841 _debug("N-gram similar results: $ngram_similar", $debug_enabled);
842
843 if ($ngram_similar > 0) {
844 $unread = 'false';
845 }
846 }
847
848 $last_marked = ($marked == 'true') ? 'NOW()' : 'NULL';
849 $last_published = ($published == 'true') ? 'NOW()' : 'NULL';
850
851 $result = db_query(
852 "INSERT INTO ttrss_user_entries
853 (ref_id, owner_uid, feed_id, unread, last_read, marked,
854 published, score, tag_cache, label_cache, uuid,
855 last_marked, last_published)
856 VALUES ('$ref_id', '$owner_uid', '$feed', $unread,
857 $last_read_qpart, $marked, $published, '$score', '', '',
858 '', $last_marked, $last_published)");
859
860 if (PUBSUBHUBBUB_HUB && $published == 'true') {
861 $rss_link = get_self_url_prefix() .
862 "/public.php?op=rss&id=-2&key=" .
863 get_feed_access_key(-2, false, $owner_uid);
864
865 $p = new Publisher(PUBSUBHUBBUB_HUB);
866
867 $pubsub_result = $p->publish_update($rss_link);
868 }
869
870 $result = db_query(
871 "SELECT int_id FROM ttrss_user_entries WHERE
872 ref_id = '$ref_id' AND owner_uid = '$owner_uid' AND
873 feed_id = '$feed' LIMIT 1");
874
875 if (db_num_rows($result) == 1) {
876 $entry_int_id = db_fetch_result($result, 0, "int_id");
877 }
878 } else {
879 _debug("user record FOUND", $debug_enabled);
880
881 $entry_ref_id = db_fetch_result($result, 0, "ref_id");
882 $entry_int_id = db_fetch_result($result, 0, "int_id");
883 }
884
885 _debug("RID: $entry_ref_id, IID: $entry_int_id", $debug_enabled);
886
887 $post_needs_update = false;
888 $update_insignificant = false;
889
890 if ($orig_num_comments != $num_comments) {
891 $post_needs_update = true;
892 $update_insignificant = true;
893 }
894
895 if ($entry_plugin_data != $orig_plugin_data) {
896 $post_needs_update = true;
897 $update_insignificant = true;
898 }
899
900 if ($content_hash != $orig_content_hash) {
901 $post_needs_update = true;
902 $update_insignificant = false;
903 }
904
905 if (db_escape_string($orig_title) != $entry_title) {
906 $post_needs_update = true;
907 $update_insignificant = false;
908 }
909
910 // if post needs update, update it and mark all user entries
911 // linking to this post as updated
912 if ($post_needs_update) {
913
914 if (defined('DAEMON_EXTENDED_DEBUG')) {
915 _debug("post $entry_guid_hashed needs update...", $debug_enabled);
916 }
917
918 // print "<!-- post $orig_title needs update : $post_needs_update -->";
919
920 db_query("UPDATE ttrss_entries
921 SET title = '$entry_title', content = '$entry_content',
922 content_hash = '$content_hash',
923 updated = '$entry_timestamp_fmt',
924 num_comments = '$num_comments',
925 plugin_data = '$entry_plugin_data'
926 WHERE id = '$ref_id'");
927
928 if (!$update_insignificant) {
929 if ($mark_unread_on_update) {
930 db_query("UPDATE ttrss_user_entries
931 SET last_read = null, unread = true WHERE ref_id = '$ref_id'");
932 }
933 }
934 }
935 }
936
937 db_query("COMMIT");
938
939 _debug("assigning labels...", $debug_enabled);
940
941 assign_article_to_label_filters($entry_ref_id, $article_filters,
942 $owner_uid, $article_labels);
943
944 _debug("looking for enclosures...", $debug_enabled);
945
946 // enclosures
947
948 $enclosures = array();
949
950 $encs = $item->get_enclosures();
951
952 if (is_array($encs)) {
953 foreach ($encs as $e) {
954 $e_item = array(
955 $e->link, $e->type, $e->length);
956 array_push($enclosures, $e_item);
957 }
958 }
959
960 if ($debug_enabled) {
961 _debug("article enclosures:", $debug_enabled);
962 print_r($enclosures);
963 }
964
965 db_query("BEGIN");
966
967 foreach ($enclosures as $enc) {
968 $enc_url = db_escape_string($enc[0]);
