]> git.wh0rd.org - tt-rss.git/blob - include/rssfuncs.php
attempt to recalculate favicon avg color if it changes
[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() {
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 '5 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 5 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 $random_qpart = sql_random_function();
117
118 // We search for feed needing update.
119 $result = db_query("SELECT DISTINCT ttrss_feeds.feed_url,$random_qpart
120 FROM
121 ttrss_feeds, ttrss_users, ttrss_user_prefs
122 WHERE
123 ttrss_feeds.owner_uid = ttrss_users.id
124 AND ttrss_users.id = ttrss_user_prefs.owner_uid
125 AND ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL'
126 $login_thresh_qpart $update_limit_qpart
127 $updstart_thresh_qpart
128 ORDER BY $random_qpart $query_limit");
129
130 if($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(',', $feeds_quoted)));
151 }
152
153 expire_cached_files($debug);
154 expire_lock_files($debug);
155 expire_error_log($debug);
156
157 $nf = 0;
158
159 // For each feed, we call the feed update function.
160 foreach ($feeds_to_update as $feed) {
161 if($debug) _debug("Base feed: $feed");
162
163 //update_rss_feed($line["id"], true);
164
165 // since we have the data cached, we can deal with other feeds with the same url
166
167 $tmp_result = db_query("SELECT DISTINCT ttrss_feeds.id,last_updated
168 FROM ttrss_feeds, ttrss_users, ttrss_user_prefs WHERE
169 ttrss_user_prefs.owner_uid = ttrss_feeds.owner_uid AND
170 ttrss_users.id = ttrss_user_prefs.owner_uid AND
171 ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL' AND
172 feed_url = '".db_escape_string($feed)."' AND
173 (ttrss_feeds.update_interval > 0 OR
174 ttrss_user_prefs.value != '-1')
175 $login_thresh_qpart
176 ORDER BY ttrss_feeds.id $query_limit");
177
178 if (db_num_rows($tmp_result) > 0) {
179 while ($tline = db_fetch_assoc($tmp_result)) {
180 if($debug) _debug(" => " . $tline["last_updated"] . ", " . $tline["id"]);
181 update_rss_feed($tline["id"], true);
182 ++$nf;
183 }
184 }
185 }
186
187 require_once "digest.php";
188
189 // Send feed digests by email if needed.
190 send_headlines_digests($debug);
191
192 return $nf;
193
194 } // function update_daemon_common
195
196 // ignore_daemon is not used
197 function update_rss_feed($feed, $ignore_daemon = false, $no_cache = false,
198 $override_url = false) {
199
200 require_once "lib/simplepie/simplepie.inc";
201
202 $debug_enabled = defined('DAEMON_EXTENDED_DEBUG') || $_REQUEST['xdebug'];
203
204 _debug("start", $debug_enabled);
205
206 $result = db_query("SELECT id,update_interval,auth_login,
207 feed_url,auth_pass,cache_images,last_updated,
208 mark_unread_on_update, owner_uid,
209 pubsub_state, auth_pass_encrypted
210 FROM ttrss_feeds WHERE id = '$feed'");
211
212 if (db_num_rows($result) == 0) {
213 _debug("feed $feed NOT FOUND/SKIPPED", $debug_enabled);
214 return false;
215 }
216
217 $last_updated = db_fetch_result($result, 0, "last_updated");
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 if ($override_url) $fetch_url = $override_url;
242
243 $date_feed_processed = date('Y-m-d H:i');
244
245 $cache_filename = CACHE_DIR . "/simplepie/" . sha1($fetch_url) . ".feed";
246
247 // Ignore cache if new feed or manual update.
248 $cache_age = ($no_cache || is_null($last_updated) || $last_updated == '1970-01-01 00:00:00') ?
249 30 : get_feed_update_interval($feed) * 60;
250
251 _debug("cache filename: $cache_filename exists: " . file_exists($cache_filename), $debug_enabled);
252 _debug("cache age: $cache_age; no cache: $no_cache", $debug_enabled);
253
254 $cached_feed_data_hash = false;
255
256 $rss = false;
257 $rss_hash = false;
258 $cache_timestamp = file_exists($cache_filename) ? filemtime($cache_filename) : 0;
259 $last_updated_timestamp = strtotime($last_updated);
260
261 $force_refetch = isset($_REQUEST["force_refetch"]);
262
263 if (file_exists($cache_filename) &&
264 is_readable($cache_filename) &&
265 !$auth_login && !$auth_pass &&
266 filemtime($cache_filename) > time() - $cache_age) {
267
268 _debug("using local cache.", $debug_enabled);
269
270 if ($cache_timestamp > $last_updated_timestamp) {
271 @$rss_data = file_get_contents($cache_filename);
272
273 if ($rss_data) {
274 $rss_hash = sha1($rss_data);
275 @$rss = unserialize($rss_data);
276 }
277 } else if (!$force_refetch) {
278 _debug("local cache valid and older than last_updated, nothing to do.", $debug_enabled);
279 return;
280 }
281 }
282
283 if (!$rss) {
284
285 if (!$feed_data) {
286 _debug("fetching [$fetch_url] (ts: $cache_timestamp/$last_updated_timestamp)", $debug_enabled);
287
288 $feed_data = fetch_file_contents($fetch_url, false,
289 $auth_login, $auth_pass, false,
290 $no_cache ? FEED_FETCH_NO_CACHE_TIMEOUT : FEED_FETCH_TIMEOUT,
291 $force_refetch ? 0 : max($last_updated_timestamp, $cache_timestamp));
