]> 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
sanitize article content when importing data from feed
[tt-rss.git] / include / rssfuncs.php
1 <?php
2 define('DAEMON_UPDATE_LOGIN_LIMIT', 30);
3 define('DAEMON_FEED_LIMIT', 100);
4 define('DAEMON_SLEEP_INTERVAL', 60);
5
6 function update_feedbrowser_cache($link) {
7
8 $result = db_query($link, "SELECT feed_url, site_url, title, COUNT(id) AS subscribers
9 FROM ttrss_feeds WHERE (SELECT COUNT(id) = 0 FROM ttrss_feeds AS tf
10 WHERE tf.feed_url = ttrss_feeds.feed_url
11 AND (private IS true OR auth_login != '' OR auth_pass != '' OR feed_url LIKE '%:%@%/%'))
12 GROUP BY feed_url, site_url, title ORDER BY subscribers DESC LIMIT 1000");
13
14 db_query($link, "BEGIN");
15
16 db_query($link, "DELETE FROM ttrss_feedbrowser_cache");
17
18 $count = 0;
19
20 while ($line = db_fetch_assoc($result)) {
21 $subscribers = db_escape_string($line["subscribers"]);
22 $feed_url = db_escape_string($line["feed_url"]);
23 $title = db_escape_string($line["title"]);
24 $site_url = db_escape_string($line["site_url"]);
25
26 $tmp_result = db_query($link, "SELECT subscribers FROM
27 ttrss_feedbrowser_cache WHERE feed_url = '$feed_url'");
28
29 if (db_num_rows($tmp_result) == 0) {
30
31 db_query($link, "INSERT INTO ttrss_feedbrowser_cache
32 (feed_url, site_url, title, subscribers) VALUES ('$feed_url',
33 '$site_url', '$title', '$subscribers')");
34
35 ++$count;
36
37 }
38
39 }
40
41 db_query($link, "COMMIT");
42
43 return $count;
44
45 }
46
47
48 /**
49 * Update a feed batch.
50 * Used by daemons to update n feeds by run.
51 * Only update feed needing a update, and not being processed
52 * by another process.
53 *
54 * @param mixed $link Database link
55 * @param integer $limit Maximum number of feeds in update batch. Default to DAEMON_FEED_LIMIT.
56 * @param boolean $from_http Set to true if you call this function from http to disable cli specific code.
57 * @param boolean $debug Set to false to disable debug output. Default to true.
58 * @return void
59 */
60 function update_daemon_common($link, $limit = DAEMON_FEED_LIMIT, $from_http = false, $debug = true) {
61 // Process all other feeds using last_updated and interval parameters
62
63 define('PREFS_NO_CACHE', true);
64
65 // Test if the user has loggued in recently. If not, it does not update its feeds.
66 if (!SINGLE_USER_MODE && DAEMON_UPDATE_LOGIN_LIMIT > 0) {
67 if (DB_TYPE == "pgsql") {
68 $login_thresh_qpart = "AND ttrss_users.last_login >= NOW() - INTERVAL '".DAEMON_UPDATE_LOGIN_LIMIT." days'";
69 } else {
70 $login_thresh_qpart = "AND ttrss_users.last_login >= DATE_SUB(NOW(), INTERVAL ".DAEMON_UPDATE_LOGIN_LIMIT." DAY)";
71 }
72 } else {
73 $login_thresh_qpart = "";
74 }
75
76 // Test if the feed need a update (update interval exceded).
77 if (DB_TYPE == "pgsql") {
78 $update_limit_qpart = "AND ((
79 ttrss_feeds.update_interval = 0
80 AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_user_prefs.value || ' minutes') AS INTERVAL)
81 ) OR (
82 ttrss_feeds.update_interval > 0
83 AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_feeds.update_interval || ' minutes') AS INTERVAL)
84 ) OR ttrss_feeds.last_updated IS NULL
85 OR last_updated = '1970-01-01 00:00:00')";
86 } else {
87 $update_limit_qpart = "AND ((
88 ttrss_feeds.update_interval = 0
89 AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL CONVERT(ttrss_user_prefs.value, SIGNED INTEGER) MINUTE)
90 ) OR (
91 ttrss_feeds.update_interval > 0
92 AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL ttrss_feeds.update_interval MINUTE)
93 ) OR ttrss_feeds.last_updated IS NULL
94 OR last_updated = '1970-01-01 00:00:00')";
95 }
96
97 // Test if feed is currently being updated by another process.
98 if (DB_TYPE == "pgsql") {
99 $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < NOW() - INTERVAL '5 minutes')";
100 } else {
101 $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < DATE_SUB(NOW(), INTERVAL 5 MINUTE))";
102 }
103
104 // Test if there is a limit to number of updated feeds
105 $query_limit = "";
106 if($limit) $query_limit = sprintf("LIMIT %d", $limit);
107
108 $random_qpart = sql_random_function();
109
110 // We search for feed needing update.
111 $result = db_query($link, "SELECT ttrss_feeds.feed_url,ttrss_feeds.id, ttrss_feeds.owner_uid,
112 ".SUBSTRING_FOR_DATE."(ttrss_feeds.last_updated,1,19) AS last_updated,
113 ttrss_feeds.update_interval
114 FROM
115 ttrss_feeds, ttrss_users, ttrss_user_prefs
116 WHERE
117 ttrss_feeds.owner_uid = ttrss_users.id
118 AND ttrss_users.id = ttrss_user_prefs.owner_uid
119 AND ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL'
120 $login_thresh_qpart $update_limit_qpart
121 $updstart_thresh_qpart
122 ORDER BY $random_qpart $query_limit");
123
124 $user_prefs_cache = array();
125
126 if($debug) _debug(sprintf("Scheduled %d feeds to update...\n", db_num_rows($result)));
127
128 // Here is a little cache magic in order to minimize risk of double feed updates.
129 $feeds_to_update = array();
130 while ($line = db_fetch_assoc($result)) {
131 $feeds_to_update[$line['id']] = $line;
132 }
133
134 // We update the feed last update started date before anything else.
