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);
+
+ 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));
+ }
+ }
+
+ return sha1(implode(",", $pluginhost->get_plugin_names()) . $tmp);
+ }
function update_feedbrowser_cache() {
$login_thresh_qpart = "";
}
- // Test if the feed need a update (update interval exceded).
+ // Test if the feed need a update (update interval exceeded).
if (DB_TYPE == "pgsql") {
$update_limit_qpart = "AND ((
ttrss_feeds.update_interval = 0
) OR (
ttrss_feeds.update_interval > 0
AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_feeds.update_interval || ' minutes') AS INTERVAL)
- ) OR ttrss_feeds.last_updated IS NULL
- OR last_updated = '1970-01-01 00:00:00')";
+ ) OR (ttrss_feeds.last_updated IS NULL
+ AND ttrss_user_prefs.value != '-1')
+ OR (last_updated = '1970-01-01 00:00:00'
+ AND ttrss_user_prefs.value != '-1'))";
} else {
$update_limit_qpart = "AND ((
ttrss_feeds.update_interval = 0
) OR (
ttrss_feeds.update_interval > 0
AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL ttrss_feeds.update_interval MINUTE)
- ) OR ttrss_feeds.last_updated IS NULL
- OR last_updated = '1970-01-01 00:00:00')";
+ ) OR (ttrss_feeds.last_updated IS NULL
+ AND ttrss_user_prefs.value != '-1')
+ OR (last_updated = '1970-01-01 00:00:00'
+ AND ttrss_user_prefs.value != '-1'))";
}
// Test if feed is currently being updated by another process.
ttrss_feeds, ttrss_users, ttrss_user_prefs
WHERE
ttrss_feeds.owner_uid = ttrss_users.id
+ AND ttrss_user_prefs.profile IS NULL
AND ttrss_users.id = ttrss_user_prefs.owner_uid
AND ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL'
$login_thresh_qpart $update_limit_qpart
}
$nf = 0;
+ $bstarted = microtime(true);
// For each feed, we call the feed update function.
foreach ($feeds_to_update as $feed) {
ttrss_user_prefs.owner_uid = ttrss_feeds.owner_uid AND
ttrss_users.id = ttrss_user_prefs.owner_uid AND
ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL' AND
+ ttrss_user_prefs.profile IS NULL AND
feed_url = '".db_escape_string($feed)."' AND
(ttrss_feeds.update_interval > 0 OR
ttrss_user_prefs.value != '-1')
ORDER BY ttrss_feeds.id $query_limit");
if (db_num_rows($tmp_result) > 0) {
+ $rss = false;
+
while ($tline = db_fetch_assoc($tmp_result)) {
if($debug) _debug(" => " . $tline["last_updated"] . ", " . $tline["id"] . " " . $tline["owner_uid"]);
- update_rss_feed($tline["id"], true);
+
+ $fstarted = microtime(true);
+ $rss = update_rss_feed($tline["id"], true, false);
+ _debug_suppress(false);
+
+ _debug(sprintf(" %.4f (sec)", microtime(true) - $fstarted));
+
++$nf;
}
}
}
+ if ($nf > 0) {
+ _debug(sprintf("Processed %d feeds in %.4f (sec), %.4f (sec/feed avg)", $nf,
+ microtime(true) - $bstarted, (microtime(true) - $bstarted) / $nf));
+ }
+
require_once "digest.php";
// Send feed digests by email if needed.
