rssfuncs.php 46 KB

12345678910111213141516171819202122232425262728293031323334353637383940414243444546474849505152535455565758596061626364656667686970717273747576777879808182838485868788899091929394959697989910010110210310410510610710810911011111211311411511611711811912012112212312412512612712812913013113213313413513613713813914014114214314414514614714814915015115215315415515615715815916016116216316416516616716816917017117217317417517617717817918018118218318418518618718818919019119219319419519619719819920020120220320420520620720820921021121221321421521621721821922022122222322422522622722822923023123223323423523623723823924024124224324424524624724824925025125225325425525625725825926026126226326426526626726826927027127227327427527627727827928028128228328428528628728828929029129229329429529629729829930030130230330430530630730830931031131231331431531631731831932032132232332432532632732832933033133233333433533633733833934034134234334434534634734834935035135235335435535635735835936036136236336436536636736836937037137237337437537637737837938038138238338438538638738838939039139239339439539639739839940040140240340440540640740840941041141241341441541641741841942042142242342442542642742842943043143243343443543643743843944044144244344444544644744844945045145245345445545645745845946046146246346446546646746846947047147247347447547647747847948048148248348448548648748848949049149249349449549649749849950050150250350450550650750850951051151251351451551651751851952052152252352452552652752852953053153253353453553653753853954054154254354454554654754854955055155255355455555655755855956056156256356456556656756856957057157257357457557657757857958058158258358458558658758858959059159259359459559659759859960060160260360460560660760860961061161261361461561661761861962062162262362462562662762862963063163263363463563663763863964064164264364464564664764864965065165265365465565665765865966066166266366466566666766866967067167267367467567667767867968068168268368468568668768868969069169269369469569669769869970070170270370470570670770870971071171271371471571671771871972072172272372472572672772872973073173273373473573673773873974074174274374474574674774874975075175275375475575675775875976076176276376476576676776876977077177277377477577677777877978078178278378478578678778878979079179279379479579679779879980080180280380480580680780880981081181281381481581681781881982082182282382482582682782882983083183283383483583683783883984084184284384484584684784884985085185285385485585685785885986086186286386486586686786886987087187287387487587687787887988088188288388488588688788888989089189289389489589689789889990090190290390490590690790890991091191291391491591691791891992092192292392492592692792892993093193293393493593693793893994094194294394494594694794894995095195295395495595695795895996096196296396496596696796896997097197297397497597697797897998098198298398498598698798898999099199299399499599699799899910001001100210031004100510061007100810091010101110121013101410151016101710181019102010211022102310241025102610271028102910301031103210331034103510361037103810391040104110421043104410451046104710481049105010511052105310541055105610571058105910601061106210631064106510661067106810691070107110721073107410751076107710781079108010811082108310841085108610871088108910901091109210931094109510961097109810991100110111021103110411051106110711081109111011111112111311141115111611171118111911201121112211231124112511261127112811291130113111321133113411351136113711381139114011411142114311441145114611471148114911501151115211531154115511561157115811591160116111621163116411651166116711681169117011711172117311741175117611771178117911801181118211831184118511861187118811891190119111921193119411951196119711981199120012011202120312041205120612071208120912101211121212131214121512161217121812191220122112221223122412251226122712281229123012311232123312341235123612371238123912401241124212431244124512461247124812491250125112521253125412551256125712581259126012611262126312641265126612671268126912701271127212731274127512761277127812791280128112821283128412851286128712881289129012911292129312941295129612971298129913001301130213031304130513061307130813091310131113121313131413151316131713181319132013211322132313241325132613271328132913301331133213331334133513361337133813391340134113421343134413451346134713481349135013511352135313541355135613571358135913601361136213631364136513661367136813691370137113721373137413751376137713781379138013811382138313841385138613871388138913901391139213931394139513961397139813991400140114021403140414051406140714081409141014111412141314141415141614171418141914201421142214231424142514261427142814291430143114321433143414351436143714381439144014411442144314441445144614471448144914501451145214531454145514561457145814591460146114621463146414651466146714681469147014711472147314741475147614771478147914801481148214831484148514861487148814891490149114921493149414951496149714981499150015011502150315041505150615071508150915101511151215131514151515161517151815191520152115221523152415251526152715281529153015311532153315341535153615371538153915401541154215431544154515461547154815491550155115521553155415551556155715581559156015611562156315641565156615671568156915701571157215731574
  1. <?php
  2. define_default('DAEMON_UPDATE_LOGIN_LIMIT', 30);
  3. define_default('DAEMON_FEED_LIMIT', 500);
  4. define_default('DAEMON_SLEEP_INTERVAL', 120);
  5. define_default('_MIN_CACHE_FILE_SIZE', 1024);
  6. // TODO: this needs to be removed from global namespace into classes/RSS.php or something
  7. function calculate_article_hash($article, $pluginhost) {
  8. $tmp = "";
  9. foreach ($article as $k => $v) {
  10. if ($k != "feed" && isset($v)) {
  11. $x = strip_tags(is_array($v) ? implode(",", $v) : $v);
  12. //_debug("$k:" . sha1($x) . ":" . htmlspecialchars($x), true);
  13. $tmp .= sha1("$k:" . sha1($x));
  14. }
  15. }
  16. return sha1(implode(",", $pluginhost->get_plugin_names()) . $tmp);
  17. }
  18. function update_feedbrowser_cache() {
  19. $result = db_query("SELECT feed_url, site_url, title, COUNT(id) AS subscribers
  20. FROM ttrss_feeds WHERE feed_url NOT IN (SELECT feed_url FROM ttrss_feeds
  21. WHERE private IS true OR auth_login != '' OR auth_pass != '' OR feed_url LIKE '%:%@%/%')
  22. GROUP BY feed_url, site_url, title ORDER BY subscribers DESC LIMIT 1000");
  23. db_query("BEGIN");
  24. db_query("DELETE FROM ttrss_feedbrowser_cache");
  25. $count = 0;
  26. while ($line = db_fetch_assoc($result)) {
  27. $subscribers = db_escape_string($line["subscribers"]);
  28. $feed_url = db_escape_string($line["feed_url"]);
  29. $title = db_escape_string($line["title"]);
  30. $site_url = db_escape_string($line["site_url"]);
  31. $tmp_result = db_query("SELECT subscribers FROM
  32. ttrss_feedbrowser_cache WHERE feed_url = '$feed_url'");
  33. if (db_num_rows($tmp_result) == 0) {
  34. db_query("INSERT INTO ttrss_feedbrowser_cache
  35. (feed_url, site_url, title, subscribers) VALUES ('$feed_url',
  36. '$site_url', '$title', '$subscribers')");
  37. ++$count;
  38. }
  39. }
  40. db_query("COMMIT");
  41. return $count;
  42. }
  43. function update_daemon_common($limit = DAEMON_FEED_LIMIT, $debug = true) {
  44. // Process all other feeds using last_updated and interval parameters
  45. $schema_version = get_schema_version();
  46. if ($schema_version != SCHEMA_VERSION) {
  47. die("Schema version is wrong, please upgrade the database.\n");
  48. }
  49. define('PREFS_NO_CACHE', true);
  50. // Test if the user has loggued in recently. If not, it does not update its feeds.
  51. if (!SINGLE_USER_MODE && DAEMON_UPDATE_LOGIN_LIMIT > 0) {
  52. if (DB_TYPE == "pgsql") {
  53. $login_thresh_qpart = "AND ttrss_users.last_login >= NOW() - INTERVAL '".DAEMON_UPDATE_LOGIN_LIMIT." days'";
  54. } else {
  55. $login_thresh_qpart = "AND ttrss_users.last_login >= DATE_SUB(NOW(), INTERVAL ".DAEMON_UPDATE_LOGIN_LIMIT." DAY)";
  56. }
  57. } else {
  58. $login_thresh_qpart = "";
  59. }
  60. // Test if the feed need a update (update interval exceeded).
