X-Git-Url: https://git.wh0rd.org/?a=blobdiff_plain;f=include%2Frssfuncs.php;h=01a11da6b535183096cb04158c82834575942160;hb=8292d05b7c9723bc91ccdf92f6acc5e46f58ce6c;hp=922b0ccb5f7f08a37a2b55a8074ddb3233b90ed2;hpb=5321e775f92bc304456e7b6ed55ad169bf21ba16;p=tt-rss.git diff --git a/include/rssfuncs.php b/include/rssfuncs.php index 922b0ccb..01a11da6 100644 --- a/include/rssfuncs.php +++ b/include/rssfuncs.php @@ -18,10 +18,10 @@ $count = 0; while ($line = db_fetch_assoc($result)) { - $subscribers = db_escape_string($line["subscribers"]); - $feed_url = db_escape_string($line["feed_url"]); - $title = db_escape_string($line["title"]); - $site_url = db_escape_string($line["site_url"]); + $subscribers = db_escape_string($link, $line["subscribers"]); + $feed_url = db_escape_string($link, $line["feed_url"]); + $title = db_escape_string($link, $line["title"]); + $site_url = db_escape_string($link, $line["site_url"]); $tmp_result = db_query($link, "SELECT subscribers FROM ttrss_feedbrowser_cache WHERE feed_url = '$feed_url'"); @@ -108,9 +108,7 @@ $random_qpart = sql_random_function(); // We search for feed needing update. - $result = db_query($link, "SELECT ttrss_feeds.feed_url,ttrss_feeds.id, ttrss_feeds.owner_uid, - ".SUBSTRING_FOR_DATE."(ttrss_feeds.last_updated,1,19) AS last_updated, - ttrss_feeds.update_interval + $result = db_query($link, "SELECT DISTINCT ttrss_feeds.feed_url FROM ttrss_feeds, ttrss_users, ttrss_user_prefs WHERE @@ -118,39 +116,62 @@ AND ttrss_users.id = ttrss_user_prefs.owner_uid AND ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL' $login_thresh_qpart $update_limit_qpart - $updstart_thresh_qpart - ORDER BY $random_qpart $query_limit"); + $updstart_thresh_qpart + ORDER BY feed_url $query_limit"); $user_prefs_cache = array(); - if($debug) _debug(sprintf("Scheduled %d feeds to update...\n", db_num_rows($result))); + if($debug) _debug(sprintf("Scheduled %d feeds to update...", db_num_rows($result))); // Here is a little cache magic in order to minimize risk of double feed updates. $feeds_to_update = array(); while ($line = db_fetch_assoc($result)) { - $feeds_to_update[$line['id']] = $line; + array_push($feeds_to_update, db_escape_string($link, $line['feed_url'])); } // We update the feed last update started date before anything else. // There is no lag due to feed contents downloads // It prevent an other process to update the same feed. - $feed_ids = array_keys($feeds_to_update); - if($feed_ids) { + + if(count($feeds_to_update) > 0) { + $feeds_quoted = array(); + + foreach ($feeds_to_update as $feed) { + array_push($feeds_quoted, "'" . db_escape_string($link, $feed) . "'"); + } + db_query($link, sprintf("UPDATE ttrss_feeds SET last_update_started = NOW() - WHERE id IN (%s)", implode(',', $feed_ids))); + WHERE feed_url IN (%s)", implode(',', $feeds_quoted))); } expire_cached_files($debug); expire_lock_files($debug); - // For each feed, we call the feed update function. - while ($line = array_pop($feeds_to_update)) { - - if($debug) _debug("Feed: " . $line["feed_url"] . ", " . $line["last_updated"]); + $nf = 0; - update_rss_feed($link, $line["id"], true); - - sleep(1); // prevent flood (FIXME make this an option?) + // For each feed, we call the feed update function. + foreach ($feeds_to_update as $feed) { + if($debug) _debug("Base feed: $feed"); + + //update_rss_feed($link, $line["id"], true); + + // since we have the data cached, we can deal with other feeds with the same url + + $tmp_result = db_query($link, "SELECT ttrss_feeds.feed_url,ttrss_feeds.id,last_updated + FROM ttrss_feeds, ttrss_users