135 // There is no lag due to feed contents downloads
136 // It prevent an other process to update the same feed.
137 $feed_ids = array_keys($feeds_to_update);
138 if($feed_ids) {
139 db_query($link, sprintf("UPDATE ttrss_feeds SET last_update_started = NOW()
140 WHERE id IN (%s)", implode(',', $feed_ids)));
141 }
142
143 expire_cached_files($debug);
144
145 // For each feed, we call the feed update function.
146 while ($line = array_pop($feeds_to_update)) {
147
148 if($debug) _debug("Feed: " . $line["feed_url"] . ", " . $line["last_updated"]);
149
150 update_rss_feed($link, $line["id"], true);
151
152 sleep(1); // prevent flood (FIXME make this an option?)
153 }
154
155 // Send feed digests by email if needed.
156 send_headlines_digests($link, $debug);
157
158 } // function update_daemon_common
159
160 function update_rss_feed($link, $feed, $ignore_daemon = false, $no_cache = false,
161 $override_url = false) {
162
163 require_once "lib/simplepie/simplepie.inc";
164 require_once "lib/magpierss/rss_fetch.inc";
165 require_once 'lib/magpierss/rss_utils.inc';
166
167 $debug_enabled = defined('DAEMON_EXTENDED_DEBUG') || $_REQUEST['xdebug'];
168
169 if (!$_REQUEST["daemon"] && !$ignore_daemon) {
170 return false;
171 }
172
173 if ($debug_enabled) {
174 _debug("update_rss_feed: start");
175 }
176
177 if (!$ignore_daemon) {
178
179 if (DB_TYPE == "pgsql") {
180 $updstart_thresh_qpart = "(ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < NOW() - INTERVAL '120 seconds')";
181 } else {
182 $updstart_thresh_qpart = "(ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < DATE_SUB(NOW(), INTERVAL 120 SECOND))";
183 }
184
185 $result = db_query($link, "SELECT id,update_interval,auth_login,
186 auth_pass,cache_images,update_method,last_updated
187 FROM ttrss_feeds WHERE id = '$feed' AND $updstart_thresh_qpart");
188
189 } else {
190
191 $result = db_query($link, "SELECT id,update_interval,auth_login,
192 feed_url,auth_pass,cache_images,update_method,last_updated,
193 mark_unread_on_update, owner_uid, update_on_checksum_change,
194 pubsub_state
195 FROM ttrss_feeds WHERE id = '$feed'");
196
197 }
198
199 if (db_num_rows($result) == 0) {
200 if ($debug_enabled) {
201 _debug("update_rss_feed: feed $feed NOT FOUND/SKIPPED");
202 }
203 return false;
204 }
205
206 $update_method = db_fetch_result($result, 0, "update_method");
207 $last_updated = db_fetch_result($result, 0, "last_updated");
208 $owner_uid = db_fetch_result($result, 0, "owner_uid");
209 $mark_unread_on_update = sql_bool_to_bool(db_fetch_result($result,
210 0, "mark_unread_on_update"));
211 $update_on_checksum_change = sql_bool_to_bool(db_fetch_result($result,
212 0, "update_on_checksum_change"));
213 $pubsub_state = db_fetch_result($result, 0, "pubsub_state");
214
215 db_query($link, "UPDATE ttrss_feeds SET last_update_started = NOW()
216 WHERE id = '$feed'");
217
218 $auth_login = db_fetch_result($result, 0, "auth_login");
219 $auth_pass = db_fetch_result($result, 0, "auth_pass");
220
221 if ($update_method == 0)
222 $update_method = DEFAULT_UPDATE_METHOD + 1;
223
224 // 1 - Magpie
225 // 2 - SimplePie
226 // 3 - Twitter OAuth
227
228 if ($update_method == 2)
229 $use_simplepie = true;
230 else
231 $use_simplepie = false;
232
233 if ($debug_enabled) {
234 _debug("update method: $update_method (feed setting: $update_method) (use simplepie: $use_simplepie)\n");
235 }
236
237 if ($update_method == 1) {
238 $auth_login = urlencode($auth_login);
239 $auth_pass = urlencode($auth_pass);
240 }
241
242 $cache_images = sql_bool_to_bool(db_fetch_result($result, 0, "cache_images"));
243 $fetch_url = db_fetch_result($result, 0, "feed_url");
244
245 $feed = db_escape_string($feed);
246
247 if ($auth_login && $auth_pass ){
248 $url_parts = array();
249 preg_match("/(^[^:]*):\/\/(.*)/", $fetch_url, $url_parts);
250
251 if ($url_parts[1] && $url_parts[2]) {
252 $fetch_url = $url_parts[1] . "://$auth_login:$auth_pass@" . $url_parts[2];
253 }
254
255 }
256
257 if ($override_url)
258 $fetch_url = $override_url;
259
260 if ($debug_enabled) {
261 _debug("update_rss_feed: fetching [$fetch_url]...");
262 }
263
264 // Ignore cache if new feed or manual update.
265 $cache_age = (is_null($last_updated) || $last_updated == '1970-01-01 00:00:00') ?
266 -1 : get_feed_update_interval($link, $feed) * 60;
267
268 if ($update_method == 1) {
269
270 define('MAGPIE_CACHE_AGE', $cache_age);
271 define('MAGPIE_CACHE_ON', !$no_cache);
272 define('MAGPIE_FETCH_TIME_OUT', $no_cache ? 15 : 60);
273 define('MAGPIE_CACHE_DIR', CACHE_DIR . "/magpie");
274
275 $rss = @fetch_rss($fetch_url);
276 } else {
277 $simplepie_cache_dir = CACHE_DIR . "/simplepie";
278
279 if (!is_dir($simplepie_cache_dir)) {
280 mkdir($simplepie_cache_dir);
281 }
282
283 $rss = new SimplePie();
284 $rss->set_useragent(SELF_USER_AGENT);
285 $rss->set_timeout($no_cache ? 15 : 60);
286 $rss->set_feed_url($fetch_url);
287 $rss->set_output_encoding('UTF-8');
288 //$rss->force_feed(true);
289
290 if ($debug_enabled) {
291 _debug("feed update interval (sec): " .