} // function update_daemon_common
+ // this is used when subscribing
+ function set_basic_feed_info($feed) {
+
+ $feed = db_escape_string($feed);
+
+ $result = db_query("SELECT feed_url,auth_pass,auth_login,auth_pass_encrypted
+ FROM ttrss_feeds WHERE id = '$feed'");
+
+ $auth_pass_encrypted = sql_bool_to_bool(db_fetch_result($result,
+ 0, "auth_pass_encrypted"));
+
+ $auth_login = db_fetch_result($result, 0, "auth_login");
+ $auth_pass = db_fetch_result($result, 0, "auth_pass");
+
+ if ($auth_pass_encrypted) {
+ require_once "crypt.php";
+ $auth_pass = decrypt_string($auth_pass);
+ }
+
+ $fetch_url = db_fetch_result($result, 0, "feed_url");
+
+ $feed_data = fetch_file_contents($fetch_url, false,
+ $auth_login, $auth_pass, false,
+ FEED_FETCH_TIMEOUT_TIMEOUT,
+ 0);
+
+ global $fetch_curl_used;
+
+ if (!$fetch_curl_used) {
+ $tmp = @gzdecode($feed_data);
+
+ if ($tmp) $feed_data = $tmp;
+ }
+
+ $feed_data = trim($feed_data);
+
+ $rss = new FeedParser($feed_data);
+ $rss->init();
+
+ if (!$rss->error()) {
+
+ $result = db_query("SELECT title, site_url FROM ttrss_feeds WHERE id = '$feed'");
+
+ $registered_title = db_fetch_result($result, 0, "title");
+ $orig_site_url = db_fetch_result($result, 0, "site_url");
+
+ $site_url = db_escape_string(mb_substr(rewrite_relative_url($fetch_url, $rss->get_link()), 0, 245));
+ $feed_title = db_escape_string(mb_substr($rss->get_title(), 0, 199));
+
+ if ($feed_title && (!$registered_title || $registered_title == "[Unknown]")) {
+ db_query("UPDATE ttrss_feeds SET
+ title = '$feed_title' WHERE id = '$feed'");
+ }
+
+ if ($site_url && $orig_site_url != $site_url) {
+ db_query("UPDATE ttrss_feeds SET
+ site_url = '$site_url' WHERE id = '$feed'");
+ }
+ }
+ }
+
// ignore_daemon is not used
- function update_rss_feed($feed, $ignore_daemon = false, $no_cache = false) {
+ function update_rss_feed($feed, $ignore_daemon = false, $no_cache = false, $rss = false) {
$debug_enabled = defined('DAEMON_EXTENDED_DEBUG') || $_REQUEST['xdebug'];
+ _debug_suppress(!$debug_enabled);
_debug("start", $debug_enabled);
+ $result = db_query("SELECT title FROM ttrss_feeds
+ WHERE id = '$feed'");
+ $title = db_fetch_result($result, 0, "title");
+
+ // feed was batch-subscribed or something, we need to get basic info
+ // this is not optimal currently as it fetches stuff separately TODO: optimize
+ if ($title == "[Unknown]") {
+ _debug("setting basic feed info for $feed...");
+ set_basic_feed_info($feed);
+ }
+
$result = db_query("SELECT id,update_interval,auth_login,
- feed_url,auth_pass,cache_images,last_updated,
+ feed_url,auth_pass,cache_images,
mark_unread_on_update, owner_uid,
pubsub_state, auth_pass_encrypted,
+ feed_language,
(SELECT max(date_entered) FROM
ttrss_entries, ttrss_user_entries where ref_id = id AND feed_id = '$feed') AS last_article_timestamp
FROM ttrss_feeds WHERE id = '$feed'");
return false;
}
- $last_updated = db_fetch_result($result, 0, "last_updated");
$last_article_timestamp = @strtotime(db_fetch_result($result, 0, "last_article_timestamp"));
if (defined('_DISABLE_HTTP_304'))
$cache_images = sql_bool_to_bool(db_fetch_result($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);
$pluginhost->load($user_plugins, PluginHost::KIND_USER, $owner_uid);
$pluginhost->load_data();
- $rss = false;
- $rss_hash = false;
-
- $force_refetch = isset($_REQUEST["force_refetch"]);
+ if ($rss && is_object($rss) && get_class($rss) == "FeedParser") {
+ _debug("using previously initialized parser object");
+ } else {
+ $rss_hash = false;
- if (file_exists($cache_filename) &&
- is_readable($cache_filename) &&
- !$auth_login && !$auth_pass &&
- filemtime($cache_filename) > time() - 30) {
+ $force_refetch = isset($_REQUEST["force_refetch"]);