WHERE + ttrss_users.id = ttrss_feeds.owner_uid AND + feed_url = '".db_escape_string($link, $feed)."' AND + ttrss_feeds.update_interval != -1 + $login_thresh_qpart + ORDER BY feed_url $query_limit"); + + if (db_num_rows($tmp_result) > 0) { + while ($tline = db_fetch_assoc($tmp_result)) { + if($debug) _debug(" => " . $tline["last_updated"] . ", " . $tline["id"]); + update_rss_feed($link, $tline["id"], true); + ++$nf; + } + } } require_once "digest.php"; @@ -158,33 +179,7 @@ // Send feed digests by email if needed. send_headlines_digests($link, $debug); - // Purge feeds with stale data OR not being updated for a while to keep DB size down - - if ($debug) _debug("Purging inactive feeds..."); - - if (DB_TYPE == "pgsql") { - $interval_qpart = "NOW() - INTERVAL '1 month'"; - } else { - $interval_qpart = "DATE_SUB(NOW(), INTERVAL 1 MONTH)"; - } - - $result = db_query($link, "SELECT id, feed_url FROM ttrss_feeds WHERE - (SELECT MAX(updated) FROM ttrss_entries, ttrss_user_entries WHERE - ttrss_entries.id = ref_id AND - ttrss_user_entries.feed_id = ttrss_feeds.id) < $interval_qpart OR - last_updated < $interval_qpart"); - - $feeds_purged = 0; - $articles_removed = 0; - - while ($line = db_fetch_assoc($result)) { - $articles_removed += purge_feed($link, $line["id"], 0, false); - ++$feeds_purged; - } - - if ($debug && $articles_removed > 0) - _debug(sprintf("Purged %d feeds (%d articles).", $feeds_purged, - $articles_removed)); + return $nf; } // function update_daemon_common @@ -228,52 +223,100 @@ $cache_images = sql_bool_to_bool(db_fetch_result($result, 0, "cache_images")); $fetch_url = db_fetch_result($result, 0, "feed_url"); - $feed = db_escape_string($feed); + $feed = db_escape_string($link, $feed); - /* if ($auth_login && $auth_pass ){ - $url_parts = array(); - preg_match("/(^[^:]*):\/\/(.*)/", $fetch_url, $url_parts); + if ($override_url) $fetch_url = $override_url; - if ($url_parts[1] && $url_parts[2]) { - $fetch_url = $url_parts[1] . "://$auth_login:$auth_pass@" . $url_parts[2]; - } - } */ + $date_feed_processed = date('Y-m-d H:i'); - if ($override_url) - $fetch_url = $override_url; + $cache_filename = CACHE_DIR . "/simplepie/" . sha1($fetch_url) . ".feed"; + + // Ignore cache if new feed or manual update. + $cache_age = ($no_cache || is_null($last_updated) || $last_updated == '1970-01-01 00:00:00') ? + 30 : get_feed_update_interval($link, $feed) * 60; if ($debug_enabled) { - _debug("update_rss_feed: fetching [$fetch_url]..."); + _debug("update_rss_feed: cache filename: $cache_filename exists: " . file_exists($cache_filename)); + _debug("update_rss_feed: cache age: $cache_age; no cache: $no_cache"); } - // Ignore cache if new feed or manual update. - $cache_age = (is_null($last_updated) || $last_updated == '1970-01-01 00:00:00') ? - -1 : get_feed_update_interval($link, $feed) * 60; + $cached_feed_data_hash = false; - $simplepie_cache_dir = CACHE_DIR . "/simplepie"; + $rss = false; + $rss_hash = false; + $cache_timestamp = file_exists($cache_filename) ? filemtime($cache_filename) : 0; + $last_updated_timestamp = strtotime($last_updated); - if (!is_dir($simplepie_cache_dir)) { - mkdir($simplepie_cache_dir); + if (file_exists($cache_filename) && + is_readable($cache_filename) && + !$auth_login && !$auth_pass && + filemtime($cache_filename) > time() - $cache_age) { + + if ($debug_enabled) { + _debug("update_rss_feed: using local cache."); + } + + if ($cache_timestamp > $last_updated_timestamp) { + @$rss_data = file_get_contents($cache_filename); + + if ($rss_data) { + $rss_hash = sha1($rss_data); + @$rss = unserialize($rss_data); + } + } else { + if ($debug_enabled) { + _debug("update_rss_feed: local cache valid and older than last_updated, nothing to do."); + } + return; + } } - $feed_data = fetch_file_contents($fetch_url, false, - $auth_login, $auth_pass, false, $no_cache ? 