292 get_feed_update_interval($link, $feed)*60);
293 }
294
295 $rss->enable_cache(!$no_cache);
296
297 if (!$no_cache) {
298 $rss->set_cache_location($simplepie_cache_dir);
299 $rss->set_cache_duration($cache_age);
300 }
301
302 $rss->init();
303 }
304
305 // print_r($rss);
306
307 if ($debug_enabled) {
308 _debug("update_rss_feed: fetch done, parsing...");
309 }
310
311 $feed = db_escape_string($feed);
312
313 if ($update_method == 2) {
314 $fetch_ok = !$rss->error();
315 } else {
316 $fetch_ok = !!$rss;
317 }
318
319 if ($fetch_ok) {
320
321 if ($debug_enabled) {
322 _debug("update_rss_feed: processing feed data...");
323 }
324
325 // db_query($link, "BEGIN");
326
327 if (DB_TYPE == "pgsql") {
328 $favicon_interval_qpart = "favicon_last_checked < NOW() - INTERVAL '12 hour'";
329 } else {
330 $favicon_interval_qpart = "favicon_last_checked < DATE_SUB(NOW(), INTERVAL 12 HOUR)";
331 }
332
333 $result = db_query($link, "SELECT title,icon_url,site_url,owner_uid,
334 (favicon_last_checked IS NULL OR $favicon_interval_qpart) AS
335 favicon_needs_check
336 FROM ttrss_feeds WHERE id = '$feed'");
337
338 $registered_title = db_fetch_result($result, 0, "title");
339 $orig_icon_url = db_fetch_result($result, 0, "icon_url");
340 $orig_site_url = db_fetch_result($result, 0, "site_url");
341 $favicon_needs_check = sql_bool_to_bool(db_fetch_result($result, 0,
342 "favicon_needs_check"));
343
344 $owner_uid = db_fetch_result($result, 0, "owner_uid");
345
346 if ($use_simplepie) {
347 $site_url = db_escape_string(trim($rss->get_link()));
348 } else {
349 $site_url = db_escape_string(trim($rss->channel["link"]));
350 }
351
352 // weird, weird Magpie
353 if (!$use_simplepie) {
354 if (!$site_url) $site_url = db_escape_string($rss->channel["link_"]);
355 }
356
357 $site_url = rewrite_relative_url($fetch_url, $site_url);
358 $site_url = substr($site_url, 0, 250);
359
360 if ($debug_enabled) {
361 _debug("update_rss_feed: checking favicon...");
362 }
363
364 if ($favicon_needs_check) {
365 check_feed_favicon($site_url, $feed, $link);
366
367 db_query($link, "UPDATE ttrss_feeds SET favicon_last_checked = NOW()
368 WHERE id = '$feed'");
369 }
370
371 if (!$registered_title || $registered_title == "[Unknown]") {
372
373 if ($use_simplepie) {
374 $feed_title = db_escape_string($rss->get_title());
375 } else {
376 $feed_title = db_escape_string($rss->channel["title"]);
377 }
378
379 if ($debug_enabled) {
380 _debug("update_rss_feed: registering title: $feed_title");
381 }
382
383 db_query($link, "UPDATE ttrss_feeds SET
384 title = '$feed_title' WHERE id = '$feed'");
385 }
386
387 if ($site_url && $orig_site_url != $site_url) {
388 db_query($link, "UPDATE ttrss_feeds SET
389 site_url = '$site_url' WHERE id = '$feed'");
390 }
391
392 // print "I: " . $rss->channel["image"]["url"];
393
394 if (!$use_simplepie) {
395 $icon_url = db_escape_string(trim($rss->image["url"]));
396 } else {
397 $icon_url = db_escape_string(trim($rss->get_image_url()));
398 }
399
400 $icon_url = rewrite_relative_url($fetch_url, $icon_url);
401 $icon_url = substr($icon_url, 0, 250);
402
403 if ($icon_url && $orig_icon_url != $icon_url) {
404 db_query($link, "UPDATE ttrss_feeds SET icon_url = '$icon_url' WHERE id = '$feed'");
405 }
406
407 if ($debug_enabled) {
408 _debug("update_rss_feed: loading filters...");
409 }
410
411 $filters = load_filters($link, $feed, $owner_uid);
412
413 if ($debug_enabled) {
414 //print_r($filters);
415 _debug("update_rss_feed: " . count($filters) . " filters loaded.");
416 }
417
418 if ($use_simplepie) {
419 $iterator = $rss->get_items();
420 } else {
421 $iterator = $rss->items;
422 if (!$iterator || !is_array($iterator)) $iterator = $rss->entries;
423 if (!$iterator || !is_array($iterator)) $iterator = $rss;