969 $enc_type = db_escape_string($enc[1]);
970 $enc_dur = db_escape_string($enc[2]);
971
972 $result = db_query("SELECT id FROM ttrss_enclosures
973 WHERE content_url = '$enc_url' AND post_id = '$entry_ref_id'");
974
975 if (db_num_rows($result) == 0) {
976 db_query("INSERT INTO ttrss_enclosures
977 (content_url, content_type, title, duration, post_id) VALUES
978 ('$enc_url', '$enc_type', '', '$enc_dur', '$entry_ref_id')");
979 }
980 }
981
982 db_query("COMMIT");
983
984 // check for manual tags (we have to do it here since they're loaded from filters)
985
986 foreach ($article_filters as $f) {
987 if ($f["type"] == "tag") {
988
989 $manual_tags = trim_array(explode(",", $f["param"]));
990
991 foreach ($manual_tags as $tag) {
992 if (tag_is_valid($tag)) {
993 array_push($entry_tags, $tag);
994 }
995 }
996 }
997 }
998
999 // Skip boring tags
1000
1001 $boring_tags = trim_array(explode(",", mb_strtolower(get_pref(
1002 'BLACKLISTED_TAGS', $owner_uid, ''), 'utf-8')));
1003
1004 $filtered_tags = array();
1005 $tags_to_cache = array();
1006
1007 if ($entry_tags && is_array($entry_tags)) {
1008 foreach ($entry_tags as $tag) {
1009 if (array_search($tag, $boring_tags) === false) {
1010 array_push($filtered_tags, $tag);
1011 }
1012 }
1013 }
1014
1015 $filtered_tags = array_unique($filtered_tags);
1016
1017 if ($debug_enabled) {
1018 _debug("filtered article tags:", $debug_enabled);
1019 print_r($filtered_tags);
1020 }
1021
1022 // Save article tags in the database
1023
1024 if (count($filtered_tags) > 0) {
1025
1026 db_query("BEGIN");
1027
1028 foreach ($filtered_tags as $tag) {
1029
1030 $tag = sanitize_tag($tag);
1031 $tag = db_escape_string($tag);
1032
1033 if (!tag_is_valid($tag)) continue;
1034
1035 $result = db_query("SELECT id FROM ttrss_tags
1036 WHERE tag_name = '$tag' AND post_int_id = '$entry_int_id' AND
1037 owner_uid = '$owner_uid' LIMIT 1");
1038
1039 if ($result && db_num_rows($result) == 0) {
1040
1041 db_query("INSERT INTO ttrss_tags
1042 (owner_uid,tag_name,post_int_id)
1043 VALUES ('$owner_uid','$tag', '$entry_int_id')");
1044 }
1045
1046 array_push($tags_to_cache, $tag);
1047 }
1048
1049 /* update the cache */
1050
1051 $tags_to_cache = array_unique($tags_to_cache);
1052
1053 $tags_str = db_escape_string(join(",", $tags_to_cache));
1054
1055 db_query("UPDATE ttrss_user_entries
1056 SET tag_cache = '$tags_str' WHERE ref_id = '$entry_ref_id'
1057 AND owner_uid = $owner_uid");
1058
1059 db_query("COMMIT");
1060 }
1061
1062 if (get_pref("AUTO_ASSIGN_LABELS", $owner_uid, false)) {
1063 _debug("auto-assigning labels...", $debug_enabled);
1064
1065 foreach ($labels as $label) {
1066 $caption = preg_quote($label["caption"]);
1067
1068 if ($caption && preg_match("/\b$caption\b/i", "$tags_str " . strip_tags($entry_content) . " $entry_title")) {
1069 if (!labels_contains_caption($article_labels, $caption)) {
1070 label_add_article($entry_ref_id, $caption, $owner_uid);
1071 }
1072 }
1073 }
1074 }
1075
1076 _debug("article processed", $debug_enabled);
1077 }
1078
1079 _debug("purging feed...", $debug_enabled);
1080
1081 purge_feed($feed, 0, $debug_enabled);
1082
1083 db_query("UPDATE ttrss_feeds
1084 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
1085
1086 // db_query("COMMIT");
1087
1088 } else {
1089
1090 $error_msg = db_escape_string(mb_substr($rss->error(), 0, 245));
1091
1092 _debug("error fetching feed: $error_msg", $debug_enabled);
1093
1094 db_query(
1095 "UPDATE ttrss_feeds SET last_error = '$error_msg',
1096 last_updated = NOW() WHERE id = '$feed'");
1097 }
1098
1099 unset($rss);