15 : 45); + if (!$rss) { - if (!$feed_data) { - global $fetch_last_error; + if (!$feed_data) { + if ($debug_enabled) { + _debug("update_rss_feed: fetching [$fetch_url] (ts: $cache_timestamp/$last_updated_timestamp)"); + } + + $feed_data = fetch_file_contents($fetch_url, false, + $auth_login, $auth_pass, false, $no_cache ? 15 : 45, + max($last_updated_timestamp, $cache_timestamp)); + + if ($debug_enabled) { + _debug("update_rss_feed: fetch done."); + } - if ($debug_enabled) { - _debug("update_rss_feed: unable to fetch: $fetch_last_error"); } - db_query($link, - "UPDATE ttrss_feeds SET last_error = '$fetch_last_error', - last_updated = NOW() WHERE id = '$feed'"); + if (!$feed_data) { + global $fetch_last_error; + global $fetch_last_error_code; + + if ($debug_enabled) { + _debug("update_rss_feed: unable to fetch: $fetch_last_error [$fetch_last_error_code]"); + } + + $error_escaped = ''; + + // If-Modified-Since + if ($fetch_last_error_code != 304) { + $error_escaped = db_escape_string($link, $fetch_last_error); + } else { + if ($debug_enabled) { + _debug("update_rss_feed: source claims data not modified, nothing to do."); + } + } + + db_query($link, + "UPDATE ttrss_feeds SET last_error = '$error_escaped', + last_updated = NOW() WHERE id = '$feed'"); - return; + return; + } } $pluginhost = new PluginHost($link); + $pluginhost->set_debug($debug_enabled); $user_plugins = get_pref($link, "_ENABLED_PLUGINS", $owner_uid); $pluginhost->load(PLUGINS, $pluginhost::KIND_ALL); @@ -284,38 +327,38 @@ $feed_data = $plugin->hook_feed_fetched($feed_data); } - if ($debug_enabled) { - _debug("update_rss_feed: fetch done, parsing..."); - } - - $rss = new SimplePie(); - $rss->set_sanitize_class("SanitizeDummy"); - // simplepie ignores the above and creates default sanitizer anyway, - // so let's override it... - $rss->sanitize = new SanitizeDummy(); - $rss->set_output_encoding('UTF-8'); - $rss->set_raw_data($feed_data); - - if ($debug_enabled) { - _debug("feed update interval (sec): " . - get_feed_update_interval($link, $feed)*60); + if (!$rss) { + $rss = new SimplePie(); + $rss->set_sanitize_class("SanitizeDummy"); + // simplepie ignores the above and creates default sanitizer anyway, + // so let's override it... + $rss->sanitize = new SanitizeDummy(); + $rss->set_output_encoding('UTF-8'); + $rss->set_raw_data($feed_data); + $rss->enable_cache(false); + + @$rss->init(); } - $rss->enable_cache(!$no_cache); - - if (!$no_cache) { - $rss->set_cache_location($simplepie_cache_dir); - $rss->set_cache_duration($cache_age); - } - - @$rss->init(); - // print_r($rss); - $feed = db_escape_string($feed); + $feed = db_escape_string($link, $feed); if (!$rss->error()) { + // cache data for later + if (!$auth_pass && !$auth_login && is_writable(CACHE_DIR . "/simplepie")) { + $rss_data = serialize($rss); + $new_rss_hash = sha1($rss_data); + + if ($new_rss_hash != $rss_hash) { + if ($debug_enabled) { + _debug("update_rss_feed: saving $cache_filename"); + } + @file_put_contents($cache_filename, serialize($rss)); + } + } + // 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); @@ -343,7 +386,7 @@ $owner_uid = db_fetch_result($result, 0, "owner_uid"); - $site_url = db_escape_string(mb_substr(rewrite_relative_url($fetch_url, $rss->get_link()), 0, 245)); + $site_url = db_escape_string($link, mb_substr(rewrite_relative_url($fetch_url, $rss->get_link()), 0, 245)); if ($debug_enabled) { _debug("update_rss_feed: checking favicon..."); @@ -358,7 +401,7 @@ if (!