  61. if (DB_TYPE == "pgsql") {
  62. $update_limit_qpart = "AND ((
  63. ttrss_feeds.update_interval = 0
  64. AND ttrss_user_prefs.value != '-1'
  65. AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_user_prefs.value || ' minutes') AS INTERVAL)
  66. ) OR (
  67. ttrss_feeds.update_interval > 0
  68. AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_feeds.update_interval || ' minutes') AS INTERVAL)
  69. ) OR (ttrss_feeds.last_updated IS NULL
  70. AND ttrss_user_prefs.value != '-1')
  71. OR (last_updated = '1970-01-01 00:00:00'
  72. AND ttrss_user_prefs.value != '-1'))";
  73. } else {
  74. $update_limit_qpart = "AND ((
  75. ttrss_feeds.update_interval = 0
  76. AND ttrss_user_prefs.value != '-1'
  77. AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL CONVERT(ttrss_user_prefs.value, SIGNED INTEGER) MINUTE)
  78. ) OR (
  79. ttrss_feeds.update_interval > 0
  80. AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL ttrss_feeds.update_interval MINUTE)
  81. ) OR (ttrss_feeds.last_updated IS NULL
  82. AND ttrss_user_prefs.value != '-1')
  83. OR (last_updated = '1970-01-01 00:00:00'
  84. AND ttrss_user_prefs.value != '-1'))";
  85. }
  86. // Test if feed is currently being updated by another process.
  87. if (DB_TYPE == "pgsql") {
  88. $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < NOW() - INTERVAL '10 minutes')";
  89. } else {
  90. $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < DATE_SUB(NOW(), INTERVAL 10 MINUTE))";
  91. }
  92. // Test if there is a limit to number of updated feeds
  93. $query_limit = "";
  94. if($limit) $query_limit = sprintf("LIMIT %d", $limit);
  95. // Update the least recently updated feeds first
  96. $query_order = "ORDER BY last_updated";
  97. if (DB_TYPE == "pgsql") $query_order .= " NULLS FIRST";
  98. $query = "SELECT DISTINCT ttrss_feeds.feed_url, ttrss_feeds.last_updated
  99. FROM
  100. ttrss_feeds, ttrss_users, ttrss_user_prefs
  101. WHERE
  102. ttrss_feeds.owner_uid = ttrss_users.id
  103. AND ttrss_user_prefs.profile IS NULL
  104. AND ttrss_users.id = ttrss_user_prefs.owner_uid
  105. AND ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL'
  106. $login_thresh_qpart $update_limit_qpart
  107. $updstart_thresh_qpart
  108. $query_order $query_limit";
  109. // We search for feed needing update.
  110. $result = db_query($query);
  111. if($debug) _debug(sprintf("Scheduled %d feeds to update...", db_num_rows($result)));
  112. // Here is a little cache magic in order to minimize risk of double feed updates.
  113. $feeds_to_update = array();
  114. while ($line = db_fetch_assoc($result)) {
  115. array_push($feeds_to_update, db_escape_string($line['feed_url']));
  116. }
  117. // We update the feed last update started date before anything else.
  118. // There is no lag due to feed contents downloads
  119. // It prevent an other process to update the same feed.
  120. if(count($feeds_to_update) > 0) {
  121. $feeds_quoted = array();
  122. foreach ($feeds_to_update as $feed) {
  123. array_push($feeds_quoted, "'" . db_escape_string($feed) . "'");
  124. }
  125. db_query(sprintf("UPDATE ttrss_feeds SET last_update_started = NOW()
  126. WHERE feed_url IN (%s)", implode(',', $feeds_quoted)));
  127. }
  128. $nf = 0;
  129. $bstarted = microtime(true);
  130. $batch_owners = array();
  131. // For each feed, we call the feed update function.
  132. foreach ($feeds_to_update as $feed) {
  133. if($debug) _debug("Base feed: $feed");
  134. //update_rss_feed($line["id"], true);
  135. // since we have the data cached, we can deal with other feeds with the same url
  136. $tmp_result = db_query("SELECT DISTINCT ttrss_feeds.id,last_updated,ttrss_feeds.owner_uid
  137. FROM ttrss_feeds, ttrss_users, ttrss_user_prefs WHERE
  138. ttrss_user_prefs.owner_uid = ttrss_feeds.owner_uid AND
  139. ttrss_users.id = ttrss_user_prefs.owner_uid AND
  140. ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL' AND
  141. ttrss_user_prefs.profile IS NULL AND
  142. feed_url = '".db_escape_string($feed)."'
  143. $update_limit_qpart
  144. $login_thresh_qpart
  145. ORDER BY ttrss_feeds.id $query_limit");
  146. if (db_num_rows($tmp_result) > 0) {
  147. while ($tline = db_fetch_assoc($tmp_result)) {
  148. if($debug) _debug(" => " . $tline["last_updated"] . ", " . $tline["id"] . " " . $tline["owner_uid"]);
  149. if (array_search($tline["owner_uid"], $batch_owners) === FALSE)
  150. array_push($batch_owners, $tline["owner_uid"]);
  151. $fstarted = microtime(true);
  152. update_rss_feed($tline["id"], true, false);
  153. _debug_suppress(false);
  154. _debug(sprintf(" %.4f (sec)", microtime(true) - $fstarted));
  155. ++$nf;
  156. }
  157. }
  158. }
  159. if ($nf > 0) {
  160. _debug(sprintf("Processed %d feeds in %.4f (sec), %.4f (sec/feed avg)", $nf,
  161. microtime(true) - $bstarted, (microtime(true) - $bstarted) / $nf));
  162. }
  163. foreach ($batch_owners as $owner_uid) {
  164. _debug("Running housekeeping tasks for user $owner_uid...");
  165. housekeeping_user($owner_uid);
  166. }
  167. // Send feed digests by email if needed.
