]>
Commit | Line | Data |
---|---|---|
1 | <?php | |
2 | class RSSUtils { | |
3 | static function calculate_article_hash($article, $pluginhost) { | |
4 | $tmp = ""; | |
5 | ||
6 | foreach ($article as $k => $v) { | |
7 | if ($k != "feed" && isset($v)) { | |
8 | $x = strip_tags(is_array($v) ? implode(",", $v) : $v); | |
9 | ||
10 | //_debug("$k:" . sha1($x) . ":" . htmlspecialchars($x), true); | |
11 | ||
12 | $tmp .= sha1("$k:" . sha1($x)); | |
13 | } | |
14 | } | |
15 | ||
16 | return sha1(implode(",", $pluginhost->get_plugin_names()) . $tmp); | |
17 | } | |
18 | ||
19 | // Strips utf8mb4 characters (i.e. emoji) for mysql | |
20 | static function strip_utf8mb4($str) { | |
21 | return preg_replace('/[\x{10000}-\x{10FFFF}]/u', "\xEF\xBF\xBD", $str); | |
22 | } | |
23 | ||
24 | static function update_feedbrowser_cache() { | |
25 | ||
26 | $pdo = Db::pdo(); | |
27 | ||
28 | $sth = $pdo->query("SELECT feed_url, site_url, title, COUNT(id) AS subscribers | |
29 | FROM ttrss_feeds WHERE feed_url NOT IN (SELECT feed_url FROM ttrss_feeds | |
30 | WHERE private IS true OR auth_login != '' OR auth_pass != '' OR feed_url LIKE '%:%@%/%') | |
31 | GROUP BY feed_url, site_url, title ORDER BY subscribers DESC LIMIT 1000"); | |
32 | ||
33 | $pdo->beginTransaction(); | |
34 | ||
35 | $pdo->query("DELETE FROM ttrss_feedbrowser_cache"); | |
36 | ||
37 | $count = 0; | |
38 | ||
39 | while ($line = $sth->fetch()) { | |
40 | ||
41 | $subscribers = $line["subscribers"]; | |
42 | $feed_url = $line["feed_url"]; | |
43 | $title = $line["title"]; | |
44 | $site_url = $line["site_url"]; | |
45 | ||
46 | $tmph = $pdo->prepare("SELECT subscribers FROM | |
47 | ttrss_feedbrowser_cache WHERE feed_url = ?"); | |
48 | $tmph->execute([$feed_url]); | |
49 | ||
50 | if (!$tmph->fetch()) { | |
51 | ||
52 | $tmph = $pdo->prepare("INSERT INTO ttrss_feedbrowser_cache | |
53 | (feed_url, site_url, title, subscribers) | |
54 | VALUES | |
55 | (?, ?, ?, ?)"); | |
56 | ||
57 | $tmph->execute([$feed_url, $site_url, $title, $subscribers]); | |
58 | ||
59 | ++$count; | |
60 | ||
61 | } | |
62 | ||
63 | } | |
64 | ||
65 | $pdo->commit(); | |
66 | ||
67 | return $count; | |
68 | ||
69 | } | |
70 | ||
71 | static function update_daemon_common($limit = DAEMON_FEED_LIMIT, $debug = true) { | |
72 | $schema_version = get_schema_version(); | |
73 | ||
74 | if ($schema_version != SCHEMA_VERSION) { | |
75 | die("Schema version is wrong, please upgrade the database.\n"); | |
76 | } | |
77 | ||
78 | $pdo = Db::pdo(); | |
79 | ||
80 | if (!SINGLE_USER_MODE && DAEMON_UPDATE_LOGIN_LIMIT > 0) { | |
81 | if (DB_TYPE == "pgsql") { | |
82 | $login_thresh_qpart = "AND ttrss_users.last_login >= NOW() - INTERVAL '".DAEMON_UPDATE_LOGIN_LIMIT." days'"; | |
83 | } else { | |
84 | $login_thresh_qpart = "AND ttrss_users.last_login >= DATE_SUB(NOW(), INTERVAL ".DAEMON_UPDATE_LOGIN_LIMIT." DAY)"; | |
85 | } | |
86 | } else { | |
87 | $login_thresh_qpart = ""; | |
88 | } | |
89 | ||
90 | if (DB_TYPE == "pgsql") { | |
91 | $update_limit_qpart = "AND (( | |
92 | ttrss_feeds.update_interval = 0 | |
93 | AND ttrss_user_prefs.value != '-1' | |
94 | AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_user_prefs.value || ' minutes') AS INTERVAL) | |
95 | ) OR ( | |
96 | ttrss_feeds.update_interval > 0 | |
97 | AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_feeds.update_interval || ' minutes') AS INTERVAL) | |
98 | ) OR (ttrss_feeds.last_updated IS NULL | |
99 | AND ttrss_user_prefs.value != '-1') | |
100 | OR (last_updated = '1970-01-01 00:00:00' | |
101 | AND ttrss_user_prefs.value != '-1'))"; | |
102 | } else { | |
103 | $update_limit_qpart = "AND (( | |
104 | ttrss_feeds.update_interval = 0 | |
105 | AND ttrss_user_prefs.value != '-1' | |
106 | AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL CONVERT(ttrss_user_prefs.value, SIGNED INTEGER) MINUTE) | |
107 | ) OR ( | |
108 | ttrss_feeds.update_interval > 0 | |
109 | AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL ttrss_feeds.update_interval MINUTE) | |
110 | ) OR (ttrss_feeds.last_updated IS NULL | |
111 | AND ttrss_user_prefs.value != '-1') | |
112 | OR (last_updated = '1970-01-01 00:00:00' | |
113 | AND ttrss_user_prefs.value != '-1'))"; | |
114 | } | |
115 | ||
116 | // Test if feed is currently being updated by another process. | |
117 | if (DB_TYPE == "pgsql") { | |
118 | $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < NOW() - INTERVAL '10 minutes')"; | |
119 | } else { | |
120 | $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < DATE_SUB(NOW(), INTERVAL 10 MINUTE))"; | |
121 | } | |
122 | ||
123 | $query_limit = $limit ? sprintf("LIMIT %d", $limit) : ""; | |
124 | ||
125 | // Update the least recently updated feeds first | |
126 | $query_order = "ORDER BY last_updated"; | |
127 | if (DB_TYPE == "pgsql") $query_order .= " NULLS FIRST"; | |
128 | ||
129 | $query = "SELECT DISTINCT ttrss_feeds.feed_url, ttrss_feeds.last_updated | |
130 | FROM | |
131 | ttrss_feeds, ttrss_users, ttrss_user_prefs | |
132 | WHERE | |
133 | ttrss_feeds.owner_uid = ttrss_users.id | |
134 | AND ttrss_user_prefs.profile IS NULL | |
135 | AND ttrss_users.id = ttrss_user_prefs.owner_uid | |
136 | AND ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL' | |
137 | $login_thresh_qpart $update_limit_qpart | |
138 | $updstart_thresh_qpart | |
139 | $query_order $query_limit"; | |
140 | ||
141 | $res = $pdo->query($query); | |
142 | ||
143 | $feeds_to_update = array(); | |
144 | while ($line = $res->fetch()) { | |
145 | array_push($feeds_to_update, $line['feed_url']); | |
146 | } | |
147 | ||
148 | if ($debug) _debug(sprintf("Scheduled %d feeds to update...", count($feeds_to_update))); | |
149 | ||
150 | // Update last_update_started before actually starting the batch | |
151 | // in order to minimize collision risk for parallel daemon tasks | |
152 | if (count($feeds_to_update) > 0) { | |
153 | $feeds_qmarks = arr_qmarks($feeds_to_update); | |
154 | ||
155 | $tmph = $pdo->prepare("UPDATE ttrss_feeds SET last_update_started = NOW() | |
156 | WHERE feed_url IN ($feeds_qmarks)"); | |
157 | $tmph->execute($feeds_to_update); | |
158 | } | |
159 | ||
160 | $nf = 0; | |
161 | $bstarted = microtime(true); | |
162 | ||
163 | $batch_owners = array(); | |
164 | ||
165 | // since we have the data cached, we can deal with other feeds with the same url | |
166 | $usth = $pdo->prepare("SELECT DISTINCT ttrss_feeds.id,last_updated,ttrss_feeds.owner_uid | |
167 | FROM ttrss_feeds, ttrss_users, ttrss_user_prefs WHERE | |
168 | ttrss_user_prefs.owner_uid = ttrss_feeds.owner_uid AND | |
169 | ttrss_users.id = ttrss_user_prefs.owner_uid AND | |
170 | ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL' AND | |
171 | ttrss_user_prefs.profile IS NULL AND | |
172 | feed_url = ? | |
173 | $update_limit_qpart | |
174 | $login_thresh_qpart | |
175 | ORDER BY ttrss_feeds.id $query_limit"); | |
176 | ||
177 | foreach ($feeds_to_update as $feed) { | |
178 | if($debug) _debug("Base feed: $feed"); | |
179 | ||
180 | $usth->execute([$feed]); | |
181 | //update_rss_feed($line["id"], true); | |
182 | ||
183 | if ($tline = $usth->fetch()) { | |
184 | if ($debug) _debug(" => " . $tline["last_updated"] . ", " . $tline["id"] . " " . $tline["owner_uid"]); | |