$registered_title || $registered_title == "[Unknown]") { - $feed_title = db_escape_string($rss->get_title()); + $feed_title = db_escape_string($link, $rss->get_title()); if ($debug_enabled) { _debug("update_rss_feed: registering title: $feed_title"); @@ -460,7 +503,7 @@ $entry_timestamp = strtotime($item->get_date()); - if ($entry_timestamp == -1 || !$entry_timestamp) { + if ($entry_timestamp == -1 || !$entry_timestamp || $entry_timestamp > time()) { $entry_timestamp = time(); $no_orig_date = 'true'; } else { @@ -500,13 +543,13 @@ $entry_author = $entry_author_item->get_name(); if (!$entry_author) $entry_author = $entry_author_item->get_email(); - $entry_author = db_escape_string($entry_author); + $entry_author = db_escape_string($link, $entry_author); } - $entry_guid = db_escape_string(mb_substr($entry_guid, 0, 245)); + $entry_guid = db_escape_string($link, mb_substr($entry_guid, 0, 245)); - $entry_comments = db_escape_string(mb_substr($entry_comments, 0, 245)); - $entry_author = db_escape_string(mb_substr($entry_author, 0, 245)); + $entry_comments = db_escape_string($link, mb_substr($entry_comments, 0, 245)); + $entry_author = db_escape_string($link, mb_substr($entry_author, 0, 245)); $num_comments = $item->get_item_tags('http://purl.org/rss/1.0/modules/slash/', 'comments'); @@ -553,7 +596,7 @@ } if ($debug_enabled) { - _debug("update_rss_feed: done collecting data [TITLE:$entry_title]"); + _debug("update_rss_feed: done collecting data."); } // TODO: less memory-hungry implementation @@ -564,7 +607,7 @@ // FIXME not sure if owner_uid is a good idea here, we may have a base entry without user entry (?) $result = db_query($link, "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)."' AND owner_uid = $owner_uid"); + WHERE ref_id = id AND guid = '".db_escape_string($link, $entry_guid)."' AND owner_uid = $owner_uid"); if (db_num_rows($result) != 0) { $entry_plugin_data = db_fetch_result($result, 0, "plugin_data"); @@ -593,19 +636,22 @@ } $entry_tags = $article["tags"]; - $entry_guid = db_escape_string($entry_guid); - $entry_content = db_escape_string($article["content"], false); - $entry_title = db_escape_string($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"]); + $entry_guid = db_escape_string($link, $entry_guid); + $entry_title = db_escape_string($link, $article["title"]); + $entry_author = db_escape_string($link, $article["author"]); + $entry_link = db_escape_string($link, $article["link"]); + $entry_plugin_data = db_escape_string($link, $article["plugin_data"]); + $entry_content = $article["content"]; // escaped below + if ($debug_enabled) { _debug("update_rss_feed: plugin data: $entry_plugin_data"); } if ($cache_images && is_writable(CACHE_DIR . '/images')) - $entry_content = cache_images($entry_content, $site_url, $debug_enabled); + cache_images($entry_content, $site_url, $debug_enabled); + + $entry_content = db_escape_string($link, $entry_content, false); $content_hash = "SHA1:" . sha1($entry_content); @@ -648,7 +694,7 @@ '', $no_orig_date, NOW(), - NOW(), + '$date_feed_processed', '$entry_comments', '$num_comments', '$entry_plugin_data', @@ -790,12 +836,17 @@ } } + $last_marked = ($marked == 'true') ? 'NOW()' : 'NULL'; + $last_published = ($published == 'true') ? 