]>
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 | static function update_feedbrowser_cache() { | |
20 | ||
21 | $result = db_query("SELECT feed_url, site_url, title, COUNT(id) AS subscribers | |
22 | FROM ttrss_feeds WHERE feed_url NOT IN (SELECT feed_url FROM ttrss_feeds | |
23 | WHERE private IS true OR auth_login != '' OR auth_pass != '' OR feed_url LIKE '%:%@%/%') | |
24 | GROUP BY feed_url, site_url, title ORDER BY subscribers DESC LIMIT 1000"); | |
25 | ||
26 | db_query("BEGIN"); | |
27 | ||
28 | db_query("DELETE FROM ttrss_feedbrowser_cache"); | |
29 | ||
30 | $count = 0; | |
31 | ||
32 | while ($line = db_fetch_assoc($result)) { | |
33 | $subscribers = db_escape_string($line["subscribers"]); | |
34 | $feed_url = db_escape_string($line["feed_url"]); | |
35 | $title = db_escape_string($line["title"]); | |
36 | $site_url = db_escape_string($line["site_url"]); | |
37 | ||
38 | $tmp_result = db_query("SELECT subscribers FROM | |
39 | ttrss_feedbrowser_cache WHERE feed_url = '$feed_url'"); | |
40 | ||
41 | if (db_num_rows($tmp_result) == 0) { | |
42 | ||
43 | db_query("INSERT INTO ttrss_feedbrowser_cache | |
44 | (feed_url, site_url, title, subscribers) VALUES ('$feed_url', | |
45 | '$site_url', '$title', '$subscribers')"); | |
46 | ||
47 | ++$count; | |
48 | ||
49 | } | |
50 | ||
51 | } | |
52 | ||
53 | db_query("COMMIT"); | |
54 | ||
55 | return $count; | |
56 | ||
57 | } | |
58 | ||
59 | static function update_daemon_common($limit = DAEMON_FEED_LIMIT, $debug = true) { | |
60 | $schema_version = get_schema_version(); | |
61 | ||
62 | if ($schema_version != SCHEMA_VERSION) { | |
63 | die("Schema version is wrong, please upgrade the database.\n"); | |
64 | } | |
65 | ||
66 | if (!SINGLE_USER_MODE && DAEMON_UPDATE_LOGIN_LIMIT > 0) { | |
67 | if (DB_TYPE == "pgsql") { | |
68 | $login_thresh_qpart = "AND ttrss_users.last_login >= NOW() - INTERVAL '".DAEMON_UPDATE_LOGIN_LIMIT." days'"; | |
69 | } else { | |
70 | $login_thresh_qpart = "AND ttrss_users.last_login >= DATE_SUB(NOW(), INTERVAL ".DAEMON_UPDATE_LOGIN_LIMIT." DAY)"; | |
71 | } | |
72 | } else { | |
73 | $login_thresh_qpart = ""; | |
74 | } | |
75 | ||
76 | if (DB_TYPE == "pgsql") { | |
77 | $update_limit_qpart = "AND (( | |
78 | ttrss_feeds.update_interval = 0 | |
79 | AND ttrss_user_prefs.value != '-1' | |
80 | AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_user_prefs.value || ' minutes') AS INTERVAL) | |
81 | ) OR ( | |
82 | ttrss_feeds.update_interval > 0 | |
83 | AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_feeds.update_interval || ' minutes') AS INTERVAL) | |
84 | ) OR (ttrss_feeds.last_updated IS NULL | |
85 | AND ttrss_user_prefs.value != '-1') | |
86 | OR (last_updated = '1970-01-01 00:00:00' | |
87 | AND ttrss_user_prefs.value != '-1'))"; | |
88 | } else { | |
89 | $update_limit_qpart = "AND (( | |
90 | ttrss_feeds.update_interval = 0 | |
91 | AND ttrss_user_prefs.value != '-1' | |
92 | AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL CONVERT(ttrss_user_prefs.value, SIGNED INTEGER) MINUTE) | |
93 | ) OR ( | |
94 | ttrss_feeds.update_interval > 0 | |
95 | AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL ttrss_feeds.update_interval MINUTE) | |
96 | ) OR (ttrss_feeds.last_updated IS NULL | |
97 | AND ttrss_user_prefs.value != '-1') | |
98 | OR (last_updated = '1970-01-01 00:00:00' | |
99 | AND ttrss_user_prefs.value != '-1'))"; | |
100 | } | |
101 | ||
102 | // Test if feed is currently being updated by another process. | |
103 | if (DB_TYPE == "pgsql") { | |
104 | $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < NOW() - INTERVAL '10 minutes')"; | |
105 | } else { | |
106 | $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < DATE_SUB(NOW(), INTERVAL 10 MINUTE))"; | |
107 | } | |
108 | ||
109 | $query_limit = $limit ? sprintf("LIMIT %d", $limit) : ""; | |
110 | ||
111 | // Update the least recently updated feeds first | |
112 | $query_order = "ORDER BY last_updated"; | |
113 | if (DB_TYPE == "pgsql") $query_order .= " NULLS FIRST"; | |
114 | ||
115 | $query = "SELECT DISTINCT ttrss_feeds.feed_url, ttrss_feeds.last_updated | |
116 | FROM | |
117 | ttrss_feeds, ttrss_users, ttrss_user_prefs | |
118 | WHERE | |
119 | ttrss_feeds.owner_uid = ttrss_users.id | |
120 | AND ttrss_user_prefs.profile IS NULL | |
121 | AND ttrss_users.id = ttrss_user_prefs.owner_uid | |
122 | AND ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL' | |
123 | $login_thresh_qpart $update_limit_qpart | |
124 | $updstart_thresh_qpart | |
125 | $query_order $query_limit"; | |
126 | ||
127 | $result = db_query($query); | |
128 | ||
129 | if ($debug) _debug(sprintf("Scheduled %d feeds to update...", db_num_rows($result))); | |
130 | ||
131 | $feeds_to_update = array(); | |
132 | while ($line = db_fetch_assoc($result)) { | |
133 | array_push($feeds_to_update, $line['feed_url']); | |
134 | } | |
135 | ||
136 | // Update last_update_started before actually starting the batch | |
137 | // in order to minimize collision risk for parallel daemon tasks | |
138 | if (count($feeds_to_update) > 0) { | |
139 | $feeds_quoted = array_map(function ($s) { return "'" . db_escape_string($s) . "'"; }, $feeds_to_update); | |
140 | ||
141 | db_query(sprintf("UPDATE ttrss_feeds SET last_update_started = NOW() | |
142 | WHERE feed_url IN (%s)", implode(',', $feeds_quoted))); | |
143 | } | |
144 | ||
145 | $nf = 0; | |
146 | $bstarted = microtime(true); | |
147 | ||
148 | $batch_owners = array(); | |
149 | ||
150 | foreach ($feeds_to_update as $feed) { | |
151 | if($debug) _debug("Base feed: $feed"); | |
152 | ||
153 | //update_rss_feed($line["id"], true); | |
154 | ||
155 | // since we have the data cached, we can deal with other feeds with the same url | |
156 | $tmp_result = db_query("SELECT DISTINCT ttrss_feeds.id,last_updated,ttrss_feeds.owner_uid | |
157 | FROM ttrss_feeds, ttrss_users, ttrss_user_prefs WHERE | |
158 | ttrss_user_prefs.owner_uid = ttrss_feeds.owner_uid AND | |
159 | ttrss_users.id = ttrss_user_prefs.owner_uid AND | |
160 | ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL' AND | |
161 | ttrss_user_prefs.profile IS NULL AND | |
162 | feed_url = '".db_escape_string($feed)."' | |
163 | $update_limit_qpart | |
164 | $login_thresh_qpart | |
165 | ORDER BY ttrss_feeds.id $query_limit"); | |
166 | ||
167 | if (db_num_rows($tmp_result) > 0) { | |
168 | while ($tline = db_fetch_assoc($tmp_result)) { | |
169 | if ($debug) _debug(" => " . $tline["last_updated"] . ", " . $tline["id"] . " " . $tline["owner_uid"]); | |
170 | ||
171 | if (array_search($tline["owner_uid"], $batch_owners) === FALSE) | |
172 | array_push($batch_owners, $tline["owner_uid"]); | |
173 | ||
174 | $fstarted = microtime(true); | |
175 | RSSUtils::update_rss_feed($tline["id"], true, false); | |
176 | _debug_suppress(false); | |
177 | ||
178 | _debug(sprintf(" %.4f (sec)", microtime(true) - $fstarted)); | |
179 | ||
180 | ++$nf; | |
181 | } | |
182 | } | |
183 | } | |