- _debug("using local cache.", $debug_enabled);
+ foreach ($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 = file_get_contents($cache_filename);
+ // try cache
+ if (!$feed_data &&
+ file_exists($cache_filename) &&
+ is_readable($cache_filename) &&
+ !$auth_login && !$auth_pass &&
+ filemtime($cache_filename) > time() - 30) {
- if ($feed_data) {
- $rss_hash = sha1($feed_data);
- }
+ _debug("using local cache [$cache_filename].", $debug_enabled);
- } else {
- _debug("local cache will not be used for this feed", $debug_enabled);
- }
+ @$feed_data = file_get_contents($cache_filename);
- if (!$rss) {
+ if ($feed_data) {
+ $rss_hash = sha1($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);
+ } else {
+ _debug("local cache will not be used for this feed", $debug_enabled);
}
+ // fetch feed from source
if (!$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);
_debug("fetch done.", $debug_enabled);
- /* if ($feed_data) {
- $error = verify_feed_xml($feed_data);
+ // cache vanilla feed data for re-use
+ if ($feed_data && !$auth_pass && !$auth_login && is_writable(CACHE_DIR . "/simplepie")) {
+ $new_rss_hash = sha1($feed_data);
- if ($error) {
- _debug("error verifying XML, code: " . $error->code, $debug_enabled);
-
- if ($error->code == 26) {
- _debug("got error 26, trying to decode entities...", $debug_enabled);
-
- $feed_data = html_entity_decode($feed_data, ENT_COMPAT, 'UTF-8');
-
- $error = verify_feed_xml($feed_data);
-
- if ($error) $feed_data = '';
- }
+ if ($new_rss_hash != $rss_hash) {
+ _debug("saving $cache_filename", $debug_enabled);
+ @file_put_contents($cache_filename, $feed_data);
}
- } */
+ }
}
if (!$feed_data) {
if (!$rss->error()) {
- // cache data for later
- if (!$auth_pass && !$auth_login && is_writable(CACHE_DIR . "/simplepie")) {
- $new_rss_hash = sha1($rss_data);
-
- if ($new_rss_hash != $rss_hash && count($rss->get_items()) > 0 ) {
- _debug("saving $cache_filename", $debug_enabled);
- @file_put_contents($cache_filename, $feed_data);
- }
- }
-
// 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");
$favicon_interval_qpart = "favicon_last_checked < DATE_SUB(NOW(), INTERVAL 12 HOUR)";
}
- $result = db_query("SELECT title,site_url,owner_uid,favicon_avg_color,
+ $result = db_query("SELECT owner_uid,favicon_avg_color,
(favicon_last_checked IS NULL OR $favicon_interval_qpart) AS
favicon_needs_check
FROM ttrss_feeds WHERE id = '$feed'");
- $registered_title = db_fetch_result($result, 0, "title");
- $orig_site_url = db_fetch_result($result, 0, "site_url");
$favicon_needs_check = sql_bool_to_bool(db_fetch_result($result, 0,
"favicon_needs_check"));
$favicon_avg_color = db_fetch_result($result, 0, "favicon_avg_color");
WHERE id = '$feed'");
}
- if (!$registered_title || $registered_title == "[Unknown]") {
-
- $feed_title = db_escape_string($rss->get_title());
-
- if ($feed_title) {
- _debug("registering title: $feed_title", $debug_enabled);
-
- db_query("UPDATE ttrss_feeds SET
- title = '$feed_title' WHERE id = '$feed'");
- }
- }
-
- if ($site_url && $orig_site_url != $site_url) {
- db_query("UPDATE ttrss_feeds SET
- site_url = '$site_url' WHERE id = '$feed'");
- }
-
_debug("loading filters & labels...", $debug_enabled);
$filters = load_filters($feed, $owner_uid);
- $labels = get_all_labels($owner_uid);
_debug("" . count($filters) . " filters loaded.", $debug_enabled);
_debug("feed hub url: $feed_hub_url", $debug_enabled);
- if ($feed_hub_url && function_exists('curl_init') &&
+ $feed_self_url = $fetch_url;
+
+ $links = $rss->get_links('self');
+
+ if ($links && is_array($links)) {
+ foreach ($links as $l) {
+ $feed_self_url = $l;
+ break;
+ }
+ }
+
+ _debug("feed self url = $feed_self_url");