292
293 global $fetch_curl_used;
294
295 if (!$fetch_curl_used) {
296 $tmp = @gzdecode($feed_data);
297
298 if ($tmp) $feed_data = $tmp;
299 }
300
301 $feed_data = trim($feed_data);
302
303 _debug("fetch done.", $debug_enabled);
304
305 if ($feed_data) {
306 $error = verify_feed_xml($feed_data);
307
308 if ($error) {
309 _debug("error verifying XML, code: " . $error->code, $debug_enabled);
310
311 if ($error->code == 26) {
312 _debug("got error 26, trying to decode entities...", $debug_enabled);
313
314 $feed_data = html_entity_decode($feed_data, ENT_COMPAT, 'UTF-8');
315
316 $error = verify_feed_xml($feed_data);
317
318 if ($error) $feed_data = '';
319 }
320 }
321 }
322 }
323
324 if (!$feed_data) {
325 global $fetch_last_error;
326 global $fetch_last_error_code;
327
328 _debug("unable to fetch: $fetch_last_error [$fetch_last_error_code]", $debug_enabled);
329
330 $error_escaped = '';
331
332 // If-Modified-Since
333 if ($fetch_last_error_code != 304) {
334 $error_escaped = db_escape_string($fetch_last_error);
335 } else {
336 _debug("source claims data not modified, nothing to do.", $debug_enabled);
337 }
338
339 db_query(
340 "UPDATE ttrss_feeds SET last_error = '$error_escaped',
341 last_updated = NOW() WHERE id = '$feed'");
342
343 return;
344 }
345 }
346
347 $pluginhost = new PluginHost();
348 $pluginhost->set_debug($debug_enabled, $debug_enabled);
349 $user_plugins = get_pref("_ENABLED_PLUGINS", $owner_uid);
350
351 $pluginhost->load(PLUGINS, PluginHost::KIND_ALL);
352 $pluginhost->load($user_plugins, PluginHost::KIND_USER, $owner_uid);
353 $pluginhost->load_data();
354
355 foreach ($pluginhost->get_hooks(PluginHost::HOOK_FEED_FETCHED) as $plugin) {
356 $feed_data = $plugin->hook_feed_fetched($feed_data);
357 }
358
359 if (!$rss) {
360 $rss = new SimplePie();
361 $rss->set_sanitize_class("SanitizeDummy");
362 // simplepie ignores the above and creates default sanitizer anyway,
363 // so let's override it...
364 $rss->sanitize = new SanitizeDummy();
365 $rss->set_output_encoding('UTF-8');
366 $rss->set_raw_data($feed_data);
367 $rss->enable_cache(false);
368
369 @$rss->init();
370 }
371
372 // print_r($rss);
373
374 $feed = db_escape_string($feed);
375
376 if (!$rss->error()) {
377
378 // cache data for later
379 if (!$auth_pass && !$auth_login && is_writable(CACHE_DIR . "/simplepie")) {
380 $rss_data = serialize($rss);
381 $new_rss_hash = sha1($rss_data);
382
383 if ($new_rss_hash != $rss_hash) {
384 _debug("saving $cache_filename", $debug_enabled);
385 @file_put_contents($cache_filename, serialize($rss));
386 }
387 }
388
389 // We use local pluginhost here because we need to load different per-user feed plugins
390 $pluginhost->run_hooks(PluginHost::HOOK_FEED_PARSED, "hook_feed_parsed", $rss);
391
392 _debug("processing feed data...", $debug_enabled);
393
394 // db_query("BEGIN");
395
396 if (DB_TYPE == "pgsql") {
397 $favicon_interval_qpart = "favicon_last_checked < NOW() - INTERVAL '12 hour'";
398 } else {
399 $favicon_interval_qpart = "favicon_last_checked < DATE_SUB(NOW(), INTERVAL 12 HOUR)";
400 }
401
402 $result = db_query("SELECT title,site_url,owner_uid,favicon_avg_color,
403 (favicon_last_checked IS NULL OR $favicon_interval_qpart) AS
404 favicon_needs_check
405 FROM ttrss_feeds WHERE id = '$feed'");
406
407 $registered_title = db_fetch_result($result, 0, "title");
408 $orig_site_url = db_fetch_result($result, 0, "site_url");
409 $favicon_needs_check = sql_bool_to_bool(db_fetch_result($result, 0,
410 "favicon_needs_check"));
411 $favicon_avg_color = db_fetch_result($result, 0, "favicon_avg_color");
412
413 $owner_uid = db_fetch_result($result, 0, "owner_uid");
414
415 $site_url = db_escape_string(mb_substr(rewrite_relative_url($fetch_url, $rss->get_link()), 0, 245));
416
417 if ($favicon_needs_check || $force_refetch) {
418
419 /* terrible hack: if we crash on floicon shit here, we won't check
420 * the icon avgcolor again (unless the icon got updated) */
421
422 db_query("UPDATE ttrss_feeds SET favicon_avg_color = 'fail' WHERE
423 id = '$feed'");
424
425 $favicon_file = ICONS_DIR . "/$feed.ico";
426 $favicon_modified = @filemtime($favicon_file);
427
428 _debug("checking favicon...", $debug_enabled);
429
430 check_feed_favicon($site_url, $feed, $link);
431 $favicon_modified_new = @filemtime($favicon_file);
432
433 if ($favicon_modified_new > $favicon_modified)
434 $favicon_avg_color = '';
435
436 if (file_exists($favicon_file) && function_exists("imagecreatefromstring") && $favicon_avg_color != 'fail') {
437 require_once "colors.php";
438
439 $favicon_color = db_escape_string(
440 calculate_avg_color($favicon_file));
441
442 $favicon_colorstring = ",favicon_avg_color = '".$favicon_color."'";
443 } else if ($favicon_avg_color == 'fail') {
444 _debug("floicon failed on this file, not trying to recalculate avg color");
