]> 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 basic feed authentication parameter encryption in the database (FEED_CRYPT_KEY)
[tt-rss.git] / include / rssfuncs.php
1 <?php
2 define('DAEMON_UPDATE_LOGIN_LIMIT', 30);
3 define('DAEMON_FEED_LIMIT', 100);
4 define('DAEMON_SLEEP_INTERVAL', 60);
5
6 function update_feedbrowser_cache($link) {
7
8 $result = db_query($link, "SELECT feed_url, site_url, title, COUNT(id) AS subscribers
9 FROM ttrss_feeds WHERE (SELECT COUNT(id) = 0 FROM ttrss_feeds AS tf
10 WHERE tf.feed_url = ttrss_feeds.feed_url
11 AND (private IS true OR auth_login != '' OR auth_pass != '' OR feed_url LIKE '%:%@%/%'))
12 GROUP BY feed_url, site_url, title ORDER BY subscribers DESC LIMIT 1000");
13
14 db_query($link, "BEGIN");
15
16 db_query($link, "DELETE FROM ttrss_feedbrowser_cache");
17
18 $count = 0;
19
20 while ($line = db_fetch_assoc($result)) {
21 $subscribers = db_escape_string($link, $line["subscribers"]);
22 $feed_url = db_escape_string($link, $line["feed_url"]);
23 $title = db_escape_string($link, $line["title"]);
24 $site_url = db_escape_string($link, $line["site_url"]);
25
26 $tmp_result = db_query($link, "SELECT subscribers FROM
27 ttrss_feedbrowser_cache WHERE feed_url = '$feed_url'");
28
29 if (db_num_rows($tmp_result) == 0) {
30
31 db_query($link, "INSERT INTO ttrss_feedbrowser_cache
32 (feed_url, site_url, title, subscribers) VALUES ('$feed_url',
33 '$site_url', '$title', '$subscribers')");
34
35 ++$count;
36
37 }
38
39 }
40
41 db_query($link, "COMMIT");
42
43 return $count;
44
45 }
46
47
48 /**
49 * Update a feed batch.
50 * Used by daemons to update n feeds by run.
51 * Only update feed needing a update, and not being processed
52 * by another process.
53 *
54 * @param mixed $link Database link
55 * @param integer $limit Maximum number of feeds in update batch. Default to DAEMON_FEED_LIMIT.
56 * @param boolean $from_http Set to true if you call this function from http to disable cli specific code.
57 * @param boolean $debug Set to false to disable debug output. Default to true.
58 * @return void
59 */
60 function update_daemon_common($link, $limit = DAEMON_FEED_LIMIT, $from_http = false, $debug = true) {
61 // Process all other feeds using last_updated and interval parameters
62
63 define('PREFS_NO_CACHE', true);
64
65 // Test if the user has loggued in recently. If not, it does not update its feeds.
66 if (!SINGLE_USER_MODE && DAEMON_UPDATE_LOGIN_LIMIT > 0) {
67 if (DB_TYPE == "pgsql") {
68 $login_thresh_qpart = "AND ttrss_users.last_login >= NOW() - INTERVAL '".DAEMON_UPDATE_LOGIN_LIMIT." days'";
69 } else {
70 $login_thresh_qpart = "AND ttrss_users.last_login >= DATE_SUB(NOW(), INTERVAL ".DAEMON_UPDATE_LOGIN_LIMIT." DAY)";
71 }
72 } else {
73 $login_thresh_qpart = "";
74 }
75
76 // Test if the feed need a update (update interval exceded).
77 if (DB_TYPE == "pgsql") {
78 $update_limit_qpart = "AND ((
79 ttrss_feeds.update_interval = 0
80 AND ttrss_user_prefs.value != '-1'
81 AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_user_prefs.value || ' minutes') AS INTERVAL)
82 ) OR (
83 ttrss_feeds.update_interval > 0
84 AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_feeds.update_interval || ' minutes') AS INTERVAL)
85 ) OR ttrss_feeds.last_updated IS NULL
86 OR last_updated = '1970-01-01 00:00:00')";
87 } else {
88 $update_limit_qpart = "AND ((
89 ttrss_feeds.update_interval = 0
90 AND ttrss_user_prefs.value != '-1'
91 AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL CONVERT(ttrss_user_prefs.value, SIGNED INTEGER) MINUTE)
92 ) OR (
93 ttrss_feeds.update_interval > 0
94 AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL ttrss_feeds.update_interval MINUTE)
95 ) OR ttrss_feeds.last_updated IS NULL
96 OR last_updated = '1970-01-01 00:00:00')";
97 }
98
99 // Test if feed is currently being updated by another process.
100 if (DB_TYPE == "pgsql") {
101 $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < NOW() - INTERVAL '5 minutes')";
102 } else {
103 $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < DATE_SUB(NOW(), INTERVAL 5 MINUTE))";
104 }
105
106 // Test if there is a limit to number of updated feeds
107 $query_limit = "";
108 if($limit) $query_limit = sprintf("LIMIT %d", $limit);
109
110 $random_qpart = sql_random_function();
111
112 // We search for feed needing update.
113 $result = db_query($link, "SELECT DISTINCT ttrss_feeds.feed_url,$random_qpart
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...", 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($result)) {
131 array_push($feeds_to_update, db_escape_string($link, $line['feed_url']));
132 }
133
134 // We update the feed last update started date before anything else.
135 // There is no lag due to feed contents downloads
136 // It prevent an other process to update the same feed.
137
138 if(count($feeds_to_update) > 0) {
139 $feeds_quoted = array();
140
141 foreach ($feeds_to_update as $feed) {
142 array_push($feeds_quoted, "'" . db_escape_string($link, $feed) . "'");
143 }
144
145 db_query($link, sprintf("UPDATE ttrss_feeds SET last_update_started = NOW()
146 WHERE feed_url IN (%s)", implode(',', $feeds_quoted)));
147 }
148
149 expire_cached_files($debug);
150 expire_lock_files($debug);
151
152 $nf = 0;
153
154 // For each feed, we call the feed update function.