  168. Digest::send_headlines_digests($debug);
  169. return $nf;
  170. }
  171. // this is used when subscribing
  172. function set_basic_feed_info($feed) {
  173. $feed = db_escape_string($feed);
  174. $result = db_query("SELECT feed_url,auth_pass,auth_login,auth_pass_encrypted
  175. FROM ttrss_feeds WHERE id = '$feed'");
  176. $auth_pass_encrypted = sql_bool_to_bool(db_fetch_result($result,
  177. 0, "auth_pass_encrypted"));
  178. $auth_login = db_fetch_result($result, 0, "auth_login");
  179. $auth_pass = db_fetch_result($result, 0, "auth_pass");
  180. if ($auth_pass_encrypted && function_exists("mcrypt_decrypt")) {
  181. require_once "crypt.php";
  182. $auth_pass = decrypt_string($auth_pass);
  183. }
  184. $fetch_url = db_fetch_result($result, 0, "feed_url");
  185. $feed_data = fetch_file_contents($fetch_url, false,
  186. $auth_login, $auth_pass, false,
  187. FEED_FETCH_TIMEOUT,
  188. 0);
  189. global $fetch_curl_used;
  190. if (!$fetch_curl_used) {
  191. $tmp = @gzdecode($feed_data);
  192. if ($tmp) $feed_data = $tmp;
  193. }
  194. $feed_data = trim($feed_data);
  195. $rss = new FeedParser($feed_data);
  196. $rss->init();
  197. if (!$rss->error()) {
  198. $result = db_query("SELECT title, site_url FROM ttrss_feeds WHERE id = '$feed'");
  199. $registered_title = db_fetch_result($result, 0, "title");
  200. $orig_site_url = db_fetch_result($result, 0, "site_url");
  201. $site_url = db_escape_string(mb_substr(rewrite_relative_url($fetch_url, $rss->get_link()), 0, 245));
  202. $feed_title = db_escape_string(mb_substr($rss->get_title(), 0, 199));
  203. if ($feed_title && (!$registered_title || $registered_title == "[Unknown]")) {
  204. db_query("UPDATE ttrss_feeds SET
  205. title = '$feed_title' WHERE id = '$feed'");
  206. }
  207. if ($site_url && $orig_site_url != $site_url) {
  208. db_query("UPDATE ttrss_feeds SET
  209. site_url = '$site_url' WHERE id = '$feed'");
  210. }
  211. }
  212. }
  213. /**
  214. * @SuppressWarnings(PHPMD.UnusedFormalParameter)
  215. */
  216. function update_rss_feed($feed, $no_cache = false) {
  217. $debug_enabled = defined('DAEMON_EXTENDED_DEBUG') || $_REQUEST['xdebug'];
  218. _debug_suppress(!$debug_enabled);
  219. _debug("start", $debug_enabled);
  220. $result = db_query("SELECT title FROM ttrss_feeds
  221. WHERE id = '$feed'");
  222. if (db_num_rows($result) == 0) {
  223. _debug("feed $feed NOT FOUND/SKIPPED", $debug_enabled);
  224. user_error("Attempt to update unknown/invalid feed $feed", E_USER_WARNING);
  225. return false;
  226. }
  227. $title = db_fetch_result($result, 0, "title");
  228. // feed was batch-subscribed or something, we need to get basic info
  229. // this is not optimal currently as it fetches stuff separately TODO: optimize
  230. if ($title == "[Unknown]") {
  231. _debug("setting basic feed info for $feed...");
  232. set_basic_feed_info($feed);
  233. }
  234. $result = db_query("SELECT id,update_interval,auth_login,
  235. feed_url,auth_pass,cache_images,
  236. mark_unread_on_update, owner_uid,
  237. pubsub_state, auth_pass_encrypted,
  238. feed_language
  239. FROM ttrss_feeds WHERE id = '$feed'");
  240. $owner_uid = db_fetch_result($result, 0, "owner_uid");
  241. $mark_unread_on_update = sql_bool_to_bool(db_fetch_result($result,
  242. 0, "mark_unread_on_update"));
  243. $pubsub_state = db_fetch_result($result, 0, "pubsub_state");
  244. $auth_pass_encrypted = sql_bool_to_bool(db_fetch_result($result,
  245. 0, "auth_pass_encrypted"));
  246. db_query("UPDATE ttrss_feeds SET last_update_started = NOW()
  247. WHERE id = '$feed'");
  248. $auth_login = db_fetch_result($result, 0, "auth_login");
  249. $auth_pass = db_fetch_result($result, 0, "auth_pass");
  250. if ($auth_pass_encrypted && function_exists("mcrypt_decrypt")) {
  251. require_once "crypt.php";
  252. $auth_pass = decrypt_string($auth_pass);
  253. }
  254. $cache_images = sql_bool_to_bool(db_fetch_result($result, 0, "cache_images"));
  255. $fetch_url = db_fetch_result($result, 0, "feed_url");
  256. $feed_language = db_escape_string(mb_strtolower(db_fetch_result($result, 0, "feed_language")));
  257. if (!$feed_language) $feed_language = 'english';
  258. $feed = db_escape_string($feed);
  259. $date_feed_processed = date('Y-m-d H:i');
  260. $cache_filename = CACHE_DIR . "/simplepie/" . sha1($fetch_url) . ".xml";
  261. $pluginhost = new PluginHost();
  262. $pluginhost->set_debug($debug_enabled);
  263. $user_plugins = get_pref("_ENABLED_PLUGINS", $owner_uid);
  264. $pluginhost->load(PLUGINS, PluginHost::KIND_ALL);
  265. $pluginhost->load($user_plugins, PluginHost::KIND_USER, $owner_uid);
  266. $pluginhost->load_data();
  267. $rss_hash = false;
  268. $force_refetch = isset($_REQUEST["force_refetch"]);
  269. $feed_data = "";
  270. foreach ($pluginhost->get_hooks(PluginHost::HOOK_FETCH_FEED) as $plugin) {
  271. $feed_data = $plugin->hook_fetch_feed($feed_data, $fetch_url, $owner_uid, $feed, 0, $auth_login, $auth_pass);
  272. }
  273. // try cache
  274. if (!$feed_data &&
  275. file_exists($cache_filename) &&
  276. is_readable($cache_filename) &&
  277. !$auth_login && !$auth_pass &&
  278. filemtime($cache_filename) > time() - 30) {
  279. _debug("using local cache [$cache_filename].", $debug_enabled);
  280. @$feed_data = file_get_contents($cache_filename);
  281. if ($feed_data) {
  282. $rss_hash = sha1($feed_data);
  283. }
  284. } else {
  285. _debug("local cache will not be used for this feed", $debug_enabled);
  286. }
  287. // fetch feed from source
  288. if (!$feed_data) {
  289. _debug("fetching [$fetch_url]...", $debug_enabled);
  290. if (ini_get("open_basedir") && function_exists("curl_init")) {
  291. _debug("not using CURL due to open_basedir restrictions");
  292. }
  293. $feed_data = fetch_file_contents($fetch_url, false,
  294. $auth_login, $auth_pass, false,
  295. $no_cache ? FEED_FETCH_NO_CACHE_TIMEOUT : FEED_FETCH_TIMEOUT,
  296. 0);
  297. global $fetch_curl_used;
  298. if (!$fetch_curl_used) {
  299. $tmp = @gzdecode($feed_data);
  300. if ($tmp) $feed_data = $tmp;
  301. }
  302. $feed_data = trim($feed_data);
  303. _debug("fetch done.", $debug_enabled);
  304. // cache vanilla feed data for re-use
  305. if ($feed_data && !$auth_pass && !$auth_login && is_writable(CACHE_DIR . "/simplepie")) {
  306. $new_rss_hash = sha1($feed_data);
  307. if ($new_rss_hash != $rss_hash) {
  308. _debug("saving $cache_filename", $debug_enabled);
  309. @file_put_contents($cache_filename, $feed_data);
  310. }
  311. }
  312. }
  313. if (!$feed_data) {
  314. global $fetch_last_error;
  315. global $fetch_last_error_code;
  316. _debug("unable to fetch: $fetch_last_error [$fetch_last_error_code]", $debug_enabled);
  317. $error_escaped = '';
  318. // If-Modified-Since
  319. if ($fetch_last_error_code != 304) {
  320. $error_escaped = db_escape_string($fetch_last_error);
  321. } else {
  322. _debug("source claims data not modified, nothing to do.", $debug_enabled);
  323. }
  324. db_query(
  325. "UPDATE ttrss_feeds SET last_error = '$error_escaped',
  326. last_updated = NOW() WHERE id = '$feed'");
  327. return;
  328. }
  329. foreach ($pluginhost->get_hooks(PluginHost::HOOK_FEED_FETCHED) as $plugin) {
  330. $feed_data = $plugin->hook_feed_fetched($feed_data, $fetch_url, $owner_uid, $feed);
  331. }
  332. $rss = new FeedParser($feed_data);
  333. $rss->init();
  334. $feed = db_escape_string($feed);
  335. if (!$rss->error()) {
  336. // We use local pluginhost here because we need to load different per-user feed plugins
  337. $pluginhost->run_hooks(PluginHost::HOOK_FEED_PARSED, "hook_feed_parsed", $rss);
  338. _debug("language: $feed_language", $debug_enabled);
  339. _debug("processing feed data...", $debug_enabled);
  340. // db_query("BEGIN");
  341. if (DB_TYPE == "pgsql") {
  342. $favicon_interval_qpart = "favicon_last_checked < NOW() - INTERVAL '12 hour'";
  343. } else {
  344. $favicon_interval_qpart = "favicon_last_checked < DATE_SUB(NOW(), INTERVAL 12 HOUR)";
  345. }
  346. $result = db_query("SELECT owner_uid,favicon_avg_color,
  347. (favicon_last_checked IS NULL OR $favicon_interval_qpart) AS
  348. favicon_needs_check
  349. FROM ttrss_feeds WHERE id = '$feed'");
  350. $favicon_needs_check = sql_bool_to_bool(db_fetch_result($result, 0,
  351. "favicon_needs_check"));
  352. $favicon_avg_color = db_fetch_result($result, 0, "favicon_avg_color");