+
+ if ($feed_hub_url && $feed_self_url && function_exists('curl_init') &&
!ini_get("open_basedir")) {
require_once 'lib/pubsubhubbub/subscriber.php';
$s = new Subscriber($feed_hub_url, $callback_url);
- $rc = $s->subscribe($fetch_url);
+ $rc = $s->subscribe($feed_self_url);
- _debug("feed hub url found, subscribe request sent.", $debug_enabled);
+ _debug("feed hub url found, subscribe request sent. [rc=$rc]", $debug_enabled);
db_query("UPDATE ttrss_feeds SET pubsub_state = 1
WHERE id = '$feed'");
_debug("processing articles...", $debug_enabled);
+ $tstart = time();
+
foreach ($items as $item) {
if ($_REQUEST['xdebug'] == 3) {
print_r($item);
}
+ if (ini_get("max_execution_time") > 0 && time() - $tstart >= ini_get("max_execution_time") * 0.7) {
+ _debug("looks like there's too many articles to process at once, breaking out", $debug_enabled);
+ break;
+ }
+
$entry_guid = $item->get_id();
if (!$entry_guid) $entry_guid = $item->get_link();
if (!$entry_guid) $entry_guid = make_guid_from_title($item->get_title());
-
- _debug("f_guid $entry_guid", $debug_enabled);
-
if (!$entry_guid) continue;
$entry_guid = "$owner_uid,$entry_guid";
if ($entry_timestamp == -1 || !$entry_timestamp || $entry_timestamp > time()) {
$entry_timestamp = time();
- $no_orig_date = 'true';
- } else {
- $no_orig_date = 'false';
}
$entry_timestamp_fmt = strftime("%Y/%m/%d %H:%M:%S", $entry_timestamp);
_debug("done collecting data.", $debug_enabled);
- // TODO: less memory-hungry implementation
-
- _debug("applying plugin filters..", $debug_enabled);
-
- // FIXME not sure if owner_uid is a good idea here, we may have a base entry without user entry (?)
- $result = db_query("SELECT plugin_data,title,content,link,tag_cache,author FROM ttrss_entries, ttrss_user_entries
- WHERE ref_id = id AND (guid = '".db_escape_string($entry_guid)."' OR guid = '$entry_guid_hashed') AND owner_uid = $owner_uid");
+ $result = db_query("SELECT id, content_hash, lang FROM ttrss_entries
+ WHERE guid = '".db_escape_string($entry_guid)."' OR guid = '$entry_guid_hashed'");
if (db_num_rows($result) != 0) {
- $entry_plugin_data = db_fetch_result($result, 0, "plugin_data");
- $stored_article = array("title" => db_fetch_result($result, 0, "title"),
- "content" => db_fetch_result($result, 0, "content"),
- "link" => db_fetch_result($result, 0, "link"),
- "tags" => explode(",", db_fetch_result($result, 0, "tag_cache")),
- "author" => db_fetch_result($result, 0, "author"));
+ $base_entry_id = db_fetch_result($result, 0, "id");
+ $entry_stored_hash = db_fetch_result($result, 0, "content_hash");
+ $article_labels = get_article_labels($base_entry_id, $owner_uid);
+ $entry_language = db_fetch_result($result, 0, "lang");
} else {
- $entry_plugin_data = "";
- $stored_article = array();
+ $base_entry_id = false;
+ $entry_stored_hash = "";
+ $article_labels = array();
+ $entry_language = "";
}
$article = array("owner_uid" => $owner_uid, // read only
"guid" => $entry_guid, // read only
+ "guid_hashed" => $entry_guid_hashed, // read only
"title" => $entry_title,
"content" => $entry_content,
"link" => $entry_link,
+ "labels" => $article_labels, // current limitation: can add labels to article, can't remove them
"tags" => $entry_tags,
- "plugin_data" => $entry_plugin_data,
"author" => $entry_author,
- "stored" => $stored_article);
+ "force_catchup" => false, // ugly hack for the time being
+ "score_modifier" => 0, // no previous value, plugin should recalculate score modifier based on content if needed
+ "language" => $entry_language,
+ "feed" => array("id" => $feed,
+ "fetch_url" => $fetch_url,
+ "site_url" => $site_url)
+ );
+
+ $entry_plugin_data = "";
+ $entry_current_hash = calculate_article_hash($article, $pluginhost);