155 foreach ($feeds_to_update as $feed) {
156 if($debug) _debug("Base feed: $feed");
157
158 //update_rss_feed($link, $line["id"], true);
159
160 // since we have the data cached, we can deal with other feeds with the same url
161
162 $tmp_result = db_query($link, "SELECT DISTINCT ttrss_feeds.id,last_updated
163 FROM ttrss_feeds, ttrss_users, ttrss_user_prefs WHERE
164 ttrss_user_prefs.owner_uid = ttrss_feeds.owner_uid AND
165 ttrss_users.id = ttrss_user_prefs.owner_uid AND
166 ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL' AND
167 feed_url = '".db_escape_string($link, $feed)."' AND
168 (ttrss_feeds.update_interval > 0 OR
169 ttrss_user_prefs.value != '-1')
170 $login_thresh_qpart
171 ORDER BY ttrss_feeds.id $query_limit");
172
173 if (db_num_rows($tmp_result) > 0) {
174 while ($tline = db_fetch_assoc($tmp_result)) {
175 if($debug) _debug(" => " . $tline["last_updated"] . ", " . $tline["id"]);
176 update_rss_feed($link, $tline["id"], true);
177 ++$nf;
178 }
179 }
180 }
181
182 require_once "digest.php";
183
184 // Send feed digests by email if needed.
185 send_headlines_digests($link, $debug);
186
187 return $nf;
188
189 } // function update_daemon_common
190
191 // ignore_daemon is not used
192 function update_rss_feed($link, $feed, $ignore_daemon = false, $no_cache = false,
193 $override_url = false) {
194
195 require_once "lib/simplepie/simplepie.inc";
196
197 $debug_enabled = defined('DAEMON_EXTENDED_DEBUG') || $_REQUEST['xdebug'];
198
199 if ($debug_enabled) {
200 _debug("update_rss_feed: start");
201 }
202
203 $result = db_query($link, "SELECT id,update_interval,auth_login,
204 feed_url,auth_pass,cache_images,last_updated,
205 mark_unread_on_update, owner_uid,
206 pubsub_state, auth_pass_encrypted
207 FROM ttrss_feeds WHERE id = '$feed'");
208
209 if (db_num_rows($result) == 0) {
210 if ($debug_enabled) {
211 _debug("update_rss_feed: feed $feed NOT FOUND/SKIPPED");
212 }
213 return false;
214 }
215
216 $last_updated = db_fetch_result($result, 0, "last_updated");
217 $owner_uid = db_fetch_result($result, 0, "owner_uid");
218 $mark_unread_on_update = sql_bool_to_bool(db_fetch_result($result,
219 0, "mark_unread_on_update"));
220 $pubsub_state = db_fetch_result($result, 0, "pubsub_state");
221 $auth_pass_encrypted = sql_bool_to_bool(db_fetch_result($result,
222 0, "auth_pass_encrypted"));
223
224 db_query($link, "UPDATE ttrss_feeds SET last_update_started = NOW()
225 WHERE id = '$feed'");
226
227 $auth_login = db_fetch_result($result, 0, "auth_login");
228 $auth_pass = db_fetch_result($result, 0, "auth_pass");
229
230 if ($auth_pass_encrypted) {
231 require_once "crypt.php";
232 $auth_pass = decrypt_string($auth_pass);
233 }
234
235 $cache_images = sql_bool_to_bool(db_fetch_result($result, 0, "cache_images"));
236 $fetch_url = db_fetch_result($result, 0, "feed_url");
237
238 $feed = db_escape_string($link, $feed);
239
240 if ($override_url) $fetch_url = $override_url;
241
242 $date_feed_processed = date('Y-m-d H:i');
243
244 $cache_filename = CACHE_DIR . "/simplepie/" . sha1($fetch_url) . ".feed";
245
246 // Ignore cache if new feed or manual update.
247 $cache_age = ($no_cache || is_null($last_updated) || $last_updated == '1970-01-01 00:00:00') ?
248 30 : get_feed_update_interval($link, $feed) * 60;
249
250 if ($debug_enabled) {
251 _debug("update_rss_feed: cache filename: $cache_filename exists: " . file_exists($cache_filename));
252 _debug("update_rss_feed: cache age: $cache_age; no cache: $no_cache");
253 }
254
255 $cached_feed_data_hash = false;
256
257 $rss = false;
258 $rss_hash = false;
259 $cache_timestamp = file_exists($cache_filename) ? filemtime($cache_filename) : 0;
260 $last_updated_timestamp = strtotime($last_updated);
261
262 if (file_exists($cache_filename) &&
263 is_readable($cache_filename) &&
264 !$auth_login && !$auth_pass &&
265 filemtime($cache_filename) > time() - $cache_age) {
266
267 if ($debug_enabled) {
268 _debug("update_rss_feed: using local cache.");
269 }
270
271 if ($cache_timestamp > $last_updated_timestamp) {
272 @$rss_data = file_get_contents($cache_filename);
273
274 if ($rss_data) {
275 $rss_hash = sha1($rss_data);
276 @$rss = unserialize($rss_data);
277 }
278 } else {
279 if ($debug_enabled) {
280 _debug("update_rss_feed: local cache valid and older than last_updated, nothing to do.");
281 }
282 return;
283 }
284 }
285
286 if (!$rss) {
287
288 if (!$feed_data) {
289 if ($debug_enabled) {
290 _debug("update_rss_feed: fetching [$fetch_url] (ts: $cache_timestamp/$last_updated_timestamp)");
291 }
292
293 $force_refetch = isset($_REQUEST["force_refetch"]);
294
295 $feed_data = fetch_file_contents($fetch_url, false,
296 $auth_login, $auth_pass, false,
297 $no_cache ? FEED_FETCH_NO_CACHE_TIMEOUT : FEED_FETCH_TIMEOUT,
298 $force_refetch ? 0 : max($last_updated_timestamp, $cache_timestamp));
299
300 if ($debug_enabled) {
301 _debug("update_rss_feed: fetch done.");
302 }
303
304 }
305
306 if (!$feed_data) {
307 global $fetch_last_error;
308 global $fetch_last_error_code;
309
310 if ($debug_enabled) {
311 _debug("update_rss_feed: unable to fetch: $fetch_last_error [$fetch_last_error_code]");
312 }
313
314 $error_escaped = '';
315
316 // If-Modified-Since
317 if ($fetch_last_error_code != 304) {
318 $error_escaped = db_escape_string($link, $fetch_last_error);
319 } else {
320 if ($debug_enabled) {
321 _debug("update_rss_feed: source claims data not modified, nothing to do.");
322 }
323 }
324
325 db_query($link,
326 "UPDATE ttrss_feeds SET last_error = '$error_escaped',
327 last_updated = NOW() WHERE id = '$feed'");
328
329 return;
330 }
331 }
332
333 $pluginhost = new PluginHost($link);
334 $pluginhost->set_debug($debug_enabled);
335 $user_plugins = get_pref($link, "_ENABLED_PLUGINS", $owner_uid);
336
337 $pluginhost->load(PLUGINS, $pluginhost::KIND_ALL);
338 $pluginhost->load($user_plugins, $pluginhost::KIND_USER, $owner_uid);
339 $pluginhost->load_data();
340
341 foreach ($pluginhost->get_hooks($pluginhost::HOOK_FEED_FETCHED) as $plugin) {
342 $feed_data = $plugin->hook_feed_fetched($feed_data);
343 }
344
345 if (!$rss) {
346 $rss = new SimplePie();
347 $rss->set_sanitize_class("SanitizeDummy");
348 // simplepie ignores the above and creates default sanitizer anyway,
349 // so let's override it...