184 | ||
185 | if ($nf > 0) { | |
186 | _debug(sprintf("Processed %d feeds in %.4f (sec), %.4f (sec/feed avg)", $nf, | |
187 | microtime(true) - $bstarted, (microtime(true) - $bstarted) / $nf)); | |
188 | } | |
189 | ||
190 | foreach ($batch_owners as $owner_uid) { | |
191 | _debug("Running housekeeping tasks for user $owner_uid..."); | |
192 | ||
193 | RSSUtils::housekeeping_user($owner_uid); | |
194 | } | |
195 | ||
196 | // Send feed digests by email if needed. | |
197 | Digest::send_headlines_digests($debug); | |
198 | ||
199 | return $nf; | |
200 | ||
201 | } | |
202 | ||
203 | // this is used when subscribing | |
204 | static function set_basic_feed_info($feed) { | |
205 | ||
206 | $feed = db_escape_string($feed); | |
207 | ||
208 | $result = db_query("SELECT feed_url,auth_pass,auth_login,auth_pass_encrypted | |
209 | FROM ttrss_feeds WHERE id = '$feed'"); | |
210 | ||
211 | $auth_pass_encrypted = sql_bool_to_bool(db_fetch_result($result, | |
212 | 0, "auth_pass_encrypted")); | |
213 | ||
214 | $auth_login = db_fetch_result($result, 0, "auth_login"); | |
215 | $auth_pass = db_fetch_result($result, 0, "auth_pass"); | |
216 | ||
217 | if ($auth_pass_encrypted && function_exists("mcrypt_decrypt")) { | |
218 | require_once "crypt.php"; | |
219 | $auth_pass = decrypt_string($auth_pass); | |
220 | } | |
221 | ||
222 | $fetch_url = db_fetch_result($result, 0, "feed_url"); | |
223 | ||
224 | $feed_data = fetch_file_contents($fetch_url, false, | |
225 | $auth_login, $auth_pass, false, | |
226 | FEED_FETCH_TIMEOUT, | |
227 | 0); | |
228 | ||
229 | global $fetch_curl_used; | |
230 | ||
231 | if (!$fetch_curl_used) { | |
232 | $tmp = @gzdecode($feed_data); | |
233 | ||
234 | if ($tmp) $feed_data = $tmp; | |
235 | } | |
236 | ||
237 | $feed_data = trim($feed_data); | |
238 | ||
239 | $rss = new FeedParser($feed_data); | |
240 | $rss->init(); | |
241 | ||
242 | if (!$rss->error()) { | |
243 | ||
244 | $result = db_query("SELECT title, site_url FROM ttrss_feeds WHERE id = '$feed'"); | |
245 | ||
246 | $registered_title = db_fetch_result($result, 0, "title"); | |
247 | $orig_site_url = db_fetch_result($result, 0, "site_url"); | |
248 | ||
249 | $site_url = db_escape_string(mb_substr(rewrite_relative_url($fetch_url, $rss->get_link()), 0, 245)); | |
250 | $feed_title = db_escape_string(mb_substr($rss->get_title(), 0, 199)); | |
251 | ||
252 | if ($feed_title && (!$registered_title || $registered_title == "[Unknown]")) { | |
253 | db_query("UPDATE ttrss_feeds SET | |
254 | title = '$feed_title' WHERE id = '$feed'"); | |
255 | } | |
256 | ||
257 | if ($site_url && $orig_site_url != $site_url) { | |
258 | db_query("UPDATE ttrss_feeds SET | |
259 | site_url = '$site_url' WHERE id = '$feed'"); | |
260 | } | |
261 | } | |
262 | } | |
263 | ||
264 | /** | |
265 | * @SuppressWarnings(PHPMD.UnusedFormalParameter) | |
266 | */ | |
267 | static function update_rss_feed($feed, $no_cache = false) { | |
268 | ||
269 | $debug_enabled = defined('DAEMON_EXTENDED_DEBUG') || $_REQUEST['xdebug']; | |
270 | ||
271 | _debug_suppress(!$debug_enabled); | |
272 | _debug("start", $debug_enabled); | |
273 | ||
274 | $result = db_query("SELECT title FROM ttrss_feeds | |
275 | WHERE id = '$feed'"); | |
276 | ||
277 | if (db_num_rows($result) == 0) { | |
278 | _debug("feed $feed NOT FOUND/SKIPPED", $debug_enabled); | |
279 | user_error("Attempt to update unknown/invalid feed $feed", E_USER_WARNING); | |
280 | return false; | |
281 | } | |
282 | ||
283 | $title = db_fetch_result($result, 0, "title"); | |
284 | ||
285 | // feed was batch-subscribed or something, we need to get basic info | |
286 | // this is not optimal currently as it fetches stuff separately TODO: optimize | |
287 | if ($title == "[Unknown]") { | |
288 | _debug("setting basic feed info for $feed..."); | |
289 | RSSUtils::set_basic_feed_info($feed); | |
290 | } | |
291 | ||
292 | $result = db_query("SELECT id,update_interval,auth_login, | |
293 | feed_url,auth_pass,cache_images, | |
294 | mark_unread_on_update, owner_uid, | |
295 | auth_pass_encrypted, feed_language, | |
296 | last_modified | |
297 | FROM ttrss_feeds WHERE id = '$feed'"); | |
298 | ||
299 | $owner_uid = db_fetch_result($result, 0, "owner_uid"); | |
300 | $mark_unread_on_update = sql_bool_to_bool(db_fetch_result($result, | |
301 | 0, "mark_unread_on_update")); | |
302 | $auth_pass_encrypted = sql_bool_to_bool(db_fetch_result($result, | |
303 | 0, "auth_pass_encrypted")); | |
304 | ||
305 | db_query("UPDATE ttrss_feeds SET last_update_started = NOW() | |
306 | WHERE id = '$feed'"); | |
307 | ||
308 | $auth_login = db_fetch_result($result, 0, "auth_login"); | |
309 | $auth_pass = db_fetch_result($result, 0, "auth_pass"); | |
310 | ||
311 | if ($auth_pass_encrypted && function_exists("mcrypt_decrypt")) { | |
312 | require_once "crypt.php"; | |
313 | $auth_pass = decrypt_string($auth_pass); | |
314 | } | |
315 | ||
316 | $stored_last_modified = db_fetch_result($result, 0, "last_modified"); | |
317 | $cache_images = sql_bool_to_bool(db_fetch_result($result, 0, "cache_images")); | |
318 | $fetch_url = db_fetch_result($result, 0, "feed_url"); | |
319 | $feed_language = db_escape_string(mb_strtolower(db_fetch_result($result, 0, "feed_language"))); | |
320 | if (!$feed_language) $feed_language = 'english'; | |
321 | ||
322 | $feed = db_escape_string($feed); | |
323 | ||
324 | $date_feed_processed = date('Y-m-d H:i'); | |
325 | ||
326 | $cache_filename = CACHE_DIR . "/simplepie/" . sha1($fetch_url) . ".xml"; | |
327 | ||
328 | $pluginhost = new PluginHost(); | |
329 | $pluginhost->set_debug($debug_enabled); | |
330 | $user_plugins = get_pref("_ENABLED_PLUGINS", $owner_uid); | |
331 | ||
332 | $pluginhost->load(PLUGINS, PluginHost::KIND_ALL); | |
333 | $pluginhost->load($user_plugins, PluginHost::KIND_USER, $owner_uid); | |
334 | $pluginhost->load_data(); | |
335 | ||
336 | $rss_hash = false; | |
337 | ||
338 | $force_refetch = isset($_REQUEST["force_refetch"]); | |
339 | $feed_data = ""; | |
340 | ||
341 | foreach ($pluginhost->get_hooks(PluginHost::HOOK_FETCH_FEED) as $plugin) { | |
342 | $feed_data = $plugin->hook_fetch_feed($feed_data, $fetch_url, $owner_uid, $feed, 0, $auth_login, $auth_pass); | |
343 | } | |
344 | ||
345 | // try cache | |
346 | if (!$feed_data && | |
347 | file_exists($cache_filename) && | |
348 | is_readable($cache_filename) && | |
349 | !$auth_login && !$auth_pass && | |
350 | filemtime($cache_filename) > time() - 30) { | |
351 | ||
352 | _debug("using local cache [$cache_filename].", $debug_enabled); | |
353 | ||
354 | @$feed_data = file_get_contents($cache_filename); | |
355 | ||
356 | if ($feed_data) { | |
357 | $rss_hash = sha1($feed_data); | |
358 | } | |
359 | ||
360 | } else { | |
361 | _debug("local cache will not be used for this feed", $debug_enabled); | |
362 | } | |
363 | ||
364 | global $fetch_last_modified; | |
365 | ||
366 | // fetch feed from source | |
367 | if (!$feed_data) { | |