445 }
446
447 db_query("UPDATE ttrss_feeds SET favicon_last_checked = NOW()
448 $favicon_colorstring
449 WHERE id = '$feed'");
450 }
451
452 if (!$registered_title || $registered_title == "[Unknown]") {
453
454 $feed_title = db_escape_string($rss->get_title());
455
456 _debug("registering title: $feed_title", $debug_enabled);
457
458 db_query("UPDATE ttrss_feeds SET
459 title = '$feed_title' WHERE id = '$feed'");
460 }
461
462 if ($site_url && $orig_site_url != $site_url) {
463 db_query("UPDATE ttrss_feeds SET
464 site_url = '$site_url' WHERE id = '$feed'");
465 }
466
467 _debug("loading filters & labels...", $debug_enabled);
468
469 $filters = load_filters($feed, $owner_uid);
470 $labels = get_all_labels($owner_uid);
471
472 _debug("" . count($filters) . " filters loaded.", $debug_enabled);
473
474 $items = $rss->get_items();
475
476 if (!is_array($items)) {
477 _debug("no articles found.", $debug_enabled);
478
479 db_query("UPDATE ttrss_feeds
480 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
481
482 return; // no articles
483 }
484
485 if ($pubsub_state != 2 && PUBSUBHUBBUB_ENABLED) {
486
487 _debug("checking for PUSH hub...", $debug_enabled);
488
489 $feed_hub_url = false;
490
491 $links = $rss->get_links('hub');
492
493 if ($links && is_array($links)) {
494 foreach ($links as $l) {
495 $feed_hub_url = $l;
496 break;
497 }
498 }
499
500 _debug("feed hub url: $feed_hub_url", $debug_enabled);
501
502 if ($feed_hub_url && function_exists('curl_init') &&
503 !ini_get("open_basedir")) {
504
505 require_once 'lib/pubsubhubbub/subscriber.php';
506
507 $callback_url = get_self_url_prefix() .
508 "/public.php?op=pubsub&id=$feed";
509
510 $s = new Subscriber($feed_hub_url, $callback_url);
511
512 $rc = $s->subscribe($fetch_url);
513
514 _debug("feed hub url found, subscribe request sent.", $debug_enabled);
515
516 db_query("UPDATE ttrss_feeds SET pubsub_state = 1
517 WHERE id = '$feed'");
518 }
519 }
520
521 _debug("processing articles...", $debug_enabled);
522
523 foreach ($items as $item) {
524 if ($_REQUEST['xdebug'] == 3) {
525 print_r($item);
526 }
527
528 $entry_guid = $item->get_id();
529 if (!$entry_guid) $entry_guid = $item->get_link();
530 if (!$entry_guid) $entry_guid = make_guid_from_title($item->get_title());
531
532 if (!$entry_guid) continue;
533
534 $entry_guid = "$owner_uid,$entry_guid";
535
536 $entry_guid_hashed = db_escape_string('SHA1:' . sha1($entry_guid));
537
538 _debug("guid $entry_guid / $entry_guid_hashed", $debug_enabled);
539
540 $entry_timestamp = "";
541
542 $entry_timestamp = strtotime($item->get_date());
543
544 if ($entry_timestamp == -1 || !$entry_timestamp || $entry_timestamp > time()) {
545 $entry_timestamp = time();
546 $no_orig_date = 'true';
547 } else {
548 $no_orig_date = 'false';
549 }
550
551 $entry_timestamp_fmt = strftime("%Y/%m/%d %H:%M:%S", $entry_timestamp);
552
553 _debug("date $entry_timestamp [$entry_timestamp_fmt]", $debug_enabled);
554
555 $entry_title = html_entity_decode($item->get_title(), ENT_COMPAT, 'UTF-8');
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_comments = $item->data["comments"];
574
575 if ($item->get_author()) {
576 $entry_author_item = $item->get_author();
577 $entry_author = $entry_author_item->get_name();
578 if (!$entry_author) $entry_author = $entry_author_item->get_email();
579 }
580
581 $entry_guid = db_escape_string(mb_substr($entry_guid, 0, 245));
582
583 $entry_comments = db_escape_string(mb_substr(trim($entry_comments), 0, 245));
584 $entry_author = db_escape_string(mb_substr(trim($entry_author), 0, 245));
585
586 $num_comments = $item->get_item_tags('http://purl.org/rss/1.0/modules/slash/', 'comments');
587
588 if (is_array($num_comments) && is_array($num_comments[0])) {
589 $num_comments = (int) $num_comments[0]["data"];
590 } else {
591 $num_comments = 0;
592 }
593
594 _debug("author $entry_author", $debug_enabled);
595 _debug("num_comments: $num_comments", $debug_enabled);
596 _debug("looking for tags [1]...", $debug_enabled);
597
598 // parse <category> entries into tags
599
600 $additional_tags = array();
601
602 $additional_tags_src = $item->get_categories();
603
604 if (is_array($additional_tags_src)) {
605 foreach ($additional_tags_src as $tobj) {
606 array_push($additional_tags, $tobj->get_term());
607 }
608 }
609
610 _debug("category tags:", $debug_enabled);
611 _debug("looking for tags [2]...", $debug_enabled);
612
613 $entry_tags = array_unique($additional_tags);
614
615 for ($i = 0; $i < count($entry_tags); $i++)
616 $entry_tags[$i] = mb_strtolower($entry_tags[$i], 'utf-8');
617
618 _debug("done collecting data.", $debug_enabled);
619
620 // TODO: less memory-hungry implementation
621
622 _debug("applying plugin filters..", $debug_enabled);
623
624 // FIXME not sure if owner_uid is a good idea here, we may have a base entry without user entry (?)