1100
1101 _debug("done", $debug_enabled);
1102 }
1103
1104 function cache_images($html, $site_url, $debug) {
1105 $cache_dir = CACHE_DIR . "/images";
1106
1107 libxml_use_internal_errors(true);
1108
1109 $charset_hack = '<head>
1110 <meta http-equiv="Content-Type" content="text/html; charset=utf-8"/>
1111 </head>';
1112
1113 $doc = new DOMDocument();
1114 $doc->loadHTML($charset_hack . $html);
1115 $xpath = new DOMXPath($doc);
1116
1117 $entries = $xpath->query('(//img[@src])');
1118
1119 foreach ($entries as $entry) {
1120 if ($entry->hasAttribute('src')) {
1121 $src = rewrite_relative_url($site_url, $entry->getAttribute('src'));
1122
1123 $local_filename = CACHE_DIR . "/images/" . sha1($src) . ".png";
1124
1125 if ($debug) _debug("cache_images: downloading: $src to $local_filename");
1126
1127 if (!file_exists($local_filename)) {
1128 $file_content = fetch_file_contents($src);
1129
1130 if ($file_content && strlen($file_content) > 1024) {
1131 file_put_contents($local_filename, $file_content);
1132 }
1133 }
1134
1135 /* if (file_exists($local_filename)) {
1136 $entry->setAttribute('src', SELF_URL_PATH . '/image.php?url=' .
1137 base64_encode($src));
1138 } */
1139 }
1140 }
1141
1142 //$node = $doc->getElementsByTagName('body')->item(0);
1143 //return $doc->saveXML($node);
1144 }
1145
1146 function expire_error_log($debug) {
1147 if ($debug) _debug("Removing old error log entries...");
1148
1149 if (DB_TYPE == "pgsql") {
1150 db_query("DELETE FROM ttrss_error_log
1151 WHERE created_at < NOW() - INTERVAL '7 days'");
1152 } else {
1153 db_query("DELETE FROM ttrss_error_log
1154 WHERE created_at < DATE_SUB(NOW(), INTERVAL 7 DAY)");
1155 }
1156
1157 }
1158
1159 function expire_lock_files($debug) {
1160 //if ($debug) _debug("Removing old lock files...");
1161
1162 $num_deleted = 0;
1163
1164 if (is_writable(LOCK_DIRECTORY)) {
1165 $files = glob(LOCK_DIRECTORY . "/*.lock");
1166
1167 if ($files) {
1168 foreach ($files as $file) {
1169 if (!file_is_locked(basename($file)) && time() - filemtime($file) > 86400*2) {
1170 unlink($file);
1171 ++$num_deleted;
1172 }
1173 }
1174 }
1175 }
1176
1177 if ($debug) _debug("Removed $num_deleted old lock files.");
1178 }
1179
1180 function expire_cached_files($debug) {
1181 foreach (array("simplepie", "images", "export", "upload") as $dir) {
1182 $cache_dir = CACHE_DIR . "/$dir";
1183
1184 // if ($debug) _debug("Expiring $cache_dir");
1185
1186 $num_deleted = 0;
1187
1188 if (is_writable($cache_dir)) {
1189 $files = glob("$cache_dir/*");
1190
1191 if ($files) {
1192 foreach ($files as $file) {
1193 if (time() - filemtime($file) > 86400*7) {
1194 unlink($file);
1195
1196 ++$num_deleted;
1197 }
1198 }
1199 }
1200 }
1201
1202 if ($debug) _debug("$cache_dir: removed $num_deleted files.");
1203 }
1204 }
1205
1206 /**
1207 * Source: http://www.php.net/manual/en/function.parse-url.php#104527
1208 * Returns the url query as associative array
1209 *
1210 * @param string query
1211 * @return array params
1212 */
1213 function convertUrlQuery($query) {
1214 $queryParts = explode('&', $query);
1215
1216 $params = array();
1217
1218 foreach ($queryParts as $param) {
1219 $item = explode('=', $param);
1220 $params[$item[0]] = $item[1];
1221 }
1222
1223 return $params;
1224 }
1225
1226 function get_article_filters($filters, $title, $content, $link, $timestamp, $author, $tags) {
1227 $matches = array();
1228
1229 foreach ($filters as $filter) {
1230 $match_any_rule = $filter["match_any_rule"];
1231 $inverse = $filter["inverse"];
1232 $filter_match = false;
1233
1234 foreach ($filter["rules"] as $rule) {