  353. $owner_uid = db_fetch_result($result, 0, "owner_uid");
  354. $site_url = db_escape_string(mb_substr(rewrite_relative_url($fetch_url, $rss->get_link()), 0, 245));
  355. _debug("site_url: $site_url", $debug_enabled);
  356. _debug("feed_title: " . $rss->get_title(), $debug_enabled);
  357. if ($favicon_needs_check || $force_refetch) {
  358. /* terrible hack: if we crash on floicon shit here, we won't check
  359. * the icon avgcolor again (unless the icon got updated) */
  360. $favicon_file = ICONS_DIR . "/$feed.ico";
  361. $favicon_modified = @filemtime($favicon_file);
  362. _debug("checking favicon...", $debug_enabled);
  363. check_feed_favicon($site_url, $feed);
  364. $favicon_modified_new = @filemtime($favicon_file);
  365. if ($favicon_modified_new > $favicon_modified)
  366. $favicon_avg_color = '';
  367. if (file_exists($favicon_file) && function_exists("imagecreatefromstring") && $favicon_avg_color == '') {
  368. require_once "colors.php";
  369. db_query("UPDATE ttrss_feeds SET favicon_avg_color = 'fail' WHERE
  370. id = '$feed'");
  371. $favicon_color = db_escape_string(
  372. calculate_avg_color($favicon_file));
  373. $favicon_colorstring = ",favicon_avg_color = '".$favicon_color."'";
  374. } else if ($favicon_avg_color == 'fail') {
  375. _debug("floicon failed on this file, not trying to recalculate avg color", $debug_enabled);
  376. }
  377. db_query("UPDATE ttrss_feeds SET favicon_last_checked = NOW()
  378. $favicon_colorstring
  379. WHERE id = '$feed'");
  380. }
  381. _debug("loading filters & labels...", $debug_enabled);
  382. $filters = load_filters($feed, $owner_uid);
  383. _debug("" . count($filters) . " filters loaded.", $debug_enabled);
  384. $items = $rss->get_items();
  385. if (!is_array($items)) {
  386. _debug("no articles found.", $debug_enabled);
  387. db_query("UPDATE ttrss_feeds
  388. SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
  389. return; // no articles
  390. }
  391. if ($pubsub_state != 2 && PUBSUBHUBBUB_ENABLED) {
  392. _debug("checking for PUSH hub...", $debug_enabled);
  393. $feed_hub_url = false;
  394. $links = $rss->get_links('hub');
  395. if ($links && is_array($links)) {
  396. foreach ($links as $l) {
  397. $feed_hub_url = $l;
  398. break;
  399. }
  400. }
  401. _debug("feed hub url: $feed_hub_url", $debug_enabled);
  402. $feed_self_url = $fetch_url;
  403. $links = $rss->get_links('self');
  404. if ($links && is_array($links)) {
  405. foreach ($links as $l) {
  406. $feed_self_url = $l;
  407. break;
  408. }
  409. }
  410. _debug("feed self url = $feed_self_url");
  411. if ($feed_hub_url && $feed_self_url && function_exists('curl_init') &&
  412. !ini_get("open_basedir")) {
  413. require_once 'lib/pubsubhubbub/Subscriber.php';
  414. $callback_url = get_self_url_prefix() .
  415. "/public.php?op=pubsub&id=$feed";
  416. $s = new Pubsubhubbub\Subscriber\Subscriber($feed_hub_url, $callback_url);
  417. $rc = $s->subscribe($feed_self_url);
  418. _debug("feed hub url found, subscribe request sent. [rc=$rc]", $debug_enabled);
  419. db_query("UPDATE ttrss_feeds SET pubsub_state = 1
  420. WHERE id = '$feed'");
  421. }
  422. }
  423. _debug("processing articles...", $debug_enabled);
  424. $tstart = time();
  425. foreach ($items as $item) {
  426. if ($_REQUEST['xdebug'] == 3) {
  427. print_r($item);
  428. }
  429. if (ini_get("max_execution_time") > 0 && time() - $tstart >= ini_get("max_execution_time") * 0.7) {
  430. _debug("looks like there's too many articles to process at once, breaking out", $debug_enabled);
  431. break;
  432. }
  433. $entry_guid = $item->get_id();
  434. if (!$entry_guid) $entry_guid = $item->get_link();
  435. if (!$entry_guid) $entry_guid = make_guid_from_title($item->get_title());
  436. if (!$entry_guid) continue;
  437. $entry_guid = "$owner_uid,$entry_guid";
  438. $entry_guid_hashed = db_escape_string('SHA1:' . sha1($entry_guid));
  439. _debug("guid $entry_guid / $entry_guid_hashed", $debug_enabled);
  440. $entry_timestamp = "";
  441. $entry_timestamp = $item->get_date();
  442. _debug("orig date: " . $item->get_date(), $debug_enabled);
  443. if ($entry_timestamp == -1 || !$entry_timestamp || $entry_timestamp > time()) {
  444. $entry_timestamp = time();
  445. }
  446. $entry_timestamp_fmt = strftime("%Y/%m/%d %H:%M:%S", $entry_timestamp);
  447. _debug("date $entry_timestamp [$entry_timestamp_fmt]", $debug_enabled);
  448. // $entry_title = html_entity_decode($item->get_title(), ENT_COMPAT, 'UTF-8');
  449. // $entry_title = decode_numeric_entities($entry_title);
  450. $entry_title = $item->get_title();
  451. $entry_link = rewrite_relative_url($site_url, $item->get_link());
  452. _debug("title $entry_title", $debug_enabled);
  453. _debug("link $entry_link", $debug_enabled);
  454. if (!$entry_title) $entry_title = date("Y-m-d H:i:s", $entry_timestamp);;
  455. $entry_content = $item->get_content();
  456. if (!$entry_content) $entry_content = $item->get_description();
  457. if ($_REQUEST["xdebug"] == 2) {
  458. print "content: ";
  459. print htmlspecialchars($entry_content);
  460. print "\n";
  461. }
  462. $entry_comments = db_escape_string(mb_substr($item->get_comments_url(), 0, 245));
  463. $num_comments = (int) $item->get_comments_count();
  464. $entry_author = $item->get_author(); // escaped later
  465. $entry_guid = db_escape_string(mb_substr($entry_guid, 0, 245));
  466. _debug("author $entry_author", $debug_enabled);
  467. _debug("num_comments: $num_comments", $debug_enabled);
  468. _debug("looking for tags...", $debug_enabled);
  469. // parse <category> entries into tags
  470. $additional_tags = array();
  471. $additional_tags_src = $item->get_categories();
  472. if (is_array($additional_tags_src)) {
  473. foreach ($additional_tags_src as $tobj) {
  474. array_push($additional_tags, $tobj);
  475. }
  476. }
  477. $entry_tags = array_unique($additional_tags);
  478. for ($i = 0; $i < count($entry_tags); $i++)
  479. $entry_tags[$i] = mb_strtolower($entry_tags[$i], 'utf-8');
  480. _debug("tags found: " . join(",", $entry_tags), $debug_enabled);
  481. _debug("done collecting data.", $debug_enabled);
  482. $result = db_query("SELECT id, content_hash, lang FROM ttrss_entries
  483. WHERE guid = '".db_escape_string($entry_guid)."' OR guid = '$entry_guid_hashed'");
  484. if (db_num_rows($result) != 0) {
  485. $base_entry_id = db_fetch_result($result, 0, "id");
  486. $entry_stored_hash = db_fetch_result($result, 0, "content_hash");
  487. $article_labels = Article::get_article_labels($base_entry_id, $owner_uid);
  488. $entry_language = db_fetch_result($result, 0, "lang");
  489. $existing_tags = Article::get_article_tags($base_entry_id, $owner_uid);
  490. $entry_tags = array_unique(array_merge($entry_tags, $existing_tags));
  491. } else {
  492. $base_entry_id = false;
  493. $entry_stored_hash = "";
  494. $article_labels = array();
  495. $entry_language = "";
  496. }
  497. $article = array("owner_uid" => $owner_uid, // read only
  498. "guid" => $entry_guid, // read only
  499. "guid_hashed" => $entry_guid_hashed, // read only
  500. "title" => $entry_title,
  501. "content" => $entry_content,
  502. "link" => $entry_link,
  503. "labels" => $article_labels, // current limitation: can add labels to article, can't remove them
  504. "tags" => $entry_tags,
  505. "author" => $entry_author,
  506. "force_catchup" => false, // ugly hack for the time being
  507. "score_modifier" => 0, // no previous value, plugin should recalculate score modifier based on content if needed
  508. "language" => $entry_language,
  509. "feed" => array("id" => $feed,
  510. "fetch_url" => $fetch_url,
  511. "site_url" => $site_url,
  512. "cache_images" => $cache_images)
  513. );
  514. $entry_plugin_data = "";
  515. $entry_current_hash = calculate_article_hash($article, $pluginhost);
  516. _debug("article hash: $entry_current_hash [stored=$entry_stored_hash]", $debug_enabled);
  517. if ($entry_current_hash == $entry_stored_hash && !isset($_REQUEST["force_rehash"])) {
  518. _debug("stored article seems up to date [IID: $base_entry_id], updating timestamp only", $debug_enabled);
  519. // we keep encountering the entry in feeds, so we need to
  520. // update date_updated column so that we don't get horrible
  521. // dupes when the entry gets purged and reinserted again e.g.