368 | _debug("stored last modified: $stored_last_modified", $debug_enabled); | |
369 | _debug("fetching [$fetch_url]...", $debug_enabled); | |
370 | ||
371 | if (ini_get("open_basedir") && function_exists("curl_init")) { | |
372 | _debug("not using CURL due to open_basedir restrictions"); | |
373 | } | |
374 | ||
375 | /*$feed_data = fetch_file_contents($fetch_url, false, | |
376 | $auth_login, $auth_pass, false, | |
377 | $no_cache ? FEED_FETCH_NO_CACHE_TIMEOUT : FEED_FETCH_TIMEOUT, | |
378 | 0);*/ | |
379 | ||
380 | // TODO: last_modified should be limited, if the feed has not been updated for a while | |
381 | // we probably should force one update without the header | |
382 | // unfortunately last_updated gets bumped on http 304 so that daemon would work properly | |
383 | ||
384 | $feed_data = fetch_file_contents([ | |
385 | "url" => $fetch_url, | |
386 | "login" => $auth_login, | |
387 | "pass" => $auth_pass, | |
388 | "timeout" => $no_cache ? FEED_FETCH_NO_CACHE_TIMEOUT : FEED_FETCH_TIMEOUT, | |
389 | "last_modified" => $force_refetch ? "" : $stored_last_modified | |
390 | ]); | |
391 | ||
392 | global $fetch_curl_used; | |
393 | ||
394 | if (!$fetch_curl_used) { | |
395 | $tmp = @gzdecode($feed_data); | |
396 | ||
397 | if ($tmp) $feed_data = $tmp; | |
398 | } | |
399 | ||
400 | $feed_data = trim($feed_data); | |
401 | ||
402 | _debug("fetch done.", $debug_enabled); | |
403 | _debug("source last modified: " . $fetch_last_modified); | |
404 | ||
405 | if ($feed_data && $fetch_last_modified != $stored_last_modified) { | |
406 | $last_modified_escaped = db_escape_string(substr($fetch_last_modified, 0, 245)); | |
407 | ||
408 | db_query("UPDATE ttrss_feeds SET last_modified = '$last_modified_escaped' WHERE id = '$feed'"); | |
409 | ||
410 | } | |
411 | ||
412 | // cache vanilla feed data for re-use | |
413 | if ($feed_data && !$auth_pass && !$auth_login && is_writable(CACHE_DIR . "/simplepie")) { | |
414 | $new_rss_hash = sha1($feed_data); | |
415 | ||
416 | if ($new_rss_hash != $rss_hash) { | |
417 | _debug("saving $cache_filename", $debug_enabled); | |
418 | @file_put_contents($cache_filename, $feed_data); | |
419 | } | |
420 | } | |
421 | } | |
422 | ||
423 | if (!$feed_data) { | |
424 | global $fetch_last_error; | |
425 | global $fetch_last_error_code; | |
426 | ||
427 | _debug("unable to fetch: $fetch_last_error [$fetch_last_error_code]", $debug_enabled); | |
428 | ||
429 | $error_escaped = ''; | |
430 | ||
431 | // If-Modified-Since | |
432 | if ($fetch_last_error_code != 304) { | |
433 | $error_escaped = db_escape_string($fetch_last_error); | |
434 | } else { | |
435 | _debug("source claims data not modified, nothing to do.", $debug_enabled); | |
436 | } | |
437 | ||
438 | db_query( | |
439 | "UPDATE ttrss_feeds SET last_error = '$error_escaped', | |
440 | last_updated = NOW() WHERE id = '$feed'"); | |
441 | ||
442 | return; | |
443 | } | |
444 | ||
445 | foreach ($pluginhost->get_hooks(PluginHost::HOOK_FEED_FETCHED) as $plugin) { | |
446 | $feed_data = $plugin->hook_feed_fetched($feed_data, $fetch_url, $owner_uid, $feed); | |
447 | } | |
448 | ||
449 | $rss = new FeedParser($feed_data); | |
450 | $rss->init(); | |
451 | ||
452 | $feed = db_escape_string($feed); | |
453 | ||
454 | if (!$rss->error()) { | |
455 | ||
456 | // We use local pluginhost here because we need to load different per-user feed plugins | |
457 | $pluginhost->run_hooks(PluginHost::HOOK_FEED_PARSED, "hook_feed_parsed", $rss); | |
458 | ||
459 | _debug("language: $feed_language", $debug_enabled); | |
460 | _debug("processing feed data...", $debug_enabled); | |
461 | ||
462 | // db_query("BEGIN"); | |
463 | ||
464 | if (DB_TYPE == "pgsql") { | |
465 | $favicon_interval_qpart = "favicon_last_checked < NOW() - INTERVAL '12 hour'"; | |
466 | } else { | |
467 | $favicon_interval_qpart = "favicon_last_checked < DATE_SUB(NOW(), INTERVAL 12 HOUR)"; | |
468 | } | |
469 | ||
470 | $result = db_query("SELECT owner_uid,favicon_avg_color, | |
471 | (favicon_last_checked IS NULL OR $favicon_interval_qpart) AS | |
472 | favicon_needs_check | |
473 | FROM ttrss_feeds WHERE id = '$feed'"); | |
474 | ||
475 | $favicon_needs_check = sql_bool_to_bool(db_fetch_result($result, 0, | |
476 | "favicon_needs_check")); | |
477 | $favicon_avg_color = db_fetch_result($result, 0, "favicon_avg_color"); | |
478 | ||
479 | $owner_uid = db_fetch_result($result, 0, "owner_uid"); | |
480 | ||
481 | $site_url = db_escape_string(mb_substr(rewrite_relative_url($fetch_url, $rss->get_link()), 0, 245)); | |
482 | ||
483 | _debug("site_url: $site_url", $debug_enabled); | |
484 | _debug("feed_title: " . $rss->get_title(), $debug_enabled); | |
485 | ||
486 | if ($favicon_needs_check || $force_refetch) { | |
487 | ||
488 | /* terrible hack: if we crash on floicon shit here, we won't check | |
489 | * the icon avgcolor again (unless the icon got updated) */ | |
490 | ||
491 | $favicon_file = ICONS_DIR . "/$feed.ico"; | |
492 | $favicon_modified = @filemtime($favicon_file); | |
493 | ||
494 | _debug("checking favicon...", $debug_enabled); | |
495 | ||
496 | RSSUtils::check_feed_favicon($site_url, $feed); | |
497 | $favicon_modified_new = @filemtime($favicon_file); | |
498 | ||
499 | if ($favicon_modified_new > $favicon_modified) | |
500 | $favicon_avg_color = ''; | |
501 | ||
502 | if (file_exists($favicon_file) && function_exists("imagecreatefromstring") && $favicon_avg_color == '') { | |
503 | require_once "colors.php"; | |
504 | ||
505 | db_query("UPDATE ttrss_feeds SET favicon_avg_color = 'fail' WHERE | |
506 | id = '$feed'"); | |
507 | ||
508 | $favicon_color = db_escape_string( | |
509 | calculate_avg_color($favicon_file)); | |
510 | ||
511 | $favicon_colorstring = ",favicon_avg_color = '".$favicon_color."'"; | |
512 | } else if ($favicon_avg_color == 'fail') { | |
513 | _debug("floicon failed on this file, not trying to recalculate avg color", $debug_enabled); | |
514 | } | |
515 | ||
516 | db_query("UPDATE ttrss_feeds SET favicon_last_checked = NOW() | |
517 | $favicon_colorstring | |
518 | WHERE id = '$feed'"); | |
519 | } | |
520 | ||
521 | _debug("loading filters & labels...", $debug_enabled); | |
522 | ||
523 | $filters = load_filters($feed, $owner_uid); | |
524 | ||
525 | if ($debug_enabled) { | |
526 | print_r($filters); | |
527 | } | |
528 | ||
529 | _debug("" . count($filters) . " filters loaded.", $debug_enabled); | |
530 | ||
531 | $items = $rss->get_items(); | |
532 | ||
533 | if (!is_array($items)) { | |
534 | _debug("no articles found.", $debug_enabled); | |
535 | ||
536 | db_query("UPDATE ttrss_feeds | |
537 | SET last_updated = NOW(), last_error = '' WHERE id = '$feed'"); | |
538 | ||
539 | return; // no articles | |
540 | } | |
541 | ||
542 | _debug("processing articles...", $debug_enabled); | |
543 | ||
544 | $tstart = time(); | |
545 | ||
546 | foreach ($items as $item) { | |
547 | if ($_REQUEST['xdebug'] == 3) { | |
548 | print_r($item); | |
549 | } | |
550 | ||