+
+ _debug("article hash: $entry_current_hash [stored=$entry_stored_hash]", $debug_enabled);
+
+ if ($entry_current_hash == $entry_stored_hash && !isset($_REQUEST["force_rehash"])) {
+ _debug("stored article seems up to date [IID: $base_entry_id], updating timestamp only", $debug_enabled);
+
+ // we keep encountering the entry in feeds, so we need to
+ // update date_updated column so that we don't get horrible
+ // dupes when the entry gets purged and reinserted again e.g.
+ // in the case of SLOW SLOW OMG SLOW updating feeds
+
+ $base_entry_id = db_fetch_result($result, 0, "id");
+
+ 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;
+ }
+ }
+
+ _debug("hash differs, applying plugin filters:", $debug_enabled);
foreach ($pluginhost->get_hooks(PluginHost::HOOK_ARTICLE_FILTER) as $plugin) {
+ _debug("... " . get_class($plugin), $debug_enabled);
+
+ $start = microtime(true);
$article = $plugin->hook_article_filter($article);
+
+ _debug("=== " . sprintf("%.4f (sec)", microtime(true) - $start), $debug_enabled);
+
+ $entry_plugin_data .= mb_strtolower(get_class($plugin)) . ",";
+ }
+
+ $entry_plugin_data = db_escape_string($entry_plugin_data);
+
+ _debug("plugin data: $entry_plugin_data", $debug_enabled);
+
+ // Workaround: 4-byte unicode requires utf8mb4 in MySQL. See https://tt-rss.org/forum/viewtopic.php?f=1&t=3377&p=20077#p20077
+ if (DB_TYPE == "mysql") {
+ foreach ($article as $k => $v) {
+
+ // i guess we'll have to take the risk of 4byte unicode labels & tags here
+ if (!is_array($article[$k])) {
+ $article[$k] = preg_replace('/[\x{10000}-\x{10FFFF}]/u', "\xEF\xBF\xBD", $v);
+ }
+ }
+ }
+
+ /* Collect article tags here so we could filter by them: */
+
+ $article_filters = get_article_filters($filters, $article["title"],
+ $article["content"], $article["link"], 0, $article["author"],
+ $article["tags"]);
+
+ if ($debug_enabled) {
+ _debug("article filters: ", $debug_enabled);
+ if (count($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"];
$entry_title = db_escape_string($article["title"]);
$entry_author = db_escape_string($article["author"]);
$entry_link = db_escape_string($article["link"]);
- $entry_plugin_data = db_escape_string($article["plugin_data"]);
$entry_content = $article["content"]; // escaped below
+ $entry_force_catchup = $article["force_catchup"];
+ $article_labels = $article["labels"];
+ $entry_score_modifier = (int) $article["score_modifier"];
+ $entry_language = db_escape_string($article["language"]);
+ if ($debug_enabled) {
+ _debug("article labels:", $debug_enabled);
+ print_r($article_labels);
+ }
- _debug("plugin data: $entry_plugin_data", $debug_enabled);
+ _debug("force catchup: $entry_force_catchup");
if ($cache_images && is_writable(CACHE_DIR . '/images'))
cache_images($entry_content, $site_url, $debug_enabled);
$entry_content = db_escape_string($entry_content, false);
- $content_hash = "SHA1:" . sha1($entry_content);
-
- db_query("BEGIN");
+ //db_query("BEGIN");
$result = db_query("SELECT id FROM ttrss_entries
WHERE (guid = '$entry_guid' OR guid = '$entry_guid_hashed')");
comments,
num_comments,
plugin_data,
- cached_content,
+ lang,
author)
VALUES
('$entry_title',
'$entry_link',
'$entry_timestamp_fmt',
'$entry_content',
- '$content_hash',
- $no_orig_date,
+ '$entry_current_hash',
+ false,
NOW(),
'$date_feed_processed',
'$entry_comments',
'$num_comments',
'$entry_plugin_data',
- '$entry_guid',
+ '$entry_language',
'$entry_author')");
- $article_labels = array();
-
} else {
- // we keep encountering the entry in feeds, so we need to
- // update date_updated column so that we don't get horrible
- // dupes when the entry gets purged and reinserted again e.g.