1235 $match = false;
1236 $reg_exp = str_replace('/', '\/', $rule["reg_exp"]);
1237 $rule_inverse = $rule["inverse"];
1238
1239 if (!$reg_exp)
1240 continue;
1241
1242 switch ($rule["type"]) {
1243 case "title":
1244 $match = @preg_match("/$reg_exp/i", $title);
1245 break;
1246 case "content":
1247 // we don't need to deal with multiline regexps
1248 $content = preg_replace("/[\r\n\t]/", "", $content);
1249
1250 $match = @preg_match("/$reg_exp/i", $content);
1251 break;
1252 case "both":
1253 // we don't need to deal with multiline regexps
1254 $content = preg_replace("/[\r\n\t]/", "", $content);
1255
1256 $match = (@preg_match("/$reg_exp/i", $title) || @preg_match("/$reg_exp/i", $content));
1257 break;
1258 case "link":
1259 $match = @preg_match("/$reg_exp/i", $link);
1260 break;
1261 case "author":
1262 $match = @preg_match("/$reg_exp/i", $author);
1263 break;
1264 case "tag":
1265 foreach ($tags as $tag) {
1266 if (@preg_match("/$reg_exp/i", $tag)) {
1267 $match = true;
1268 break;
1269 }
1270 }
1271 break;
1272 }
1273
1274 if ($rule_inverse) $match = !$match;
1275
1276 if ($match_any_rule) {
1277 if ($match) {
1278 $filter_match = true;
1279 break;
1280 }
1281 } else {
1282 $filter_match = $match;
1283 if (!$match) {
1284 break;
1285 }
1286 }
1287 }
1288
1289 if ($inverse) $filter_match = !$filter_match;
1290
1291 if ($filter_match) {
1292 foreach ($filter["actions"] AS $action) {
1293 array_push($matches, $action);
1294
1295 // if Stop action encountered, perform no further processing
1296 if ($action["type"] == "stop") return $matches;
1297 }
1298 }
1299 }
1300
1301 return $matches;
1302 }
1303
1304 function find_article_filter($filters, $filter_name) {
1305 foreach ($filters as $f) {
1306 if ($f["type"] == $filter_name) {
1307 return $f;
1308 };
1309 }
1310 return false;
1311 }
1312
1313 function find_article_filters($filters, $filter_name) {
1314 $results = array();
1315
1316 foreach ($filters as $f) {
1317 if ($f["type"] == $filter_name) {
1318 array_push($results, $f);
1319 };
1320 }
1321 return $results;
1322 }
1323
1324 function calculate_article_score($filters) {
1325 $score = 0;
1326
1327 foreach ($filters as $f) {
1328 if ($f["type"] == "score") {
1329 $score += $f["param"];
1330 };
1331 }
1332 return $score;
1333 }
1334
1335 function labels_contains_caption($labels, $caption) {
1336 foreach ($labels as $label) {
1337 if ($label[1] == $caption) {
1338 return true;
1339 }
1340 }
1341
1342 return false;
1343 }
1344
1345 function assign_article_to_label_filters($id, $filters, $owner_uid, $article_labels) {
1346 foreach ($filters as $f) {
1347 if ($f["type"] == "label") {
1348 if (!labels_contains_caption($article_labels, $f["param"])) {
1349 label_add_article($id, $f["param"], $owner_uid);
1350 }
1351 }
1352 }
1353 }
1354
1355 function make_guid_from_title($title) {
1356 return preg_replace("/[ \"\',.:;]/", "-",
1357 mb_strtolower(strip_tags($title), 'utf-8'));
1358 }
1359
1360 /* function verify_feed_xml($feed_data) {
1361 libxml_use_internal_errors(true);
1362 $doc = new DOMDocument();
1363 $doc->loadXML($feed_data);
1364 $error = libxml_get_last_error();
1365 libxml_clear_errors();
1366 return $error;
1367 } */
1368
1369 function housekeeping_common($debug) {
1370 expire_cached_files($debug);
1371 expire_lock_files($debug);
1372 expire_error_log($debug);
1373
1374 $count = update_feedbrowser_cache();
1375 _debug("Feedbrowser updated, $count feeds processed.");
1376
1377 purge_orphans( true);
1378 $rc = cleanup_tags( 14, 50000);
1379
1380 _debug("Cleaned $rc cached tags.");
1381 }
1382 ?>