define_default('DAEMON_UPDATE_LOGIN_LIMIT', 30);
define_default('DAEMON_FEED_LIMIT', 500);
define_default('DAEMON_SLEEP_INTERVAL', 120);
- define_default('_MIN_CACHE_IMAGE_SIZE', 1024);
+ define_default('_MIN_CACHE_FILE_SIZE', 1024);
function calculate_article_hash($article, $pluginhost) {
$tmp = "";
foreach ($article as $k => $v) {
if ($k != "feed" && isset($v)) {
- $tmp .= sha1("$k:" . (is_array($v) ? implode(",", $v) : $v));
+ $x = strip_tags(is_array($v) ? implode(",", $v) : $v);
+
+ //_debug("$k:" . sha1($x) . ":" . htmlspecialchars($x), true);
+
+ $tmp .= sha1("$k:" . sha1($x));
}
}
function update_feedbrowser_cache() {
$result = db_query("SELECT feed_url, site_url, title, COUNT(id) AS subscribers
- FROM ttrss_feeds WHERE (SELECT COUNT(id) = 0 FROM ttrss_feeds AS tf
- WHERE tf.feed_url = ttrss_feeds.feed_url
- AND (private IS true OR auth_login != '' OR auth_pass != '' OR feed_url LIKE '%:%@%/%'))
+ FROM ttrss_feeds WHERE feed_url NOT IN (SELECT feed_url FROM ttrss_feeds
+ WHERE private IS true OR auth_login != '' OR auth_pass != '' OR feed_url LIKE '%:%@%/%')
GROUP BY feed_url, site_url, title ORDER BY subscribers DESC LIMIT 1000");
db_query("BEGIN");
$query_limit = "";
if($limit) $query_limit = sprintf("LIMIT %d", $limit);
+ // Update the least recently updated feeds first
+ $query_order = "ORDER BY last_updated";
+ if (DB_TYPE == "pgsql") $query_order .= " NULLS FIRST";
+
$query = "SELECT DISTINCT ttrss_feeds.feed_url, ttrss_feeds.last_updated
FROM
ttrss_feeds, ttrss_users, ttrss_user_prefs
AND ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL'
$login_thresh_qpart $update_limit_qpart
$updstart_thresh_qpart
- ORDER BY last_updated $query_limit";
+ $query_order $query_limit";
// We search for feed needing update.
$result = db_query($query);
$nf = 0;
$bstarted = microtime(true);
+ $batch_owners = array();
+
// For each feed, we call the feed update function.
foreach ($feeds_to_update as $feed) {
if($debug) _debug("Base feed: $feed");
ttrss_users.id = ttrss_user_prefs.owner_uid AND
ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL' AND
ttrss_user_prefs.profile IS NULL AND
- feed_url = '".db_escape_string($feed)."' AND
- (ttrss_feeds.update_interval > 0 OR
- ttrss_user_prefs.value != '-1')
+ feed_url = '".db_escape_string($feed)."'
+ $update_limit_qpart
$login_thresh_qpart
ORDER BY ttrss_feeds.id $query_limit");
while ($tline = db_fetch_assoc($tmp_result)) {
if($debug) _debug(" => " . $tline["last_updated"] . ", " . $tline["id"] . " " . $tline["owner_uid"]);
+ if (array_search($tline["owner_uid"], $batch_owners) === FALSE)
+ array_push($batch_owners, $tline["owner_uid"]);
+
$fstarted = microtime(true);
$rss = update_rss_feed($tline["id"], true, false);
_debug_suppress(false);
microtime(true) - $bstarted, (microtime(true) - $bstarted) / $nf));
}
+ foreach ($batch_owners as $owner_uid) {
+ _debug("Running housekeeping tasks for user $owner_uid...");
+
+ housekeeping_user($owner_uid);
+ }
+
require_once "digest.php";
// Send feed digests by email if needed.