551 | if (ini_get("max_execution_time") > 0 && time() - $tstart >= ini_get("max_execution_time") * 0.7) { | |
552 | _debug("looks like there's too many articles to process at once, breaking out", $debug_enabled); | |
553 | break; | |
554 | } | |
555 | ||
556 | $entry_guid = $item->get_id(); | |
557 | if (!$entry_guid) $entry_guid = $item->get_link(); | |
558 | if (!$entry_guid) $entry_guid = RSSUtils::make_guid_from_title($item->get_title()); | |
559 | if (!$entry_guid) continue; | |
560 | ||
561 | $entry_guid = "$owner_uid,$entry_guid"; | |
562 | ||
563 | $entry_guid_hashed = db_escape_string('SHA1:' . sha1($entry_guid)); | |
564 | ||
565 | _debug("guid $entry_guid / $entry_guid_hashed", $debug_enabled); | |
566 | ||
567 | $entry_timestamp = ""; | |
568 | ||
569 | $entry_timestamp = $item->get_date(); | |
570 | ||
571 | _debug("orig date: " . $item->get_date(), $debug_enabled); | |
572 | ||
573 | if ($entry_timestamp == -1 || !$entry_timestamp || $entry_timestamp > time()) { | |
574 | $entry_timestamp = time(); | |
575 | } | |
576 | ||
577 | $entry_timestamp_fmt = strftime("%Y/%m/%d %H:%M:%S", $entry_timestamp); | |
578 | ||
579 | _debug("date $entry_timestamp [$entry_timestamp_fmt]", $debug_enabled); | |
580 | ||
581 | // $entry_title = html_entity_decode($item->get_title(), ENT_COMPAT, 'UTF-8'); | |
582 | // $entry_title = decode_numeric_entities($entry_title); | |
583 | $entry_title = $item->get_title(); | |
584 | ||
585 | $entry_link = rewrite_relative_url($site_url, $item->get_link()); | |
586 | ||
587 | _debug("title $entry_title", $debug_enabled); | |
588 | _debug("link $entry_link", $debug_enabled); | |
589 | ||
590 | if (!$entry_title) $entry_title = date("Y-m-d H:i:s", $entry_timestamp);; | |
591 | ||
592 | $entry_content = $item->get_content(); | |
593 | if (!$entry_content) $entry_content = $item->get_description(); | |
594 | ||
595 | if ($_REQUEST["xdebug"] == 2) { | |
596 | print "content: "; | |
597 | print htmlspecialchars($entry_content); | |
598 | print "\n"; | |
599 | } | |
600 | ||
601 | $entry_comments = db_escape_string(mb_substr($item->get_comments_url(), 0, 245)); | |
602 | $num_comments = (int) $item->get_comments_count(); | |
603 | ||
604 | $entry_author = $item->get_author(); // escaped later | |
605 | $entry_guid = db_escape_string(mb_substr($entry_guid, 0, 245)); | |
606 | ||
607 | _debug("author $entry_author", $debug_enabled); | |
608 | _debug("num_comments: $num_comments", $debug_enabled); | |
609 | _debug("looking for tags...", $debug_enabled); | |
610 | ||
611 | // parse <category> entries into tags | |
612 | ||
613 | $additional_tags = array(); | |
614 | ||
615 | $additional_tags_src = $item->get_categories(); | |
616 | ||
617 | if (is_array($additional_tags_src)) { | |
618 | foreach ($additional_tags_src as $tobj) { | |
619 | array_push($additional_tags, $tobj); | |
620 | } | |
621 | } | |
622 | ||
623 | $entry_tags = array_unique($additional_tags); | |
624 | ||
625 | for ($i = 0; $i < count($entry_tags); $i++) | |
626 | $entry_tags[$i] = mb_strtolower($entry_tags[$i], 'utf-8'); | |
627 | ||
628 | _debug("tags found: " . join(",", $entry_tags), $debug_enabled); | |
629 | ||
630 | _debug("done collecting data.", $debug_enabled); | |
631 | ||
632 | $result = db_query("SELECT id, content_hash, lang FROM ttrss_entries | |
633 | WHERE guid = '".db_escape_string($entry_guid)."' OR guid = '$entry_guid_hashed'"); | |
634 | ||
635 | if (db_num_rows($result) != 0) { | |
636 | $base_entry_id = db_fetch_result($result, 0, "id"); | |
637 | $entry_stored_hash = db_fetch_result($result, 0, "content_hash"); | |
638 | $article_labels = Article::get_article_labels($base_entry_id, $owner_uid); | |
639 | $entry_language = db_fetch_result($result, 0, "lang"); | |
640 | ||
641 | $existing_tags = Article::get_article_tags($base_entry_id, $owner_uid); | |
642 | $entry_tags = array_unique(array_merge($entry_tags, $existing_tags)); | |
643 | ||
644 | } else { | |
645 | $base_entry_id = false; | |
646 | $entry_stored_hash = ""; | |
647 | $article_labels = array(); | |
648 | $entry_language = ""; | |
649 | } | |
650 | ||
651 | $article = array("owner_uid" => $owner_uid, // read only | |
652 | "guid" => $entry_guid, // read only | |
653 | "guid_hashed" => $entry_guid_hashed, // read only | |
654 | "title" => $entry_title, | |
655 | "content" => $entry_content, | |
656 | "link" => $entry_link, | |
657 | "labels" => $article_labels, // current limitation: can add labels to article, can't remove them | |
658 | "tags" => $entry_tags, | |
659 | "author" => $entry_author, | |
660 | "force_catchup" => false, // ugly hack for the time being | |
661 | "score_modifier" => 0, // no previous value, plugin should recalculate score modifier based on content if needed | |
662 | "language" => $entry_language, | |
663 | "num_comments" => $num_comments, // read only | |
664 | "feed" => array("id" => $feed, | |
665 | "fetch_url" => $fetch_url, | |
666 | "site_url" => $site_url, | |
667 | "cache_images" => $cache_images) | |
668 | ); | |
669 | ||
670 | $entry_plugin_data = ""; | |
671 | $entry_current_hash = RSSUtils::calculate_article_hash($article, $pluginhost); | |
672 | ||
673 | _debug("article hash: $entry_current_hash [stored=$entry_stored_hash]", $debug_enabled); | |
674 | ||
675 | if ($entry_current_hash == $entry_stored_hash && !isset($_REQUEST["force_rehash"])) { | |
676 | _debug("stored article seems up to date [IID: $base_entry_id], updating timestamp only", $debug_enabled); | |
677 | ||
678 | // we keep encountering the entry in feeds, so we need to | |
679 | // update date_updated column so that we don't get horrible | |
680 | // dupes when the entry gets purged and reinserted again e.g. | |
681 | // in the case of SLOW SLOW OMG SLOW updating feeds | |
682 | ||
683 | $base_entry_id = db_fetch_result($result, 0, "id"); | |
684 | ||
685 | db_query("UPDATE ttrss_entries SET date_updated = NOW() | |
686 | WHERE id = '$base_entry_id'"); | |
687 | ||
688 | continue; | |
689 | } | |
690 | ||
691 | _debug("hash differs, applying plugin filters:", $debug_enabled); | |
692 | ||
693 | foreach ($pluginhost->get_hooks(PluginHost::HOOK_ARTICLE_FILTER) as $plugin) { | |
694 | _debug("... " . get_class($plugin), $debug_enabled); | |
695 | ||
696 | $start = microtime(true); | |
697 | $article = $plugin->hook_article_filter($article); | |
698 | ||
699 | _debug("=== " . sprintf("%.4f (sec)", microtime(true) - $start), $debug_enabled); | |
700 | ||
701 | $entry_plugin_data .= mb_strtolower(get_class($plugin)) . ","; | |
702 | } | |
703 | ||
704 | if ($_REQUEST["xdebug"] == 2) { | |
705 | print "processed content: "; | |
706 | print htmlspecialchars($article["content"]); | |
707 | print "\n"; | |
708 | } | |
709 | ||
710 | $entry_plugin_data = db_escape_string($entry_plugin_data); | |
711 | ||
712 | _debug("plugin data: $entry_plugin_data", $debug_enabled); | |
713 | ||
714 | // Workaround: 4-byte unicode requires utf8mb4 in MySQL. See https://tt-rss.org/forum/viewtopic.php?f=1&t=3377&p=20077#p20077 | |
