FiveFilters.org: Full-Text RSS http://fivefilters.org/content-only/ CHANGELOG ------------------------------------ 3.0 (2012-09-04) - Multi-page support - next_page_link now supported in site config (enable/disable with $options->multipage) - HTML5 parser available - use parser: html5lib in site config, also see $options->allowed_parsers - Updated site patterns for better extraction - New global site config to be applied to all sites (global.txt) - APC caching of site config files to improve performance, if APC available - see $options->apc - Site config editor in admin/ - easily find, edit, test, and test site config files, or add new ones - Debug mode to see what's happening behind the scenes - see $options->debug - Removed deprecated config options: restrict, message_to_prepend_with_key, message_to_append_with_key, error_message_with_key - Removed extraction with CSS via querystring - Removed config option: $options->alternative_url - Bug fix: allow extraction of a single element - Bug fix: redirect handling improved - Strip 'http://' prefix when API key is supplied - Site config merging (custom + standard + fingerprint + global) - Site config command replace_string(find): replace can now be split over two lines: find_string: find, replace_string: replace - YouTube and Vimeo URLs now return iframe embed code - We now look for OpenGraph title and date elements - Improved extraction from AJAX pages - we now look for AJAX triggers embedded in HTML, per Google spec - JSONP support - use &format=json&callback=functionName in querystring - New config option to enable Cross-Origin Resource Sharing (CORS): $option->cors - New config option to enable XSS filtering, if required: $option->xss_filter - Zend_Cache updated - Smart caching - experimental feature to store cache IDs in APC first, and write output to disk on subsequent request (see $options->smart_cache) - Easier cloud deploy - manifest.yml added for AppFog - Override most config options with environment variables, e.g. ftr_max_entries: 3 2.9.5 (2012-04-29) - Language detection using Text_LanguageDetect or PHP-CLD (dc:language field in output and $options->detect_language in config) - New site patterns added and old ones updated - Experimental tool for simpler site pattern updates (access admin/ folder) - Plus other fixes/improvements 2.9.1 (2011-11-02) - Fix: Character encoding issue affecting some non-English articles (makefulltextfeed.php and SimplePie/Misc.php changed) 2.9 (2011-11-01) - New site patterns added and old ones updated - New config option: require_key - restrict access to those with password/key - New config option: rewrite_url - URL rewrite rules to be applied before HTTP request - New site config options to extract author(s) and publication date (matches included in feed item as and ) - New site config option: replace_string([string to find]): [replacement string] - New site identification method: site fingerprints (HTML fragments linked to site config) - Update check now also checks for new site patterns - Effective URL (URL after redirects/rewrites) now included in feed item as - Prevent indexing of generated feeds by search engines - Enclosure support (enclosures preserved as elements) - Better handling of non-HTML content types - Sending custom User-Agent HTTP header for matching sites now supported - CSS extraction deprecated in favour of site patterns (still works, but form field removed and feature may disappear in 3.0) - Fix: Improved character-encoding detection - Fix: URL parsing issues for certain URLs (SimplePie updated) - Fix: Author and other Dublin Core () elements now appear in JSON output - Fix: Minor fixes for PHP Readability - Plus other minor fixes/improvements 2.8 (2011-05-30) - Tidy no longer stripping HTML5 elements - JSON output (pass &format=json in querystring) - New site patterns added and old ones updated - New site config option to force full-page retrieval on multi-page articles: single_page_link - User Guide (PDF) now included (although still a work in progress) - URL placeholders now accepted in message_to_prepend/append config options - Plus minor fixes... 2.7 (2011-03-21) - Site patterns for better control over extraction (see site_config/README.txt) - hNews support (improves content extraction for sites using hNews microformatting) - Cookie Jar now used to store and sends cookies when following HTTP redirects - Better handling of certain cases where HTML Tidy fails to clean up properly - Bug fix: curl_multi_select() timing out in certain environments (fixed in HumbleHttpAgent.php) - Bug fix: broken HTTP header parsing in some environments (fixed in SimplePie_HumbleHttpAgent.php) - Bug fix: invalid API URL shown (fixed in index.php) - Plus other minor fixes... 2.6 (2011-03-02) - Rewriting of hash-bang (#!) URLs (see http://www.tbray.org/ongoing/When/201x/2011/02/09/Hash-Blecch for an explanation) - Improved parallel fetching support (HumbleHttpAgent uses curl_multi_* functions if PECL HTTP extension is not present) - Improved HTTP redirect support (now handled in HumbleHttpAgent, no longer relies on PHP) - Improved performance for single page (non-feed) requests: (SimplePie connected to HumbleHttpAgent) - Improved memory use for processing large feeds (HumbleHttpAgent's stored responses cleared as they're retrieved) - Bug fix: exclude on fail option no longer requires valid key - Bug fix: workaround for PHP bug http://bugs.php.net/51192 (fixed in makefulltextfeed.php) - Plus other minor changes... 2.5 (2011-01-08) - New option: custom extraction pattern (CSS selectors) - New option: allowed URLs (restrict service to pre-defined feeds/domains) - New option: exclude items on fail (remove items from feed if content extraction fails) - Remove 'http://' from URL before form submission (prevents errors on hosts which have overly vigilant security software) - Allow overriding of index.php with custom_index.php - config.php now required (override with custom_config.php) - index.php now uses config.php to determine what to display - Bug fix: occasional fatal error in IRI::__toString() (IRI updated) - Bug fix: workaround for PHP bug http://bugs.php.net/51192 (fixed in HumbleHttpAgent.php) 2.2 (2010-10-30) - Character-encoding detection improved (minor change) - Rewriting of relative URLs improved (tracks redirect URLs) - Minor changes to prevent errors in certain hosting environments - Compatibility test file updated with more tests 2.1 (2010-09-13) - Better content extraction (using PHP Readability 1.7.1) - Parallel HTTP requests (using Humble HTTP Agent) - Auto loading of necessary classes - Rewriting of relative URLs (using IRI) - Added compatibility test file (to check if server meets requirements) - Character-encoding support improved (using SimplePie) 1.5 (2010-05-30) - Support for PHP 5.3 (thanks Murilo!) - Character-encoding support improved (favours iconv over mb_convert_encoding) 1.0 (2010-03-05) - Better support for different character-encodings - Auto-cleanup of cache files - Very basic option for load distribution (if you're planning on installing the code on multiple servers) - Separate config file (see config-sample.php)