350 $rss->sanitize = new SanitizeDummy();
351 $rss->set_output_encoding('UTF-8');
352 $rss->set_raw_data($feed_data);
353 $rss->enable_cache(false);
354
355 @$rss->init();
356 }
357
358 // print_r($rss);
359
360 $feed = db_escape_string($link, $feed);
361
362 if (!$rss->error()) {
363
364 // cache data for later
365 if (!$auth_pass && !$auth_login && is_writable(CACHE_DIR . "/simplepie")) {
366 $rss_data = serialize($rss);
367 $new_rss_hash = sha1($rss_data);
368
369 if ($new_rss_hash != $rss_hash) {
370 if ($debug_enabled) {
371 _debug("update_rss_feed: saving $cache_filename");
372 }
373 @file_put_contents($cache_filename, serialize($rss));
374 }
375 }
376
377 // We use local pluginhost here because we need to load different per-user feed plugins
378 $pluginhost->run_hooks($pluginhost::HOOK_FEED_PARSED, "hook_feed_parsed", $rss);
379
380 if ($debug_enabled) {
381 _debug("update_rss_feed: processing feed data...");
382 }
383
384 // db_query($link, "BEGIN");
385
386 if (DB_TYPE == "pgsql") {
387 $favicon_interval_qpart = "favicon_last_checked < NOW() - INTERVAL '12 hour'";
388 } else {
389 $favicon_interval_qpart = "favicon_last_checked < DATE_SUB(NOW(), INTERVAL 12 HOUR)";
390 }
391
392 $result = db_query($link, "SELECT title,site_url,owner_uid,
393 (favicon_last_checked IS NULL OR $favicon_interval_qpart) AS
394 favicon_needs_check
395 FROM ttrss_feeds WHERE id = '$feed'");
396
397 $registered_title = db_fetch_result($result, 0, "title");
398 $orig_site_url = db_fetch_result($result, 0, "site_url");
399 $favicon_needs_check = sql_bool_to_bool(db_fetch_result($result, 0,
400 "favicon_needs_check"));
401
402 $owner_uid = db_fetch_result($result, 0, "owner_uid");
403
404 $site_url = db_escape_string($link, mb_substr(rewrite_relative_url($fetch_url, $rss->get_link()), 0, 245));
405
406 if ($debug_enabled) {
407 _debug("update_rss_feed: checking favicon...");
408 }
409
410 if ($favicon_needs_check) {
411 check_feed_favicon($site_url, $feed, $link);
412
413 db_query($link, "UPDATE ttrss_feeds SET favicon_last_checked = NOW()
414 WHERE id = '$feed'");
415 }
416
417 if (!$registered_title || $registered_title == "[Unknown]") {
418
419 $feed_title = db_escape_string($link, $rss->get_title());
420
421 if ($debug_enabled) {
422 _debug("update_rss_feed: registering title: $feed_title");
423 }
424
425 db_query($link, "UPDATE ttrss_feeds SET
426 title = '$feed_title' WHERE id = '$feed'");
427 }
428
429 if ($site_url && $orig_site_url != $site_url) {
430 db_query($link, "UPDATE ttrss_feeds SET
431 site_url = '$site_url' WHERE id = '$feed'");
432 }
433
434 if ($debug_enabled) {
435 _debug("update_rss_feed: loading filters & labels...");
436 }
437
438 $filters = load_filters($link, $feed, $owner_uid);
439 $labels = get_all_labels($link, $owner_uid);
440
441 if ($debug_enabled) {
442 //print_r($filters);
443 _debug("update_rss_feed: " . count($filters) . " filters loaded.");
444 }
445
446 $items = $rss->get_items();
447
448 if (!is_array($items)) {
449 if ($debug_enabled) {
450 _debug("update_rss_feed: no articles found.");
451 }
452
453 db_query($link, "UPDATE ttrss_feeds
454 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
455
456 return; // no articles
457 }
458
459 if ($pubsub_state != 2 && PUBSUBHUBBUB_ENABLED) {
460
461 if ($debug_enabled) _debug("update_rss_feed: checking for PUSH hub...");
462
463 $feed_hub_url = false;
464
465 $links = $rss->get_links('hub');
466
467 if ($links && is_array($links)) {
468 foreach ($links as $l) {
469 $feed_hub_url = $l;
470 break;
471 }
472 }
473
474 if ($debug_enabled) _debug("update_rss_feed: feed hub url: $feed_hub_url");
475
476 if ($feed_hub_url && function_exists('curl_init') &&
477 !ini_get("open_basedir")) {
478
479 require_once 'lib/pubsubhubbub/subscriber.php';
480
481 $callback_url = get_self_url_prefix() .