185 | ||
186 | if (array_search($tline["owner_uid"], $batch_owners) === FALSE) | |
187 | array_push($batch_owners, $tline["owner_uid"]); | |
188 | ||
189 | $fstarted = microtime(true); | |
190 | ||
191 | try { | |
192 | RSSUtils::update_rss_feed($tline["id"], true, false); | |
193 | } catch (PDOException $e) { | |
194 | Logger::get()->log_error(E_USER_NOTICE, $e->getMessage(), $e->getFile(), $e->getLine(), $e->getTraceAsString()); | |
195 | ||
196 | try { | |
197 | $pdo->rollback(); | |
198 | } catch (PDOException $e) { | |
199 | // it doesn't matter if there wasn't actually anything to rollback, PDO Exception can be | |
200 | // thrown outside of an active transaction during feed update | |
201 | } | |
202 | } | |
203 | _debug_suppress(false); | |
204 | ||
205 | _debug(sprintf(" %.4f (sec)", microtime(true) - $fstarted)); | |
206 | ||
207 | ++$nf; | |
208 | } | |
209 | } | |
210 | ||
211 | if ($nf > 0) { | |
212 | _debug(sprintf("Processed %d feeds in %.4f (sec), %.4f (sec/feed avg)", $nf, | |
213 | microtime(true) - $bstarted, (microtime(true) - $bstarted) / $nf)); | |
214 | } | |
215 | ||
216 | foreach ($batch_owners as $owner_uid) { | |
217 | _debug("Running housekeeping tasks for user $owner_uid..."); | |
218 | ||
219 | RSSUtils::housekeeping_user($owner_uid); | |
220 | } | |
221 | ||
222 | // Send feed digests by email if needed. | |
223 | Digest::send_headlines_digests($debug); | |
224 | ||
225 | return $nf; | |
226 | } | |
227 | ||
228 | // this is used when subscribing | |
229 | static function set_basic_feed_info($feed) { | |
230 | ||
231 | $pdo = Db::pdo(); | |
232 | ||
233 | $sth = $pdo->prepare("SELECT owner_uid,feed_url,auth_pass,auth_login | |
234 | FROM ttrss_feeds WHERE id = ?"); | |
235 | $sth->execute([$feed]); | |
236 | ||
237 | if ($row = $sth->fetch()) { | |
238 | ||
239 | $owner_uid = $row["owner_uid"]; | |
240 | $auth_login = $row["auth_login"]; | |
241 | $auth_pass = $row["auth_pass"]; | |
242 | $fetch_url = $row["feed_url"]; | |
243 | ||
244 | $pluginhost = new PluginHost(); | |
245 | $user_plugins = get_pref("_ENABLED_PLUGINS", $owner_uid); | |
246 | ||
247 | $pluginhost->load(PLUGINS, PluginHost::KIND_ALL); | |
248 | $pluginhost->load($user_plugins, PluginHost::KIND_USER, $owner_uid); | |
249 | $pluginhost->load_data(); | |
250 | ||
251 | $basic_info = array(); | |
252 | foreach ($pluginhost->get_hooks(PluginHost::HOOK_FEED_BASIC_INFO) as $plugin) { | |
253 | $basic_info = $plugin->hook_feed_basic_info($basic_info, $fetch_url, $owner_uid, $feed, $auth_login, $auth_pass); | |
254 | } | |
255 | ||
256 | if (!$basic_info) { | |
257 | $feed_data = fetch_file_contents($fetch_url, false, | |
258 | $auth_login, $auth_pass, false, | |
259 | FEED_FETCH_TIMEOUT, | |
260 | 0); | |
261 | ||
262 | global $fetch_curl_used; | |
263 | ||
264 | if (!$fetch_curl_used) { | |
265 | $tmp = @gzdecode($feed_data); | |
266 | ||
267 | if ($tmp) $feed_data = $tmp; | |
268 | } | |
269 | ||
270 | $feed_data = trim($feed_data); | |
271 | ||
272 | $rss = new FeedParser($feed_data); | |
273 | $rss->init(); | |
274 | ||
275 | if (!$rss->error()) { | |
276 | $basic_info = array( | |
277 | 'title' => mb_substr($rss->get_title(), 0, 199), | |
278 | 'site_url' => mb_substr(rewrite_relative_url($fetch_url, $rss->get_link()), 0, 245) | |
279 | ); | |
280 | } | |
281 | } | |
282 | ||
283 | if ($basic_info && is_array($basic_info)) { | |
284 | $sth = $pdo->prepare("SELECT title, site_url FROM ttrss_feeds WHERE id = ?"); | |
285 | $sth->execute([$feed]); | |
286 | ||
287 | if ($row = $sth->fetch()) { | |
288 | ||
289 | $registered_title = $row["title"]; | |
290 | $orig_site_url = $row["site_url"]; | |
291 | ||
292 | if ($basic_info['title'] && (!$registered_title || $registered_title == "[Unknown]")) { | |
293 | ||
294 | $sth = $pdo->prepare("UPDATE ttrss_feeds SET | |
295 | title = ? WHERE id = ?"); | |
296 | $sth->execute([$basic_info['title'], $feed]); | |
297 | } | |
298 | ||
299 | if ($basic_info['site_url'] && $orig_site_url != $basic_info['site_url']) { | |
300 | $sth = $pdo->prepare("UPDATE ttrss_feeds SET | |
301 | site_url = ? WHERE id = ?"); | |
302 | $sth->execute([$basic_info['site_url'], $feed]); | |
303 | } | |
304 | ||
305 | } | |
306 | } | |
307 | } | |
308 | } | |
309 | ||
310 | /** | |
311 | * @SuppressWarnings(PHPMD.UnusedFormalParameter) | |
312 | */ | |
313 | static function update_rss_feed($feed, $no_cache = false) { | |
314 | ||
315 | $debug_enabled = defined('DAEMON_EXTENDED_DEBUG') || clean($_REQUEST['xdebug']); | |
316 | ||
317 | _debug_suppress(!$debug_enabled); | |
318 | _debug("start", $debug_enabled); | |
319 | ||
320 | $pdo = Db::pdo(); | |
321 | ||
322 | $sth = $pdo->prepare("SELECT title FROM ttrss_feeds WHERE id = ?"); | |
323 | $sth->execute([$feed]); | |
324 | ||
325 | if (!$row = $sth->fetch()) { | |
326 | _debug("feed $feed NOT FOUND/SKIPPED", $debug_enabled); | |
327 | user_error("Attempt to update unknown/invalid feed $feed", E_USER_WARNING); | |
328 | return false; | |
329 | } | |
330 | ||
331 | $title = $row["title"]; | |
332 | ||
333 | // feed was batch-subscribed or something, we need to get basic info | |
334 | // this is not optimal currently as it fetches stuff separately TODO: optimize | |
335 | if ($title == "[Unknown]") { | |
336 | _debug("setting basic feed info for $feed..."); | |
337 | RSSUtils::set_basic_feed_info($feed); | |
338 | } | |
339 | ||
340 | $sth = $pdo->prepare("SELECT id,update_interval,auth_login, | |
341 | feed_url,auth_pass,cache_images, | |
342 | mark_unread_on_update, owner_uid, | |
343 | auth_pass_encrypted, feed_language, | |
344 | last_modified, | |
345 | ".SUBSTRING_FOR_DATE."(last_unconditional, 1, 19) AS last_unconditional | |
346 | FROM ttrss_feeds WHERE id = ?"); | |
347 | $sth->execute([$feed]); | |
348 | ||
349 | if ($row = $sth->fetch()) { | |
350 | ||
351 | $owner_uid = $row["owner_uid"]; | |
352 | $mark_unread_on_update = $row["mark_unread_on_update"]; | |
353 | ||
354 | $sth = $pdo->prepare("UPDATE ttrss_feeds SET last_update_started = NOW() | |
355 | WHERE id = ?"); | |
356 | $sth->execute([$feed]); | |
357 | ||
358 | $auth_login = $row["auth_login"]; | |
359 | $auth_pass = $row["auth_pass"]; | |
360 | $stored_last_modified = $row["last_modified"]; | |
361 | $last_unconditional = $row["last_unconditional"]; | |
362 | $cache_images = $row["cache_images"]; | |
363 | $fetch_url = $row["feed_url"]; | |
364 | ||
365 | $feed_language = mb_strtolower($row["feed_language"]); | |
366 | if (!$feed_language) $feed_language = 'english'; | |
367 | ||
368 | } else { | |
369 | return false; | |
370 | } | |
371 | ||
372 | $date_feed_processed = date('Y-m-d H:i'); | |
373 | ||
374 | $cache_filename = CACHE_DIR . "/feeds/" . sha1($fetch_url) . ".xml"; | |
375 | ||
376 | $pluginhost = new PluginHost(); | |
377 | $pluginhost->set_debug($debug_enabled); | |
378 | $user_plugins = get_pref("_ENABLED_PLUGINS", $owner_uid); | |
379 | ||
380 | $pluginhost->load(PLUGINS, PluginHost::KIND_ALL); | |
381 | $pluginhost->load($user_plugins, PluginHost::KIND_USER, $owner_uid); | |
382 | $pluginhost->load_data(); | |
383 | ||
384 | $rss_hash = false; | |
385 | ||
386 | $force_refetch = isset($_REQUEST["force_refetch"]); | |