$feed = db_escape_string($feed);
- $result = db_query("SELECT feed_url,auth_pass,auth_pass_encrypted
+ $result = db_query("SELECT feed_url,auth_pass,auth_login,auth_pass_encrypted
FROM ttrss_feeds WHERE id = '$feed'");
$auth_pass_encrypted = sql_bool_to_bool(db_fetch_result($result,
$auth_login = db_fetch_result($result, 0, "auth_login");
$auth_pass = db_fetch_result($result, 0, "auth_pass");
- if ($auth_pass_encrypted) {
+ if ($auth_pass_encrypted && function_exists("mcrypt_decrypt")) {
require_once "crypt.php";
$auth_pass = decrypt_string($auth_pass);
}
$feed_data = fetch_file_contents($fetch_url, false,
$auth_login, $auth_pass, false,
- FEED_FETCH_TIMEOUT_TIMEOUT,
+ FEED_FETCH_TIMEOUT,
0);
global $fetch_curl_used;
$result = db_query("SELECT title FROM ttrss_feeds
WHERE id = '$feed'");
+
+ if (db_num_rows($result) == 0) {
+ _debug("feed $feed NOT FOUND/SKIPPED", $debug_enabled);
+ user_error("Attempt to update unknown/invalid feed $feed", E_USER_WARNING);
+ return false;
+ }
+
$title = db_fetch_result($result, 0, "title");
// feed was batch-subscribed or something, we need to get basic info
feed_url,auth_pass,cache_images,
mark_unread_on_update, owner_uid,
pubsub_state, auth_pass_encrypted,
- (SELECT max(date_entered) FROM
- ttrss_entries, ttrss_user_entries where ref_id = id AND feed_id = '$feed') AS last_article_timestamp
+ feed_language
FROM ttrss_feeds WHERE id = '$feed'");
- if (db_num_rows($result) == 0) {
- _debug("feed $feed NOT FOUND/SKIPPED", $debug_enabled);
- return false;
- }
-
- $last_article_timestamp = @strtotime(db_fetch_result($result, 0, "last_article_timestamp"));
-
- if (defined('_DISABLE_HTTP_304'))
- $last_article_timestamp = 0;
-
$owner_uid = db_fetch_result($result, 0, "owner_uid");
$mark_unread_on_update = sql_bool_to_bool(db_fetch_result($result,
0, "mark_unread_on_update"));
$auth_login = db_fetch_result($result, 0, "auth_login");
$auth_pass = db_fetch_result($result, 0, "auth_pass");
- if ($auth_pass_encrypted) {
+ if ($auth_pass_encrypted && function_exists("mcrypt_decrypt")) {
require_once "crypt.php";
$auth_pass = decrypt_string($auth_pass);
}
$cache_images = sql_bool_to_bool(db_fetch_result($result, 0, "cache_images"));
$fetch_url = db_fetch_result($result, 0, "feed_url");
+ $feed_language = db_escape_string(mb_strtolower(db_fetch_result($result, 0, "feed_language")));
+ if (!$feed_language) $feed_language = 'english';
$feed = db_escape_string($feed);
$rss_hash = false;
$force_refetch = isset($_REQUEST["force_refetch"]);
+ $feed_data = "";
foreach ($pluginhost->get_hooks(PluginHost::HOOK_FETCH_FEED) as $plugin) {
- $feed_data = $plugin->hook_fetch_feed($feed_data, $fetch_url, $owner_uid, $feed, $last_article_timestamp, $auth_login, $auth_pass);
+ $feed_data = $plugin->hook_fetch_feed($feed_data, $fetch_url, $owner_uid, $feed, 0, $auth_login, $auth_pass);
}
// try cache
// fetch feed from source
if (!$feed_data) {
_debug("fetching [$fetch_url]...", $debug_enabled);
- _debug("If-Modified-Since: ".gmdate('D, d M Y H:i:s \G\M\T', $last_article_timestamp), $debug_enabled);
+
+ if (ini_get("open_basedir") && function_exists("curl_init")) {
+ _debug("not using CURL due to open_basedir restrictions");
+ }
$feed_data = fetch_file_contents($fetch_url, false,
$auth_login, $auth_pass, false,
$no_cache ? FEED_FETCH_NO_CACHE_TIMEOUT : FEED_FETCH_TIMEOUT,
- $force_refetch ? 0 : $last_article_timestamp);
+ 0);
global $fetch_curl_used;
// We use local pluginhost here because we need to load different per-user feed plugins
$pluginhost->run_hooks(PluginHost::HOOK_FEED_PARSED, "hook_feed_parsed", $rss);
+ _debug("language: $feed_language", $debug_enabled);
_debug("processing feed data...", $debug_enabled);
// db_query("BEGIN");
if ($feed_hub_url && $feed_self_url && function_exists('curl_init') &&
!ini_get("open_basedir")) {
- require_once 'lib/pubsubhubbub/subscriber.php';
+ require_once 'lib/pubsubhubbub/Subscriber.php';
$callback_url = get_self_url_prefix() .