482 "/public.php?op=pubsub&id=$feed";
483
484 $s = new Subscriber($feed_hub_url, $callback_url);
485
486 $rc = $s->subscribe($fetch_url);
487
488 if ($debug_enabled)
489 _debug("update_rss_feed: feed hub url found, subscribe request sent.");
490
491 db_query($link, "UPDATE ttrss_feeds SET pubsub_state = 1
492 WHERE id = '$feed'");
493 }
494 }
495
496 if ($debug_enabled) {
497 _debug("update_rss_feed: processing articles...");
498 }
499
500 foreach ($items as $item) {
501 if ($_REQUEST['xdebug'] == 3) {
502 print_r($item);
503 }
504
505 $entry_guid = $item->get_id();
506 if (!$entry_guid) $entry_guid = $item->get_link();
507 if (!$entry_guid) $entry_guid = make_guid_from_title($item->get_title());
508
509 if ($debug_enabled) {
510 _debug("update_rss_feed: guid $entry_guid");
511 }
512
513 if (!$entry_guid) continue;
514
515 $entry_guid = "$owner_uid,$entry_guid";
516
517 $entry_timestamp = "";
518
519 $entry_timestamp = strtotime($item->get_date());
520
521 if ($entry_timestamp == -1 || !$entry_timestamp || $entry_timestamp > time()) {
522 $entry_timestamp = time();
523 $no_orig_date = 'true';
524 } else {
525 $no_orig_date = 'false';
526 }
527
528 $entry_timestamp_fmt = strftime("%Y/%m/%d %H:%M:%S", $entry_timestamp);
529
530 if ($debug_enabled) {
531 _debug("update_rss_feed: date $entry_timestamp [$entry_timestamp_fmt]");
532 }
533
534 $entry_title = html_entity_decode($item->get_title());
535
536 $entry_link = rewrite_relative_url($site_url, $item->get_link());
537
538 if ($debug_enabled) {
539 _debug("update_rss_feed: title $entry_title");
540 _debug("update_rss_feed: link $entry_link");
541 }
542
543 if (!$entry_title) $entry_title = date("Y-m-d H:i:s", $entry_timestamp);;
544
545 $entry_content = $item->get_content();
546 if (!$entry_content) $entry_content = $item->get_description();
547
548 if ($_REQUEST["xdebug"] == 2) {
549 print "update_rss_feed: content: ";
550 print $entry_content;
551 print "\n";
552 }
553
554 $entry_comments = $item->data["comments"];
555
556 if ($item->get_author()) {
557 $entry_author_item = $item->get_author();
558 $entry_author = $entry_author_item->get_name();
559 if (!$entry_author) $entry_author = $entry_author_item->get_email();
560
561 $entry_author = db_escape_string($link, $entry_author);
562 }
563
564 $entry_guid = db_escape_string($link, mb_substr($entry_guid, 0, 245));
565
566 $entry_comments = db_escape_string($link, mb_substr($entry_comments, 0, 245));
567 $entry_author = db_escape_string($link, mb_substr($entry_author, 0, 245));
568
569 $num_comments = $item->get_item_tags('http://purl.org/rss/1.0/modules/slash/', 'comments');
570
571 if (is_array($num_comments) && is_array($num_comments[0])) {
572 $num_comments = (int) $num_comments[0]["data"];
573 } else {
574 $num_comments = 0;
575 }
576
577 if ($debug_enabled) {
578 _debug("update_rss_feed: num_comments: $num_comments");
579 _debug("update_rss_feed: looking for tags [1]...");
580 }
581
582 // parse <category> entries into tags
583
584 $additional_tags = array();
585
586 $additional_tags_src = $item->get_categories();
587
588 if (is_array($additional_tags_src)) {
589 foreach ($additional_tags_src as $tobj) {
590 array_push($additional_tags, $tobj->get_term());
591 }
592 }
593
594 if ($debug_enabled) {
595 _debug("update_rss_feed: category tags:");
596 print_r($additional_tags);
597 }
598
599 if ($debug_enabled) {
600 _debug("update_rss_feed: looking for tags [2]...");
601 }
602
603 $entry_tags = array_unique($additional_tags);
604
605 for ($i = 0; $i < count($entry_tags); $i++)
606 $entry_tags[$i] = mb_strtolower($entry_tags[$i], 'utf-8');
607
608 if ($debug_enabled) {
609 //_debug("update_rss_feed: unfiltered tags found:");
610 //print_r($entry_tags);
611 }
612
613 if ($debug_enabled) {
614 _debug("update_rss_feed: done collecting data.");
615 }
616
617 // TODO: less memory-hungry implementation
618
619 if ($debug_enabled) {
620 _debug("update_rss_feed: applying plugin filters..");
621 }
622
623 // FIXME not sure if owner_uid is a good idea here, we may have a base entry without user entry (?)