387 | $feed_data = ""; | |
388 | ||
389 | foreach ($pluginhost->get_hooks(PluginHost::HOOK_FETCH_FEED) as $plugin) { | |
390 | $feed_data = $plugin->hook_fetch_feed($feed_data, $fetch_url, $owner_uid, $feed, 0, $auth_login, $auth_pass); | |
391 | } | |
392 | ||
393 | // try cache | |
394 | if (!$feed_data && | |
395 | file_exists($cache_filename) && | |
396 | is_readable($cache_filename) && | |
397 | !$auth_login && !$auth_pass && | |
398 | filemtime($cache_filename) > time() - 30) { | |
399 | ||
400 | _debug("using local cache [$cache_filename].", $debug_enabled); | |
401 | ||
402 | @$feed_data = file_get_contents($cache_filename); | |
403 | ||
404 | if ($feed_data) { | |
405 | $rss_hash = sha1($feed_data); | |
406 | } | |
407 | ||
408 | } else { | |
409 | _debug("local cache will not be used for this feed", $debug_enabled); | |
410 | } | |
411 | ||
412 | global $fetch_last_modified; | |
413 | ||
414 | // fetch feed from source | |
415 | if (!$feed_data) { | |
416 | _debug("last unconditional update request: $last_unconditional"); | |
417 | ||
418 | if (ini_get("open_basedir") && function_exists("curl_init")) { | |
419 | _debug("not using CURL due to open_basedir restrictions"); | |
420 | } | |
421 | ||
422 | if (time() - strtotime($last_unconditional) > MAX_CONDITIONAL_INTERVAL) { | |
423 | _debug("maximum allowed interval for conditional requests exceeded, forcing refetch"); | |
424 | ||
425 | $force_refetch = true; | |
426 | } else { | |
427 | _debug("stored last modified for conditional request: $stored_last_modified", $debug_enabled); | |
428 | } | |
429 | ||
430 | _debug("fetching [$fetch_url] (force_refetch: $force_refetch)...", $debug_enabled); | |
431 | ||
432 | $feed_data = fetch_file_contents([ | |
433 | "url" => $fetch_url, | |
434 | "login" => $auth_login, | |
435 | "pass" => $auth_pass, | |
436 | "timeout" => $no_cache ? FEED_FETCH_NO_CACHE_TIMEOUT : FEED_FETCH_TIMEOUT, | |
437 | "last_modified" => $force_refetch ? "" : $stored_last_modified | |
438 | ]); | |
439 | ||
440 | global $fetch_curl_used; | |
441 | ||
442 | if (!$fetch_curl_used) { | |
443 | $tmp = @gzdecode($feed_data); | |
444 | ||
445 | if ($tmp) $feed_data = $tmp; | |
446 | } | |
447 | ||
448 | $feed_data = trim($feed_data); | |
449 | ||
450 | _debug("fetch done.", $debug_enabled); | |
451 | _debug("source last modified: " . $fetch_last_modified, $debug_enabled); | |
452 | ||
453 | if ($feed_data && $fetch_last_modified != $stored_last_modified) { | |
454 | $sth = $pdo->prepare("UPDATE ttrss_feeds SET last_modified = ? WHERE id = ?"); | |
455 | $sth->execute([substr($fetch_last_modified, 0, 245), $feed]); | |
456 | } | |
457 | ||
458 | // cache vanilla feed data for re-use | |
459 | if ($feed_data && !$auth_pass && !$auth_login && is_writable(CACHE_DIR . "/feeds")) { | |
460 | $new_rss_hash = sha1($feed_data); | |
461 | ||
462 | if ($new_rss_hash != $rss_hash) { | |
463 | _debug("saving $cache_filename", $debug_enabled); | |
464 | @file_put_contents($cache_filename, $feed_data); | |
465 | } | |
466 | } | |
467 | } | |
468 | ||
469 | if (!$feed_data) { | |
470 | global $fetch_last_error; | |
471 | global $fetch_last_error_code; | |
472 | ||
473 | _debug("unable to fetch: $fetch_last_error [$fetch_last_error_code]", $debug_enabled); | |
474 | ||
475 | // If-Modified-Since | |
476 | if ($fetch_last_error_code != 304) { | |
477 | $error_message = $fetch_last_error; | |
478 | } else { | |
479 | _debug("source claims data not modified, nothing to do.", $debug_enabled); | |
480 | $error_message = ""; | |
481 | } | |
482 | ||
483 | $sth = $pdo->prepare("UPDATE ttrss_feeds SET last_error = ?, | |
484 | last_updated = NOW() WHERE id = ?"); | |
485 | $sth->execute([$error_message, $feed]); | |
486 | ||
487 | return; | |
488 | } | |
489 | ||
490 | foreach ($pluginhost->get_hooks(PluginHost::HOOK_FEED_FETCHED) as $plugin) { | |
491 | $feed_data = $plugin->hook_feed_fetched($feed_data, $fetch_url, $owner_uid, $feed); | |
492 | } | |
493 | ||
494 | $rss = new FeedParser($feed_data); | |
495 | $rss->init(); | |
496 | ||
497 | if (!$rss->error()) { | |
498 | ||
499 | // We use local pluginhost here because we need to load different per-user feed plugins | |
500 | $pluginhost->run_hooks(PluginHost::HOOK_FEED_PARSED, "hook_feed_parsed", $rss); | |
501 | ||
502 | _debug("language: $feed_language", $debug_enabled); | |
503 | _debug("processing feed data...", $debug_enabled); | |
504 | ||
505 | if (DB_TYPE == "pgsql") { | |
506 | $favicon_interval_qpart = "favicon_last_checked < NOW() - INTERVAL '12 hour'"; | |
507 | } else { | |
508 | $favicon_interval_qpart = "favicon_last_checked < DATE_SUB(NOW(), INTERVAL 12 HOUR)"; | |
509 | } | |
510 | ||
511 | $sth = $pdo->prepare("SELECT owner_uid,favicon_avg_color, | |
512 | (favicon_last_checked IS NULL OR $favicon_interval_qpart) AS | |
513 | favicon_needs_check | |
514 | FROM ttrss_feeds WHERE id = ?"); | |
515 | $sth->execute([$feed]); | |
516 | ||
517 | if ($row = $sth->fetch()) { | |
518 | $favicon_needs_check = $row["favicon_needs_check"]; | |
519 | $favicon_avg_color = $row["favicon_avg_color"]; | |
520 | $owner_uid = $row["owner_uid"]; | |
521 | } else { | |
522 | return false; | |
523 | } | |
524 | ||
525 | $site_url = mb_substr(rewrite_relative_url($fetch_url, $rss->get_link()), 0, 245); | |
526 | ||
527 | _debug("site_url: $site_url", $debug_enabled); | |
528 | _debug("feed_title: " . $rss->get_title(), $debug_enabled); | |
529 | ||
530 | if ($favicon_needs_check || $force_refetch) { | |
531 | ||
532 | /* terrible hack: if we crash on floicon shit here, we won't check | |
533 | * the icon avgcolor again (unless the icon got updated) */ | |
534 | ||
535 | $favicon_file = ICONS_DIR . "/$feed.ico"; | |
536 | $favicon_modified = @filemtime($favicon_file); | |
537 | ||
538 | _debug("checking favicon...", $debug_enabled); | |
539 | ||
540 | RSSUtils::check_feed_favicon($site_url, $feed); | |
541 | $favicon_modified_new = @filemtime($favicon_file); | |
542 | ||
543 | if ($favicon_modified_new > $favicon_modified) | |
544 | $favicon_avg_color = ''; | |
545 | ||
546 | $favicon_colorstring = ""; | |
547 | if (file_exists($favicon_file) && function_exists("imagecreatefromstring") && $favicon_avg_color == '') { | |
548 | require_once "colors.php"; | |
549 | ||
550 | $sth = $pdo->prepare("UPDATE ttrss_feeds SET favicon_avg_color = 'fail' WHERE | |
551 | id = ?"); | |
552 | $sth->execute([$feed]); | |
553 | ||
554 | $favicon_color = calculate_avg_color($favicon_file); | |
555 | ||
556 | $favicon_colorstring = ",favicon_avg_color = " . $pdo->quote($favicon_color); | |
557 | ||
558 | } else if ($favicon_avg_color == 'fail') { | |
559 | _debug("floicon failed on this file, not trying to recalculate avg color", $debug_enabled); | |
560 | } | |
561 | ||
562 | $sth = $pdo->prepare("UPDATE ttrss_feeds SET favicon_last_checked = NOW() | |
563 | $favicon_colorstring WHERE id = ?"); | |
564 | $sth->execute([$feed]); | |
565 | } | |
566 | ||
567 | _debug("loading filters & labels...", $debug_enabled); | |
568 | ||
569 | $filters = load_filters($feed, $owner_uid); | |
570 | ||
571 | if ($debug_enabled) { | |
572 | print_r($filters); | |
573 | } | |
574 | ||