625 $result = db_query("SELECT plugin_data,title,content,link,tag_cache,author FROM ttrss_entries, ttrss_user_entries
626 WHERE ref_id = id AND (guid = '".db_escape_string($entry_guid)."' OR guid = '$entry_guid_hashed') AND owner_uid = $owner_uid");
627
628 if (db_num_rows($result) != 0) {
629 $entry_plugin_data = db_fetch_result($result, 0, "plugin_data");
630 $stored_article = array("title" => db_fetch_result($result, 0, "title"),
631 "content" => db_fetch_result($result, 0, "content"),
632 "link" => db_fetch_result($result, 0, "link"),
633 "tags" => explode(",", db_fetch_result($result, 0, "tag_cache")),
634 "author" => db_fetch_result($result, 0, "author"));
635 } else {
636 $entry_plugin_data = "";
637 $stored_article = array();
638 }
639
640 $article = array("owner_uid" => $owner_uid, // read only
641 "guid" => $entry_guid, // read only
642 "title" => $entry_title,
643 "content" => $entry_content,
644 "link" => $entry_link,
645 "tags" => $entry_tags,
646 "plugin_data" => $entry_plugin_data,
647 "author" => $entry_author,
648 "stored" => $stored_article);
649
650 foreach ($pluginhost->get_hooks(PluginHost::HOOK_ARTICLE_FILTER) as $plugin) {
651 $article = $plugin->hook_article_filter($article);
652 }
653
654 $entry_tags = $article["tags"];
655 $entry_guid = db_escape_string($entry_guid);
656 $entry_title = db_escape_string($article["title"]);
657 $entry_author = db_escape_string($article["author"]);
658 $entry_link = db_escape_string($article["link"]);
659 $entry_plugin_data = db_escape_string($article["plugin_data"]);
660 $entry_content = $article["content"]; // escaped below
661
662
663 _debug("plugin data: $entry_plugin_data", $debug_enabled);
664
665 if ($cache_images && is_writable(CACHE_DIR . '/images'))
666 cache_images($entry_content, $site_url, $debug_enabled);
667
668 $entry_content = db_escape_string($entry_content, false);
669
670 $content_hash = "SHA1:" . sha1($entry_content);
671
672 db_query("BEGIN");
673
674 $result = db_query("SELECT id FROM ttrss_entries
675 WHERE (guid = '$entry_guid' OR guid = '$entry_guid_hashed')");
676
677 if (db_num_rows($result) == 0) {
678
679 _debug("base guid [$entry_guid] not found", $debug_enabled);
680
681 // base post entry does not exist, create it
682
683 $result = db_query(
684 "INSERT INTO ttrss_entries
685 (title,
686 guid,
687 link,
688 updated,
689 content,
690 content_hash,
691 cached_content,
692 no_orig_date,
693 date_updated,
694 date_entered,
695 comments,
696 num_comments,
697 plugin_data,
698 author)
699 VALUES
700 ('$entry_title',
701 '$entry_guid_hashed',
702 '$entry_link',
703 '$entry_timestamp_fmt',
704 '$entry_content',
705 '$content_hash',
706 '',
707 $no_orig_date,
708 NOW(),
709 '$date_feed_processed',
710 '$entry_comments',
711 '$num_comments',
712 '$entry_plugin_data',
713 '$entry_author')");
714
715 $article_labels = array();
716
717 } else {
718 // we keep encountering the entry in feeds, so we need to
719 // update date_updated column so that we don't get horrible
720 // dupes when the entry gets purged and reinserted again e.g.