"/public.php?op=pubsub&id=$feed";
- $s = new Subscriber($feed_hub_url, $callback_url);
+ $s = new Pubsubhubbub\Subscriber\Subscriber($feed_hub_url, $callback_url);
$rc = $s->subscribe($feed_self_url);
if ($_REQUEST["xdebug"] == 2) {
print "content: ";
- print $entry_content;
+ print htmlspecialchars($entry_content);
print "\n";
}
- $entry_comments = $item->get_comments_url();
- $entry_author = $item->get_author();
+ $entry_comments = db_escape_string(mb_substr($item->get_comments_url(), 0, 245));
+ $num_comments = (int) $item->get_comments_count();
+ $entry_author = $item->get_author(); // escaped later
$entry_guid = db_escape_string(mb_substr($entry_guid, 0, 245));
- $entry_comments = db_escape_string(mb_substr(trim($entry_comments), 0, 245));
- $entry_author = db_escape_string(mb_substr(trim($entry_author), 0, 245));
-
- $num_comments = (int) $item->get_comments_count();
-
_debug("author $entry_author", $debug_enabled);
_debug("num_comments: $num_comments", $debug_enabled);
_debug("looking for tags...", $debug_enabled);
$entry_stored_hash = db_fetch_result($result, 0, "content_hash");
$article_labels = get_article_labels($base_entry_id, $owner_uid);
$entry_language = db_fetch_result($result, 0, "lang");
+
+ $existing_tags = get_article_tags($base_entry_id, $owner_uid);
+ $entry_tags = array_unique(array_merge($entry_tags, $existing_tags));
+
} else {
$base_entry_id = false;
$entry_stored_hash = "";
"language" => $entry_language,
"feed" => array("id" => $feed,
"fetch_url" => $fetch_url,
- "site_url" => $site_url)
+ "site_url" => $site_url,
+ "cache_images" => $cache_images)
);
$entry_plugin_data = "";
db_query("UPDATE ttrss_entries SET date_updated = NOW()
WHERE id = '$base_entry_id'");
- // if we allow duplicate posts, we have to continue to
- // create the user entries for this feed
- if (!get_pref("ALLOW_DUPLICATE_POSTS", $owner_uid, false)) {
- continue;
- }
+ continue;
}
_debug("hash differs, applying plugin filters:", $debug_enabled);
$entry_plugin_data .= mb_strtolower(get_class($plugin)) . ",";
}
+ if ($_REQUEST["xdebug"] == 2) {
+ print "processed content: ";
+ print htmlspecialchars($article["content"]);
+ print "\n";
+ }
+
$entry_plugin_data = db_escape_string($entry_plugin_data);
_debug("plugin data: $entry_plugin_data", $debug_enabled);
foreach ($article as $k => $v) {
// i guess we'll have to take the risk of 4byte unicode labels & tags here
- if (!is_array($article[$k])) {
+ if (is_string($article[$k])) {
$article[$k] = preg_replace('/[\x{10000}-\x{10FFFF}]/u', "\xEF\xBF\xBD", $v);
}
}
}
+ /* Collect article tags here so we could filter by them: */
+
+ $matched_rules = array();
+
+ $article_filters = get_article_filters($filters, $article["title"],
+ $article["content"], $article["link"], 0, $article["author"],
+ $article["tags"], $matched_rules);
+
+ if ($debug_enabled) {
+ _debug("matched filter rules: ", $debug_enabled);
+
+ if (count($matched_rules) != 0) {
+ print_r($matched_rules);
+ }
+
+ _debug("filter actions: ", $debug_enabled);
+
+ if (count($article_filters) != 0) {
+ print_r($article_filters);
+ }
+ }
+
+ $plugin_filter_names = find_article_filters($article_filters, "plugin");
+ $plugin_filter_actions = $pluginhost->get_filter_actions();
+
+ if (count($plugin_filter_names) > 0) {