575 | _debug("" . count($filters) . " filters loaded.", $debug_enabled); | |
576 | ||
577 | $items = $rss->get_items(); | |
578 | ||
579 | if (!is_array($items)) { | |
580 | _debug("no articles found.", $debug_enabled); | |
581 | ||
582 | $sth = $pdo->prepare("UPDATE ttrss_feeds | |
583 | SET last_updated = NOW(), last_unconditional = NOW(), last_error = '' WHERE id = ?"); | |
584 | $sth->execute([$feed]); | |
585 | ||
586 | return true; // no articles | |
587 | } | |
588 | ||
589 | _debug("processing articles...", $debug_enabled); | |
590 | ||
591 | $tstart = time(); | |
592 | ||
593 | foreach ($items as $item) { | |
594 | $pdo->beginTransaction(); | |
595 | ||
596 | if (clean($_REQUEST['xdebug']) == 3) { | |
597 | print_r($item); | |
598 | } | |
599 | ||
600 | if (ini_get("max_execution_time") > 0 && time() - $tstart >= ini_get("max_execution_time") * 0.7) { | |
601 | _debug("looks like there's too many articles to process at once, breaking out", $debug_enabled); | |
602 | $pdo->commit(); | |
603 | break; | |
604 | } | |
605 | ||
606 | $entry_guid = strip_tags($item->get_id()); | |
607 | if (!$entry_guid) $entry_guid = strip_tags($item->get_link()); | |
608 | if (!$entry_guid) $entry_guid = RSSUtils::make_guid_from_title($item->get_title()); | |
609 | ||
610 | if (!$entry_guid) { | |
611 | $pdo->commit(); | |
612 | continue; | |
613 | } | |
614 | ||
615 | $entry_guid = "$owner_uid,$entry_guid"; | |
616 | ||
617 | $entry_guid_hashed = 'SHA1:' . sha1($entry_guid); | |
618 | ||
619 | _debug("guid $entry_guid / $entry_guid_hashed", $debug_enabled); | |
620 | ||
621 | $entry_timestamp = strip_tags($item->get_date()); | |
622 | ||
623 | _debug("orig date: " . $item->get_date(), $debug_enabled); | |
624 | ||
625 | if ($entry_timestamp == -1 || !$entry_timestamp || $entry_timestamp > time()) { | |
626 | $entry_timestamp = time(); | |
627 | } | |
628 | ||
629 | $entry_timestamp_fmt = strftime("%Y/%m/%d %H:%M:%S", $entry_timestamp); | |
630 | ||
631 | _debug("date $entry_timestamp [$entry_timestamp_fmt]", $debug_enabled); | |
632 | ||
633 | $entry_title = strip_tags($item->get_title()); | |
634 | ||
635 | $entry_link = rewrite_relative_url($site_url, $item->get_link()); | |
636 | ||
637 | $entry_language = mb_substr(trim($item->get_language()), 0, 2); | |
638 | ||
639 | _debug("title $entry_title", $debug_enabled); | |
640 | _debug("link $entry_link", $debug_enabled); | |
641 | _debug("language $entry_language", $debug_enabled); | |
642 | ||
643 | if (!$entry_title) $entry_title = date("Y-m-d H:i:s", $entry_timestamp);; | |
644 | ||
645 | $entry_content = $item->get_content(); | |
646 | if (!$entry_content) $entry_content = $item->get_description(); | |
647 | ||
648 | if (clean($_REQUEST["xdebug"]) == 2) { | |
649 | print "content: "; | |
650 | print htmlspecialchars($entry_content); | |
651 | print "\n"; | |
652 | } | |
653 | ||
654 | $entry_comments = mb_substr(strip_tags($item->get_comments_url()), 0, 245); | |
655 | $num_comments = (int) $item->get_comments_count(); | |
656 | ||
657 | $entry_author = strip_tags($item->get_author()); | |
658 | $entry_guid = mb_substr($entry_guid, 0, 245); | |
659 | ||
660 | _debug("author $entry_author", $debug_enabled); | |
661 | _debug("num_comments: $num_comments", $debug_enabled); | |
662 | _debug("looking for tags...", $debug_enabled); | |
663 | ||
664 | // parse <category> entries into tags | |
665 | ||
666 | $additional_tags = array(); | |
667 | ||
668 | $additional_tags_src = $item->get_categories(); | |
669 | ||
670 | if (is_array($additional_tags_src)) { | |
671 | foreach ($additional_tags_src as $tobj) { | |
672 | array_push($additional_tags, $tobj); | |
673 | } | |
674 | } | |
675 | ||
676 | $entry_tags = array_unique($additional_tags); | |
677 | ||
678 | for ($i = 0; $i < count($entry_tags); $i++) { | |
679 | $entry_tags[$i] = mb_strtolower($entry_tags[$i], 'utf-8'); | |
680 | ||
681 | // we don't support numeric tags, let's prefix them | |
682 | if (is_numeric($entry_tags[$i])) $entry_tags[$i] = 't:' . $entry_tags[$i]; | |
683 | } | |
684 | ||
685 | _debug("tags found: " . join(",", $entry_tags), $debug_enabled); | |
686 | ||
687 | _debug("done collecting data.", $debug_enabled); | |
688 | ||
689 | $sth = $pdo->prepare("SELECT id, content_hash, lang FROM ttrss_entries | |
690 | WHERE guid = ? OR guid = ?"); | |
691 | $sth->execute([$entry_guid, $entry_guid_hashed]); | |
692 | ||
693 | if ($row = $sth->fetch()) { | |
694 | $base_entry_id = $row["id"]; | |
695 | $entry_stored_hash = $row["content_hash"]; | |
696 | $article_labels = Article::get_article_labels($base_entry_id, $owner_uid); | |
697 | ||
698 | $existing_tags = Article::get_article_tags($base_entry_id, $owner_uid); | |
699 | $entry_tags = array_unique(array_merge($entry_tags, $existing_tags)); | |
700 | } else { | |
701 | $base_entry_id = false; | |
702 | $entry_stored_hash = ""; | |
703 | $article_labels = array(); | |
704 | } | |
705 | ||
706 | $article = array("owner_uid" => $owner_uid, // read only | |
707 | "guid" => $entry_guid, // read only | |
708 | "guid_hashed" => $entry_guid_hashed, // read only | |
709 | "title" => $entry_title, | |
710 | "content" => $entry_content, | |
711 | "link" => $entry_link, | |
712 | "labels" => $article_labels, // current limitation: can add labels to article, can't remove them | |
713 | "tags" => $entry_tags, | |
714 | "author" => $entry_author, | |
715 | "force_catchup" => false, // ugly hack for the time being | |
716 | "score_modifier" => 0, // no previous value, plugin should recalculate score modifier based on content if needed | |
717 | "language" => $entry_language, | |
718 | "num_comments" => $num_comments, // read only | |
719 | "feed" => array("id" => $feed, | |
720 | "fetch_url" => $fetch_url, | |
721 | "site_url" => $site_url, | |
722 | "cache_images" => $cache_images) | |
723 | ); | |
724 | ||
725 | $entry_plugin_data = ""; | |
726 | $entry_current_hash = RSSUtils::calculate_article_hash($article, $pluginhost); | |
727 | ||
728 | _debug("article hash: $entry_current_hash [stored=$entry_stored_hash]", $debug_enabled); | |
729 | ||
730 | if ($entry_current_hash == $entry_stored_hash && !isset($_REQUEST["force_rehash"])) { | |
731 | _debug("stored article seems up to date [IID: $base_entry_id], updating timestamp only", $debug_enabled); | |
732 | ||
733 | // we keep encountering the entry in feeds, so we need to | |
734 | // update date_updated column so that we don't get horrible | |
735 | // dupes when the entry gets purged and reinserted again e.g. | |
736 | // in the case of SLOW SLOW OMG SLOW updating feeds | |
737 | ||
738 | $sth = $pdo->prepare("UPDATE ttrss_entries SET date_updated = NOW() | |
739 | WHERE id = ?"); | |
740 | $sth->execute([$base_entry_id]); | |
741 | ||
742 | $pdo->commit(); | |
743 | continue; | |
744 | } | |
745 | ||
746 | _debug("hash differs, applying plugin filters:", $debug_enabled); | |
747 | ||
748 | foreach ($pluginhost->get_hooks(PluginHost::HOOK_ARTICLE_FILTER) as $plugin) { | |
749 | _debug("... " . get_class($plugin), $debug_enabled); | |
750 | ||
751 | $start = microtime(true); | |
752 | $article = $plugin->hook_article_filter($article); | |
753 | ||