715 | if (DB_TYPE == "mysql") { | |
716 | foreach ($article as $k => $v) { | |
717 | ||
718 | // i guess we'll have to take the risk of 4byte unicode labels & tags here | |
719 | if (is_string($article[$k])) { | |
720 | $article[$k] = preg_replace('/[\x{10000}-\x{10FFFF}]/u', "\xEF\xBF\xBD", $v); | |
721 | } | |
722 | } | |
723 | } | |
724 | ||
725 | /* Collect article tags here so we could filter by them: */ | |
726 | ||
727 | $matched_rules = array(); | |
728 | ||
729 | $article_filters = RSSUtils::get_article_filters($filters, $article["title"], | |
730 | $article["content"], $article["link"], $article["author"], | |
731 | $article["tags"], $matched_rules); | |
732 | ||
733 | if ($debug_enabled) { | |
734 | _debug("matched filter rules: ", $debug_enabled); | |
735 | ||
736 | if (count($matched_rules) != 0) { | |
737 | print_r($matched_rules); | |
738 | } | |
739 | ||
740 | _debug("filter actions: ", $debug_enabled); | |
741 | ||
742 | if (count($article_filters) != 0) { | |
743 | print_r($article_filters); | |
744 | } | |
745 | } | |
746 | ||
747 | $plugin_filter_names = RSSUtils::find_article_filters($article_filters, "plugin"); | |
748 | $plugin_filter_actions = $pluginhost->get_filter_actions(); | |
749 | ||
750 | if (count($plugin_filter_names) > 0) { | |
751 | _debug("applying plugin filter actions...", $debug_enabled); | |
752 | ||
753 | foreach ($plugin_filter_names as $pfn) { | |
754 | list($pfclass,$pfaction) = explode(":", $pfn["param"]); | |
755 | ||
756 | if (isset($plugin_filter_actions[$pfclass])) { | |
757 | $plugin = $pluginhost->get_plugin($pfclass); | |
758 | ||
759 | _debug("... $pfclass: $pfaction", $debug_enabled); | |
760 | ||
761 | if ($plugin) { | |
762 | $start = microtime(true); | |
763 | $article = $plugin->hook_article_filter_action($article, $pfaction); | |
764 | ||
765 | _debug("=== " . sprintf("%.4f (sec)", microtime(true) - $start), $debug_enabled); | |
766 | } else { | |
767 | _debug("??? $pfclass: plugin object not found."); | |
768 | } | |
769 | } else { | |
770 | _debug("??? $pfclass: filter plugin not registered."); | |
771 | } | |
772 | } | |
773 | } | |
774 | ||
775 | $entry_tags = $article["tags"]; | |
776 | $entry_guid = db_escape_string($entry_guid); | |
777 | $entry_title = db_escape_string($article["title"]); | |
778 | $entry_author = db_escape_string(mb_substr($article["author"], 0, 245)); | |
779 | $entry_link = db_escape_string($article["link"]); | |
780 | $entry_content = $article["content"]; // escaped below | |
781 | $entry_force_catchup = $article["force_catchup"]; | |
782 | $article_labels = $article["labels"]; | |
783 | $entry_score_modifier = (int) $article["score_modifier"]; | |
784 | $entry_language = db_escape_string($article["language"]); | |
785 | ||
786 | if ($debug_enabled) { | |
787 | _debug("article labels:", $debug_enabled); | |
788 | ||
789 | if (count($article_labels) != 0) { | |
790 | print_r($article_labels); | |
791 | } | |
792 | } | |
793 | ||
794 | _debug("force catchup: $entry_force_catchup"); | |
795 | ||
796 | if ($cache_images && is_writable(CACHE_DIR . '/images')) | |
797 | RSSUtils::cache_media($entry_content, $site_url, $debug_enabled); | |
798 | ||
799 | $entry_content = db_escape_string($entry_content, false); | |
800 | ||
801 | //db_query("BEGIN"); | |
802 | ||
803 | $result = db_query("SELECT id FROM ttrss_entries | |
804 | WHERE (guid = '$entry_guid' OR guid = '$entry_guid_hashed')"); | |
805 | ||
806 | if (db_num_rows($result) == 0) { | |
807 | ||
808 | _debug("base guid [$entry_guid or $entry_guid_hashed] not found, creating...", $debug_enabled); | |
809 | ||
810 | // base post entry does not exist, create it | |
811 | ||
812 | db_query( | |
813 | "INSERT INTO ttrss_entries | |
814 | (title, | |
815 | guid, | |
816 | link, | |
817 | updated, | |
818 | content, | |
819 | content_hash, | |
820 | no_orig_date, | |
821 | date_updated, | |
822 | date_entered, | |
823 | comments, | |
824 | num_comments, | |
825 | plugin_data, | |
826 | lang, | |
827 | author) | |
828 | VALUES | |
829 | ('$entry_title', | |
830 | '$entry_guid_hashed', | |
831 | '$entry_link', | |
832 | '$entry_timestamp_fmt', | |
833 | '$entry_content', | |
834 | '$entry_current_hash', | |
835 | false, | |
836 | NOW(), | |
837 | '$date_feed_processed', | |
838 | '$entry_comments', | |
839 | '$num_comments', | |
840 | '$entry_plugin_data', | |
841 | '$entry_language', | |
842 | '$entry_author')"); | |
843 | ||
844 | } | |
845 | ||
846 | // now it should exist, if not - bad luck then | |
847 | ||
848 | $result = db_query("SELECT id FROM ttrss_entries | |
849 | WHERE guid = '$entry_guid' OR guid = '$entry_guid_hashed'"); | |
850 | ||
851 | $entry_ref_id = 0; | |
852 | $entry_int_id = 0; | |
853 | ||
854 | if (db_num_rows($result) == 1) { | |
855 | ||
856 | _debug("base guid found, checking for user record", $debug_enabled); | |
857 | ||
858 | $ref_id = db_fetch_result($result, 0, "id"); | |
859 | $entry_ref_id = $ref_id; | |
860 | ||
861 | /* $stored_guid = db_fetch_result($result, 0, "guid"); | |
862 | if ($stored_guid != $entry_guid_hashed) { | |
863 | if ($debug_enabled) _debug("upgrading compat guid to hashed one", $debug_enabled); | |
864 | ||
865 | db_query("UPDATE ttrss_entries SET guid = '$entry_guid_hashed' WHERE | |
866 | id = '$ref_id'"); | |
867 | } */ | |
868 | ||
869 | if (RSSUtils::find_article_filter($article_filters, "filter")) { | |
870 | //db_query("COMMIT"); // close transaction in progress | |
871 | continue; | |
872 | } | |
873 | ||
874 | $score = RSSUtils::calculate_article_score($article_filters) + $entry_score_modifier; | |
875 | ||
876 | _debug("initial score: $score [including plugin modifier: $entry_score_modifier]", $debug_enabled); | |
877 | ||
878 | // check for user post link to main table | |
879 | ||
880 | $query = "SELECT ref_id, int_id FROM ttrss_user_entries WHERE | |
881 | ref_id = '$ref_id' AND owner_uid = '$owner_uid'"; | |
882 | ||
883 | // if ($_REQUEST["xdebug"]) print "$query\n"; | |
884 | ||
885 | $result = db_query($query); | |
886 | ||
887 | // okay it doesn't exist - create user entry | |
888 | if (db_num_rows($result) == 0) { | |
889 | ||
890 | _debug("user record not found, creating...", $debug_enabled); | |
891 | ||
892 | if ($score >= -500 && !RSSUtils::find_article_filter($article_filters, 'catchup') && !$entry_force_catchup) { | |
893 | $unread = 'true'; | |
894 | $last_read_qpart = 'NULL'; | |
895 | } else { | |
896 | $unread = 'false'; | |
897 | $last_read_qpart = 'NOW()'; | |
898 | } | |
899 | ||
900 | if (RSSUtils::find_article_filter($article_filters, 'mark') || $score > 1000) { | |
901 | $marked = 'true'; | |
902 | } else { | |
903 | $marked = 'false'; | |
904 | } | |
905 | ||
906 | if (RSSUtils::find_article_filter($article_filters, 'publish')) { | |
907 | $published = 'true'; | |
908 | } else { | |
909 | $published = 'false'; | |
910 | } | |
911 | ||