624 $result = db_query($link, "SELECT plugin_data,title,content,link,tag_cache,author FROM ttrss_entries, ttrss_user_entries
625 WHERE ref_id = id AND guid = '".db_escape_string($link, $entry_guid)."' AND owner_uid = $owner_uid");
626
627 if (db_num_rows($result) != 0) {
628 $entry_plugin_data = db_fetch_result($result, 0, "plugin_data");
629 $stored_article = array("title" => db_fetch_result($result, 0, "title"),
630 "content" => db_fetch_result($result, 0, "content"),
631 "link" => db_fetch_result($result, 0, "link"),
632 "tags" => explode(",", db_fetch_result($result, 0, "tag_cache")),
633 "author" => db_fetch_result($result, 0, "author"));
634 } else {
635 $entry_plugin_data = "";
636 $stored_article = array();
637 }
638
639 $article = array("owner_uid" => $owner_uid, // read only
640 "guid" => $entry_guid, // read only
641 "title" => $entry_title,
642 "content" => $entry_content,
643 "link" => $entry_link,
644 "tags" => $entry_tags,
645 "plugin_data" => $entry_plugin_data,
646 "author" => $entry_author,
647 "stored" => $stored_article);
648
649 foreach ($pluginhost->get_hooks($pluginhost::HOOK_ARTICLE_FILTER) as $plugin) {
650 $article = $plugin->hook_article_filter($article);
651 }
652
653 $entry_tags = $article["tags"];
654 $entry_guid = db_escape_string($link, $entry_guid);
655 $entry_title = db_escape_string($link, $article["title"]);
656 $entry_author = db_escape_string($link, $article["author"]);
657 $entry_link = db_escape_string($link, $article["link"]);
658 $entry_plugin_data = db_escape_string($link, $article["plugin_data"]);
659 $entry_content = $article["content"]; // escaped below
660
661
662 if ($debug_enabled) {
663 _debug("update_rss_feed: plugin data: $entry_plugin_data");
664 }
665
666 if ($cache_images && is_writable(CACHE_DIR . '/images'))
667 cache_images($entry_content, $site_url, $debug_enabled);
668
669 $entry_content = db_escape_string($link, $entry_content, false);
670
671 $content_hash = "SHA1:" . sha1($entry_content);
672
673 db_query($link, "BEGIN");
674
675 $result = db_query($link, "SELECT id FROM ttrss_entries
676 WHERE guid = '$entry_guid'");
677
678 if (db_num_rows($result) == 0) {
679
680 if ($debug_enabled) {
681 _debug("update_rss_feed: base guid [$entry_guid] not found");
682 }
683
684 // base post entry does not exist, create it
685
686 $result = db_query($link,
687 "INSERT INTO ttrss_entries
688 (title,
689 guid,
690 link,
691 updated,
692 content,
693 content_hash,
694 cached_content,
695 no_orig_date,
696 date_updated,
697 date_entered,
698 comments,
699 num_comments,
700 plugin_data,
701 author)
702 VALUES
703 ('$entry_title',
704 '$entry_guid',
705 '$entry_link',
706 '$entry_timestamp_fmt',
707 '$entry_content',
708 '$content_hash',
709 '',
710 $no_orig_date,
711 NOW(),
712 '$date_feed_processed',
713 '$entry_comments',
714 '$num_comments',
715 '$entry_plugin_data',
716 '$entry_author')");
717
718 $article_labels = array();
719
720 } else {
721 // we keep encountering the entry in feeds, so we need to
722 // update date_updated column so that we don't get horrible
723 // dupes when the entry gets purged and reinserted again e.g.
724 // in the case of SLOW SLOW OMG SLOW updating feeds
725
726 $base_entry_id = db_fetch_result($result, 0, "id");
727
728 db_query($link, "UPDATE ttrss_entries SET date_updated = NOW()
729 WHERE id = '$base_entry_id'");
730
731 $article_labels = get_article_labels($link, $base_entry_id, $owner_uid);
732 }
733
734 // now it should exist, if not - bad luck then
735
736 $result = db_query($link, "SELECT
737 id,content_hash,no_orig_date,title,plugin_data,
738 ".SUBSTRING_FOR_DATE."(date_updated,1,19) as date_updated,
739 ".SUBSTRING_FOR_DATE."(updated,1,19) as updated,
740 num_comments
741 FROM
742 ttrss_entries
743 WHERE guid = '$entry_guid'");
744
745 $entry_ref_id = 0;
746 $entry_int_id = 0;
747
748 if (db_num_rows($result) == 1) {
749
750 if ($debug_enabled) {
751 _debug("update_rss_feed: base guid [$entry_guid] found, checking for user record");
752 }
753
754 // this will be used below in update handler
755 $orig_content_hash = db_fetch_result($result, 0, "content_hash");
756 $orig_title = db_fetch_result($result, 0, "title");
757 $orig_num_comments = db_fetch_result($result, 0, "num_comments");
758 $orig_date_updated = strtotime(db_fetch_result($result,
759 0, "date_updated"));
760 $orig_plugin_data = db_fetch_result($result, 0, "plugin_data");
761
762 $ref_id = db_fetch_result($result, 0, "id");
763 $entry_ref_id = $ref_id;
764
765 // check for user post link to main table
766
767 // do we allow duplicate posts with same GUID in different feeds?
768 if (get_pref($link, "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("update_rss_feed: article filters: ");
782 if (count($article_filters) != 0) {
783 print_r($article_filters);
784 }
785 }
786
787 if (find_article_filter($article_filters, "filter")) {
788 db_query($link, "COMMIT"); // close transaction in progress
789 continue;
790 }
791
792 $score = calculate_article_score($article_filters);
793
794 if ($debug_enabled) {
795 _debug("update_rss_feed: initial score: $score");
796 }
797
798 $query = "SELECT ref_id, int_id FROM ttrss_user_entries WHERE
799 ref_id = '$ref_id' AND owner_uid = '$owner_uid'
800 $dupcheck_qpart";
801
802 // if ($_REQUEST["xdebug"]) print "$query\n";
803
804 $result = db_query($link, $query);
805
806 // okay it doesn't exist - create user entry
807 if (db_num_rows($result) == 0) {
808
809 if ($debug_enabled) {
810 _debug("update_rss_feed: user record not found, creating...");
811 }
812
813 if ($score >= -500 && !find_article_filter($article_filters, 'catchup')) {
814 $unread = 'true';
815 $last_read_qpart = 'NULL';
816 } else {
817 $unread = 'false';
818 $last_read_qpart = 'NOW()';
819 }
820
821 if (find_article_filter($article_filters, 'mark') || $score > 1000) {
822 $marked = 'true';
823 } else {
824 $marked = 'false';
825 }
826
827 if (find_article_filter($article_filters, 'publish')) {
828 $published = 'true';
829 } else {
830 $published = 'false';