754 | _debug("=== " . sprintf("%.4f (sec)", microtime(true) - $start), $debug_enabled); | |
755 | ||
756 | $entry_plugin_data .= mb_strtolower(get_class($plugin)) . ","; | |
757 | } | |
758 | ||
759 | if (clean($_REQUEST["xdebug"]) == 2) { | |
760 | print "processed content: "; | |
761 | print htmlspecialchars($article["content"]); | |
762 | print "\n"; | |
763 | } | |
764 | ||
765 | _debug("plugin data: $entry_plugin_data", $debug_enabled); | |
766 | ||
767 | // Workaround: 4-byte unicode requires utf8mb4 in MySQL. See https://tt-rss.org/forum/viewtopic.php?f=1&t=3377&p=20077#p20077 | |
768 | if (DB_TYPE == "mysql" && MYSQL_CHARSET != "UTF8MB4") { | |
769 | foreach ($article as $k => $v) { | |
770 | // i guess we'll have to take the risk of 4byte unicode labels & tags here | |
771 | if (is_string($article[$k])) { | |
772 | $article[$k] = RSSUtils::strip_utf8mb4($v); | |
773 | } | |
774 | } | |
775 | } | |
776 | ||
777 | /* Collect article tags here so we could filter by them: */ | |
778 | ||
779 | $matched_rules = array(); | |
780 | ||
781 | $article_filters = RSSUtils::get_article_filters($filters, $article["title"], | |
782 | $article["content"], $article["link"], $article["author"], | |
783 | $article["tags"], $matched_rules); | |
784 | ||
785 | if ($debug_enabled) { | |
786 | _debug("matched filter rules: ", $debug_enabled); | |
787 | ||
788 | if (count($matched_rules) != 0) { | |
789 | print_r($matched_rules); | |
790 | } | |
791 | ||
792 | _debug("filter actions: ", $debug_enabled); | |
793 | ||
794 | if (count($article_filters) != 0) { | |
795 | print_r($article_filters); | |
796 | } | |
797 | } | |
798 | ||
799 | $plugin_filter_names = RSSUtils::find_article_filters($article_filters, "plugin"); | |
800 | $plugin_filter_actions = $pluginhost->get_filter_actions(); | |
801 | ||
802 | if (count($plugin_filter_names) > 0) { | |
803 | _debug("applying plugin filter actions...", $debug_enabled); | |
804 | ||
805 | foreach ($plugin_filter_names as $pfn) { | |
806 | list($pfclass,$pfaction) = explode(":", $pfn["param"]); | |
807 | ||
808 | if (isset($plugin_filter_actions[$pfclass])) { | |
809 | $plugin = $pluginhost->get_plugin($pfclass); | |
810 | ||
811 | _debug("... $pfclass: $pfaction", $debug_enabled); | |
812 | ||
813 | if ($plugin) { | |
814 | $start = microtime(true); | |
815 | $article = $plugin->hook_article_filter_action($article, $pfaction); | |
816 | ||
817 | _debug("=== " . sprintf("%.4f (sec)", microtime(true) - $start), $debug_enabled); | |
818 | } else { | |
819 | _debug("??? $pfclass: plugin object not found."); | |
820 | } | |
821 | } else { | |
822 | _debug("??? $pfclass: filter plugin not registered."); | |
823 | } | |
824 | } | |
825 | } | |
826 | ||
827 | $entry_tags = $article["tags"]; | |
828 | $entry_title = strip_tags($article["title"]); | |
829 | $entry_author = mb_substr(strip_tags($article["author"]), 0, 245); | |
830 | $entry_link = strip_tags($article["link"]); | |
831 | $entry_content = $article["content"]; // escaped below | |
832 | $entry_force_catchup = $article["force_catchup"]; | |
833 | $article_labels = $article["labels"]; | |
834 | $entry_score_modifier = (int) $article["score_modifier"]; | |
835 | $entry_language = $article["language"]; | |
836 | ||
837 | if ($debug_enabled) { | |
838 | _debug("article labels:", $debug_enabled); | |
839 | ||
840 | if (count($article_labels) != 0) { | |
841 | print_r($article_labels); | |
842 | } | |
843 | } | |
844 | ||
845 | _debug("force catchup: $entry_force_catchup"); | |
846 | ||
847 | if ($cache_images && is_writable(CACHE_DIR . '/images')) | |
848 | RSSUtils::cache_media($entry_content, $site_url, $debug_enabled); | |
849 | ||
850 | $csth = $pdo->prepare("SELECT id FROM ttrss_entries | |
851 | WHERE guid = ? OR guid = ?"); | |
852 | $csth->execute([$entry_guid, $entry_guid_hashed]); | |
853 | ||
854 | if (!$row = $csth->fetch()) { | |
855 | ||
856 | _debug("base guid [$entry_guid or $entry_guid_hashed] not found, creating...", $debug_enabled); | |
857 | ||
858 | // base post entry does not exist, create it | |
859 | ||
860 | $usth = $pdo->prepare( | |
861 | "INSERT INTO ttrss_entries | |
862 | (title, | |
863 | guid, | |
864 | link, | |
865 | updated, | |
866 | content, | |
867 | content_hash, | |
868 | no_orig_date, | |
869 | date_updated, | |
870 | date_entered, | |
871 | comments, | |
872 | num_comments, | |
873 | plugin_data, | |
874 | lang, | |
875 | author) | |
876 | VALUES | |
877 | (?, ?, ?, ?, ?, ?, | |
878 | false, | |
879 | NOW(), | |
880 | ?, ?, ?, ?, ?, ?)"); | |
881 | ||
882 | $usth->execute([$entry_title, | |
883 | $entry_guid_hashed, | |
884 | $entry_link, | |
885 | $entry_timestamp_fmt, | |
886 | "$entry_content", | |
887 | $entry_current_hash, | |
888 | $date_feed_processed, | |
889 | $entry_comments, | |
890 | (int)$num_comments, | |
891 | $entry_plugin_data, | |
892 | "$entry_language", | |
893 | "$entry_author"]); | |
894 | ||
895 | } | |
896 | ||
897 | $csth->execute([$entry_guid, $entry_guid_hashed]); | |
898 | ||
899 | $entry_ref_id = 0; | |
900 | $entry_int_id = 0; | |
901 | ||
902 | if ($row = $csth->fetch()) { | |
903 | ||
904 | _debug("base guid found, checking for user record", $debug_enabled); | |
905 | ||
906 | $ref_id = $row['id']; | |
907 | $entry_ref_id = $ref_id; | |
908 | ||
909 | if (RSSUtils::find_article_filter($article_filters, "filter")) { | |
910 | $pdo->commit(); | |
911 | continue; | |
912 | } | |
913 | ||
914 | $score = RSSUtils::calculate_article_score($article_filters) + $entry_score_modifier; | |
915 | ||
916 | _debug("initial score: $score [including plugin modifier: $entry_score_modifier]", $debug_enabled); | |
917 | ||
918 | // check for user post link to main table | |
919 | ||
920 | $sth = $pdo->prepare("SELECT ref_id, int_id FROM ttrss_user_entries WHERE | |
921 | ref_id = ? AND owner_uid = ?"); | |
922 | $sth->execute([$ref_id, $owner_uid]); | |
923 | ||
924 | // okay it doesn't exist - create user entry | |
925 | if ($row = $sth->fetch()) { | |
926 | $entry_ref_id = $row["ref_id"]; | |
927 | $entry_int_id = $row["int_id"]; | |
928 | ||
929 | _debug("user record FOUND: RID: $entry_ref_id, IID: $entry_int_id", $debug_enabled); | |
930 | } else { | |
931 | ||
932 | _debug("user record not found, creating...", $debug_enabled); | |
933 | ||
934 | if ($score >= -500 && !RSSUtils::find_article_filter($article_filters, 'catchup') && !$entry_force_catchup) { | |
935 | $unread = 1; | |
936 | $last_read_qpart = null; | |
937 | } else { | |
938 | $unread = 0; | |
939 | $last_read_qpart = date("Y-m-d H:i"); // we can't use NOW() here because it gets quoted | |
940 | } | |
941 | ||
942 | if (RSSUtils::find_article_filter($article_filters, 'mark') || $score > 1000) { | |
943 | $marked = 1; | |
944 | } else { | |
945 | $marked = 0; | |
946 | } | |
947 | ||
948 | if (RSSUtils::find_article_filter($article_filters, 'publish')) { | |
949 | $published = 1; | |
950 | } else { | |
951 | $published = 0; | |
952 | } | |
953 | ||
954 | $last_marked = ($marked == 1) ? 'NOW()' : 'NULL'; | |
955 | $last_published = ($published == 1) ? 'NOW()' : 'NULL'; | |
956 | ||
957 | $sth = $pdo->prepare( | |
958 | "INSERT INTO ttrss_user_entries | |
959 | (ref_id, owner_uid, feed_id, unread, last_read, marked, | |