424 }
425
426 if (!is_array($iterator)) {
427 /* db_query($link, "UPDATE ttrss_feeds
428 SET last_error = 'Parse error: can\'t find any articles.'
429 WHERE id = '$feed'"); */
430
431 // clear any errors and mark feed as updated if fetched okay
432 // even if it's blank
433
434 if ($debug_enabled) {
435 _debug("update_rss_feed: entry iterator is not an array, no articles?");
436 }
437
438 db_query($link, "UPDATE ttrss_feeds
439 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
440
441 return; // no articles
442 }
443
444 if ($pubsub_state != 2 && PUBSUBHUBBUB_ENABLED) {
445
446 if ($debug_enabled) _debug("update_rss_feed: checking for PUSH hub...");
447
448 $feed_hub_url = false;
449 if ($use_simplepie) {
450 $links = $rss->get_links('hub');
451
452 if ($links && is_array($links)) {
453 foreach ($links as $l) {
454 $feed_hub_url = $l;
455 break;
456 }
457 }
458
459 } else {
460 $atom = $rss->channel['atom'];
461
462 if ($atom) {
463 if ($atom['link@rel'] == 'hub') {
464 $feed_hub_url = $atom['link@href'];
465 }
466
467 if (!$feed_hub_url && $atom['link#'] > 1) {
468 for ($i = 2; $i <= $atom['link#']; $i++) {
469 if ($atom["link#$i@rel"] == 'hub') {
470 $feed_hub_url = $atom["link#$i@href"];
471 break;
472 }
473 }
474 }
475 } else {
476 $feed_hub_url = $rss->channel['link_hub'];
477 }
478 }
479
480 if ($debug_enabled) _debug("update_rss_feed: feed hub url: $feed_hub_url");
481
482 if ($feed_hub_url && function_exists('curl_init') &&
483 !ini_get("open_basedir")) {
484
485 require_once 'lib/pubsubhubbub/subscriber.php';
486
487 $callback_url = get_self_url_prefix() .
488 "/public.php?op=pubsub&id=$feed";
489
490 $s = new Subscriber($feed_hub_url, $callback_url);
491
492 $rc = $s->subscribe($fetch_url);
493
494 if ($debug_enabled)
495 _debug("update_rss_feed: feed hub url found, subscribe request sent.");
496
497 db_query($link, "UPDATE ttrss_feeds SET pubsub_state = 1
498 WHERE id = '$feed'");
499 }
500 }
501
502 if ($debug_enabled) {
503 _debug("update_rss_feed: processing articles...");
504 }
505
506 foreach ($iterator as $item) {
507 if ($_REQUEST['xdebug'] == 2) {
508 print_r($item);
509 }
510
511 if ($use_simplepie) {
512 $entry_guid = $item->get_id();
513 if (!$entry_guid) $entry_guid = $item->get_link();
514 if (!$entry_guid) $entry_guid = make_guid_from_title($item->get_title());
515
516 } else {
517
518 $entry_guid = $item["id"];
519
520 if (!$entry_guid) $entry_guid = $item["guid"];
521 if (!$entry_guid) $entry_guid = $item["about"];
522 if (!$entry_guid) $entry_guid = $item["link"];
523 if (!$entry_guid) $entry_guid = make_guid_from_title($item["title"]);
524 }
525
526 if ($debug_enabled) {
527 _debug("update_rss_feed: guid $entry_guid");
528 }
529
530 if (!$entry_guid) continue;
531
532 $entry_timestamp = "";
533
534 if ($use_simplepie) {
535 $entry_timestamp = strtotime($item->get_date());
536 } else {
537 $rss_2_date = $item['pubdate'];
538 $rss_1_date = $item['dc']['date'];
539 $atom_date = $item['issued'];
540 if (!$atom_date) $atom_date = $item['updated'];
541
542 if ($atom_date != "") $entry_timestamp = parse_w3cdtf($atom_date);
543 if ($rss_1_date != "") $entry_timestamp = parse_w3cdtf($rss_1_date);
544 if ($rss_2_date != "") $entry_timestamp = strtotime($rss_2_date);
545
546 }
547
548 if ($entry_timestamp == "" || $entry_timestamp == -1 || !$entry_timestamp) {
549 $entry_timestamp = time();
550 $no_orig_date = 'true';
551 } else {
552 $no_orig_date = 'false';
553 }
554
555 $entry_timestamp_fmt = strftime("%Y/%m/%d %H:%M:%S", $entry_timestamp);
556
557 if ($debug_enabled) {
558 _debug("update_rss_feed: date $entry_timestamp [$entry_timestamp_fmt]");
559 }
560
561 if ($use_simplepie) {
562 $entry_title = $item->get_title();
563 } else {
564 $entry_title = trim(strip_tags($item["title"]));
565 }
566
567 if ($use_simplepie) {
568 $entry_link = $item->get_link();
569 } else {
570 // strange Magpie workaround
571 $entry_link = $item["link_"];
572 if (!$entry_link) $entry_link = $item["link"];
573 }
574
575 $entry_link = rewrite_relative_url($site_url, $entry_link);
576
577 if ($debug_enabled) {
578 _debug("update_rss_feed: title $entry_title");
579 _debug("update_rss_feed: link $entry_link");
580 }
581
582 if (!$entry_title) $entry_title = date("Y-m-d H:i:s", $entry_timestamp);;
583
584 $entry_link = strip_tags($entry_link);
585
586 if ($use_simplepie) {
587 $entry_content = $item->get_content();
588 if (!$entry_content) $entry_content = $item->get_description();
589 } else {
590 $entry_content = $item["content:escaped"];
591
592 if (!$entry_content) $entry_content = $item["content:encoded"];
593 if (!$entry_content && is_array($entry_content)) $entry_content = $item["content"]["encoded"];
594 if (!$entry_content) $entry_content = $item["content"];
595
596 if (is_array($entry_content)) $entry_content = $entry_content[0];
597
598 // Magpie bugs are getting ridiculous
599 if (trim($entry_content) == "Array") $entry_content = false;
600
601 if (!$entry_content) $entry_content = $item["atom_content"];
602 if (!$entry_content) $entry_content = $item["summary"];
603
604 if (!$entry_content ||
605 strlen($entry_content) < strlen($item["description"])) {
606 $entry_content = $item["description"];
607 };
608
609 // WTF
610 if (is_array($entry_content)) {
611 $entry_content = $entry_content["encoded"];
612 if (!$entry_content) $entry_content = $entry_content["escaped"];