  522. // in the case of SLOW SLOW OMG SLOW updating feeds
  523. $base_entry_id = db_fetch_result($result, 0, "id");
  524. db_query("UPDATE ttrss_entries SET date_updated = NOW()
  525. WHERE id = '$base_entry_id'");
  526. continue;
  527. }
  528. _debug("hash differs, applying plugin filters:", $debug_enabled);
  529. foreach ($pluginhost->get_hooks(PluginHost::HOOK_ARTICLE_FILTER) as $plugin) {
  530. _debug("... " . get_class($plugin), $debug_enabled);
  531. $start = microtime(true);
  532. $article = $plugin->hook_article_filter($article);
  533. _debug("=== " . sprintf("%.4f (sec)", microtime(true) - $start), $debug_enabled);
  534. $entry_plugin_data .= mb_strtolower(get_class($plugin)) . ",";
  535. }
  536. if ($_REQUEST["xdebug"] == 2) {
  537. print "processed content: ";
  538. print htmlspecialchars($article["content"]);
  539. print "\n";
  540. }
  541. $entry_plugin_data = db_escape_string($entry_plugin_data);
  542. _debug("plugin data: $entry_plugin_data", $debug_enabled);
  543. // Workaround: 4-byte unicode requires utf8mb4 in MySQL. See https://tt-rss.org/forum/viewtopic.php?f=1&t=3377&p=20077#p20077
  544. if (DB_TYPE == "mysql") {
  545. foreach ($article as $k => $v) {
  546. // i guess we'll have to take the risk of 4byte unicode labels & tags here
  547. if (is_string($article[$k])) {
  548. $article[$k] = preg_replace('/[\x{10000}-\x{10FFFF}]/u', "\xEF\xBF\xBD", $v);
  549. }
  550. }
  551. }
  552. /* Collect article tags here so we could filter by them: */
  553. $matched_rules = array();
  554. $article_filters = get_article_filters($filters, $article["title"],
  555. $article["content"], $article["link"], $article["author"],
  556. $article["tags"], $matched_rules);
  557. if ($debug_enabled) {
  558. _debug("matched filter rules: ", $debug_enabled);
  559. if (count($matched_rules) != 0) {
  560. print_r($matched_rules);
  561. }
  562. _debug("filter actions: ", $debug_enabled);
  563. if (count($article_filters) != 0) {
  564. print_r($article_filters);
  565. }
  566. }
  567. $plugin_filter_names = find_article_filters($article_filters, "plugin");
  568. $plugin_filter_actions = $pluginhost->get_filter_actions();
  569. if (count($plugin_filter_names) > 0) {
  570. _debug("applying plugin filter actions...", $debug_enabled);
  571. foreach ($plugin_filter_names as $pfn) {
  572. list($pfclass,$pfaction) = explode(":", $pfn["param"]);
  573. if (isset($plugin_filter_actions[$pfclass])) {
  574. $plugin = $pluginhost->get_plugin($pfclass);
  575. _debug("... $pfclass: $pfaction", $debug_enabled);
  576. if ($plugin) {
  577. $start = microtime(true);
  578. $article = $plugin->hook_article_filter_action($article, $pfaction);
  579. _debug("=== " . sprintf("%.4f (sec)", microtime(true) - $start), $debug_enabled);
  580. } else {
  581. _debug("??? $pfclass: plugin object not found.");
  582. }
  583. } else {
  584. _debug("??? $pfclass: filter plugin not registered.");
  585. }
  586. }
  587. }
  588. $entry_tags = $article["tags"];
  589. $entry_guid = db_escape_string($entry_guid);
  590. $entry_title = db_escape_string($article["title"]);
  591. $entry_author = db_escape_string(mb_substr($article["author"], 0, 245));
  592. $entry_link = db_escape_string($article["link"]);
  593. $entry_content = $article["content"]; // escaped below
  594. $entry_force_catchup = $article["force_catchup"];
  595. $article_labels = $article["labels"];
  596. $entry_score_modifier = (int) $article["score_modifier"];
  597. $entry_language = db_escape_string($article["language"]);
  598. if ($debug_enabled) {
  599. _debug("article labels:", $debug_enabled);
  600. if (count($article_labels) != 0) {
  601. print_r($article_labels);
  602. }
  603. }
  604. _debug("force catchup: $entry_force_catchup");
  605. if ($cache_images && is_writable(CACHE_DIR . '/images'))
  606. cache_media($entry_content, $site_url, $debug_enabled);
  607. $entry_content = db_escape_string($entry_content, false);
  608. //db_query("BEGIN");
  609. $result = db_query("SELECT id FROM ttrss_entries
  610. WHERE (guid = '$entry_guid' OR guid = '$entry_guid_hashed')");
  611. if (db_num_rows($result) == 0) {
  612. _debug("base guid [$entry_guid or $entry_guid_hashed] not found, creating...", $debug_enabled);
  613. // base post entry does not exist, create it
  614. db_query(
  615. "INSERT INTO ttrss_entries
  616. (title,
  617. guid,
  618. link,
  619. updated,
  620. content,
  621. content_hash,
  622. no_orig_date,
  623. date_updated,
  624. date_entered,
  625. comments,
  626. num_comments,
  627. plugin_data,
  628. lang,
  629. author)
  630. VALUES
  631. ('$entry_title',
  632. '$entry_guid_hashed',
  633. '$entry_link',
  634. '$entry_timestamp_fmt',
  635. '$entry_content',
  636. '$entry_current_hash',
  637. false,
  638. NOW(),
  639. '$date_feed_processed',
  640. '$entry_comments',
  641. '$num_comments',
  642. '$entry_plugin_data',
  643. '$entry_language',
  644. '$entry_author')");
  645. }
  646. // now it should exist, if not - bad luck then
  647. $result = db_query("SELECT id FROM ttrss_entries
  648. WHERE guid = '$entry_guid' OR guid = '$entry_guid_hashed'");
  649. $entry_ref_id = 0;
  650. $entry_int_id = 0;
  651. if (db_num_rows($result) == 1) {
  652. _debug("base guid found, checking for user record", $debug_enabled);
  653. $ref_id = db_fetch_result($result, 0, "id");
  654. $entry_ref_id = $ref_id;
  655. /* $stored_guid = db_fetch_result($result, 0, "guid");
  656. if ($stored_guid != $entry_guid_hashed) {
  657. if ($debug_enabled) _debug("upgrading compat guid to hashed one", $debug_enabled);
  658. db_query("UPDATE ttrss_entries SET guid = '$entry_guid_hashed' WHERE
  659. id = '$ref_id'");
  660. } */
  661. if (find_article_filter($article_filters, "filter")) {
  662. //db_query("COMMIT"); // close transaction in progress
  663. continue;
  664. }
  665. $score = calculate_article_score($article_filters) + $entry_score_modifier;
  666. _debug("initial score: $score [including plugin modifier: $entry_score_modifier]", $debug_enabled);
  667. // check for user post link to main table
  668. $query = "SELECT ref_id, int_id FROM ttrss_user_entries WHERE