- // in the case of SLOW SLOW OMG SLOW updating feeds
-
$base_entry_id = db_fetch_result($result, 0, "id");
-
- db_query("UPDATE ttrss_entries SET date_updated = NOW()
- WHERE id = '$base_entry_id'");
-
- $article_labels = get_article_labels($base_entry_id, $owner_uid);
}
// now it should exist, if not - bad luck then
- $result = db_query("SELECT
- id,content_hash,no_orig_date,title,plugin_data,guid,
- ".SUBSTRING_FOR_DATE."(date_updated,1,19) as date_updated,
- ".SUBSTRING_FOR_DATE."(updated,1,19) as updated,
- num_comments
- FROM
- ttrss_entries
+ $result = db_query("SELECT id FROM ttrss_entries
WHERE guid = '$entry_guid' OR guid = '$entry_guid_hashed'");
$entry_ref_id = 0;
_debug("base guid found, checking for user record", $debug_enabled);
- // this will be used below in update handler
- $orig_content_hash = db_fetch_result($result, 0, "content_hash");
- $orig_title = db_fetch_result($result, 0, "title");
- $orig_num_comments = db_fetch_result($result, 0, "num_comments");
- $orig_date_updated = strtotime(db_fetch_result($result,
- 0, "date_updated"));
- $orig_plugin_data = db_fetch_result($result, 0, "plugin_data");
-
$ref_id = db_fetch_result($result, 0, "id");
$entry_ref_id = $ref_id;
$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;
}
- $score = calculate_article_score($article_filters);
+ $score = calculate_article_score($article_filters) + $entry_score_modifier;
- _debug("initial score: $score", $debug_enabled);
+ _debug("initial score: $score [including plugin modifier: $entry_score_modifier]", $debug_enabled);
$query = "SELECT ref_id, int_id FROM ttrss_user_entries WHERE
ref_id = '$ref_id' AND owner_uid = '$owner_uid'
_debug("user record not found, creating...", $debug_enabled);
- if ($score >= -500 && !find_article_filter($article_filters, 'catchup')) {
+ if ($score >= -500 && !find_article_filter($article_filters, 'catchup') && !$entry_force_catchup) {
$unread = 'true';
$last_read_qpart = 'NULL';
} else {
// N-grams
- if (DB_TYPE == "pgsql" and defined('_NGRAM_TITLE_DUPLICATE_THRESHOLD')) {
+ /* if (DB_TYPE == "pgsql" and defined('_NGRAM_TITLE_DUPLICATE_THRESHOLD')) {
$result = db_query("SELECT COUNT(*) AS similar FROM
ttrss_entries,ttrss_user_entries
if ($ngram_similar > 0) {
$unread = 'false';
}
- }
+ } */
$last_marked = ($marked == 'true') ? 'NOW()' : 'NULL';
$last_published = ($published == 'true') ? 'NOW()' : 'NULL';
$p = new Publisher(PUBSUBHUBBUB_HUB);
- $pubsub_result = $p->publish_update($rss_link);
+ /* $pubsub_result = */ $p->publish_update($rss_link);
}
$result = db_query(
_debug("RID: $entry_ref_id, IID: $entry_int_id", $debug_enabled);
- $post_needs_update = false;
- $update_insignificant = false;
+ if (DB_TYPE == "pgsql") {
+ $tsvector_combined = db_escape_string(mb_substr($entry_title . ' ' . strip_tags(str_replace('<', ' <', $entry_content)),
+ 0, 1000000));
- if ($orig_num_comments != $num_comments) {
- $post_needs_update = true;
- $update_insignificant = true;
- }
+ $tsvector_qpart = "tsvector_combined = to_tsvector('$feed_language', '$tsvector_combined'),";