613 }
614 }
615
616 if ($cache_images && is_writable(CACHE_DIR . '/images'))
617 $entry_content = cache_images($entry_content, $site_url, $debug_enabled);
618
619 if ($_REQUEST["xdebug"] == 2) {
620 print "update_rss_feed: content: ";
621 print $entry_content;
622 print "\n";
623 }
624
625 $entry_content_unescaped = $entry_content;
626
627 if ($use_simplepie) {
628 $entry_comments = strip_tags($item->data["comments"]);
629 if ($item->get_author()) {
630 $entry_author_item = $item->get_author();
631 $entry_author = $entry_author_item->get_name();
632 if (!$entry_author) $entry_author = $entry_author_item->get_email();
633
634 $entry_author = db_escape_string($entry_author);
635 }
636 } else {
637 $entry_comments = strip_tags($item["comments"]);
638
639 $entry_author = db_escape_string(strip_tags($item['dc']['creator']));
640
641 if ($item['author']) {
642
643 if (is_array($item['author'])) {
644
645 if (!$entry_author) {
646 $entry_author = db_escape_string(strip_tags($item['author']['name']));
647 }
648
649 if (!$entry_author) {
650 $entry_author = db_escape_string(strip_tags($item['author']['email']));
651 }
652 }
653
654 if (!$entry_author) {
655 $entry_author = db_escape_string(strip_tags($item['author']));
656 }
657 }
658 }
659
660 if (preg_match('/^[\t\n\r ]*$/', $entry_author)) $entry_author = '';
661
662 $entry_guid = db_escape_string(strip_tags($entry_guid));
663 $entry_guid = mb_substr($entry_guid, 0, 250);
664
665 $result = db_query($link, "SELECT id FROM ttrss_entries
666 WHERE guid = '$entry_guid'");
667
668 $entry_content = db_escape_string($entry_content, false);
669
670 $content_hash = "SHA1:" . sha1(strip_tags($entry_content));
671
672 $entry_title = db_escape_string($entry_title);
673 $entry_link = db_escape_string($entry_link);
674 $entry_comments = mb_substr(db_escape_string($entry_comments), 0, 250);
675 $entry_author = mb_substr($entry_author, 0, 250);
676
677 if ($use_simplepie) {
678 $num_comments = 0; #FIXME#
679 } else {
680 $num_comments = db_escape_string($item["slash"]["comments"]);
681 }
682
683 if (!$num_comments) $num_comments = 0;
684
685 if ($debug_enabled) {
686 _debug("update_rss_feed: looking for tags [1]...");
687 }
688
689 // parse <category> entries into tags
690
691 $additional_tags = array();
692
693 if ($use_simplepie) {
694
695 $additional_tags_src = $item->get_categories();
696
697 if (is_array($additional_tags_src)) {
698 foreach ($additional_tags_src as $tobj) {
699 array_push($additional_tags, $tobj->get_term());
700 }
701 }
702
703 if ($debug_enabled) {
704 _debug("update_rss_feed: category tags:");
705 print_r($additional_tags);
706 }
707
708 } else {
709
710 $t_ctr = $item['category#'];
711
712 if ($t_ctr == 0) {
713 $additional_tags = array();
714 } else if ($t_ctr > 0) {
715 $additional_tags = array($item['category']);
716
717 if ($item['category@term']) {
718 array_push($additional_tags, $item['category@term']);
719 }
720
721 for ($i = 0; $i <= $t_ctr; $i++ ) {
722 if ($item["category#$i"]) {
723 array_push($additional_tags, $item["category#$i"]);
724 }
725
726 if ($item["category#$i@term"]) {
727 array_push($additional_tags, $item["category#$i@term"]);
728 }
729 }
730 }
731
732 // parse <dc:subject> elements
733
734 $t_ctr = $item['dc']['subject#'];
735
736 if ($t_ctr > 0) {
737 array_push($additional_tags, $item['dc']['subject']);
738
739 for ($i = 0; $i <= $t_ctr; $i++ ) {
740 if ($item['dc']["subject#$i"]) {
741 array_push($additional_tags, $item['dc']["subject#$i"]);
742 }
743 }
744 }
745 }
746
747 if ($debug_enabled) {
748 _debug("update_rss_feed: looking for tags [2]...");
749 }
750
751 /* taaaags */
752 // <a href="..." rel="tag">Xorg</a>, //
753
754 $entry_tags = null;
755
756 preg_match_all("/<a.*?rel=['\"]tag['\"].*?\>([^<]+)<\/a>/i",
757 $entry_content_unescaped, $entry_tags);
758
759 $entry_tags = $entry_tags[1];
760
761 $entry_tags = array_merge($entry_tags, $additional_tags);
762 $entry_tags = array_unique($entry_tags);
763
764 for ($i = 0; $i < count($entry_tags); $i++)
765 $entry_tags[$i] = mb_strtolower($entry_tags[$i], 'utf-8');
766
767 if ($debug_enabled) {
768 //_debug("update_rss_feed: unfiltered tags found:");
769 //print_r($entry_tags);
770 }
771
772 # sanitize content
773 $entry_content = sanitize($link, $entry_content, $owner_uid, $site_url);
774 $entry_title = strip_tags($entry_title);
775
776 if ($debug_enabled) {
777 _debug("update_rss_feed: done collecting data [TITLE:$entry_title]");
778 }
779
780 db_query($link, "BEGIN");
781
782 if (db_num_rows($result) == 0) {
783
784 if ($debug_enabled) {
785 _debug("update_rss_feed: base guid not found");
786 }
787
788 // base post entry does not exist, create it
789
790 $result = db_query($link,
791 "INSERT INTO ttrss_entries
792 (title,
793 guid,
794 link,
795 updated,
796 content,
797 content_hash,
798 no_orig_date,
799 date_updated,
800 date_entered,
801 comments,
802 num_comments,
803 author)
804 VALUES
805 ('$entry_title',
806 '$entry_guid',
807 '$entry_link',
808 '$entry_timestamp_fmt',
809 '$entry_content',
810 '$content_hash',
811 $no_orig_date,
812 NOW(),
813 NOW(),
814 '$entry_comments',
815 '$num_comments',
816 '$entry_author')");
817 } else {
818 // we keep encountering the entry in feeds, so we need to
819 // update date_updated column so that we don't get horrible
820 // dupes when the entry gets purged and reinserted again e.g.