'NOW()' : 'NULL'; + $result = db_query($link, "INSERT INTO ttrss_user_entries (ref_id, owner_uid, feed_id, unread, last_read, marked, - published, score, tag_cache, label_cache, uuid) + published, score, tag_cache, label_cache, uuid, + last_marked, last_published) VALUES ('$ref_id', '$owner_uid', '$feed', $unread, - $last_read_qpart, $marked, $published, '$score', '', '', '')"); + $last_read_qpart, $marked, $published, '$score', '', '', + '', $last_marked, $last_published)"); if (PUBSUBHUBBUB_HUB && $published == 'true') { $rss_link = get_self_url_prefix() . @@ -846,7 +897,7 @@ $update_insignificant = false; } - if (db_escape_string($orig_title) != $entry_title) { + if (db_escape_string($link, $orig_title) != $entry_title) { $post_needs_update = true; $update_insignificant = false; } @@ -913,9 +964,9 @@ db_query($link, "BEGIN"); 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]); + $enc_url = db_escape_string($link, $enc[0]); + $enc_type = db_escape_string($link, $enc[1]); + $enc_dur = db_escape_string($link, $enc[2]); $result = db_query($link, "SELECT id FROM ttrss_enclosures WHERE content_url = '$enc_url' AND post_id = '$entry_ref_id'"); @@ -976,7 +1027,7 @@ foreach ($filtered_tags as $tag) { $tag = sanitize_tag($tag); - $tag = db_escape_string($tag); + $tag = db_escape_string($link, $tag); if (!tag_is_valid($tag)) continue; @@ -998,7 +1049,7 @@ $tags_to_cache = array_unique($tags_to_cache); - $tags_str = db_escape_string(join(",", $tags_to_cache)); + $tags_str = db_escape_string($link, join(",", $tags_to_cache)); db_query($link, "UPDATE ttrss_user_entries SET tag_cache = '$tags_str' WHERE ref_id = '$entry_ref_id' @@ -1013,9 +1064,9 @@ } foreach ($labels as $label) { - $caption = $label["caption"]; + $caption = preg_quote($label["caption"]); - if (preg_match("/\b$caption\b/i", "$tags_str " . strip_tags($entry_content) . " $entry_title")) { + if ($caption && preg_match("/\b$caption\b/i", "$tags_str " . strip_tags($entry_content) . " $entry_title")) { if (!labels_contains_caption($article_labels, $caption)) { label_add_article($link, $entry_ref_id, $caption, $owner_uid); } @@ -1048,7 +1099,7 @@ } else { - $error_msg = db_escape_string(mb_substr($rss->error(), 0, 245)); + $error_msg = db_escape_string($link, mb_substr($rss->error(), 0, 245)); if ($debug_enabled) { _debug("update_rss_feed: error fetching feed: $error_msg"); @@ -1107,7 +1158,7 @@ $node = $doc->getElementsByTagName('body')->item(0); - return $doc->saveXML($node, LIBXML_NOEMPTYTAG); + return $doc->saveXML($node); } function expire_lock_files($debug) { @@ -1182,11 +1233,13 @@ foreach ($filters as $filter) { $match_any_rule = $filter["match_any_rule"]; + $inverse = $filter["inverse"]; $filter_match = false; foreach ($filter["rules"] as $rule) { $match = false; $reg_exp = $rule["reg_exp"]; + $rule_inverse = $rule["inverse"]; if (!$reg_exp) continue; @@ -1219,6 +1272,8 @@ break; } + if ($rule_inverse) $match = !$match; + if ($match_any_rule) { if ($match) { $filter_match = true; @@ -1232,6 +1287,8 @@ } } + if ($inverse) $filter_match = !$filter_match; + if ($filter_match) { foreach ($filter["actions"] AS $action) { array_push($matches, $action); @@ -1293,27 +1350,6 @@ } } - function cache_content($link, $url, $login, $pass) { - - $content = fetch_file_contents($url, $login, $pass); - - if ($content) { - $doc = new DOMDocument(); - @$doc->loadHTML($content); - $xpath = new DOMXPath($doc); - - $node = $doc->getElementsByTagName('body')->item(0); - - if ($node) { - $content = $doc->saveXML($node, LIBXML_NOEMPTYTAG); - - return $content; - } - } - - return ""; - } - function make_guid_from_title($title) { return preg_replace("/[ \"\',.:;]/", "-", mb_strtolower(strip_tags($title), 'utf-8'));