- if ($entry_plugin_data != $orig_plugin_data) {
- $post_needs_update = true;
- $update_insignificant = true;
+ } else {
+ $tsvector_qpart = "";
}
- if ($content_hash != $orig_content_hash) {
- $post_needs_update = true;
- $update_insignificant = false;
- }
+ db_query("UPDATE ttrss_entries
+ SET title = '$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',
+ lang = '$entry_language'
+ WHERE id = '$ref_id'");
+
+ // update aux data
+ db_query("UPDATE ttrss_user_entries
+ SET score = '$score' WHERE ref_id = '$ref_id'");
- if (db_escape_string($orig_title) != $entry_title) {
- $post_needs_update = true;
- $update_insignificant = false;
+ if ($mark_unread_on_update) {
+ db_query("UPDATE ttrss_user_entries
+ SET last_read = null, unread = true WHERE ref_id = '$ref_id'");
}
+ }
- // if post needs update, update it and mark all user entries
- // linking to this post as updated
- if ($post_needs_update) {
-
- if (defined('DAEMON_EXTENDED_DEBUG')) {
- _debug("post $entry_guid_hashed needs update...", $debug_enabled);
- }
-
-// print "<!-- post $orig_title needs update : $post_needs_update -->";
+ //db_query("COMMIT");
- db_query("UPDATE ttrss_entries
- SET title = '$entry_title', content = '$entry_content',
- content_hash = '$content_hash',
- updated = '$entry_timestamp_fmt',
- num_comments = '$num_comments',
- plugin_data = '$entry_plugin_data'
- WHERE id = '$ref_id'");
+ _debug("assigning labels [other]...", $debug_enabled);
- if (!$update_insignificant) {
- if ($mark_unread_on_update) {
- db_query("UPDATE ttrss_user_entries
- SET last_read = null, unread = true WHERE ref_id = '$ref_id'");
- }
- }
- }
+ foreach ($article_labels as $label) {
+ label_add_article($entry_ref_id, $label[1], $owner_uid);
}
- db_query("COMMIT");
-
- _debug("assigning labels...", $debug_enabled);
+ _debug("assigning labels [filters]...", $debug_enabled);
assign_article_to_label_filters($entry_ref_id, $article_filters,
$owner_uid, $article_labels);
if (is_array($encs)) {
foreach ($encs as $e) {
$e_item = array(
- $e->link, $e->type, $e->length);
+ $e->link, $e->type, $e->length, $e->title, $e->width, $e->height);
array_push($enclosures, $e_item);
}
}
print_r($enclosures);
}
- db_query("BEGIN");
+ //db_query("BEGIN");
+
+// debugging
+// db_query("DELETE FROM ttrss_enclosures WHERE post_id = '$entry_ref_id'");
foreach ($enclosures as $enc) {
$enc_url = db_escape_string($enc[0]);
$enc_type = db_escape_string($enc[1]);
$enc_dur = db_escape_string($enc[2]);
+ $enc_title = db_escape_string($enc[3]);
+ $enc_width = intval($enc[4]);
+ $enc_height = intval($enc[5]);
$result = db_query("SELECT id FROM ttrss_enclosures
WHERE content_url = '$enc_url' AND post_id = '$entry_ref_id'");
if (db_num_rows($result) == 0) {
db_query("INSERT INTO ttrss_enclosures
- (content_url, content_type, title, duration, post_id) VALUES
- ('$enc_url', '$enc_type', '', '$enc_dur', '$entry_ref_id')");
+ (content_url, content_type, title, duration, post_id, width, height) VALUES