721 // in the case of SLOW SLOW OMG SLOW updating feeds
722
723 $base_entry_id = db_fetch_result($result, 0, "id");
724
725 db_query("UPDATE ttrss_entries SET date_updated = NOW()
726 WHERE id = '$base_entry_id'");
727
728 $article_labels = get_article_labels($base_entry_id, $owner_uid);
729 }
730
731 // now it should exist, if not - bad luck then
732
733 $result = db_query("SELECT
734 id,content_hash,no_orig_date,title,plugin_data,guid,
735 ".SUBSTRING_FOR_DATE."(date_updated,1,19) as date_updated,
736 ".SUBSTRING_FOR_DATE."(updated,1,19) as updated,
737 num_comments
738 FROM
739 ttrss_entries
740 WHERE guid = '$entry_guid' OR guid = '$entry_guid_hashed'");
741
742 $entry_ref_id = 0;
743 $entry_int_id = 0;
744
745 if (db_num_rows($result) == 1) {
746
747 _debug("base guid found, checking for user record", $debug_enabled);
748
749 // this will be used below in update handler
750 $orig_content_hash = db_fetch_result($result, 0, "content_hash");
751 $orig_title = db_fetch_result($result, 0, "title");
752 $orig_num_comments = db_fetch_result($result, 0, "num_comments");
753 $orig_date_updated = strtotime(db_fetch_result($result,
754 0, "date_updated"));
755 $orig_plugin_data = db_fetch_result($result, 0, "plugin_data");
756
757 $ref_id = db_fetch_result($result, 0, "id");
758 $entry_ref_id = $ref_id;
759
760 /* $stored_guid = db_fetch_result($result, 0, "guid");
761 if ($stored_guid != $entry_guid_hashed) {
762 if ($debug_enabled) _debug("upgrading compat guid to hashed one", $debug_enabled);
763
764 db_query("UPDATE ttrss_entries SET guid = '$entry_guid_hashed' WHERE
765 id = '$ref_id'");
766 } */
767
768 // check for user post link to main table
769
770 // do we allow duplicate posts with same GUID in different feeds?
771 if (get_pref("ALLOW_DUPLICATE_POSTS", $owner_uid, false)) {
772 $dupcheck_qpart = "AND (feed_id = '$feed' OR feed_id IS NULL)";
773 } else {
774 $dupcheck_qpart = "";
775 }
776
777 /* Collect article tags here so we could filter by them: */
778
779 $article_filters = get_article_filters($filters, $entry_title,
780 $entry_content, $entry_link, $entry_timestamp, $entry_author,
781 $entry_tags);
782
783 if ($debug_enabled) {
784 _debug("article filters: ", $debug_enabled);
785 if (count($article_filters) != 0) {
786 print_r($article_filters);
787 }
788 }
789
790 if (find_article_filter($article_filters, "filter")) {
791 db_query("COMMIT"); // close transaction in progress
792 continue;
793 }
794
795 $score = calculate_article_score($article_filters);
796
797 _debug("initial score: $score", $debug_enabled);
798
799 $query = "SELECT ref_id, int_id FROM ttrss_user_entries WHERE
800 ref_id = '$ref_id' AND owner_uid = '$owner_uid'
801 $dupcheck_qpart";
802
803 // if ($_REQUEST["xdebug"]) print "$query\n";
804
805 $result = db_query($query);
806
807 // okay it doesn't exist - create user entry
808 if (db_num_rows($result) == 0) {
809
810 _debug("user record not found, creating...", $debug_enabled);
811
812 if ($score >= -500 && !find_article_filter($article_filters, 'catchup')) {
813 $unread = 'true';
814 $last_read_qpart = 'NULL';
815 } else {
816 $unread = 'false';
817 $last_read_qpart = 'NOW()';
818 }
819
820 if (find_article_filter($article_filters, 'mark') || $score > 1000) {
821 $marked = 'true';
822 } else {
823 $marked = 'false';
824 }
825
826 if (find_article_filter($article_filters, 'publish')) {
827 $published = 'true';
828 } else {
829 $published = 'false';
830 }
831
832 // N-grams
833
834 if (DB_TYPE == "pgsql" and defined('_NGRAM_TITLE_DUPLICATE_THRESHOLD')) {
835
836 $result = db_query("SELECT COUNT(*) AS similar FROM
837 ttrss_entries,ttrss_user_entries
838 WHERE ref_id = id AND updated >= NOW() - INTERVAL '7 day'
839 AND similarity(title, '$entry_title') >= "._NGRAM_TITLE_DUPLICATE_THRESHOLD."
840 AND owner_uid = $owner_uid");
841
842 $ngram_similar = db_fetch_result($result, 0, "similar");
843
844 _debug("N-gram similar results: $ngram_similar", $debug_enabled);
845
846 if ($ngram_similar > 0) {
847 $unread = 'false';
848 }
849 }
850
851 $last_marked = ($marked == 'true') ? 'NOW()' : 'NULL';
852 $last_published = ($published == 'true') ? 'NOW()' : 'NULL';
853
854 $result = db_query(
855 "INSERT INTO ttrss_user_entries
856 (ref_id, owner_uid, feed_id, unread, last_read, marked,
857 published, score, tag_cache, label_cache, uuid,
858 last_marked, last_published)
859 VALUES ('$ref_id', '$owner_uid', '$feed', $unread,
860 $last_read_qpart, $marked, $published, '$score', '', '',
861 '', $last_marked, $last_published)");
862
863 if (PUBSUBHUBBUB_HUB && $published == 'true') {
864 $rss_link = get_self_url_prefix() .
865 "/public.php?op=rss&id=-2&key=" .