+ _debug("applying plugin filter actions...", $debug_enabled);
+
+ foreach ($plugin_filter_names as $pfn) {
+ list($pfclass,$pfaction) = explode(":", $pfn["param"]);
+
+ if (isset($plugin_filter_actions[$pfclass])) {
+ $plugin = $pluginhost->get_plugin($pfclass);
+
+ _debug("... $pfclass: $pfaction", $debug_enabled);
+
+ if ($plugin) {
+ $start = microtime(true);
+ $article = $plugin->hook_article_filter_action($article, $pfaction);
+
+ _debug("=== " . sprintf("%.4f (sec)", microtime(true) - $start), $debug_enabled);
+ } else {
+ _debug("??? $pfclass: plugin object not found.");
+ }
+ } else {
+ _debug("??? $pfclass: filter plugin not registered.");
+ }
+ }
+ }
+
$entry_tags = $article["tags"];
$entry_guid = db_escape_string($entry_guid);
$entry_title = db_escape_string($article["title"]);
- $entry_author = db_escape_string($article["author"]);
+ $entry_author = db_escape_string(mb_substr($article["author"], 0, 245));
$entry_link = db_escape_string($article["link"]);
$entry_content = $article["content"]; // escaped below
$entry_force_catchup = $article["force_catchup"];
if ($debug_enabled) {
_debug("article labels:", $debug_enabled);
- print_r($article_labels);
+
+ if (count($article_labels) != 0) {
+ print_r($article_labels);
+ }
}
_debug("force catchup: $entry_force_catchup");
if ($cache_images && is_writable(CACHE_DIR . '/images'))
- cache_images($entry_content, $site_url, $debug_enabled);
+ cache_media($entry_content, $site_url, $debug_enabled);
$entry_content = db_escape_string($entry_content, false);
- db_query("BEGIN");
+ //db_query("BEGIN");
$result = db_query("SELECT id FROM ttrss_entries
WHERE (guid = '$entry_guid' OR guid = '$entry_guid_hashed')");
id = '$ref_id'");
} */
- // check for user post link to main table
-
- // do we allow duplicate posts with same GUID in different feeds?
- if (get_pref("ALLOW_DUPLICATE_POSTS", $owner_uid, false)) {
- $dupcheck_qpart = "AND (feed_id = '$feed' OR feed_id IS NULL)";
- } else {
- $dupcheck_qpart = "";
- }
-
- /* Collect article tags here so we could filter by them: */
-
- $article_filters = get_article_filters($filters, $entry_title,
- $entry_content, $entry_link, $entry_timestamp, $entry_author,
- $entry_tags);
-
- if ($debug_enabled) {
- _debug("article filters: ", $debug_enabled);
- if (count($article_filters) != 0) {
- print_r($article_filters);
- }
- }
-
if (find_article_filter($article_filters, "filter")) {
- db_query("COMMIT"); // close transaction in progress
+ //db_query("COMMIT"); // close transaction in progress
continue;
}
_debug("initial score: $score [including plugin modifier: $entry_score_modifier]", $debug_enabled);
+ // check for user post link to main table
+
$query = "SELECT ref_id, int_id FROM ttrss_user_entries WHERE
- ref_id = '$ref_id' AND owner_uid = '$owner_uid'
- $dupcheck_qpart";
+ ref_id = '$ref_id' AND owner_uid = '$owner_uid'";
// if ($_REQUEST["xdebug"]) print "$query\n";
$published = 'false';
}
- // N-grams
-
- /* if (DB_TYPE == "pgsql" and defined('_NGRAM_TITLE_DUPLICATE_THRESHOLD')) {
-
- $result = db_query("SELECT COUNT(*) AS similar FROM
- ttrss_entries,ttrss_user_entries
- WHERE ref_id = id AND updated >= NOW() - INTERVAL '7 day'
- AND similarity(title, '$entry_title') >= "._NGRAM_TITLE_DUPLICATE_THRESHOLD."