960 | published, score, tag_cache, label_cache, uuid, | |
961 | last_marked, last_published) | |
962 | VALUES (?, ?, ?, ?, ?, ?, ?, ?, '', '', '', ".$last_marked.", ".$last_published.")"); | |
963 | ||
964 | $sth->execute([$ref_id, $owner_uid, $feed, $unread, $last_read_qpart, $marked, | |
965 | $published, $score]); | |
966 | ||
967 | $sth = $pdo->prepare("SELECT int_id FROM ttrss_user_entries WHERE | |
968 | ref_id = ? AND owner_uid = ? AND | |
969 | feed_id = ? LIMIT 1"); | |
970 | ||
971 | $sth->execute([$ref_id, $owner_uid, $feed]); | |
972 | ||
973 | if ($row = $sth->fetch()) | |
974 | $entry_int_id = $row['int_id']; | |
975 | } | |
976 | ||
977 | _debug("resulting RID: $entry_ref_id, IID: $entry_int_id", $debug_enabled); | |
978 | ||
979 | if (DB_TYPE == "pgsql") | |
980 | $tsvector_qpart = "tsvector_combined = to_tsvector(:ts_lang, :ts_content),"; | |
981 | else | |
982 | $tsvector_qpart = ""; | |
983 | ||
984 | $sth = $pdo->prepare("UPDATE ttrss_entries | |
985 | SET title = :title, | |
986 | $tsvector_qpart | |
987 | content = :content, | |
988 | content_hash = :content_hash, | |
989 | updated = :updated, | |
990 | date_updated = NOW(), | |
991 | num_comments = :num_comments, | |
992 | plugin_data = :plugin_data, | |
993 | author = :author, | |
994 | lang = :lang | |
995 | WHERE id = :id"); | |
996 | ||
997 | $params = [":title" => $entry_title, | |
998 | ":content" => "$entry_content", | |
999 | ":content_hash" => $entry_current_hash, | |
1000 | ":updated" => $entry_timestamp_fmt, | |
1001 | ":num_comments" => (int)$num_comments, | |
1002 | ":plugin_data" => $entry_plugin_data, | |
1003 | ":author" => "$entry_author", | |
1004 | ":lang" => $entry_language, | |
1005 | ":id" => $ref_id]; | |
1006 | ||
1007 | if (DB_TYPE == "pgsql") { | |
1008 | $params[":ts_lang"] = $feed_language; | |
1009 | $params[":ts_content"] = mb_substr(strip_tags($entry_title . " " . $entry_content), 0, 900000); | |
1010 | } | |
1011 | ||
1012 | $sth->execute($params); | |
1013 | ||
1014 | // update aux data | |
1015 | $sth = $pdo->prepare("UPDATE ttrss_user_entries | |
1016 | SET score = ? WHERE ref_id = ?"); | |
1017 | $sth->execute([$score, $ref_id]); | |
1018 | ||
1019 | if ($mark_unread_on_update) { | |
1020 | _debug("article updated, marking unread as requested.", $debug_enabled); | |
1021 | ||
1022 | $sth = $pdo->prepare("UPDATE ttrss_user_entries | |
1023 | SET last_read = null, unread = true WHERE ref_id = ?"); | |
1024 | $sth->execute([$ref_id]); | |
1025 | } | |
1026 | } | |
1027 | ||
1028 | _debug("assigning labels [other]...", $debug_enabled); | |
1029 | ||
1030 | foreach ($article_labels as $label) { | |
1031 | Labels::add_article($entry_ref_id, $label[1], $owner_uid); | |
1032 | } | |
1033 | ||
1034 | _debug("assigning labels [filters]...", $debug_enabled); | |
1035 | ||
1036 | RSSUtils::assign_article_to_label_filters($entry_ref_id, $article_filters, | |
1037 | $owner_uid, $article_labels); | |
1038 | ||
1039 | _debug("looking for enclosures...", $debug_enabled); | |
1040 | ||
1041 | // enclosures | |
1042 | ||
1043 | $enclosures = array(); | |
1044 | ||
1045 | $encs = $item->get_enclosures(); | |
1046 | ||
1047 | if (is_array($encs)) { | |
1048 | foreach ($encs as $e) { | |
1049 | $e_item = array( | |
1050 | rewrite_relative_url($site_url, $e->link), | |
1051 | $e->type, $e->length, $e->title, $e->width, $e->height); | |
1052 | ||
1053 | // Yet another episode of "mysql utf8_general_ci is gimped" | |
1054 | if (DB_TYPE == "mysql" && MYSQL_CHARSET != "UTF8MB4") { | |
1055 | for ($i = 0; $i < count($e_item); $i++) { | |
1056 | if (is_string($e_item[$i])) { | |
1057 | $e_item[$i] = RSSUtils::strip_utf8mb4($e_item[$i]); | |
1058 | } | |
1059 | } | |
1060 | } | |
1061 | ||
1062 | array_push($enclosures, $e_item); | |
1063 | } | |
1064 | } | |
1065 | ||
1066 | if ($cache_images && is_writable(CACHE_DIR . '/images')) | |
1067 | RSSUtils::cache_enclosures($enclosures, $site_url, $debug_enabled); | |
1068 | ||
1069 | if ($debug_enabled) { | |
1070 | _debug("article enclosures:", $debug_enabled); | |
1071 | print_r($enclosures); | |
1072 | } | |
1073 | ||
1074 | $esth = $pdo->prepare("SELECT id FROM ttrss_enclosures | |
1075 | WHERE content_url = ? AND content_type = ? AND post_id = ?"); | |
1076 | ||
1077 | $usth = $pdo->prepare("INSERT INTO ttrss_enclosures | |
1078 | (content_url, content_type, title, duration, post_id, width, height) VALUES | |
1079 | (?, ?, ?, ?, ?, ?, ?)"); | |
1080 | ||
1081 | foreach ($enclosures as $enc) { | |
1082 | $enc_url = $enc[0]; | |
1083 | $enc_type = $enc[1]; | |
1084 | $enc_dur = (int)$enc[2]; | |
1085 | $enc_title = $enc[3]; | |
1086 | $enc_width = intval($enc[4]); | |
1087 | $enc_height = intval($enc[5]); | |
1088 | ||
1089 | $esth->execute([$enc_url, $enc_type, $entry_ref_id]); | |
1090 | ||
1091 | if (!$esth->fetch()) { | |
1092 | $usth->execute([$enc_url, $enc_type, (string)$enc_title, $enc_dur, $entry_ref_id, $enc_width, $enc_height]); | |
1093 | } | |
1094 | } | |
1095 | ||
1096 | // check for manual tags (we have to do it here since they're loaded from filters) | |
1097 | ||
1098 | foreach ($article_filters as $f) { | |
1099 | if ($f["type"] == "tag") { | |
1100 | ||
1101 | $manual_tags = trim_array(explode(",", $f["param"])); | |
1102 | ||
1103 | foreach ($manual_tags as $tag) { | |
1104 | if (tag_is_valid($tag)) { | |
1105 | array_push($entry_tags, $tag); | |
1106 | } | |
1107 | } | |
1108 | } | |
1109 | } | |
1110 | ||
1111 | // Skip boring tags | |
1112 | ||
1113 | $boring_tags = trim_array(explode(",", mb_strtolower(get_pref( | |
1114 | 'BLACKLISTED_TAGS', $owner_uid, ''), 'utf-8'))); | |
1115 | ||
1116 | $filtered_tags = array(); | |
1117 | $tags_to_cache = array(); | |
1118 | ||
1119 | if ($entry_tags && is_array($entry_tags)) { | |
1120 | foreach ($entry_tags as $tag) { | |
1121 | if (array_search($tag, $boring_tags) === false) { | |
1122 | array_push($filtered_tags, $tag); | |
1123 | } | |
1124 | } | |
1125 | } | |
1126 | ||
1127 | $filtered_tags = array_unique($filtered_tags); | |
1128 | ||
1129 | if ($debug_enabled) { | |
1130 | _debug("filtered article tags:", $debug_enabled); | |
1131 | print_r($filtered_tags); | |
1132 | } | |
1133 | ||
1134 | // Save article tags in the database | |
1135 | ||
1136 | if (count($filtered_tags) > 0) { | |
1137 | ||
1138 | $tsth = $pdo->prepare("SELECT id FROM ttrss_tags | |
1139 | WHERE tag_name = ? AND post_int_id = ? AND | |
1140 | owner_uid = ? LIMIT 1"); | |
1141 | ||
1142 | $usth = $pdo->prepare("INSERT INTO ttrss_tags | |
1143 | (owner_uid,tag_name,post_int_id) | |
1144 | VALUES (?, ?, ?)"); | |
1145 | ||
1146 | foreach ($filtered_tags as $tag) { | |
1147 | ||
1148 | $tag = sanitize_tag($tag); | |
1149 | ||
1150 | if (!tag_is_valid($tag)) continue; | |
1151 | ||
1152 | $tsth->execute([$tag, $entry_int_id, $owner_uid]); | |
1153 | ||
1154 | if (!$tsth->fetch()) { | |
1155 | $usth->execute([$owner_uid, $tag, $entry_int_id]); | |
1156 | } | |
1157 | ||
1158 | array_push($tags_to_cache, $tag); | |
1159 | } | |
1160 | ||
1161 | /* update the cache */ | |
1162 | ||
1163 | $tags_to_cache = array_unique($tags_to_cache); | |
1164 | ||
1165 | $tags_str = join(",", $tags_to_cache); | |