912 | $last_marked = ($marked == 'true') ? 'NOW()' : 'NULL'; | |
913 | $last_published = ($published == 'true') ? 'NOW()' : 'NULL'; | |
914 | ||
915 | $result = db_query( | |
916 | "INSERT INTO ttrss_user_entries | |
917 | (ref_id, owner_uid, feed_id, unread, last_read, marked, | |
918 | published, score, tag_cache, label_cache, uuid, | |
919 | last_marked, last_published) | |
920 | VALUES ('$ref_id', '$owner_uid', '$feed', $unread, | |
921 | $last_read_qpart, $marked, $published, '$score', '', '', | |
922 | '', $last_marked, $last_published)"); | |
923 | ||
924 | $result = db_query( | |
925 | "SELECT int_id FROM ttrss_user_entries WHERE | |
926 | ref_id = '$ref_id' AND owner_uid = '$owner_uid' AND | |
927 | feed_id = '$feed' LIMIT 1"); | |
928 | ||
929 | if (db_num_rows($result) == 1) { | |
930 | $entry_int_id = db_fetch_result($result, 0, "int_id"); | |
931 | } | |
932 | } else { | |
933 | _debug("user record FOUND", $debug_enabled); | |
934 | ||
935 | $entry_ref_id = db_fetch_result($result, 0, "ref_id"); | |
936 | $entry_int_id = db_fetch_result($result, 0, "int_id"); | |
937 | } | |
938 | ||
939 | _debug("RID: $entry_ref_id, IID: $entry_int_id", $debug_enabled); | |
940 | ||
941 | if (DB_TYPE == "pgsql") { | |
942 | $tsvector_combined = db_escape_string(mb_substr($entry_title . ' ' . strip_tags(str_replace('<', ' <', $entry_content)), | |
943 | 0, 1000000)); | |
944 | ||
945 | $tsvector_qpart = "tsvector_combined = to_tsvector('$feed_language', '$tsvector_combined'),"; | |
946 | ||
947 | } else { | |
948 | $tsvector_qpart = ""; | |
949 | } | |
950 | ||
951 | db_query("UPDATE ttrss_entries | |
952 | SET title = '$entry_title', | |
953 | content = '$entry_content', | |
954 | content_hash = '$entry_current_hash', | |
955 | updated = '$entry_timestamp_fmt', | |
956 | $tsvector_qpart | |
957 | num_comments = '$num_comments', | |
958 | plugin_data = '$entry_plugin_data', | |
959 | author = '$entry_author', | |
960 | lang = '$entry_language' | |
961 | WHERE id = '$ref_id'"); | |
962 | ||
963 | // update aux data | |
964 | db_query("UPDATE ttrss_user_entries | |
965 | SET score = '$score' WHERE ref_id = '$ref_id'"); | |
966 | ||
967 | if ($mark_unread_on_update) { | |
968 | _debug("article updated, marking unread as requested.", $debug_enabled); | |
969 | ||
970 | db_query("UPDATE ttrss_user_entries | |
971 | SET last_read = null, unread = true WHERE ref_id = '$ref_id'"); | |
972 | } | |
973 | } | |
974 | ||
975 | //db_query("COMMIT"); | |
976 | ||
977 | _debug("assigning labels [other]...", $debug_enabled); | |
978 | ||
979 | foreach ($article_labels as $label) { | |
980 | Labels::add_article($entry_ref_id, $label[1], $owner_uid); | |
981 | } | |
982 | ||
983 | _debug("assigning labels [filters]...", $debug_enabled); | |
984 | ||
985 | RSSUtils::assign_article_to_label_filters($entry_ref_id, $article_filters, | |
986 | $owner_uid, $article_labels); | |
987 | ||
988 | _debug("looking for enclosures...", $debug_enabled); | |
989 | ||
990 | // enclosures | |
991 | ||
992 | $enclosures = array(); | |
993 | ||
994 | $encs = $item->get_enclosures(); | |
995 | ||
996 | if (is_array($encs)) { | |
997 | foreach ($encs as $e) { | |
998 | $e_item = array( | |
999 | rewrite_relative_url($site_url, $e->link), | |
1000 | $e->type, $e->length, $e->title, $e->width, $e->height); | |
1001 | array_push($enclosures, $e_item); | |
1002 | } | |
1003 | } | |
1004 | ||
1005 | if ($cache_images && is_writable(CACHE_DIR . '/images')) | |
1006 | RSSUtils::cache_enclosures($enclosures, $site_url, $debug_enabled); | |
1007 | ||
1008 | if ($debug_enabled) { | |
1009 | _debug("article enclosures:", $debug_enabled); | |
1010 | print_r($enclosures); | |
1011 | } | |
1012 | ||
1013 | //db_query("BEGIN"); | |
1014 | ||
1015 | // debugging | |
1016 | // db_query("DELETE FROM ttrss_enclosures WHERE post_id = '$entry_ref_id'"); | |
1017 | ||
1018 | foreach ($enclosures as $enc) { | |
1019 | $enc_url = db_escape_string($enc[0]); | |
1020 | $enc_type = db_escape_string($enc[1]); | |
1021 | $enc_dur = db_escape_string($enc[2]); | |
1022 | $enc_title = db_escape_string($enc[3]); | |
1023 | $enc_width = intval($enc[4]); | |
1024 | $enc_height = intval($enc[5]); | |
1025 | ||
1026 | $result = db_query("SELECT id FROM ttrss_enclosures | |
1027 | WHERE content_url = '$enc_url' AND post_id = '$entry_ref_id'"); | |
1028 | ||
1029 | if (db_num_rows($result) == 0) { | |
1030 | db_query("INSERT INTO ttrss_enclosures | |
1031 | (content_url, content_type, title, duration, post_id, width, height) VALUES | |
1032 | ('$enc_url', '$enc_type', '$enc_title', '$enc_dur', '$entry_ref_id', $enc_width, $enc_height)"); | |
1033 | } | |
1034 | } | |
1035 | ||
1036 | //db_query("COMMIT"); | |
1037 | ||
1038 | // check for manual tags (we have to do it here since they're loaded from filters) | |
1039 | ||
1040 | foreach ($article_filters as $f) { | |
1041 | if ($f["type"] == "tag") { | |
1042 | ||
1043 | $manual_tags = trim_array(explode(",", $f["param"])); | |
1044 | ||
1045 | foreach ($manual_tags as $tag) { | |
1046 | if (tag_is_valid($tag)) { | |
1047 | array_push($entry_tags, $tag); | |
1048 | } | |
1049 | } | |
1050 | } | |
1051 | } | |
1052 | ||
1053 | // Skip boring tags | |
1054 | ||
1055 | $boring_tags = trim_array(explode(",", mb_strtolower(get_pref( | |
1056 | 'BLACKLISTED_TAGS', $owner_uid, ''), 'utf-8'))); | |
1057 | ||
1058 | $filtered_tags = array(); | |
1059 | $tags_to_cache = array(); | |
1060 | ||
1061 | if ($entry_tags && is_array($entry_tags)) { | |
1062 | foreach ($entry_tags as $tag) { | |
1063 | if (array_search($tag, $boring_tags) === false) { | |
1064 | array_push($filtered_tags, $tag); | |
1065 | } | |
1066 | } | |
1067 | } | |
1068 | ||
1069 | $filtered_tags = array_unique($filtered_tags); | |
1070 | ||
1071 | if ($debug_enabled) { | |
1072 | _debug("filtered article tags:", $debug_enabled); | |
1073 | print_r($filtered_tags); | |
1074 | } | |
1075 | ||
1076 | // Save article tags in the database | |
1077 | ||
1078 | if (count($filtered_tags) > 0) { | |
1079 | ||
1080 | //db_query("BEGIN"); | |
1081 | ||
1082 | foreach ($filtered_tags as $tag) { | |
1083 | ||
1084 | $tag = sanitize_tag($tag); | |
1085 | $tag = db_escape_string($tag); | |
1086 | ||
1087 | if (!tag_is_valid($tag)) continue; | |
1088 | ||
1089 | $result = db_query("SELECT id FROM ttrss_tags | |
1090 | WHERE tag_name = '$tag' AND post_int_id = '$entry_int_id' AND | |
1091 | owner_uid = '$owner_uid' LIMIT 1"); | |
1092 | ||
1093 | if ($result && db_num_rows($result) == 0) { | |
1094 | ||
1095 | db_query("INSERT INTO ttrss_tags | |
1096 | (owner_uid,tag_name,post_int_id) | |
1097 | VALUES ('$owner_uid','$tag', '$entry_int_id')"); | |
1098 | } | |
1099 | ||
1100 | array_push($tags_to_cache, $tag); | |
1101 | } | |
1102 | ||
1103 | /* update the cache */ | |
1104 | ||
1105 | $tags_to_cache = array_unique($tags_to_cache); | |
1106 | ||
1107 | $tags_str = db_escape_string(join(",", $tags_to_cache)); | |