- AND owner_uid = $owner_uid");
-
- $ngram_similar = db_fetch_result($result, 0, "similar");
-
- _debug("N-gram similar results: $ngram_similar", $debug_enabled);
-
- if ($ngram_similar > 0) {
- $unread = 'false';
- }
- } */
-
$last_marked = ($marked == 'true') ? 'NOW()' : 'NULL';
$last_published = ($published == 'true') ? 'NOW()' : 'NULL';
"/public.php?op=rss&id=-2&key=" .
get_feed_access_key(-2, false, $owner_uid);
- $p = new Publisher(PUBSUBHUBBUB_HUB);
+ $p = new pubsubhubbub\publisher\Publisher(PUBSUBHUBBUB_HUB);
/* $pubsub_result = */ $p->publish_update($rss_link);
}
_debug("RID: $entry_ref_id, IID: $entry_int_id", $debug_enabled);
+ if (DB_TYPE == "pgsql") {
+ $tsvector_combined = db_escape_string(mb_substr($entry_title . ' ' . strip_tags(str_replace('<', ' <', $entry_content)),
+ 0, 1000000));
+
+ $tsvector_qpart = "tsvector_combined = to_tsvector('$feed_language', '$tsvector_combined'),";
+
+ } else {
+ $tsvector_qpart = "";
+ }
+
db_query("UPDATE ttrss_entries
SET title = '$entry_title',
content = '$entry_content',
content_hash = '$entry_current_hash',
updated = '$entry_timestamp_fmt',
+ $tsvector_qpart
num_comments = '$num_comments',
plugin_data = '$entry_plugin_data',
author = '$entry_author',
SET score = '$score' WHERE ref_id = '$ref_id'");
if ($mark_unread_on_update) {
+ _debug("article updated, marking unread as requested.", $debug_enabled);
+
db_query("UPDATE ttrss_user_entries
SET last_read = null, unread = true WHERE ref_id = '$ref_id'");
}
}
- db_query("COMMIT");
+ //db_query("COMMIT");
_debug("assigning labels [other]...", $debug_enabled);
}
}
+ if ($cache_images && is_writable(CACHE_DIR . '/images'))
+ cache_enclosures($enclosures, $site_url, $debug_enabled);
+
if ($debug_enabled) {
_debug("article enclosures:", $debug_enabled);
print_r($enclosures);
}
- db_query("BEGIN");
+ //db_query("BEGIN");
// debugging
// db_query("DELETE FROM ttrss_enclosures WHERE post_id = '$entry_ref_id'");
}
}
- db_query("COMMIT");
+ //db_query("COMMIT");
// check for manual tags (we have to do it here since they're loaded from filters)
if (count($filtered_tags) > 0) {
- db_query("BEGIN");
+ //db_query("BEGIN");
foreach ($filtered_tags as $tag) {
SET tag_cache = '$tags_str' WHERE ref_id = '$entry_ref_id'
AND owner_uid = $owner_uid");
- db_query("COMMIT");
+ //db_query("COMMIT");
}
_debug("article processed", $debug_enabled);
return $rss;
}
- function cache_images($html, $site_url, $debug) {
+ function cache_enclosures($enclosures, $site_url, $debug) {
+ foreach ($enclosures as $enc) {
+
+ if (preg_match("/(image|audio|video)/", $enc[1])) {
+
+ $src = rewrite_relative_url($site_url, $enc[0]);
+
+ $local_filename = CACHE_DIR . "/images/" . sha1($src);
+
+ if ($debug) _debug("cache_enclosures: downloading: $src to $local_filename");
+
+ if (!file_exists($local_filename)) {
+ $file_content = fetch_file_contents($src);
+
+ if ($file_content && strlen($file_content) > _MIN_CACHE_FILE_SIZE) {
+ file_put_contents($local_filename, $file_content);
+ }
+ } else {
+ touch($local_filename);
+ }
+ }
+ }
+ }
+
+ function cache_media($html, $site_url, $debug) {
libxml_use_internal_errors(true);
$charset_hack = '<head>
$doc->loadHTML($charset_hack . $html);
$xpath = new DOMXPath($doc);
- $entries = $xpath->query('(//img[@src])');
+ $entries = $xpath->query('(//img[@src])|(//video/source[@src])|(//audio/source[@src])');