1166 | ||
1167 | $tsth = $pdo->prepare("UPDATE ttrss_user_entries | |
1168 | SET tag_cache = ? WHERE ref_id = ? | |
1169 | AND owner_uid = ?"); | |
1170 | $tsth->execute([$tags_str, $entry_ref_id, $owner_uid]); | |
1171 | } | |
1172 | ||
1173 | _debug("article processed", $debug_enabled); | |
1174 | ||
1175 | $pdo->commit(); | |
1176 | } | |
1177 | ||
1178 | _debug("purging feed...", $debug_enabled); | |
1179 | ||
1180 | purge_feed($feed, 0, $debug_enabled); | |
1181 | ||
1182 | $sth = $pdo->prepare("UPDATE ttrss_feeds | |
1183 | SET last_updated = NOW(), last_unconditional = NOW(), last_error = '' WHERE id = ?"); | |
1184 | $sth->execute([$feed]); | |
1185 | ||
1186 | } else { | |
1187 | ||
1188 | $error_msg = mb_substr($rss->error(), 0, 245); | |
1189 | ||
1190 | _debug("fetch error: $error_msg", $debug_enabled); | |
1191 | ||
1192 | if (count($rss->errors()) > 1) { | |
1193 | foreach ($rss->errors() as $error) { | |
1194 | _debug("+ $error"); | |
1195 | } | |
1196 | } | |
1197 | ||
1198 | $sth = $pdo->prepare("UPDATE ttrss_feeds SET last_error = ?, | |
1199 | last_updated = NOW(), last_unconditional = NOW() WHERE id = ?"); | |
1200 | $sth->execute([$error_msg, $feed]); | |
1201 | ||
1202 | unset($rss); | |
1203 | return false; | |
1204 | } | |
1205 | ||
1206 | _debug("done", $debug_enabled); | |
1207 | ||
1208 | return true; | |
1209 | } | |
1210 | ||
1211 | static function cache_enclosures($enclosures, $site_url, $debug) { | |
1212 | foreach ($enclosures as $enc) { | |
1213 | ||
1214 | if (preg_match("/(image|audio|video)/", $enc[1])) { | |
1215 | ||
1216 | $src = rewrite_relative_url($site_url, $enc[0]); | |
1217 | ||
1218 | $local_filename = CACHE_DIR . "/images/" . sha1($src); | |
1219 | ||
1220 | if ($debug) _debug("cache_enclosures: downloading: $src to $local_filename"); | |
1221 | ||
1222 | if (!file_exists($local_filename)) { | |
1223 | $file_content = fetch_file_contents($src); | |
1224 | ||
1225 | if ($file_content && strlen($file_content) > MIN_CACHE_FILE_SIZE) { | |
1226 | file_put_contents($local_filename, $file_content); | |
1227 | } | |
1228 | } else if (is_writable($local_filename)) { | |
1229 | touch($local_filename); | |
1230 | } | |
1231 | } | |
1232 | } | |
1233 | } | |
1234 | ||
1235 | static function cache_media($html, $site_url, $debug) { | |
1236 | libxml_use_internal_errors(true); | |
1237 | ||
1238 | $charset_hack = '<head> | |
1239 | <meta http-equiv="Content-Type" content="text/html; charset=utf-8"/> | |
1240 | </head>'; | |
1241 | ||
1242 | $doc = new DOMDocument(); | |
1243 | $doc->loadHTML($charset_hack . $html); | |
1244 | $xpath = new DOMXPath($doc); | |
1245 | ||
1246 | $entries = $xpath->query('(//img[@src])|(//video/source[@src])|(//audio/source[@src])'); | |
1247 | ||
1248 | foreach ($entries as $entry) { | |
1249 | if ($entry->hasAttribute('src') && strpos($entry->getAttribute('src'), "data:") !== 0) { | |
1250 | $src = rewrite_relative_url($site_url, $entry->getAttribute('src')); | |
1251 | ||
1252 | $local_filename = CACHE_DIR . "/images/" . sha1($src); | |
1253 | ||
1254 | if ($debug) _debug("cache_media: checking $src"); | |
1255 | ||
1256 | if (!file_exists($local_filename)) { | |
1257 | if ($debug) _debug("cache_media: downloading: $src to $local_filename"); | |
1258 | ||
1259 | $file_content = fetch_file_contents($src); | |
1260 | ||
1261 | if ($file_content && strlen($file_content) > MIN_CACHE_FILE_SIZE) { | |
1262 | file_put_contents($local_filename, $file_content); | |
1263 | } | |
1264 | } else if (is_writable($local_filename)) { | |
1265 | touch($local_filename); | |
1266 | } | |
1267 | } | |
1268 | } | |
1269 | } | |
1270 | ||
1271 | static function expire_error_log($debug) { | |
1272 | if ($debug) _debug("Removing old error log entries..."); | |
1273 | ||
1274 | $pdo = Db::pdo(); | |
1275 | ||
1276 | if (DB_TYPE == "pgsql") { | |
1277 | $pdo->query("DELETE FROM ttrss_error_log | |
1278 | WHERE created_at < NOW() - INTERVAL '7 days'"); | |
1279 | } else { | |
1280 | $pdo->query("DELETE FROM ttrss_error_log | |
1281 | WHERE created_at < DATE_SUB(NOW(), INTERVAL 7 DAY)"); | |
1282 | } | |
1283 | } | |
1284 | ||
1285 | static function expire_lock_files($debug) { | |
1286 | //if ($debug) _debug("Removing old lock files..."); | |
1287 | ||
1288 | $num_deleted = 0; | |
1289 | ||
1290 | if (is_writable(LOCK_DIRECTORY)) { | |
1291 | $files = glob(LOCK_DIRECTORY . "/*.lock"); | |
1292 | ||
1293 | if ($files) { | |
1294 | foreach ($files as $file) { | |
1295 | if (!file_is_locked(basename($file)) && time() - filemtime($file) > 86400*2) { | |
1296 | unlink($file); | |
1297 | ++$num_deleted; | |
1298 | } | |
1299 | } | |
1300 | } | |
1301 | } | |
1302 | ||
1303 | if ($debug) _debug("Removed $num_deleted old lock files."); | |
1304 | } | |
1305 | ||
1306 | static function expire_cached_files($debug) { | |
1307 | foreach (array("feeds", "images", "export", "upload") as $dir) { | |
1308 | $cache_dir = CACHE_DIR . "/$dir"; | |
1309 | ||
1310 | // if ($debug) _debug("Expiring $cache_dir"); | |
1311 | ||
1312 | $num_deleted = 0; | |
1313 | ||
1314 | if (is_writable($cache_dir)) { | |
1315 | $files = glob("$cache_dir/*"); | |
1316 | ||
1317 | if ($files) { | |
1318 | foreach ($files as $file) { | |
1319 | if (time() - filemtime($file) > 86400*CACHE_MAX_DAYS) { | |
1320 | unlink($file); | |
1321 | ||
1322 | ++$num_deleted; | |
1323 | } | |
1324 | } | |
1325 | } | |
1326 | } | |
1327 | ||
1328 | if ($debug) _debug("$cache_dir: removed $num_deleted files."); | |
1329 | } | |
1330 | } | |
1331 | ||
1332 | /** | |
1333 | * Source: http://www.php.net/manual/en/function.parse-url.php#104527 | |
1334 | * Returns the url query as associative array | |
1335 | * | |
1336 | * @param string query | |
1337 | * @return array params | |
1338 | */ | |
1339 | static function convertUrlQuery($query) { | |
1340 | $queryParts = explode('&', $query); | |
1341 | ||
1342 | $params = array(); | |
1343 | ||
1344 | foreach ($queryParts as $param) { | |
1345 | $item = explode('=', $param); | |
1346 | $params[$item[0]] = $item[1]; | |
1347 | } | |
1348 | ||
1349 | return $params; | |
1350 | } | |
1351 | ||
1352 | static function get_article_filters($filters, $title, $content, $link, $author, $tags, &$matched_rules = false) { | |
1353 | $matches = array(); | |
1354 | ||
1355 | foreach ($filters as $filter) { | |
1356 | $match_any_rule = $filter["match_any_rule"]; | |
1357 | $inverse = $filter["inverse"]; | |
1358 | $filter_match = false; | |
1359 | ||
1360 | foreach ($filter["rules"] as $rule) { | |
1361 | $match = false; | |
1362 | $reg_exp = str_replace('/', '\/', $rule["reg_exp"]); | |
1363 | $rule_inverse = $rule["inverse"]; | |
1364 | ||
1365 | if (!$reg_exp) | |
1366 | continue; | |
1367 | ||
1368 | switch ($rule["type"]) { | |
1369 | case "title": | |
1370 | $match = @preg_match("/$reg_exp/iu", $title); | |
1371 | break; | |
1372 | case "content": | |
1373 | // we don't need to deal with multiline regexps | |
1374 | $content = preg_replace("/[\r\n\t]/", "", $content); | |
1375 | ||
1376 | $match = @preg_match("/$reg_exp/iu", $content); | |
1377 | break; | |
1378 | case "both": | |
1379 | // we don't need to deal with multiline regexps | |