866 get_feed_access_key(-2, false, $owner_uid);
867
868 $p = new Publisher(PUBSUBHUBBUB_HUB);
869
870 $pubsub_result = $p->publish_update($rss_link);
871 }
872
873 $result = db_query(
874 "SELECT int_id FROM ttrss_user_entries WHERE
875 ref_id = '$ref_id' AND owner_uid = '$owner_uid' AND
876 feed_id = '$feed' LIMIT 1");
877
878 if (db_num_rows($result) == 1) {
879 $entry_int_id = db_fetch_result($result, 0, "int_id");
880 }
881 } else {
882 _debug("user record FOUND", $debug_enabled);
883
884 $entry_ref_id = db_fetch_result($result, 0, "ref_id");
885 $entry_int_id = db_fetch_result($result, 0, "int_id");
886 }
887
888 _debug("RID: $entry_ref_id, IID: $entry_int_id", $debug_enabled);
889
890 $post_needs_update = false;
891 $update_insignificant = false;
892
893 if ($orig_num_comments != $num_comments) {
894 $post_needs_update = true;
895 $update_insignificant = true;
896 }
897
898 if ($entry_plugin_data != $orig_plugin_data) {
899 $post_needs_update = true;
900 $update_insignificant = true;
901 }
902
903 if ($content_hash != $orig_content_hash) {
904 $post_needs_update = true;
905 $update_insignificant = false;
906 }
907
908 if (db_escape_string($orig_title) != $entry_title) {
909 $post_needs_update = true;
910 $update_insignificant = false;
911 }
912
913 // if post needs update, update it and mark all user entries
914 // linking to this post as updated
915 if ($post_needs_update) {
916
917 if (defined('DAEMON_EXTENDED_DEBUG')) {
918 _debug("post $entry_guid_hashed needs update...", $debug_enabled);
919 }
920
921 // print "<!-- post $orig_title needs update : $post_needs_update -->";
922
923 db_query("UPDATE ttrss_entries
924 SET title = '$entry_title', content = '$entry_content',
925 content_hash = '$content_hash',
926 updated = '$entry_timestamp_fmt',
927 num_comments = '$num_comments',
928 plugin_data = '$entry_plugin_data'
929 WHERE id = '$ref_id'");
930
931 if (!$update_insignificant) {
932 if ($mark_unread_on_update) {
933 db_query("UPDATE ttrss_user_entries
934 SET last_read = null, unread = true WHERE ref_id = '$ref_id'");
935 }
936 }
937 }
938 }
939
940 db_query("COMMIT");
941
942 _debug("assigning labels...", $debug_enabled);
943
944 assign_article_to_label_filters($entry_ref_id, $article_filters,
945 $owner_uid, $article_labels);
946
947 _debug("looking for enclosures...", $debug_enabled);
948
949 // enclosures
950
951 $enclosures = array();
952
953 $encs = $item->get_enclosures();
954
955 if (is_array($encs)) {
956 foreach ($encs as $e) {
957 $e_item = array(
958 $e->link, $e->type, $e->length);
959 array_push($enclosures, $e_item);
960 }
961 }
962
963 if ($debug_enabled) {
964 _debug("article enclosures:", $debug_enabled);
965 print_r($enclosures);
966 }
967
968 db_query("BEGIN");
969
970 foreach ($enclosures as $enc) {
971 $enc_url = db_escape_string($enc[0]);
972 $enc_type = db_escape_string($enc[1]);
973 $enc_dur = db_escape_string($enc[2]);
974
975 $result = db_query("SELECT id FROM ttrss_enclosures
976 WHERE content_url = '$enc_url' AND post_id = '$entry_ref_id'");
977
978 if (db_num_rows($result) == 0) {
979 db_query("INSERT INTO ttrss_enclosures
980 (content_url, content_type, title, duration, post_id) VALUES
981 ('$enc_url', '$enc_type', '', '$enc_dur', '$entry_ref_id')");
982 }
983 }
984
985 db_query("COMMIT");
986
987 // check for manual tags (we have to do it here since they're loaded from filters)
988
989 foreach ($article_filters as $f) {
990 if ($f["type"] == "tag") {
991
992 $manual_tags = trim_array(explode(",", $f["param"]));
993
994 foreach ($manual_tags as $tag) {
995 if (tag_is_valid($tag)) {
996 array_push($entry_tags, $tag);
997 }
998 }
999 }
1000 }
1001
1002 // Skip boring tags
1003
1004 $boring_tags = trim_array(explode(",", mb_strtolower(get_pref(
1005 'BLACKLISTED_TAGS', $owner_uid, ''), 'utf-8')));
1006
1007 $filtered_tags = array();