1108 | ||
1109 | db_query("UPDATE ttrss_user_entries | |
1110 | SET tag_cache = '$tags_str' WHERE ref_id = '$entry_ref_id' | |
1111 | AND owner_uid = $owner_uid"); | |
1112 | ||
1113 | //db_query("COMMIT"); | |
1114 | } | |
1115 | ||
1116 | _debug("article processed", $debug_enabled); | |
1117 | } | |
1118 | ||
1119 | _debug("purging feed...", $debug_enabled); | |
1120 | ||
1121 | purge_feed($feed, 0, $debug_enabled); | |
1122 | ||
1123 | db_query("UPDATE ttrss_feeds | |
1124 | SET last_updated = NOW(), last_error = '' WHERE id = '$feed'"); | |
1125 | ||
1126 | // db_query("COMMIT"); | |
1127 | ||
1128 | } else { | |
1129 | ||
1130 | $error_msg = db_escape_string(mb_substr($rss->error(), 0, 245)); | |
1131 | ||
1132 | _debug("fetch error: $error_msg", $debug_enabled); | |
1133 | ||
1134 | if (count($rss->errors()) > 1) { | |
1135 | foreach ($rss->errors() as $error) { | |
1136 | _debug("+ $error"); | |
1137 | } | |
1138 | } | |
1139 | ||
1140 | db_query( | |
1141 | "UPDATE ttrss_feeds SET last_error = '$error_msg', | |
1142 | last_updated = NOW() WHERE id = '$feed'"); | |
1143 | ||
1144 | unset($rss); | |
1145 | return; | |
1146 | } | |
1147 | ||
1148 | _debug("done", $debug_enabled); | |
1149 | ||
1150 | return true; | |
1151 | } | |
1152 | ||
1153 | static function cache_enclosures($enclosures, $site_url, $debug) { | |
1154 | foreach ($enclosures as $enc) { | |
1155 | ||
1156 | if (preg_match("/(image|audio|video)/", $enc[1])) { | |
1157 | ||
1158 | $src = rewrite_relative_url($site_url, $enc[0]); | |
1159 | ||
1160 | $local_filename = CACHE_DIR . "/images/" . sha1($src); | |
1161 | ||
1162 | if ($debug) _debug("cache_enclosures: downloading: $src to $local_filename"); | |
1163 | ||
1164 | if (!file_exists($local_filename)) { | |
1165 | $file_content = fetch_file_contents($src); | |
1166 | ||
1167 | if ($file_content && strlen($file_content) > MIN_CACHE_FILE_SIZE) { | |
1168 | file_put_contents($local_filename, $file_content); | |
1169 | } | |
1170 | } else { | |
1171 | touch($local_filename); | |
1172 | } | |
1173 | } | |
1174 | } | |
1175 | } | |
1176 | ||
1177 | static function cache_media($html, $site_url, $debug) { | |
1178 | libxml_use_internal_errors(true); | |
1179 | ||
1180 | $charset_hack = '<head> | |
1181 | <meta http-equiv="Content-Type" content="text/html; charset=utf-8"/> | |
1182 | </head>'; | |
1183 | ||
1184 | $doc = new DOMDocument(); | |
1185 | $doc->loadHTML($charset_hack . $html); | |
1186 | $xpath = new DOMXPath($doc); | |
1187 | ||
1188 | $entries = $xpath->query('(//img[@src])|(//video/source[@src])|(//audio/source[@src])'); | |
1189 | ||
1190 | foreach ($entries as $entry) { | |
1191 | if ($entry->hasAttribute('src') && strpos($entry->getAttribute('src'), "data:") !== 0) { | |
1192 | $src = rewrite_relative_url($site_url, $entry->getAttribute('src')); | |
1193 | ||
1194 | $local_filename = CACHE_DIR . "/images/" . sha1($src); | |
1195 | ||
1196 | if ($debug) _debug("cache_media: downloading: $src to $local_filename"); | |
1197 | ||
1198 | if (!file_exists($local_filename)) { | |
1199 | $file_content = fetch_file_contents($src); | |
1200 | ||
1201 | if ($file_content && strlen($file_content) > MIN_CACHE_FILE_SIZE) { | |
1202 | file_put_contents($local_filename, $file_content); | |
1203 | } | |
1204 | } else { | |
1205 | touch($local_filename); | |
1206 | } | |
1207 | } | |
1208 | } | |
1209 | } | |
1210 | ||
1211 | static function expire_error_log($debug) { | |
1212 | if ($debug) _debug("Removing old error log entries..."); | |
1213 | ||
1214 | if (DB_TYPE == "pgsql") { | |
1215 | db_query("DELETE FROM ttrss_error_log | |
1216 | WHERE created_at < NOW() - INTERVAL '7 days'"); | |
1217 | } else { | |
1218 | db_query("DELETE FROM ttrss_error_log | |
1219 | WHERE created_at < DATE_SUB(NOW(), INTERVAL 7 DAY)"); | |
1220 | } | |
1221 | ||
1222 | } | |
1223 | ||
1224 | static function expire_lock_files($debug) { | |
1225 | //if ($debug) _debug("Removing old lock files..."); | |
1226 | ||
1227 | $num_deleted = 0; | |
1228 | ||
1229 | if (is_writable(LOCK_DIRECTORY)) { | |
1230 | $files = glob(LOCK_DIRECTORY . "/*.lock"); | |
1231 | ||
1232 | if ($files) { | |
1233 | foreach ($files as $file) { | |
1234 | if (!file_is_locked(basename($file)) && time() - filemtime($file) > 86400*2) { | |
1235 | unlink($file); | |
1236 | ++$num_deleted; | |
1237 | } | |
1238 | } | |
1239 | } | |
1240 | } | |
1241 | ||
1242 | if ($debug) _debug("Removed $num_deleted old lock files."); | |
1243 | } | |
1244 | ||
1245 | static function expire_cached_files($debug) { | |
1246 | foreach (array("simplepie", "images", "export", "upload") as $dir) { | |
1247 | $cache_dir = CACHE_DIR . "/$dir"; | |
1248 | ||
1249 | // if ($debug) _debug("Expiring $cache_dir"); | |
1250 | ||
1251 | $num_deleted = 0; | |
1252 | ||
1253 | if (is_writable($cache_dir)) { | |
1254 | $files = glob("$cache_dir/*"); | |
1255 | ||
1256 | if ($files) { | |
1257 | foreach ($files as $file) { | |
1258 | if (time() - filemtime($file) > 86400*CACHE_MAX_DAYS) { | |
1259 | unlink($file); | |
1260 | ||
1261 | ++$num_deleted; | |
1262 | } | |
1263 | } | |
1264 | } | |
1265 | } | |
1266 | ||
1267 | if ($debug) _debug("$cache_dir: removed $num_deleted files."); | |
1268 | } | |
1269 | } | |
1270 | ||
1271 | /** | |
1272 | * Source: http://www.php.net/manual/en/function.parse-url.php#104527 | |
1273 | * Returns the url query as associative array | |
1274 | * | |
1275 | * @param string query | |
1276 | * @return array params | |
1277 | */ | |
1278 | static function convertUrlQuery($query) { | |
1279 | $queryParts = explode('&', $query); | |
1280 | ||
1281 | $params = array(); | |
1282 | ||
1283 | foreach ($queryParts as $param) { | |
1284 | $item = explode('=', $param); | |
1285 | $params[$item[0]] = $item[1]; | |
1286 | } | |
1287 | ||
1288 | return $params; | |
1289 | } | |
1290 | ||
1291 | static function get_article_filters($filters, $title, $content, $link, $author, $tags, &$matched_rules = false) { | |
1292 | $matches = array(); | |
1293 | ||
1294 | foreach ($filters as $filter) { | |
1295 | $match_any_rule = $filter["match_any_rule"]; | |
1296 | $inverse = $filter["inverse"]; | |
1297 | $filter_match = false; | |
1298 | ||
1299 | foreach ($filter["rules"] as $rule) { | |
1300 | $match = false; | |
1301 | $reg_exp = str_replace('/', '\/', $rule["reg_exp"]); | |
1302 | $rule_inverse = $rule["inverse"]; | |
1303 | ||
1304 | if (!$reg_exp) | |
1305 | continue; | |
1306 | ||
1307 | switch ($rule["type"]) { | |
1308 | case "title": | |
1309 | $match = @preg_match("/$reg_exp/iu", $title); | |
1310 | break; | |
1311 | case "content": | |
1312 | // we don't need to deal with multiline regexps | |
1313 | $content = preg_replace("/[\r\n\t]/", "", $content); | |
1314 | ||
1315 | $match = @preg_match("/$reg_exp/iu", $content); | |
1316 | break; | |
1317 | case "both": | |
1318 | // we don't need to deal with multiline regexps | |