831 }
832
833 // N-grams
834
835 if (DB_TYPE == "pgsql" and defined('_NGRAM_TITLE_DUPLICATE_THRESHOLD')) {
836
837 $result = db_query($link, "SELECT COUNT(*) AS similar FROM
838 ttrss_entries,ttrss_user_entries
839 WHERE ref_id = id AND updated >= NOW() - INTERVAL '7 day'
840 AND similarity(title, '$entry_title') >= "._NGRAM_TITLE_DUPLICATE_THRESHOLD."
841 AND owner_uid = $owner_uid");
842
843 $ngram_similar = db_fetch_result($result, 0, "similar");
844
845 if ($debug_enabled) {
846 _debug("update_rss_feed: N-gram similar results: $ngram_similar");
847 }
848
849 if ($ngram_similar > 0) {
850 $unread = 'false';
851 }
852 }
853
854 $last_marked = ($marked == 'true') ? 'NOW()' : 'NULL';
855 $last_published = ($published == 'true') ? 'NOW()' : 'NULL';
856
857 $result = db_query($link,
858 "INSERT INTO ttrss_user_entries
859 (ref_id, owner_uid, feed_id, unread, last_read, marked,
860 published, score, tag_cache, label_cache, uuid,
861 last_marked, last_published)
862 VALUES ('$ref_id', '$owner_uid', '$feed', $unread,
863 $last_read_qpart, $marked, $published, '$score', '', '',
864 '', $last_marked, $last_published)");
865
866 if (PUBSUBHUBBUB_HUB && $published == 'true') {
867 $rss_link = get_self_url_prefix() .
868 "/public.php?op=rss&id=-2&key=" .
869 get_feed_access_key($link, -2, false, $owner_uid);
870
871 $p = new Publisher(PUBSUBHUBBUB_HUB);
872
873 $pubsub_result = $p->publish_update($rss_link);
874 }
875
876 $result = db_query($link,
877 "SELECT int_id FROM ttrss_user_entries WHERE
878 ref_id = '$ref_id' AND owner_uid = '$owner_uid' AND
879 feed_id = '$feed' LIMIT 1");
880
881 if (db_num_rows($result) == 1) {
882 $entry_int_id = db_fetch_result($result, 0, "int_id");
883 }
884 } else {
885 if ($debug_enabled) {
886 _debug("update_rss_feed: user record FOUND");
887 }
888
889 $entry_ref_id = db_fetch_result($result, 0, "ref_id");
890 $entry_int_id = db_fetch_result($result, 0, "int_id");
891 }
892
893 if ($debug_enabled) {
894 _debug("update_rss_feed: RID: $entry_ref_id, IID: $entry_int_id");
895 }
896
897 $post_needs_update = false;
898 $update_insignificant = false;
899
900 if ($orig_num_comments != $num_comments) {
901 $post_needs_update = true;
902 $update_insignificant = true;
903 }
904
905 if ($entry_plugin_data != $orig_plugin_data) {
906 $post_needs_update = true;
907 $update_insignificant = true;
908 }
909
910 if ($content_hash != $orig_content_hash) {
911 $post_needs_update = true;
912 $update_insignificant = false;
913 }
914
915 if (db_escape_string($link, $orig_title) != $entry_title) {
916 $post_needs_update = true;
917 $update_insignificant = false;
918 }
919
920 // if post needs update, update it and mark all user entries
921 // linking to this post as updated
922 if ($post_needs_update) {
923
924 if (defined('DAEMON_EXTENDED_DEBUG')) {
925 _debug("update_rss_feed: post $entry_guid needs update...");
926 }
927
928 // print "<!-- post $orig_title needs update : $post_needs_update -->";
929
930 db_query($link, "UPDATE ttrss_entries
931 SET title = '$entry_title', content = '$entry_content',
932 content_hash = '$content_hash',
933 updated = '$entry_timestamp_fmt',
934 num_comments = '$num_comments',
935 plugin_data = '$entry_plugin_data'
936 WHERE id = '$ref_id'");
937
938 if (!$update_insignificant) {
939 if ($mark_unread_on_update) {
940 db_query($link, "UPDATE ttrss_user_entries
941 SET last_read = null, unread = true WHERE ref_id = '$ref_id'");
942 }
943 }
944 }
945 }
946
947 db_query($link, "COMMIT");
948
949 if ($debug_enabled) {
950 _debug("update_rss_feed: assigning labels...");
951 }
952
953 assign_article_to_label_filters($link, $entry_ref_id, $article_filters,
954 $owner_uid, $article_labels);
955
956 if ($debug_enabled) {
957 _debug("update_rss_feed: looking for enclosures...");
958 }
959
960 // enclosures
961
962 $enclosures = array();
963
964 $encs = $item->get_enclosures();
965
966 if (is_array($encs)) {
967 foreach ($encs as $e) {
968 $e_item = array(
969 $e->link, $e->type, $e->length);
970 array_push($enclosures, $e_item);
971 }
972 }
973
974 if ($debug_enabled) {
975 _debug("update_rss_feed: article enclosures:");
976 print_r($enclosures);
977 }
978
979 db_query($link, "BEGIN");
980
981 foreach ($enclosures as $enc) {
982 $enc_url = db_escape_string($link, $enc[0]);
983 $enc_type = db_escape_string($link, $enc[1]);
984 $enc_dur = db_escape_string($link, $enc[2]);
985
986 $result = db_query($link, "SELECT id FROM ttrss_enclosures
987 WHERE content_url = '$enc_url' AND post_id = '$entry_ref_id'");
988
989 if (db_num_rows($result) == 0) {
990 db_query($link, "INSERT INTO ttrss_enclosures
991 (content_url, content_type, title, duration, post_id) VALUES
992 ('$enc_url', '$enc_type', '', '$enc_dur', '$entry_ref_id')");
993 }