1008 $tags_to_cache = array();
1009
1010 if ($entry_tags && is_array($entry_tags)) {
1011 foreach ($entry_tags as $tag) {
1012 if (array_search($tag, $boring_tags) === false) {
1013 array_push($filtered_tags, $tag);
1014 }
1015 }
1016 }
1017
1018 $filtered_tags = array_unique($filtered_tags);
1019
1020 if ($debug_enabled) {
1021 _debug("filtered article tags:", $debug_enabled);
1022 print_r($filtered_tags);
1023 }
1024
1025 // Save article tags in the database
1026
1027 if (count($filtered_tags) > 0) {
1028
1029 db_query("BEGIN");
1030
1031 foreach ($filtered_tags as $tag) {
1032
1033 $tag = sanitize_tag($tag);
1034 $tag = db_escape_string($tag);
1035
1036 if (!tag_is_valid($tag)) continue;
1037
1038 $result = db_query("SELECT id FROM ttrss_tags
1039 WHERE tag_name = '$tag' AND post_int_id = '$entry_int_id' AND
1040 owner_uid = '$owner_uid' LIMIT 1");
1041
1042 if ($result && db_num_rows($result) == 0) {
1043
1044 db_query("INSERT INTO ttrss_tags
1045 (owner_uid,tag_name,post_int_id)
1046 VALUES ('$owner_uid','$tag', '$entry_int_id')");
1047 }
1048
1049 array_push($tags_to_cache, $tag);
1050 }
1051
1052 /* update the cache */
1053
1054 $tags_to_cache = array_unique($tags_to_cache);
1055
1056 $tags_str = db_escape_string(join(",", $tags_to_cache));
1057
1058 db_query("UPDATE ttrss_user_entries
1059 SET tag_cache = '$tags_str' WHERE ref_id = '$entry_ref_id'
1060 AND owner_uid = $owner_uid");
1061
1062 db_query("COMMIT");
1063 }
1064
1065 if (get_pref("AUTO_ASSIGN_LABELS", $owner_uid, false)) {
1066 _debug("auto-assigning labels...", $debug_enabled);
1067
1068 foreach ($labels as $label) {
1069 $caption = preg_quote($label["caption"]);
1070
1071 if ($caption && preg_match("/\b$caption\b/i", "$tags_str " . strip_tags($entry_content) . " $entry_title")) {
1072 if (!labels_contains_caption($article_labels, $caption)) {
1073 label_add_article($entry_ref_id, $caption, $owner_uid);
1074 }
1075 }
1076 }
1077 }
1078
1079 _debug("article processed", $debug_enabled);
1080 }
1081
1082 if (!$last_updated) {
1083 _debug("new feed, catching it up...", $debug_enabled);
1084 catchup_feed($feed, false, $owner_uid);
1085 }
1086
1087 _debug("purging feed...", $debug_enabled);
1088
1089 purge_feed($feed, 0, $debug_enabled);
1090
1091 db_query("UPDATE ttrss_feeds
1092 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
1093
1094 // db_query("COMMIT");
1095
1096 } else {
1097
1098 $error_msg = db_escape_string(mb_substr($rss->error(), 0, 245));
1099
1100 _debug("error fetching feed: $error_msg", $debug_enabled);
1101
1102 db_query(
1103 "UPDATE ttrss_feeds SET last_error = '$error_msg',
1104 last_updated = NOW() WHERE id = '$feed'");
1105 }
1106
1107 unset($rss);
1108
1109 _debug("done", $debug_enabled);
1110 }
1111
1112 function cache_images($html, $site_url, $debug) {
1113 $cache_dir = CACHE_DIR . "/images";
1114
1115 libxml_use_internal_errors(true);
1116
1117 $charset_hack = '<head>
1118 <meta http-equiv="Content-Type" content="text/html; charset=utf-8"/>
1119 </head>';
1120
1121 $doc = new DOMDocument();
1122 $doc->loadHTML($charset_hack . $html);
1123 $xpath = new DOMXPath($doc);
1124
1125 $entries = $xpath->query('(//img[@src])');
1126
1127 foreach ($entries as $entry) {
1128 if ($entry->hasAttribute('src')) {
1129 $src = rewrite_relative_url($site_url, $entry->getAttribute('src'));
1130
1131 $local_filename = CACHE_DIR . "/images/" . sha1($src) . ".png";
1132
1133 if ($debug) _debug("cache_images: downloading: $src to $local_filename");
1134
1135 if (!file_exists($local_filename)) {
1136 $file_content = fetch_file_contents($src);
1137
1138 if ($file_content && strlen($file_content) > 1024) {
1139 file_put_contents($local_filename, $file_content);
1140 }
1141 }
1142
1143 if (file_exists($local_filename)) {
1144 $entry->setAttribute('src', SELF_URL_PATH . '/image.php?url=' .