+ ('$enc_url', '$enc_type', '$enc_title', '$enc_dur', '$entry_ref_id', $enc_width, $enc_height)");
}
}
- db_query("COMMIT");
+ //db_query("COMMIT");
// check for manual tags (we have to do it here since they're loaded from filters)
if (count($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");
- }
-
- if (get_pref("AUTO_ASSIGN_LABELS", $owner_uid, false)) {
- _debug("auto-assigning labels...", $debug_enabled);
-
- foreach ($labels as $label) {
- $caption = preg_quote($label["caption"]);
-
- if ($caption && preg_match("/\b$caption\b/i", "$tags_str " . strip_tags($entry_content) . " $entry_title")) {
- if (!labels_contains_caption($article_labels, $caption)) {
- label_add_article($entry_ref_id, $caption, $owner_uid);
- }
- }
- }
+ //db_query("COMMIT");
}
_debug("article processed", $debug_enabled);
$error_msg = db_escape_string(mb_substr($rss->error(), 0, 245));
- _debug("error fetching feed: $error_msg", $debug_enabled);
+ _debug("fetch error: $error_msg", $debug_enabled);
+
+ if (count($rss->errors()) > 1) {
+ foreach ($rss->errors() as $error) {
+ _debug("+ $error");
+ }
+ }
db_query(
"UPDATE ttrss_feeds SET last_error = '$error_msg',
- last_updated = NOW() WHERE id = '$feed'");
- }
+ last_updated = NOW() WHERE id = '$feed'");
- unset($rss);
+ unset($rss);
+ }
_debug("done", $debug_enabled);
+
+ return $rss;
}
function cache_images($html, $site_url, $debug) {
- $cache_dir = CACHE_DIR . "/images";
-
libxml_use_internal_errors(true);
$charset_hack = '<head>
if (!file_exists($local_filename)) {
$file_content = fetch_file_contents($src);
- if ($file_content && strlen($file_content) > 1024) {
+ if ($file_content && strlen($file_content) > _MIN_CACHE_IMAGE_SIZE) {
file_put_contents($local_filename, $file_content);
}
- }
-
- if (file_exists($local_filename)) {
- $entry->setAttribute('src', SELF_URL_PATH . '/image.php?url=' .
- base64_encode($src));
+ } else {
+ touch($local_filename);
}
}
}
-
- $node = $doc->getElementsByTagName('body')->item(0);
-
- return $doc->saveXML($node);
}
function expire_error_log($debug) {
array_push($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;
}
}
}
return $error;
} */
+ function cleanup_counters_cache($debug) {
+ $result = db_query("DELETE FROM ttrss_counters_cache
+ WHERE feed_id > 0 AND
+ (SELECT COUNT(id) FROM ttrss_feeds WHERE
+ id = feed_id AND
+ ttrss_counters_cache.owner_uid = ttrss_feeds.owner_uid) = 0");
+ $frows = db_affected_rows($result);
+
+ $result = db_query("DELETE FROM ttrss_cat_counters_cache
+ WHERE feed_id > 0 AND
+ (SELECT COUNT(id) FROM ttrss_feed_categories WHERE
+ id = feed_id AND
+ ttrss_cat_counters_cache.owner_uid = ttrss_feed_categories.owner_uid) = 0");
+ $crows = db_affected_rows($result);
+
+ _debug("Removed $frows (feeds) $crows (cats) orphaned counter cache entries.");
+ }
+
function housekeeping_common($debug) {
expire_cached_files($debug);
expire_lock_files($debug);
_debug("Feedbrowser updated, $count feeds processed.");
purge_orphans( true);
+ cleanup_counters_cache($debug);
$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