  669. ref_id = '$ref_id' AND owner_uid = '$owner_uid'";
  670. // if ($_REQUEST["xdebug"]) print "$query\n";
  671. $result = db_query($query);
  672. // okay it doesn't exist - create user entry
  673. if (db_num_rows($result) == 0) {
  674. _debug("user record not found, creating...", $debug_enabled);
  675. if ($score >= -500 && !find_article_filter($article_filters, 'catchup') && !$entry_force_catchup) {
  676. $unread = 'true';
  677. $last_read_qpart = 'NULL';
  678. } else {
  679. $unread = 'false';
  680. $last_read_qpart = 'NOW()';
  681. }
  682. if (find_article_filter($article_filters, 'mark') || $score > 1000) {
  683. $marked = 'true';
  684. } else {
  685. $marked = 'false';
  686. }
  687. if (find_article_filter($article_filters, 'publish')) {
  688. $published = 'true';
  689. } else {
  690. $published = 'false';
  691. }
  692. $last_marked = ($marked == 'true') ? 'NOW()' : 'NULL';
  693. $last_published = ($published == 'true') ? 'NOW()' : 'NULL';
  694. $result = db_query(
  695. "INSERT INTO ttrss_user_entries
  696. (ref_id, owner_uid, feed_id, unread, last_read, marked,
  697. published, score, tag_cache, label_cache, uuid,
  698. last_marked, last_published)
  699. VALUES ('$ref_id', '$owner_uid', '$feed', $unread,
  700. $last_read_qpart, $marked, $published, '$score', '', '',
  701. '', $last_marked, $last_published)");
  702. if (PUBSUBHUBBUB_HUB && $published == 'true') {
  703. $rss_link = get_self_url_prefix() .
  704. "/public.php?op=rss&id=-2&key=" .
  705. get_feed_access_key(-2, false, $owner_uid);
  706. $p = new pubsubhubbub\publisher\Publisher(PUBSUBHUBBUB_HUB);
  707. /* $pubsub_result = */ $p->publish_update($rss_link);
  708. }
  709. $result = db_query(
  710. "SELECT int_id FROM ttrss_user_entries WHERE
  711. ref_id = '$ref_id' AND owner_uid = '$owner_uid' AND
  712. feed_id = '$feed' LIMIT 1");
  713. if (db_num_rows($result) == 1) {
  714. $entry_int_id = db_fetch_result($result, 0, "int_id");
  715. }
  716. } else {
  717. _debug("user record FOUND", $debug_enabled);
  718. $entry_ref_id = db_fetch_result($result, 0, "ref_id");
  719. $entry_int_id = db_fetch_result($result, 0, "int_id");
  720. }
  721. _debug("RID: $entry_ref_id, IID: $entry_int_id", $debug_enabled);
  722. if (DB_TYPE == "pgsql") {
  723. $tsvector_combined = db_escape_string(mb_substr($entry_title . ' ' . strip_tags(str_replace('<', ' <', $entry_content)),
  724. 0, 1000000));
  725. $tsvector_qpart = "tsvector_combined = to_tsvector('$feed_language', '$tsvector_combined'),";
  726. } else {
  727. $tsvector_qpart = "";
  728. }
  729. db_query("UPDATE ttrss_entries
  730. SET title = '$entry_title',
  731. content = '$entry_content',
  732. content_hash = '$entry_current_hash',
  733. updated = '$entry_timestamp_fmt',
  734. $tsvector_qpart
  735. num_comments = '$num_comments',
  736. plugin_data = '$entry_plugin_data',
  737. author = '$entry_author',
  738. lang = '$entry_language'
  739. WHERE id = '$ref_id'");
  740. // update aux data
  741. db_query("UPDATE ttrss_user_entries
  742. SET score = '$score' WHERE ref_id = '$ref_id'");
  743. if ($mark_unread_on_update) {
  744. _debug("article updated, marking unread as requested.", $debug_enabled);
  745. db_query("UPDATE ttrss_user_entries
  746. SET last_read = null, unread = true WHERE ref_id = '$ref_id'");
  747. }
  748. }
  749. //db_query("COMMIT");
  750. _debug("assigning labels [other]...", $debug_enabled);
  751. foreach ($article_labels as $label) {
  752. label_add_article($entry_ref_id, $label[1], $owner_uid);
  753. }
  754. _debug("assigning labels [filters]...", $debug_enabled);
  755. assign_article_to_label_filters($entry_ref_id, $article_filters,
  756. $owner_uid, $article_labels);
  757. _debug("looking for enclosures...", $debug_enabled);
  758. // enclosures
  759. $enclosures = array();
  760. $encs = $item->get_enclosures();
  761. if (is_array($encs)) {
  762. foreach ($encs as $e) {
  763. $e_item = array(
  764. rewrite_relative_url($site_url, $e->link),
  765. $e->type, $e->length, $e->title, $e->width, $e->height);
  766. array_push($enclosures, $e_item);
  767. }
  768. }
  769. if ($cache_images && is_writable(CACHE_DIR . '/images'))
  770. cache_enclosures($enclosures, $site_url, $debug_enabled);
  771. if ($debug_enabled) {
  772. _debug("article enclosures:", $debug_enabled);
  773. print_r($enclosures);
  774. }
  775. //db_query("BEGIN");
  776. // debugging
  777. // db_query("DELETE FROM ttrss_enclosures WHERE post_id = '$entry_ref_id'");
  778. foreach ($enclosures as $enc) {
  779. $enc_url = db_escape_string($enc[0]);
  780. $enc_type = db_escape_string($enc[1]);
  781. $enc_dur = db_escape_string($enc[2]);
  782. $enc_title = db_escape_string($enc[3]);
  783. $enc_width = intval($enc[4]);
  784. $enc_height = intval($enc[5]);
  785. $result = db_query("SELECT id FROM ttrss_enclosures
  786. WHERE content_url = '$enc_url' AND post_id = '$entry_ref_id'");
  787. if (db_num_rows($result) == 0) {
  788. db_query("INSERT INTO ttrss_enclosures
  789. (content_url, content_type, title, duration, post_id, width, height) VALUES
  790. ('$enc_url', '$enc_type', '$enc_title', '$enc_dur', '$entry_ref_id', $enc_width, $enc_height)");
  791. }
  792. }
  793. //db_query("COMMIT");
  794. // check for manual tags (we have to do it here since they're loaded from filters)
  795. foreach ($article_filters as $f) {
  796. if ($f["type"] == "tag") {
  797. $manual_tags = trim_array(explode(",", $f["param"]));
  798. foreach ($manual_tags as $tag) {
  799. if (tag_is_valid($tag)) {
  800. array_push($entry_tags, $tag);
  801. }
  802. }
  803. }
  804. }
  805. // Skip boring tags
  806. $boring_tags = trim_array(explode(",", mb_strtolower(get_pref(
  807. 'BLACKLISTED_TAGS', $owner_uid, ''), 'utf-8')));
  808. $filtered_tags = array();
  809. $tags_to_cache = array();
  810. if ($entry_tags && is_array($entry_tags)) {
  811. foreach ($entry_tags as $tag) {
  812. if (array_search($tag, $boring_tags) === false) {
  813. array_push($filtered_tags, $tag);
  814. }
  815. }
  816. }
  817. $filtered_tags = array_unique($filtered_tags);
  818. if ($debug_enabled) {
  819. _debug("filtered article tags:", $debug_enabled);
  820. print_r($filtered_tags);