1380 | $content = preg_replace("/[\r\n\t]/", "", $content); | |
1381 | ||
1382 | $match = (@preg_match("/$reg_exp/iu", $title) || @preg_match("/$reg_exp/iu", $content)); | |
1383 | break; | |
1384 | case "link": | |
1385 | $match = @preg_match("/$reg_exp/iu", $link); | |
1386 | break; | |
1387 | case "author": | |
1388 | $match = @preg_match("/$reg_exp/iu", $author); | |
1389 | break; | |
1390 | case "tag": | |
1391 | foreach ($tags as $tag) { | |
1392 | if (@preg_match("/$reg_exp/iu", $tag)) { | |
1393 | $match = true; | |
1394 | break; | |
1395 | } | |
1396 | } | |
1397 | break; | |
1398 | } | |
1399 | ||
1400 | if ($rule_inverse) $match = !$match; | |
1401 | ||
1402 | if ($match_any_rule) { | |
1403 | if ($match) { | |
1404 | $filter_match = true; | |
1405 | break; | |
1406 | } | |
1407 | } else { | |
1408 | $filter_match = $match; | |
1409 | if (!$match) { | |
1410 | break; | |
1411 | } | |
1412 | } | |
1413 | } | |
1414 | ||
1415 | if ($inverse) $filter_match = !$filter_match; | |
1416 | ||
1417 | if ($filter_match) { | |
1418 | if (is_array($matched_rules)) array_push($matched_rules, $rule); | |
1419 | ||
1420 | foreach ($filter["actions"] AS $action) { | |
1421 | array_push($matches, $action); | |
1422 | ||
1423 | // if Stop action encountered, perform no further processing | |
1424 | if (isset($action["type"]) && $action["type"] == "stop") return $matches; | |
1425 | } | |
1426 | } | |
1427 | } | |
1428 | ||
1429 | return $matches; | |
1430 | } | |
1431 | ||
1432 | static function find_article_filter($filters, $filter_name) { | |
1433 | foreach ($filters as $f) { | |
1434 | if ($f["type"] == $filter_name) { | |
1435 | return $f; | |
1436 | }; | |
1437 | } | |
1438 | return false; | |
1439 | } | |
1440 | ||
1441 | static function find_article_filters($filters, $filter_name) { | |
1442 | $results = array(); | |
1443 | ||
1444 | foreach ($filters as $f) { | |
1445 | if ($f["type"] == $filter_name) { | |
1446 | array_push($results, $f); | |
1447 | }; | |
1448 | } | |
1449 | return $results; | |
1450 | } | |
1451 | ||
1452 | static function calculate_article_score($filters) { | |
1453 | $score = 0; | |
1454 | ||
1455 | foreach ($filters as $f) { | |
1456 | if ($f["type"] == "score") { | |
1457 | $score += $f["param"]; | |
1458 | }; | |
1459 | } | |
1460 | return $score; | |
1461 | } | |
1462 | ||
1463 | static function labels_contains_caption($labels, $caption) { | |
1464 | foreach ($labels as $label) { | |
1465 | if ($label[1] == $caption) { | |
1466 | return true; | |
1467 | } | |
1468 | } | |
1469 | ||
1470 | return false; | |
1471 | } | |
1472 | ||
1473 | static function assign_article_to_label_filters($id, $filters, $owner_uid, $article_labels) { | |
1474 | foreach ($filters as $f) { | |
1475 | if ($f["type"] == "label") { | |
1476 | if (!RSSUtils::labels_contains_caption($article_labels, $f["param"])) { | |
1477 | Labels::add_article($id, $f["param"], $owner_uid); | |
1478 | } | |
1479 | } | |
1480 | } | |
1481 | } | |
1482 | ||
1483 | static function make_guid_from_title($title) { | |
1484 | return preg_replace("/[ \"\',.:;]/", "-", | |
1485 | mb_strtolower(strip_tags($title), 'utf-8')); | |
1486 | } | |
1487 | ||
1488 | static function cleanup_counters_cache($debug) { | |
1489 | $pdo = Db::pdo(); | |
1490 | ||
1491 | $res = $pdo->query("DELETE FROM ttrss_counters_cache | |
1492 | WHERE feed_id > 0 AND | |
1493 | (SELECT COUNT(id) FROM ttrss_feeds WHERE | |
1494 | id = feed_id AND | |
1495 | ttrss_counters_cache.owner_uid = ttrss_feeds.owner_uid) = 0"); | |
1496 | ||
1497 | $frows = $res->rowCount(); | |
1498 | ||
1499 | $res = $pdo->query("DELETE FROM ttrss_cat_counters_cache | |
1500 | WHERE feed_id > 0 AND | |
1501 | (SELECT COUNT(id) FROM ttrss_feed_categories WHERE | |
1502 | id = feed_id AND | |
1503 | ttrss_cat_counters_cache.owner_uid = ttrss_feed_categories.owner_uid) = 0"); | |
1504 | ||
1505 | $crows = $res->rowCount(); | |
1506 | ||
1507 | if ($debug) _debug("Removed $frows (feeds) $crows (cats) orphaned counter cache entries."); | |
1508 | } | |
1509 | ||
1510 | static function housekeeping_user($owner_uid) { | |
1511 | $tmph = new PluginHost(); | |
1512 | ||
1513 | load_user_plugins($owner_uid, $tmph); | |
1514 | ||
1515 | $tmph->run_hooks(PluginHost::HOOK_HOUSE_KEEPING, "hook_house_keeping", ""); | |
1516 | } | |
1517 | ||
1518 | static function housekeeping_common($debug) { | |
1519 | RSSUtils::expire_cached_files($debug); | |
1520 | RSSUtils::expire_lock_files($debug); | |
1521 | RSSUtils::expire_error_log($debug); | |
1522 | ||
1523 | $count = RSSUtils::update_feedbrowser_cache(); | |
1524 | _debug("Feedbrowser updated, $count feeds processed."); | |
1525 | ||
1526 | Article::purge_orphans( true); | |
1527 | RSSUtils::cleanup_counters_cache($debug); | |
1528 | ||
1529 | //$rc = cleanup_tags( 14, 50000); | |
1530 | //_debug("Cleaned $rc cached tags."); | |
1531 | ||
1532 | PluginHost::getInstance()->run_hooks(PluginHost::HOOK_HOUSE_KEEPING, "hook_house_keeping", ""); | |
1533 | } | |
1534 | ||
1535 | static function check_feed_favicon($site_url, $feed) { | |
1536 | # print "FAVICON [$site_url]: $favicon_url\n"; | |
1537 | ||
1538 | $icon_file = ICONS_DIR . "/$feed.ico"; | |
1539 | ||
1540 | if (!file_exists($icon_file)) { | |
1541 | $favicon_url = get_favicon_url($site_url); | |
1542 | ||
1543 | if ($favicon_url) { | |
1544 | // Limiting to "image" type misses those served with text/plain | |
1545 | $contents = fetch_file_contents($favicon_url); // , "image"); | |
1546 | ||
1547 | if ($contents) { | |
1548 | // Crude image type matching. | |
1549 | // Patterns gleaned from the file(1) source code. | |
1550 | if (preg_match('/^\x00\x00\x01\x00/', $contents)) { | |
1551 | // 0 string \000\000\001\000 MS Windows icon resource | |
1552 | //error_log("check_feed_favicon: favicon_url=$favicon_url isa MS Windows icon resource"); | |
1553 | } | |
1554 | elseif (preg_match('/^GIF8/', $contents)) { | |
1555 | // 0 string GIF8 GIF image data | |
1556 | //error_log("check_feed_favicon: favicon_url=$favicon_url isa GIF image"); | |
1557 | } | |
1558 | elseif (preg_match('/^\x89PNG\x0d\x0a\x1a\x0a/', $contents)) { | |
1559 | // 0 string \x89PNG\x0d\x0a\x1a\x0a PNG image data | |
1560 | //error_log("check_feed_favicon: favicon_url=$favicon_url isa PNG image"); | |
1561 | } | |
1562 | elseif (preg_match('/^\xff\xd8/', $contents)) { | |
1563 | // 0 beshort 0xffd8 JPEG image data | |
1564 | //error_log("check_feed_favicon: favicon_url=$favicon_url isa JPG image"); | |
1565 | } | |
1566 | elseif (preg_match('/^BM/', $contents)) { | |
1567 | // 0 string BM PC bitmap (OS2, Windows BMP files) | |
1568 | //error_log("check_feed_favicon, favicon_url=$favicon_url isa BMP image"); | |
1569 | } | |
1570 | else { | |
1571 | //error_log("check_feed_favicon: favicon_url=$favicon_url isa UNKNOWN type"); | |
1572 | $contents = ""; | |
1573 | } | |
1574 | } | |
1575 | ||
1576 | if ($contents) { | |
1577 | $fp = @fopen($icon_file, "w"); | |
1578 | ||
1579 | if ($fp) { | |
1580 | fwrite($fp, $contents); | |
1581 | fclose($fp); | |
1582 | chmod($icon_file, 0644); | |
1583 | } | |
1584 | } | |
1585 | } | |
1586 | return $icon_file; | |
1587 | } | |
1588 | } | |
1589 | ||
1590 | ||
1591 | ||
1592 | } |