994 }
995
996 db_query($link, "COMMIT");
997
998 // check for manual tags (we have to do it here since they're loaded from filters)
999
1000 foreach ($article_filters as $f) {
1001 if ($f["type"] == "tag") {
1002
1003 $manual_tags = trim_array(explode(",", $f["param"]));
1004
1005 foreach ($manual_tags as $tag) {
1006 if (tag_is_valid($tag)) {
1007 array_push($entry_tags, $tag);
1008 }
1009 }
1010 }
1011 }
1012
1013 // Skip boring tags
1014
1015 $boring_tags = trim_array(explode(",", mb_strtolower(get_pref($link,
1016 'BLACKLISTED_TAGS', $owner_uid, ''), 'utf-8')));
1017
1018 $filtered_tags = array();
1019 $tags_to_cache = array();
1020
1021 if ($entry_tags && is_array($entry_tags)) {
1022 foreach ($entry_tags as $tag) {
1023 if (array_search($tag, $boring_tags) === false) {
1024 array_push($filtered_tags, $tag);
1025 }
1026 }
1027 }
1028
1029 $filtered_tags = array_unique($filtered_tags);
1030
1031 if ($debug_enabled) {
1032 _debug("update_rss_feed: filtered article tags:");
1033 print_r($filtered_tags);
1034 }
1035
1036 // Save article tags in the database
1037
1038 if (count($filtered_tags) > 0) {
1039
1040 db_query($link, "BEGIN");
1041
1042 foreach ($filtered_tags as $tag) {
1043
1044 $tag = sanitize_tag($tag);
1045 $tag = db_escape_string($link, $tag);
1046
1047 if (!tag_is_valid($tag)) continue;
1048
1049 $result = db_query($link, "SELECT id FROM ttrss_tags
1050 WHERE tag_name = '$tag' AND post_int_id = '$entry_int_id' AND
1051 owner_uid = '$owner_uid' LIMIT 1");
1052
1053 if ($result && db_num_rows($result) == 0) {
1054
1055 db_query($link, "INSERT INTO ttrss_tags
1056 (owner_uid,tag_name,post_int_id)
1057 VALUES ('$owner_uid','$tag', '$entry_int_id')");
1058 }
1059
1060 array_push($tags_to_cache, $tag);
1061 }
1062
1063 /* update the cache */
1064
1065 $tags_to_cache = array_unique($tags_to_cache);
1066
1067 $tags_str = db_escape_string($link, join(",", $tags_to_cache));
1068
1069 db_query($link, "UPDATE ttrss_user_entries
1070 SET tag_cache = '$tags_str' WHERE ref_id = '$entry_ref_id'
1071 AND owner_uid = $owner_uid");
1072
1073 db_query($link, "COMMIT");
1074 }
1075
1076 if (get_pref($link, "AUTO_ASSIGN_LABELS", $owner_uid, false)) {
1077 if ($debug_enabled) {
1078 _debug("update_rss_feed: auto-assigning labels...");
1079 }
1080
1081 foreach ($labels as $label) {
1082 $caption = preg_quote($label["caption"]);
1083
1084 if ($caption && preg_match("/\b$caption\b/i", "$tags_str " . strip_tags($entry_content) . " $entry_title")) {
1085 if (!labels_contains_caption($article_labels, $caption)) {
1086 label_add_article($link, $entry_ref_id, $caption, $owner_uid);
1087 }
1088 }
1089 }
1090 }
1091
1092 if ($debug_enabled) {
1093 _debug("update_rss_feed: article processed");
1094 }
1095 }
1096
1097 if (!$last_updated) {
1098 if ($debug_enabled) {
1099 _debug("update_rss_feed: new feed, catching it up...");
1100 }
1101 catchup_feed($link, $feed, false, $owner_uid);
1102 }
1103
1104 if ($debug_enabled) {
1105 _debug("purging feed...");
1106 }
1107
1108 purge_feed($link, $feed, 0, $debug_enabled);
1109
1110 db_query($link, "UPDATE ttrss_feeds
1111 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
1112
1113 // db_query($link, "COMMIT");
1114
1115 } else {
1116
1117 $error_msg = db_escape_string($link, mb_substr($rss->error(), 0, 245));
1118
1119 if ($debug_enabled) {
1120 _debug("update_rss_feed: error fetching feed: $error_msg");
1121 }
1122
1123 db_query($link,
1124 "UPDATE ttrss_feeds SET last_error = '$error_msg',
1125 last_updated = NOW() WHERE id = '$feed'");
1126 }
1127
1128 unset($rss);
1129
1130 if ($debug_enabled) {
1131 _debug("update_rss_feed: done");
1132 }
1133
1134 }
1135
1136 function cache_images($html, $site_url, $debug) {
1137 $cache_dir = CACHE_DIR . "/images";
1138
1139 libxml_use_internal_errors(true);
1140
1141 $charset_hack = '<head>
1142 <meta http-equiv="Content-Type" content="text/html; charset=utf-8"/>
1143 </head>';
1144
1145 $doc = new DOMDocument();
1146 $doc->loadHTML($charset_hack . $html);
1147 $xpath = new DOMXPath($doc);
1148
1149 $entries = $xpath->query('(//img[@src])');
1150
1151 foreach ($entries as $entry) {
1152 if ($entry->hasAttribute('src')) {
1153 $src = rewrite_relative_url($site_url, $entry->getAttribute('src'));
1154
1155 $local_filename = CACHE_DIR . "/images/" . sha1($src) . ".png";
1156
1157 if ($debug) _debug("cache_images: downloading: $src to $local_filename");
1158
1159 if (!file_exists($local_filename)) {
1160 $file_content = fetch_file_contents($src);
1161
1162 if ($file_content && strlen($file_content) > 1024) {
1163 file_put_contents($local_filename, $file_content);
1164 }
1165 }
1166
1167 if (file_exists($local_filename)) {
1168 $entry->setAttribute('src', SELF_URL_PATH . '/image.php?url=' .