foreach ($entries as $entry) {
- if ($entry->hasAttribute('src')) {
+ if ($entry->hasAttribute('src') && strpos($entry->getAttribute('src'), "data:") !== 0) {
$src = rewrite_relative_url($site_url, $entry->getAttribute('src'));
- $local_filename = CACHE_DIR . "/images/" . sha1($src) . ".png";
+ $local_filename = CACHE_DIR . "/images/" . sha1($src);
- if ($debug) _debug("cache_images: downloading: $src to $local_filename");
+ if ($debug) _debug("cache_media: downloading: $src to $local_filename");
if (!file_exists($local_filename)) {
$file_content = fetch_file_contents($src);
- if ($file_content && strlen($file_content) > _MIN_CACHE_IMAGE_SIZE) {
+ if ($file_content && strlen($file_content) > _MIN_CACHE_FILE_SIZE) {
file_put_contents($local_filename, $file_content);
}
+ } else {
+ touch($local_filename);
}
}
}
return $params;
}
- function get_article_filters($filters, $title, $content, $link, $timestamp, $author, $tags) {
+ function get_article_filters($filters, $title, $content, $link, $timestamp, $author, $tags, &$matched_rules = false) {
$matches = array();
foreach ($filters as $filter) {
switch ($rule["type"]) {
case "title":
- $match = @preg_match("/$reg_exp/i", $title);
+ $match = @preg_match("/$reg_exp/iu", $title);
break;
case "content":
// we don't need to deal with multiline regexps
$content = preg_replace("/[\r\n\t]/", "", $content);
- $match = @preg_match("/$reg_exp/i", $content);
+ $match = @preg_match("/$reg_exp/iu", $content);
break;
case "both":
// we don't need to deal with multiline regexps
$content = preg_replace("/[\r\n\t]/", "", $content);
- $match = (@preg_match("/$reg_exp/i", $title) || @preg_match("/$reg_exp/i", $content));
+ $match = (@preg_match("/$reg_exp/iu", $title) || @preg_match("/$reg_exp/iu", $content));
break;
case "link":
- $match = @preg_match("/$reg_exp/i", $link);
+ $match = @preg_match("/$reg_exp/iu", $link);
break;
case "author":
- $match = @preg_match("/$reg_exp/i", $author);
+ $match = @preg_match("/$reg_exp/iu", $author);
break;
case "tag":
foreach ($tags as $tag) {
- if (@preg_match("/$reg_exp/i", $tag)) {
+ if (@preg_match("/$reg_exp/iu", $tag)) {
$match = true;
break;
}
if ($inverse) $filter_match = !$filter_match;
if ($filter_match) {
+ if (is_array($matched_rules)) array_push($matched_rules, $rule);
+
foreach ($filter["actions"] AS $action) {
array_push($matches, $action);
// if Stop action encountered, perform no further processing
- if ($action["type"] == "stop") return $matches;
+ if (isset($action["type"]) && $action["type"] == "stop") return $matches;
}
}
}
_debug("Removed $frows (feeds) $crows (cats) orphaned counter cache entries.");
}
+ function housekeeping_user($owner_uid) {
+ $tmph = new PluginHost();
+
+ load_user_plugins($owner_uid, $tmph);
+
+ $tmph->run_hooks(PluginHost::HOOK_HOUSE_KEEPING, "hook_house_keeping", "");
+ }
+
function housekeeping_common($debug) {
expire_cached_files($debug);
expire_lock_files($debug);
purge_orphans( true);
cleanup_counters_cache($debug);
- $rc = cleanup_tags( 14, 50000);
- _debug("Cleaned $rc cached tags.");
+ //$rc = cleanup_tags( 14, 50000);
+ //_debug("Cleaned $rc cached tags.");
PluginHost::getInstance()->run_hooks(PluginHost::HOOK_HOUSE_KEEPING, "hook_house_keeping", "");
-
}
?>
git.wh0rd.org / tt-rss.git / blobdiff