821 // in the case of SLOW SLOW OMG SLOW updating feeds
822
823 $base_entry_id = db_fetch_result($result, 0, "id");
824
825 db_query($link, "UPDATE ttrss_entries SET date_updated = NOW()
826 WHERE id = '$base_entry_id'");
827 }
828
829 // now it should exist, if not - bad luck then
830
831 $result = db_query($link, "SELECT
832 id,content_hash,no_orig_date,title,
833 ".SUBSTRING_FOR_DATE."(date_updated,1,19) as date_updated,
834 ".SUBSTRING_FOR_DATE."(updated,1,19) as updated,
835 num_comments
836 FROM
837 ttrss_entries
838 WHERE guid = '$entry_guid'");
839
840 $entry_ref_id = 0;
841 $entry_int_id = 0;
842
843 if (db_num_rows($result) == 1) {
844
845 if ($debug_enabled) {
846 _debug("update_rss_feed: base guid found, checking for user record");
847 }
848
849 // this will be used below in update handler
850 $orig_content_hash = db_fetch_result($result, 0, "content_hash");
851 $orig_title = db_fetch_result($result, 0, "title");
852 $orig_num_comments = db_fetch_result($result, 0, "num_comments");
853 $orig_date_updated = strtotime(db_fetch_result($result,
854 0, "date_updated"));
855
856 $ref_id = db_fetch_result($result, 0, "id");
857 $entry_ref_id = $ref_id;
858
859 // check for user post link to main table
860
861 // do we allow duplicate posts with same GUID in different feeds?
862 if (get_pref($link, "ALLOW_DUPLICATE_POSTS", $owner_uid, false)) {
863 $dupcheck_qpart = "AND (feed_id = '$feed' OR feed_id IS NULL)";
864 } else {
865 $dupcheck_qpart = "";
866 }
867
868 /* Collect article tags here so we could filter by them: */
869
870 $article_filters = get_article_filters($filters, $entry_title,
871 $entry_content, $entry_link, $entry_timestamp, $entry_author,
872 $entry_tags);
873
874 if ($debug_enabled) {
875 _debug("update_rss_feed: article filters: ");
876 if (count($article_filters) != 0) {
877 print_r($article_filters);
878 }
879 }
880
881 if (find_article_filter($article_filters, "filter")) {
882 db_query($link, "COMMIT"); // close transaction in progress
883 continue;
884 }
885
886 $score = calculate_article_score($article_filters);
887
888 if ($debug_enabled) {
889 _debug("update_rss_feed: initial score: $score");
890 }
891
892 $query = "SELECT ref_id, int_id FROM ttrss_user_entries WHERE
893 ref_id = '$ref_id' AND owner_uid = '$owner_uid'
894 $dupcheck_qpart";
895
896 // if ($_REQUEST["xdebug"]) print "$query\n";
897
898 $result = db_query($link, $query);
899
900 // okay it doesn't exist - create user entry
901 if (db_num_rows($result) == 0) {
902
903 if ($debug_enabled) {
904 _debug("update_rss_feed: user record not found, creating...");
905 }
906
907 if ($score >= -500 && !find_article_filter($article_filters, 'catchup')) {
908 $unread = 'true';
909 $last_read_qpart = 'NULL';
910 } else {
911 $unread = 'false';
912 $last_read_qpart = 'NOW()';
913 }
914
915 if (find_article_filter($article_filters, 'mark') || $score > 1000) {
916 $marked = 'true';
917 } else {
918 $marked = 'false';
919 }
920
921 if (find_article_filter($article_filters, 'publish')) {
922 $published = 'true';
923 } else {
924 $published = 'false';
925 }
926
927 // N-grams
928
929 if (DB_TYPE == "pgsql" and defined('_NGRAM_TITLE_DUPLICATE_THRESHOLD')) {
930
931 $result = db_query($link, "SELECT COUNT(*) AS similar FROM
932 ttrss_entries,ttrss_user_entries
933 WHERE ref_id = id AND updated >= NOW() - INTERVAL '7 day'
934 AND similarity(title, '$entry_title') >= "._NGRAM_TITLE_DUPLICATE_THRESHOLD."
935 AND owner_uid = $owner_uid");
936
937 $ngram_similar = db_fetch_result($result, 0, "similar");
938
939 if ($debug_enabled) {
940 _debug("update_rss_feed: N-gram similar results: $ngram_similar");
941 }
942
943 if ($ngram_similar > 0) {
944 $unread = 'false';
945 }
946 }
947
948 $result = db_query($link,
949 "INSERT INTO ttrss_user_entries
950 (ref_id, owner_uid, feed_id, unread, last_read, marked,
951 published, score, tag_cache, label_cache, uuid)
952 VALUES ('$ref_id', '$owner_uid', '$feed', $unread,
953 $last_read_qpart, $marked, $published, '$score', '', '', '')");
954
955 if (PUBSUBHUBBUB_HUB && $published == 'true') {
956 $rss_link = get_self_url_prefix() .
957 "/public.php?op=rss&id=-2&key=" .