  821. }
  822. // Save article tags in the database
  823. if (count($filtered_tags) > 0) {
  824. //db_query("BEGIN");
  825. foreach ($filtered_tags as $tag) {
  826. $tag = sanitize_tag($tag);
  827. $tag = db_escape_string($tag);
  828. if (!tag_is_valid($tag)) continue;
  829. $result = db_query("SELECT id FROM ttrss_tags
  830. WHERE tag_name = '$tag' AND post_int_id = '$entry_int_id' AND
  831. owner_uid = '$owner_uid' LIMIT 1");
  832. if ($result && db_num_rows($result) == 0) {
  833. db_query("INSERT INTO ttrss_tags
  834. (owner_uid,tag_name,post_int_id)
  835. VALUES ('$owner_uid','$tag', '$entry_int_id')");
  836. }
  837. array_push($tags_to_cache, $tag);
  838. }
  839. /* update the cache */
  840. $tags_to_cache = array_unique($tags_to_cache);
  841. $tags_str = db_escape_string(join(",", $tags_to_cache));
  842. db_query("UPDATE ttrss_user_entries
  843. SET tag_cache = '$tags_str' WHERE ref_id = '$entry_ref_id'
  844. AND owner_uid = $owner_uid");
  845. //db_query("COMMIT");
  846. }
  847. _debug("article processed", $debug_enabled);
  848. }
  849. _debug("purging feed...", $debug_enabled);
  850. purge_feed($feed, 0, $debug_enabled);
  851. db_query("UPDATE ttrss_feeds
  852. SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
  853. // db_query("COMMIT");
  854. } else {
  855. $error_msg = db_escape_string(mb_substr($rss->error(), 0, 245));
  856. _debug("fetch error: $error_msg", $debug_enabled);
  857. if (count($rss->errors()) > 1) {
  858. foreach ($rss->errors() as $error) {
  859. _debug("+ $error");
  860. }
  861. }
  862. db_query(
  863. "UPDATE ttrss_feeds SET last_error = '$error_msg',
  864. last_updated = NOW() WHERE id = '$feed'");
  865. unset($rss);
  866. return;
  867. }
  868. _debug("done", $debug_enabled);
  869. return true;
  870. }
  871. function cache_enclosures($enclosures, $site_url, $debug) {
  872. foreach ($enclosures as $enc) {
  873. if (preg_match("/(image|audio|video)/", $enc[1])) {
  874. $src = rewrite_relative_url($site_url, $enc[0]);
  875. $local_filename = CACHE_DIR . "/images/" . sha1($src);
  876. if ($debug) _debug("cache_enclosures: downloading: $src to $local_filename");
  877. if (!file_exists($local_filename)) {
  878. $file_content = fetch_file_contents($src);
  879. if ($file_content && strlen($file_content) > _MIN_CACHE_FILE_SIZE) {
  880. file_put_contents($local_filename, $file_content);
  881. }
  882. } else {
  883. touch($local_filename);
  884. }
  885. }
  886. }
  887. }
  888. function cache_media($html, $site_url, $debug) {
  889. libxml_use_internal_errors(true);
  890. $charset_hack = '<head>
  891. <meta http-equiv="Content-Type" content="text/html; charset=utf-8"/>
  892. </head>';
  893. $doc = new DOMDocument();
  894. $doc->loadHTML($charset_hack . $html);
  895. $xpath = new DOMXPath($doc);
  896. $entries = $xpath->query('(//img[@src])|(//video/source[@src])|(//audio/source[@src])');
  897. foreach ($entries as $entry) {
  898. if ($entry->hasAttribute('src') && strpos($entry->getAttribute('src'), "data:") !== 0) {
  899. $src = rewrite_relative_url($site_url, $entry->getAttribute('src'));
  900. $local_filename = CACHE_DIR . "/images/" . sha1($src);
  901. if ($debug) _debug("cache_media: downloading: $src to $local_filename");
  902. if (!file_exists($local_filename)) {
  903. $file_content = fetch_file_contents($src);
  904. if ($file_content && strlen($file_content) > _MIN_CACHE_FILE_SIZE) {
  905. file_put_contents($local_filename, $file_content);
  906. }
  907. } else {
  908. touch($local_filename);
  909. }
  910. }
  911. }
  912. }
  913. function expire_error_log($debug) {
  914. if ($debug) _debug("Removing old error log entries...");
  915. if (DB_TYPE == "pgsql") {
  916. db_query("DELETE FROM ttrss_error_log
  917. WHERE created_at < NOW() - INTERVAL '7 days'");
  918. } else {
  919. db_query("DELETE FROM ttrss_error_log
  920. WHERE created_at < DATE_SUB(NOW(), INTERVAL 7 DAY)");
  921. }
  922. }
  923. function expire_lock_files($debug) {
  924. //if ($debug) _debug("Removing old lock files...");
  925. $num_deleted = 0;
  926. if (is_writable(LOCK_DIRECTORY)) {
  927. $files = glob(LOCK_DIRECTORY . "/*.lock");
  928. if ($files) {
  929. foreach ($files as $file) {
  930. if (!file_is_locked(basename($file)) && time() - filemtime($file) > 86400*2) {
  931. unlink($file);
  932. ++$num_deleted;
  933. }
  934. }
  935. }
  936. }
  937. if ($debug) _debug("Removed $num_deleted old lock files.");
  938. }
  939. function expire_cached_files($debug) {
  940. foreach (array("simplepie", "images", "export", "upload") as $dir) {
  941. $cache_dir = CACHE_DIR . "/$dir";
  942. // if ($debug) _debug("Expiring $cache_dir");
  943. $num_deleted = 0;
  944. if (is_writable($cache_dir)) {
  945. $files = glob("$cache_dir/*");
  946. if ($files) {
  947. foreach ($files as $file) {
  948. if (time() - filemtime($file) > 86400*7) {
  949. unlink($file);
  950. ++$num_deleted;
  951. }
  952. }
  953. }
  954. }
  955. if ($debug) _debug("$cache_dir: removed $num_deleted files.");
  956. }
  957. }
  958. /**
  959. * Source: http://www.php.net/manual/en/function.parse-url.php#104527
  960. * Returns the url query as associative array
  961. *
  962. * @param string query
  963. * @return array params
  964. */
  965. function convertUrlQuery($query) {
  966. $queryParts = explode('&', $query);
  967. $params = array();
  968. foreach ($queryParts as $param) {
  969. $item = explode('=', $param);
  970. $params[$item[0]] = $item[1];
  971. }
  972. return $params;
  973. }
  974. function get_article_filters($filters, $title, $content, $link, $author, $tags, &$matched_rules = false) {
  975. $matches = array();
  976. foreach ($filters as $filter) {
  977. $match_any_rule = $filter["match_any_rule"];
  978. $inverse = $filter["inverse"];
  979. $filter_match = false;
  980. foreach ($filter["rules"] as $rule) {
  981. $match = false;
  982. $reg_exp = str_replace('/', '\/', $rule["reg_exp"]);
  983. $rule_inverse = $rule["inverse"];
  984. if (!$reg_exp)
  985. continue;
  986. switch ($rule["type"]) {
  987. case "title":
  988. $match = @preg_match("/$reg_exp/iu", $title);
  989. break;