1145 base64_encode($src));
1146 }
1147 }
1148 }
1149
1150 $node = $doc->getElementsByTagName('body')->item(0);
1151
1152 return $doc->saveXML($node);
1153 }
1154
1155 function expire_error_log($debug) {
1156 if ($debug) _debug("Removing old error log entries...");
1157
1158 if (DB_TYPE == "pgsql") {
1159 db_query("DELETE FROM ttrss_error_log
1160 WHERE created_at < NOW() - INTERVAL '7 days'");
1161 } else {
1162 db_query("DELETE FROM ttrss_error_log
1163 WHERE created_at < DATE_SUB(NOW(), INTERVAL 7 DAY)");
1164 }
1165
1166 }
1167
1168 function expire_lock_files($debug) {
1169 if ($debug) _debug("Removing old lock files...");
1170
1171 $num_deleted = 0;
1172
1173 if (is_writable(LOCK_DIRECTORY)) {
1174 $files = glob(LOCK_DIRECTORY . "/*.lock");
1175
1176 if ($files) {
1177 foreach ($files as $file) {
1178 if (!file_is_locked(basename($file)) && time() - filemtime($file) > 86400*2) {
1179 unlink($file);
1180 ++$num_deleted;
1181 }
1182 }
1183 }
1184 }
1185
1186 if ($debug) _debug("Removed $num_deleted files.");
1187 }
1188
1189 function expire_cached_files($debug) {
1190 foreach (array("simplepie", "images", "export", "upload") as $dir) {
1191 $cache_dir = CACHE_DIR . "/$dir";
1192
1193 if ($debug) _debug("Expiring $cache_dir");
1194
1195 $num_deleted = 0;
1196
1197 if (is_writable($cache_dir)) {
1198 $files = glob("$cache_dir/*");
1199
1200 if ($files) {
1201 foreach ($files as $file) {
1202 if (time() - filemtime($file) > 86400*7) {
1203 unlink($file);
1204
1205 ++$num_deleted;
1206 }
1207 }
1208 }
1209 }
1210
1211 if ($debug) _debug("Removed $num_deleted files.");
1212 }
1213 }
1214
1215 /**
1216 * Source: http://www.php.net/manual/en/function.parse-url.php#104527
1217 * Returns the url query as associative array
1218 *
1219 * @param string query
1220 * @return array params
1221 */
1222 function convertUrlQuery($query) {
1223 $queryParts = explode('&', $query);
1224
1225 $params = array();
1226
1227 foreach ($queryParts as $param) {
1228 $item = explode('=', $param);
1229 $params[$item[0]] = $item[1];
1230 }
1231
1232 return $params;
1233 }
1234
1235 function get_article_filters($filters, $title, $content, $link, $timestamp, $author, $tags) {
1236 $matches = array();
1237
1238 foreach ($filters as $filter) {
1239 $match_any_rule = $filter["match_any_rule"];
1240 $inverse = $filter["inverse"];
1241 $filter_match = false;
1242
1243 foreach ($filter["rules"] as $rule) {
1244 $match = false;
1245 $reg_exp = $rule["reg_exp"];
1246 $rule_inverse = $rule["inverse"];
1247
1248 if (!$reg_exp)
1249 continue;
1250
1251 switch ($rule["type"]) {
1252 case "title":
1253 $match = @preg_match("/$reg_exp/i", $title);
1254 break;
1255 case "content":
1256 // we don't need to deal with multiline regexps
1257 $content = preg_replace("/[\r\n\t]/", "", $content);
1258
1259 $match = @preg_match("/$reg_exp/i", $content);
1260 break;
1261 case "both":
1262 // we don't need to deal with multiline regexps
1263 $content = preg_replace("/[\r\n\t]/", "", $content);
1264
1265 $match = (@preg_match("/$reg_exp/i", $title) || @preg_match("/$reg_exp/i", $content));
1266 break;
1267 case "link":
1268 $match = @preg_match("/$reg_exp/i", $link);
1269 break;
1270 case "author":
1271 $match = @preg_match("/$reg_exp/i", $author);
1272 break;
1273 case "tag":
1274 $tag_string = join(",", $tags);
1275 $match = @preg_match("/$reg_exp/i", $tag_string);
1276 break;
1277 }
1278
1279 if ($rule_inverse) $match = !$match;
1280
1281 if ($match_any_rule) {
1282 if ($match) {
1283 $filter_match = true;
1284 break;
1285 }
1286 } else {
1287 $filter_match = $match;
1288 if (!$match) {
1289 break;
1290 }
1291 }
1292 }
1293
1294 if ($inverse) $filter_match = !$filter_match;
1295
1296 if ($filter_match) {
1297 foreach ($filter["actions"] AS $action) {
1298 array_push($matches, $action);
1299
1300 // if Stop action encountered, perform no further processing
1301 if ($action["type"] == "stop") return $matches;
1302 }
1303 }
1304 }
1305
1306 return $matches;
1307 }
1308
1309 function find_article_filter($filters, $filter_name) {
1310 foreach ($filters as $f) {
1311 if ($f["type"] == $filter_name) {
1312 return $f;
1313 };
1314 }
1315 return false;
1316 }
1317
1318 function find_article_filters($filters, $filter_name) {
1319 $results = array();
1320
1321 foreach ($filters as $f) {
1322 if ($f["type"] == $filter_name) {
1323 array_push($results, $f);
1324 };
1325 }
1326 return $results;
1327 }
1328
1329 function calculate_article_score($filters) {
1330 $score = 0;
1331
1332 foreach ($filters as $f) {
1333 if ($f["type"] == "score") {
1334 $score += $f["param"];
1335 };
1336 }
1337 return $score;
1338 }
1339
1340 function labels_contains_caption($labels, $caption) {
1341 foreach ($labels as $label) {
1342 if ($label[1] == $caption) {
1343 return true;
1344 }
1345 }
1346
1347 return false;
1348 }
1349
1350 function assign_article_to_label_filters($id, $filters, $owner_uid, $article_labels) {
1351 foreach ($filters as $f) {
1352 if ($f["type"] == "label") {
1353 if (!labels_contains_caption($article_labels, $f["param"])) {
1354 label_add_article($id, $f["param"], $owner_uid);
1355 }
1356 }
1357 }
1358 }
1359
1360 function make_guid_from_title($title) {
1361 return preg_replace("/[ \"\',.:;]/", "-",
1362 mb_strtolower(strip_tags($title), 'utf-8'));
1363 }
1364
1365 function verify_feed_xml($feed_data) {
1366 libxml_use_internal_errors(true);
1367 $doc = new DOMDocument();
1368 $doc->loadXML($feed_data);
1369 $error = libxml_get_last_error();
1370 libxml_clear_errors();
1371 return $error;
1372 }
1373
1374 ?>