958 get_feed_access_key($link, -2, false, $owner_uid);
959
960 $p = new Publisher(PUBSUBHUBBUB_HUB);
961
962 $pubsub_result = $p->publish_update($rss_link);
963 }
964
965 $result = db_query($link,
966 "SELECT int_id FROM ttrss_user_entries WHERE
967 ref_id = '$ref_id' AND owner_uid = '$owner_uid' AND
968 feed_id = '$feed' LIMIT 1");
969
970 if (db_num_rows($result) == 1) {
971 $entry_int_id = db_fetch_result($result, 0, "int_id");
972 }
973 } else {
974 if ($debug_enabled) {
975 _debug("update_rss_feed: user record FOUND");
976 }
977
978 $entry_ref_id = db_fetch_result($result, 0, "ref_id");
979 $entry_int_id = db_fetch_result($result, 0, "int_id");
980 }
981
982 if ($debug_enabled) {
983 _debug("update_rss_feed: RID: $entry_ref_id, IID: $entry_int_id");
984 }
985
986 $post_needs_update = false;
987 $update_insignificant = false;
988
989 if ($orig_num_comments != $num_comments) {
990 $post_needs_update = true;
991 $update_insignificant = true;
992 }
993
994 if ($content_hash != $orig_content_hash) {
995 $post_needs_update = true;
996 $update_insignificant = false;
997 }
998
999 if (db_escape_string($orig_title) != $entry_title) {
1000 $post_needs_update = true;
1001 $update_insignificant = false;
1002 }
1003
1004 // if post needs update, update it and mark all user entries
1005 // linking to this post as updated
1006 if ($post_needs_update) {
1007
1008 if (defined('DAEMON_EXTENDED_DEBUG')) {
1009 _debug("update_rss_feed: post $entry_guid needs update...");
1010 }
1011
1012 // print "<!-- post $orig_title needs update : $post_needs_update -->";
1013
1014 db_query($link, "UPDATE ttrss_entries
1015 SET title = '$entry_title', content = '$entry_content',
1016 content_hash = '$content_hash',
1017 updated = '$entry_timestamp_fmt',
1018 num_comments = '$num_comments'
1019 WHERE id = '$ref_id'");
1020
1021 if (!$update_insignificant) {
1022 if ($mark_unread_on_update) {
1023 db_query($link, "UPDATE ttrss_user_entries
1024 SET last_read = null, unread = true WHERE ref_id = '$ref_id'");
1025 } else if ($update_on_checksum_change) {
1026 db_query($link, "UPDATE ttrss_user_entries
1027 SET last_read = null WHERE ref_id = '$ref_id'
1028 AND unread = false");
1029 }
1030 }
1031 }
1032 }
1033
1034 db_query($link, "COMMIT");
1035
1036 if ($debug_enabled) {
1037 _debug("update_rss_feed: assigning labels...");
1038 }
1039
1040 assign_article_to_labels($link, $entry_ref_id, $article_filters,
1041 $owner_uid);
1042
1043 if ($debug_enabled) {
1044 _debug("update_rss_feed: looking for enclosures...");
1045 }
1046
1047 // enclosures
1048
1049 $enclosures = array();
1050
1051 if ($use_simplepie) {
1052 $encs = $item->get_enclosures();
1053
1054 if (is_array($encs)) {
1055 foreach ($encs as $e) {
1056 $e_item = array(
1057 $e->link, $e->type, $e->length);
1058
1059 array_push($enclosures, $e_item);
1060 }
1061 }
1062
1063 } else {
1064 // <enclosure>
1065
1066 $e_ctr = $item['enclosure#'];
1067
1068 if ($e_ctr > 0) {
1069 $e_item = array($item['enclosure@url'],
1070 $item['enclosure@type'],
1071 $item['enclosure@length']);
1072
1073 array_push($enclosures, $e_item);
1074
1075 for ($i = 0; $i <= $e_ctr; $i++ ) {
1076
1077 if ($item["enclosure#$i@url"]) {
1078 $e_item = array($item["enclosure#$i@url"],
1079 $item["enclosure#$i@type"],
1080 $item["enclosure#$i@length"]);
1081 array_push($enclosures, $e_item);
1082 }
1083 }
1084 }
1085
1086 // <media:content>
1087 // can there be many of those? yes -fox
1088
1089 $m_ctr = $item['media']['content#'];
1090
1091 if ($m_ctr > 0) {
1092 $e_item = array($item['media']['content@url'],
1093 $item['media']['content@medium'],
1094 $item['media']['content@length']);
1095
1096 array_push($enclosures, $e_item);
1097
1098 for ($i = 0; $i <= $m_ctr; $i++ ) {
1099
1100 if ($item["media"]["content#$i@url"]) {
1101 $e_item = array($item["media"]["content#$i@url"],
1102 $item["media"]["content#$i@medium"],
1103 $item["media"]["content#$i@length"]);
1104 array_push($enclosures, $e_item);
1105 }
1106 }
1107
1108 }
1109 }
1110
1111
1112 if ($debug_enabled) {
1113 _debug("update_rss_feed: article enclosures:");
1114 print_r($enclosures);
1115 }
1116
1117 db_query($link, "BEGIN");
1118
1119 foreach ($enclosures as $enc) {
1120 $enc_url = db_escape_string($enc[0]);
1121 $enc_type = db_escape_string($enc[1]);
1122 $enc_dur = db_escape_string($enc[2]);
1123
1124 $result = db_query($link, "SELECT id FROM ttrss_enclosures
1125 WHERE content_url = '$enc_url' AND post_id = '$entry_ref_id'");
1126
1127 if (db_num_rows($result) == 0) {
1128 db_query($link, "INSERT INTO ttrss_enclosures
1129 (content_url, content_type, title, duration, post_id) VALUES
1130 ('$enc_url', '$enc_type', '', '$enc_dur', '$entry_ref_id')");
1131 }
1132 }
1133
1134 db_query($link, "COMMIT");
1135
1136 // check for manual tags (we have to do it here since they're loaded from filters)
1137
1138 foreach ($article_filters as $f) {
1139 if ($f["type"] == "tag") {
1140
1141 $manual_tags = trim_array(explode(",", $f["param"]));
1142
1143 foreach ($manual_tags as $tag) {
1144 if (tag_is_valid($tag)) {
1145 array_push($entry_tags, $tag);
1146 }
1147 }
1148 }
1149 }
1150
1151 // Skip boring tags
1152