  990. case "content":
  991. // we don't need to deal with multiline regexps
  992. $content = preg_replace("/[\r\n\t]/", "", $content);
  993. $match = @preg_match("/$reg_exp/iu", $content);
  994. break;
  995. case "both":
  996. // we don't need to deal with multiline regexps
  997. $content = preg_replace("/[\r\n\t]/", "", $content);
  998. $match = (@preg_match("/$reg_exp/iu", $title) || @preg_match("/$reg_exp/iu", $content));
  999. break;
  1000. case "link":
  1001. $match = @preg_match("/$reg_exp/iu", $link);
  1002. break;
  1003. case "author":
  1004. $match = @preg_match("/$reg_exp/iu", $author);
  1005. break;
  1006. case "tag":
  1007. foreach ($tags as $tag) {
  1008. if (@preg_match("/$reg_exp/iu", $tag)) {
  1009. $match = true;
  1010. break;
  1011. }
  1012. }
  1013. break;
  1014. }
  1015. if ($rule_inverse) $match = !$match;
  1016. if ($match_any_rule) {
  1017. if ($match) {
  1018. $filter_match = true;
  1019. break;
  1020. }
  1021. } else {
  1022. $filter_match = $match;
  1023. if (!$match) {
  1024. break;
  1025. }
  1026. }
  1027. }
  1028. if ($inverse) $filter_match = !$filter_match;
  1029. if ($filter_match) {
  1030. if (is_array($matched_rules)) array_push($matched_rules, $rule);
  1031. foreach ($filter["actions"] AS $action) {
  1032. array_push($matches, $action);
  1033. // if Stop action encountered, perform no further processing
  1034. if (isset($action["type"]) && $action["type"] == "stop") return $matches;
  1035. }
  1036. }
  1037. }
  1038. return $matches;
  1039. }
  1040. function find_article_filter($filters, $filter_name) {
  1041. foreach ($filters as $f) {
  1042. if ($f["type"] == $filter_name) {
  1043. return $f;
  1044. };
  1045. }
  1046. return false;
  1047. }
  1048. function find_article_filters($filters, $filter_name) {
  1049. $results = array();
  1050. foreach ($filters as $f) {
  1051. if ($f["type"] == $filter_name) {
  1052. array_push($results, $f);
  1053. };
  1054. }
  1055. return $results;
  1056. }
  1057. function calculate_article_score($filters) {
  1058. $score = 0;
  1059. foreach ($filters as $f) {
  1060. if ($f["type"] == "score") {
  1061. $score += $f["param"];
  1062. };
  1063. }
  1064. return $score;
  1065. }
  1066. function labels_contains_caption($labels, $caption) {
  1067. foreach ($labels as $label) {
  1068. if ($label[1] == $caption) {
  1069. return true;
  1070. }
  1071. }
  1072. return false;
  1073. }
  1074. function assign_article_to_label_filters($id, $filters, $owner_uid, $article_labels) {
  1075. foreach ($filters as $f) {
  1076. if ($f["type"] == "label") {
  1077. if (!labels_contains_caption($article_labels, $f["param"])) {
  1078. label_add_article($id, $f["param"], $owner_uid);
  1079. }
  1080. }
  1081. }
  1082. }
  1083. function make_guid_from_title($title) {
  1084. return preg_replace("/[ \"\',.:;]/", "-",
  1085. mb_strtolower(strip_tags($title), 'utf-8'));
  1086. }
  1087. function cleanup_counters_cache($debug) {
  1088. $result = db_query("DELETE FROM ttrss_counters_cache
  1089. WHERE feed_id > 0 AND
  1090. (SELECT COUNT(id) FROM ttrss_feeds WHERE
  1091. id = feed_id AND
  1092. ttrss_counters_cache.owner_uid = ttrss_feeds.owner_uid) = 0");
  1093. $frows = db_affected_rows($result);
  1094. $result = db_query("DELETE FROM ttrss_cat_counters_cache
  1095. WHERE feed_id > 0 AND
  1096. (SELECT COUNT(id) FROM ttrss_feed_categories WHERE
  1097. id = feed_id AND
  1098. ttrss_cat_counters_cache.owner_uid = ttrss_feed_categories.owner_uid) = 0");
  1099. $crows = db_affected_rows($result);
  1100. if ($debug) _debug("Removed $frows (feeds) $crows (cats) orphaned counter cache entries.");
  1101. }
  1102. function housekeeping_user($owner_uid) {
  1103. $tmph = new PluginHost();
  1104. load_user_plugins($owner_uid, $tmph);
  1105. $tmph->run_hooks(PluginHost::HOOK_HOUSE_KEEPING, "hook_house_keeping", "");
  1106. }
  1107. function housekeeping_common($debug) {
  1108. expire_cached_files($debug);
  1109. expire_lock_files($debug);
  1110. expire_error_log($debug);
  1111. $count = update_feedbrowser_cache();
  1112. _debug("Feedbrowser updated, $count feeds processed.");
  1113. Article::purge_orphans( true);
  1114. cleanup_counters_cache($debug);
  1115. //$rc = cleanup_tags( 14, 50000);
  1116. //_debug("Cleaned $rc cached tags.");
  1117. PluginHost::getInstance()->run_hooks(PluginHost::HOOK_HOUSE_KEEPING, "hook_house_keeping", "");
  1118. }
  1119. function check_feed_favicon($site_url, $feed) {
  1120. # print "FAVICON [$site_url]: $favicon_url\n";
  1121. $icon_file = ICONS_DIR . "/$feed.ico";
  1122. if (!file_exists($icon_file)) {
  1123. $favicon_url = get_favicon_url($site_url);
  1124. if ($favicon_url) {
  1125. // Limiting to "image" type misses those served with text/plain
  1126. $contents = fetch_file_contents($favicon_url); // , "image");
  1127. if ($contents) {
  1128. // Crude image type matching.
  1129. // Patterns gleaned from the file(1) source code.
  1130. if (preg_match('/^\x00\x00\x01\x00/', $contents)) {
  1131. // 0 string \000\000\001\000 MS Windows icon resource
  1132. //error_log("check_feed_favicon: favicon_url=$favicon_url isa MS Windows icon resource");
  1133. }
  1134. elseif (preg_match('/^GIF8/', $contents)) {
  1135. // 0 string GIF8 GIF image data
  1136. //error_log("check_feed_favicon: favicon_url=$favicon_url isa GIF image");
  1137. }
  1138. elseif (preg_match('/^\x89PNG\x0d\x0a\x1a\x0a/', $contents)) {
  1139. // 0 string \x89PNG\x0d\x0a\x1a\x0a PNG image data
  1140. //error_log("check_feed_favicon: favicon_url=$favicon_url isa PNG image");
  1141. }
  1142. elseif (preg_match('/^\xff\xd8/', $contents)) {
  1143. // 0 beshort 0xffd8 JPEG image data
  1144. //error_log("check_feed_favicon: favicon_url=$favicon_url isa JPG image");
  1145. }
  1146. else {
  1147. //error_log("check_feed_favicon: favicon_url=$favicon_url isa UNKNOWN type");
  1148. $contents = "";
  1149. }
  1150. }
  1151. if ($contents) {
  1152. $fp = @fopen($icon_file, "w");
  1153. if ($fp) {
  1154. fwrite($fp, $contents);
  1155. fclose($fp);
  1156. chmod($icon_file, 0644);
  1157. }
  1158. }
  1159. }
  1160. return $icon_file;
  1161. }
  1162. }