1169 base64_encode($src));
1170 }
1171 }
1172 }
1173
1174 $node = $doc->getElementsByTagName('body')->item(0);
1175
1176 return $doc->saveXML($node);
1177 }
1178
1179 function expire_lock_files($debug) {
1180 if ($debug) _debug("Removing old lock files...");
1181
1182 $num_deleted = 0;
1183
1184 if (is_writable(LOCK_DIRECTORY)) {
1185 $files = glob(LOCK_DIRECTORY . "/*.lock");
1186
1187 if ($files) {
1188 foreach ($files as $file) {
1189 if (!file_is_locked($file) && time() - filemtime($file) > 86400*2) {
1190 unlink($file);
1191 ++$num_deleted;
1192 }
1193 }
1194 }
1195 }
1196
1197 if ($debug) _debug("Removed $num_deleted files.");
1198 }
1199
1200 function expire_cached_files($debug) {
1201 foreach (array("simplepie", "images", "export", "upload") as $dir) {
1202 $cache_dir = CACHE_DIR . "/$dir";
1203
1204 if ($debug) _debug("Expiring $cache_dir");
1205
1206 $num_deleted = 0;
1207
1208 if (is_writable($cache_dir)) {
1209 $files = glob("$cache_dir/*");
1210
1211 if ($files) {
1212 foreach ($files as $file) {
1213 if (time() - filemtime($file) > 86400*7) {
1214 unlink($file);
1215
1216 ++$num_deleted;
1217 }
1218 }
1219 }
1220 }
1221
1222 if ($debug) _debug("Removed $num_deleted files.");
1223 }
1224 }
1225
1226 /**
1227 * Source: http://www.php.net/manual/en/function.parse-url.php#104527
1228 * Returns the url query as associative array
1229 *
1230 * @param string query
1231 * @return array params
1232 */
1233 function convertUrlQuery($query) {
1234 $queryParts = explode('&', $query);
1235
1236 $params = array();
1237
1238 foreach ($queryParts as $param) {
1239 $item = explode('=', $param);
1240 $params[$item[0]] = $item[1];
1241 }
1242
1243 return $params;
1244 }
1245
1246 function get_article_filters($filters, $title, $content, $link, $timestamp, $author, $tags) {
1247 $matches = array();
1248
1249 foreach ($filters as $filter) {
1250 $match_any_rule = $filter["match_any_rule"];
1251 $inverse = $filter["inverse"];
1252 $filter_match = false;
1253
1254 foreach ($filter["rules"] as $rule) {
1255 $match = false;
1256 $reg_exp = $rule["reg_exp"];
1257 $rule_inverse = $rule["inverse"];
1258
1259 if (!$reg_exp)
1260 continue;
1261
1262 switch ($rule["type"]) {
1263 case "title":
1264 $match = @preg_match("/$reg_exp/i", $title);
1265 break;
1266 case "content":
1267 // we don't need to deal with multiline regexps
1268 $content = preg_replace("/[\r\n\t]/", "", $content);
1269
1270 $match = @preg_match("/$reg_exp/i", $content);
1271 break;
1272 case "both":
1273 // we don't need to deal with multiline regexps
1274 $content = preg_replace("/[\r\n\t]/", "", $content);
1275
1276 $match = (@preg_match("/$reg_exp/i", $title) || @preg_match("/$reg_exp/i", $content));
1277 break;
1278 case "link":
1279 $match = @preg_match("/$reg_exp/i", $link);
1280 break;
1281 case "author":
1282 $match = @preg_match("/$reg_exp/i", $author);
1283 break;
1284 case "tag":
1285 $tag_string = join(",", $tags);
1286 $match = @preg_match("/$reg_exp/i", $tag_string);
1287 break;
1288 }
1289
1290 if ($rule_inverse) $match = !$match;
1291
1292 if ($match_any_rule) {
1293 if ($match) {
1294 $filter_match = true;
1295 break;
1296 }
1297 } else {
1298 $filter_match = $match;
1299 if (!$match) {
1300 break;
1301 }
1302 }
1303 }
1304
1305 if ($inverse) $filter_match = !$filter_match;
1306
1307 if ($filter_match) {
1308 foreach ($filter["actions"] AS $action) {
1309 array_push($matches, $action);
1310
1311 // if Stop action encountered, perform no further processing
1312 if ($action["type"] == "stop") return $matches;
1313 }
1314 }
1315 }
1316
1317 return $matches;
1318 }
1319
1320 function find_article_filter($filters, $filter_name) {
1321 foreach ($filters as $f) {
1322 if ($f["type"] == $filter_name) {
1323 return $f;
1324 };
1325 }
1326 return false;
1327 }
1328
1329 function find_article_filters($filters, $filter_name) {
1330 $results = array();
1331
1332 foreach ($filters as $f) {
1333 if ($f["type"] == $filter_name) {
1334 array_push($results, $f);
1335 };
1336 }
1337 return $results;
1338 }
1339
1340 function calculate_article_score($filters) {
1341 $score = 0;
1342
1343 foreach ($filters as $f) {
1344 if ($f["type"] == "score") {
1345 $score += $f["param"];
1346 };
1347 }
1348 return $score;
1349 }
1350
1351 function labels_contains_caption($labels, $caption) {
1352 foreach ($labels as $label) {
1353 if ($label[1] == $caption) {
1354 return true;
1355 }
1356 }
1357
1358 return false;
1359 }
1360
1361 function assign_article_to_label_filters($link, $id, $filters, $owner_uid, $article_labels) {
1362 foreach ($filters as $f) {
1363 if ($f["type"] == "label") {
1364 if (!labels_contains_caption($article_labels, $f["param"])) {
1365 label_add_article($link, $id, $f["param"], $owner_uid);
1366 }
1367 }
1368 }
1369 }
1370
1371 function make_guid_from_title($title) {
1372 return preg_replace("/[ \"\',.:;]/", "-",
1373 mb_strtolower(strip_tags($title), 'utf-8'));
1374 }
1375
1376
1377 ?>
web-based news feed (RSS/Atom) reader and aggregator; see https://tt-rss.org/