1319 | $content = preg_replace("/[\r\n\t]/", "", $content); | |
1320 | ||
1321 | $match = (@preg_match("/$reg_exp/iu", $title) || @preg_match("/$reg_exp/iu", $content)); | |
1322 | break; | |
1323 | case "link": | |
1324 | $match = @preg_match("/$reg_exp/iu", $link); | |
1325 | break; | |
1326 | case "author": | |
1327 | $match = @preg_match("/$reg_exp/iu", $author); | |
1328 | break; | |
1329 | case "tag": | |
1330 | foreach ($tags as $tag) { | |
1331 | if (@preg_match("/$reg_exp/iu", $tag)) { | |
1332 | $match = true; | |
1333 | break; | |
1334 | } | |
1335 | } | |
1336 | break; | |
1337 | } | |
1338 | ||
1339 | if ($rule_inverse) $match = !$match; | |
1340 | ||
1341 | if ($match_any_rule) { | |
1342 | if ($match) { | |
1343 | $filter_match = true; | |
1344 | break; | |
1345 | } | |
1346 | } else { | |
1347 | $filter_match = $match; | |
1348 | if (!$match) { | |
1349 | break; | |
1350 | } | |
1351 | } | |
1352 | } | |
1353 | ||
1354 | if ($inverse) $filter_match = !$filter_match; | |
1355 | ||
1356 | if ($filter_match) { | |
1357 | if (is_array($matched_rules)) array_push($matched_rules, $rule); | |
1358 | ||
1359 | foreach ($filter["actions"] AS $action) { | |
1360 | array_push($matches, $action); | |
1361 | ||
1362 | // if Stop action encountered, perform no further processing | |
1363 | if (isset($action["type"]) && $action["type"] == "stop") return $matches; | |
1364 | } | |
1365 | } | |
1366 | } | |
1367 | ||
1368 | return $matches; | |
1369 | } | |
1370 | ||
1371 | static function find_article_filter($filters, $filter_name) { | |
1372 | foreach ($filters as $f) { | |
1373 | if ($f["type"] == $filter_name) { | |
1374 | return $f; | |
1375 | }; | |
1376 | } | |
1377 | return false; | |
1378 | } | |
1379 | ||
1380 | static function find_article_filters($filters, $filter_name) { | |
1381 | $results = array(); | |
1382 | ||
1383 | foreach ($filters as $f) { | |
1384 | if ($f["type"] == $filter_name) { | |
1385 | array_push($results, $f); | |
1386 | }; | |
1387 | } | |
1388 | return $results; | |
1389 | } | |
1390 | ||
1391 | static function calculate_article_score($filters) { | |
1392 | $score = 0; | |
1393 | ||
1394 | foreach ($filters as $f) { | |
1395 | if ($f["type"] == "score") { | |
1396 | $score += $f["param"]; | |
1397 | }; | |
1398 | } | |
1399 | return $score; | |
1400 | } | |
1401 | ||
1402 | static function labels_contains_caption($labels, $caption) { | |
1403 | foreach ($labels as $label) { | |
1404 | if ($label[1] == $caption) { | |
1405 | return true; | |
1406 | } | |
1407 | } | |
1408 | ||
1409 | return false; | |
1410 | } | |
1411 | ||
1412 | static function assign_article_to_label_filters($id, $filters, $owner_uid, $article_labels) { | |
1413 | foreach ($filters as $f) { | |
1414 | if ($f["type"] == "label") { | |
1415 | if (!RSSUtils::labels_contains_caption($article_labels, $f["param"])) { | |
1416 | Labels::add_article($id, $f["param"], $owner_uid); | |
1417 | } | |
1418 | } | |
1419 | } | |
1420 | } | |
1421 | ||
1422 | static function make_guid_from_title($title) { | |
1423 | return preg_replace("/[ \"\',.:;]/", "-", | |
1424 | mb_strtolower(strip_tags($title), 'utf-8')); | |
1425 | } | |
1426 | ||
1427 | static function cleanup_counters_cache($debug) { | |
1428 | $result = db_query("DELETE FROM ttrss_counters_cache | |
1429 | WHERE feed_id > 0 AND | |
1430 | (SELECT COUNT(id) FROM ttrss_feeds WHERE | |
1431 | id = feed_id AND | |
1432 | ttrss_counters_cache.owner_uid = ttrss_feeds.owner_uid) = 0"); | |
1433 | $frows = db_affected_rows($result); | |
1434 | ||
1435 | $result = db_query("DELETE FROM ttrss_cat_counters_cache | |
1436 | WHERE feed_id > 0 AND | |
1437 | (SELECT COUNT(id) FROM ttrss_feed_categories WHERE | |
1438 | id = feed_id AND | |
1439 | ttrss_cat_counters_cache.owner_uid = ttrss_feed_categories.owner_uid) = 0"); | |
1440 | $crows = db_affected_rows($result); | |
1441 | ||
1442 | if ($debug) _debug("Removed $frows (feeds) $crows (cats) orphaned counter cache entries."); | |
1443 | } | |
1444 | ||
1445 | static function housekeeping_user($owner_uid) { | |
1446 | $tmph = new PluginHost(); | |
1447 | ||
1448 | load_user_plugins($owner_uid, $tmph); | |
1449 | ||
1450 | $tmph->run_hooks(PluginHost::HOOK_HOUSE_KEEPING, "hook_house_keeping", ""); | |
1451 | } | |
1452 | ||
1453 | static function housekeeping_common($debug) { | |
1454 | RSSUtils::expire_cached_files($debug); | |
1455 | RSSUtils::expire_lock_files($debug); | |
1456 | RSSUtils::expire_error_log($debug); | |
1457 | ||
1458 | $count = RSSUtils::update_feedbrowser_cache(); | |
1459 | _debug("Feedbrowser updated, $count feeds processed."); | |
1460 | ||
1461 | Article::purge_orphans( true); | |
1462 | RSSUtils::cleanup_counters_cache($debug); | |
1463 | ||
1464 | //$rc = cleanup_tags( 14, 50000); | |
1465 | //_debug("Cleaned $rc cached tags."); | |
1466 | ||
1467 | PluginHost::getInstance()->run_hooks(PluginHost::HOOK_HOUSE_KEEPING, "hook_house_keeping", ""); | |
1468 | } | |
1469 | ||
1470 | static function check_feed_favicon($site_url, $feed) { | |
1471 | # print "FAVICON [$site_url]: $favicon_url\n"; | |
1472 | ||
1473 | $icon_file = ICONS_DIR . "/$feed.ico"; | |
1474 | ||
1475 | if (!file_exists($icon_file)) { | |
1476 | $favicon_url = get_favicon_url($site_url); | |
1477 | ||
1478 | if ($favicon_url) { | |
1479 | // Limiting to "image" type misses those served with text/plain | |
1480 | $contents = fetch_file_contents($favicon_url); // , "image"); | |
1481 | ||
1482 | if ($contents) { | |
1483 | // Crude image type matching. | |
1484 | // Patterns gleaned from the file(1) source code. | |
1485 | if (preg_match('/^\x00\x00\x01\x00/', $contents)) { | |
1486 | // 0 string \000\000\001\000 MS Windows icon resource | |
1487 | //error_log("check_feed_favicon: favicon_url=$favicon_url isa MS Windows icon resource"); | |
1488 | } | |
1489 | elseif (preg_match('/^GIF8/', $contents)) { | |
1490 | // 0 string GIF8 GIF image data | |
1491 | //error_log("check_feed_favicon: favicon_url=$favicon_url isa GIF image"); | |
1492 | } | |
1493 | elseif (preg_match('/^\x89PNG\x0d\x0a\x1a\x0a/', $contents)) { | |
1494 | // 0 string \x89PNG\x0d\x0a\x1a\x0a PNG image data | |
1495 | //error_log("check_feed_favicon: favicon_url=$favicon_url isa PNG image"); | |
1496 | } | |
1497 | elseif (preg_match('/^\xff\xd8/', $contents)) { | |
1498 | // 0 beshort 0xffd8 JPEG image data | |
1499 | //error_log("check_feed_favicon: favicon_url=$favicon_url isa JPG image"); | |
1500 | } | |
1501 | else { | |
1502 | //error_log("check_feed_favicon: favicon_url=$favicon_url isa UNKNOWN type"); | |
1503 | $contents = ""; | |
1504 | } | |
1505 | } | |
1506 | ||
1507 | if ($contents) { | |
1508 | $fp = @fopen($icon_file, "w"); | |
1509 | ||
1510 | if ($fp) { | |
1511 | fwrite($fp, $contents); | |
1512 | fclose($fp); | |
1513 | chmod($icon_file, 0644); | |
1514 | } | |
1515 | } | |
1516 | } | |
1517 | return $icon_file; | |
1518 | } | |
1519 | } | |
1520 | ||
1521 | ||
1522 | ||
1523 | } |