1153 $boring_tags = trim_array(explode(",", mb_strtolower(get_pref($link,
1154 'BLACKLISTED_TAGS', $owner_uid, ''), 'utf-8')));
1155
1156 $filtered_tags = array();
1157 $tags_to_cache = array();
1158
1159 if ($entry_tags && is_array($entry_tags)) {
1160 foreach ($entry_tags as $tag) {
1161 if (array_search($tag, $boring_tags) === false) {
1162 array_push($filtered_tags, $tag);
1163 }
1164 }
1165 }
1166
1167 $filtered_tags = array_unique($filtered_tags);
1168
1169 if ($debug_enabled) {
1170 _debug("update_rss_feed: filtered article tags:");
1171 print_r($filtered_tags);
1172 }
1173
1174 // Save article tags in the database
1175
1176 if (count($filtered_tags) > 0) {
1177
1178 db_query($link, "BEGIN");
1179
1180 foreach ($filtered_tags as $tag) {
1181
1182 $tag = sanitize_tag($tag);
1183 $tag = db_escape_string($tag);
1184
1185 if (!tag_is_valid($tag)) continue;
1186
1187 $result = db_query($link, "SELECT id FROM ttrss_tags
1188 WHERE tag_name = '$tag' AND post_int_id = '$entry_int_id' AND
1189 owner_uid = '$owner_uid' LIMIT 1");
1190
1191 if ($result && db_num_rows($result) == 0) {
1192
1193 db_query($link, "INSERT INTO ttrss_tags
1194 (owner_uid,tag_name,post_int_id)
1195 VALUES ('$owner_uid','$tag', '$entry_int_id')");
1196 }
1197
1198 array_push($tags_to_cache, $tag);
1199 }
1200
1201 /* update the cache */
1202
1203 $tags_to_cache = array_unique($tags_to_cache);
1204
1205 $tags_str = db_escape_string(join(",", $tags_to_cache));
1206
1207 db_query($link, "UPDATE ttrss_user_entries
1208 SET tag_cache = '$tags_str' WHERE ref_id = '$entry_ref_id'
1209 AND owner_uid = $owner_uid");
1210
1211 db_query($link, "COMMIT");
1212 }
1213
1214 if ($debug_enabled) {
1215 _debug("update_rss_feed: article processed");
1216 }
1217 }
1218
1219 if (!$last_updated) {
1220 if ($debug_enabled) {
1221 _debug("update_rss_feed: new feed, catching it up...");
1222 }
1223 catchup_feed($link, $feed, false, $owner_uid);
1224 }
1225
1226 if ($debug_enabled) {
1227 _debug("purging feed...");
1228 }
1229
1230 purge_feed($link, $feed, 0, $debug_enabled);
1231
1232 db_query($link, "UPDATE ttrss_feeds
1233 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
1234
1235 // db_query($link, "COMMIT");
1236
1237 } else {
1238
1239 if ($use_simplepie) {
1240 $error_msg = mb_substr($rss->error(), 0, 250);
1241 } else {
1242 $error_msg = mb_substr(magpie_error(), 0, 250);
1243 }
1244
1245 if ($debug_enabled) {
1246 _debug("update_rss_feed: error fetching feed: $error_msg");
1247 }
1248
1249 $error_msg = db_escape_string($error_msg);
1250
1251 db_query($link,
1252 "UPDATE ttrss_feeds SET last_error = '$error_msg',
1253 last_updated = NOW() WHERE id = '$feed'");
1254 }
1255
1256 if ($use_simplepie) {
1257 unset($rss);
1258 }
1259
1260 if ($debug_enabled) {
1261 _debug("update_rss_feed: done");
1262 }
1263
1264 }
1265
1266 function cache_images($html, $site_url, $debug) {
1267 $cache_dir = CACHE_DIR . "/images";
1268
1269 libxml_use_internal_errors(true);
1270
1271 $charset_hack = '<head>
1272 <meta http-equiv="Content-Type" content="text/html; charset=utf-8"/>
1273 </head>';
1274
1275 $doc = new DOMDocument();
1276 $doc->loadHTML($charset_hack . $html);
1277 $xpath = new DOMXPath($doc);
1278
1279 $entries = $xpath->query('(//img[@src])');
1280
1281 foreach ($entries as $entry) {
1282 if ($entry->hasAttribute('src')) {
1283 $src = rewrite_relative_url($site_url, $entry->getAttribute('src'));
1284
1285 $local_filename = CACHE_DIR . "/images/" . sha1($src) . ".png";
1286
1287 if ($debug) _debug("cache_images: downloading: $src to $local_filename");
1288
1289 if (!file_exists($local_filename)) {
1290 $file_content = fetch_file_contents($src);
1291
1292 if ($file_content && strlen($file_content) > 1024) {
1293 file_put_contents($local_filename, $file_content);
1294 }
1295 }
1296
1297 if (file_exists($local_filename)) {
1298 $entry->setAttribute('src', SELF_URL_PATH . '/image.php?url=' .
1299 base64_encode($src));
1300 }
1301 }
1302 }
1303
1304 $node = $doc->getElementsByTagName('body')->item(0);
1305
1306 return $doc->saveXML($node, LIBXML_NOEMPTYTAG);
1307 }
1308
1309 function expire_cached_files($debug) {
1310 foreach (array("magpie", "simplepie", "images", "export") as $dir) {
1311 $cache_dir = CACHE_DIR . "/$dir";
1312
1313 if ($debug) _debug("Expiring $cache_dir");
1314
1315 $num_deleted = 0;
1316
1317 if (is_writable($cache_dir)) {
1318 $files = glob("$cache_dir/*");
1319
1320 if ($files)
1321 foreach ($files as $file) {
1322 if (time() - filemtime($file) > 86400*7) {
1323 unlink($file);
1324
1325 ++$num_deleted;
1326 }
1327 }
1328 }
1329
1330 if ($debug) _debug("Removed $num_deleted files.");
1331 }
1332 }
1333
1334 /**
1335 * Source: http://www.php.net/manual/en/function.parse-url.php#104527
1336 * Returns the url query as associative array
1337 *
1338 * @param string query
1339 * @return array params
1340 */
1341 function convertUrlQuery($query) {
1342 $queryParts = explode('&', $query);
1343
1344 $params = array();
1345
1346 foreach ($queryParts as $param) {
1347 $item = explode('=', $param);
1348 $params[$item[0]] = $item[1];
1349 }
1350
1351 return $params;
1352 }
1353 ?>
web-based news feed (RSS/Atom) reader and aggregator; see https://tt-rss.org/