diff --git a/rss/magpie/AUTHORS b/rss/magpie/AUTHORS new file mode 100644 index 00000000000..7d7f3f53ebe --- /dev/null +++ b/rss/magpie/AUTHORS @@ -0,0 +1 @@ +kellan diff --git a/rss/magpie/README b/rss/magpie/README new file mode 100644 index 00000000000..6af7edb07c8 --- /dev/null +++ b/rss/magpie/README @@ -0,0 +1,48 @@ +NAME + + MagpieRSS - a simple RSS integration tool + +SYNOPSIS + + require_once(rss_fetch.inc); + $url = $_GET['url']; + $rss = fetch_rss( $url ); + + echo "Channel Title: " . $rss->channel['title'] . "

"; + echo "

"; + +DESCRIPTION + + MapieRSS is an XML-based RSS parser in PHP. It attempts to be "PHP-like", + and simple to use. + + Some features include: + + * supports RSS 0.9 - 1.0, with limited RSS 2.0 support + * supports namespaces, and modules, including mod_content and mod_event + * open minded [1] + * simple, functional interface, to object oriented backend parser + * automatic caching of parsed RSS objects makes its easy to integrate + * supports conditional GET with Last-Modified, and ETag + * uses constants for easy override of default behaviour + * heavily commented + + +1. By open minded I mean Magpie will accept any tag it finds in good faith that + it was supposed to be here. For strict validation, look elsewhere. + + +GETTING STARTED + + + +COPYRIGHT: + Copyright(c) 2002 kellan@protest.net. All rights reserved. + This software is released under the GNU General Public License. + Please read the disclaimer at the top of the Snoopy.class.inc file. diff --git a/rss/magpie/README_MOODLE.txt b/rss/magpie/README_MOODLE.txt new file mode 100644 index 00000000000..3bc9d8c900f --- /dev/null +++ b/rss/magpie/README_MOODLE.txt @@ -0,0 +1,3 @@ +This folder is the MagpieRSS news feed client library +http://magpierss.sourceforge.net/ +Moodle's rss_client block uses these libraries to download, parse and cache remote new feeds. diff --git a/rss/magpie/TROUBLESHOOTING b/rss/magpie/TROUBLESHOOTING new file mode 100644 index 00000000000..89068d382c1 --- /dev/null +++ b/rss/magpie/TROUBLESHOOTING @@ -0,0 +1,152 @@ +TROUBLESHOOTING + + +Trouble Installing MagpieRSS: + +1. Fatal error: Failed opening required '/path/to/script/rss_fetch.inc' + (include_path='.:/usr/local/lib/php:/usr/local/lib/php/pear') + +2. Cache couldn't make dir './cache'. + +3. Fatal error: Failed to load PHP's XML Extension. + http://www.php.net/manual/en/ref.xml.php + +Trouble Using MagpieRSS + +4. Warning: MagpieRSS: Failed to fetch example.com/index.rdf. + (HTTP Error: Invalid protocol "") + +5. Warning: MagpieRSS: Failed to parse RSS file. + (not well-formed (invalid token) at line 19, column 98) + +6. Warning: MagpieRSS: Failed to fetch http://localhost/rss/features.1-0.rss. + (HTTP Response: HTTP/1.1 404 Not Found) + +If you would rather provide a custom error, see the COOKBOOK +(http://magpierss.sf.net/cookbook.html) recipe 2. + +************************************************************************* +1. Fatal error: Failed opening required '/path/to/script/rss_fetch.inc' + (include_path='.:/usr/local/lib/php:/usr/local/lib/php/pear') + + This could mean that: + + a) PHP can't find the MagpieRSS files. + b) PHP found them the MagpieRSS files, but can't read them. + + a. Telling PHP where to look for MagpieRSS file. + + This might mean your PHP program can't find the MagpieRSS libraries. + Magpie relies on 4 include files, rss_fetch.inc, rss_parse.inc, + rss_cache.inc, rss_util.inc, and for normal use you'll need all 4 (see the + cookbook for exceptions). + + This can be fixed by making sure the MagpieRSS files are in your include + path. + + If you can edit your include path (for example your on a shared host) then + you need to replace: + + require_once('rss_fetch.inc'); + + -with- + + define('MAGPIE_DIR', '/path/to/magpierss/'); + require_once(MAGPIE_DIR.'rss_fetch.inc'); + + b. PHP can't read the MagpieRSS files + + All PHP libraries need to be readable by your webserver. + + On Unix you can accomplish this with: + + chmod 755 rss_fetch.inc rss_parse.inc rss_cache.inc rss_util.inc + +************************************************************************* +2. Cache couldn't make dir './cache'. + + MagpieRSS caches the results of fetched and parsed RSS to reduce the load on + both your server, and the remote server providing the RSS. It does this by + writing files to a cache directory. + + This error means the webserver doesn't have write access to the current + directory. + + a. Make a webserver writeable cache directory + + Find the webserver's group. (on my system it is 'www') + + mkdir ./cache + chgrp www directory_name + chmod g+w directory_name + + (this is the best, and desired solution) + + b. Tell MagpieRSS to create the cache directory somewhere the webserver can + write to. + + define('MAGPIE_CACHE_DIR', '/tmp/magpierss'); + + (this is not a great solution, and might have security considerations) + + c. Turn off cacheing. + + Magpie can work fine with cacheing, but it will be slower, and you might + become a nuiance to the RSS provider, but it is an option. + + define('MAGPIE_CACHE_ON', 0); + + d. And lastly, do NOT + + chmod 777 ./cache + + Any of the above solutions are better then this. + + NOTE: If none of this works for you, let me know. I've got root, and a + custom compiled Apache on almost any box I ever touch, so I can be a little + out of touch with reality. But I won't know that if I don't feedback. + +************************************************************************* 3. +3. Fatal error: Failed to load PHP's XML Extension. + http://www.php.net/manual/en/ref.xml.php + + -or- + + Fatal error: Failed to create an instance of PHP's XML parser. + http://www.php.net/manual/en/ref.xml.php + + Make sure your PHP was built with --with-xml + + This has been turned on by default for several versions of PHP, but it might + be turned off in your build. + + See php.net for details on building and configuring PHP. + + +************************************************************************* +4. Warning: MagpieRSS: Failed to fetch index.rdf. + (HTTP Error: Invalid protocol "") + + You need to put http:// in front of your the URL to your RSS feed + +************************************************************************* +5. Warning: MagpieRSS: Failed to parse RSS file. + (not well-formed (invalid token) at line 19, column 98) + + There is a problem with the RSS feed you are trying to read. + MagpieRSS is an XML parser, and therefore can't parse RSS feed with invalid + characters. Some RSS parser are based on regular expressions, and can + parse invalid RSS but they have their own problems. + + You could try contacting the author of the RSS feed, and pointing them to + the online RSS validator at: + + http://feeds.archive.org/validator/ + +************************************************************************* +6. Warning: MagpieRSS: Failed to fetch http://example.com/index.rdf + (HTTP Response: HTTP/1.1 404 Not Found) + + Its a 404! The RSS file ain't there. + + diff --git a/rss/magpie/cookbook b/rss/magpie/cookbook new file mode 100644 index 00000000000..45dda98bcd9 --- /dev/null +++ b/rss/magpie/cookbook @@ -0,0 +1,125 @@ +MAGPIERSS RECIPES: Cooking with Corbies + + "Four and twenty blackbirds baked in a pie." + +1. LIMIT THE NUMBER OF HEADLINES(AKA ITEMS) RETURNED. + +PROBLEM: + +You want to display the 10 (or 3) most recent headlines, but the RSS feed +contains 15. + +SOLUTION: + +$num_items = 10; +$rss = fetch_rss($url); + +$items = array_slice($rss->items, 0, $num_items); + +DISCUSSION: + +Rather then trying to limit the number of items Magpie parses, a much simpler, +and more flexible approach is to take a "slice" of the array of items. And +array_slice() is smart enough to do the right thing if the feed has less items +then $num_items. + +See: http://www.php.net/array_slice + + +2. DISPLAY A CUSTOM ERROR MESSAGE IF SOMETHING GOES WRONG + +PROBLEM: + +You don't want Magpie's error messages showing up if something goes wrong. + +SOLUTION: + +# Magpie throws USER_WARNINGS only +# so you can cloak these, by only showing ERRORs +error_reporting(E_ERROR); + +# check the return value of fetch_rss() + +$rss = fetch_rss($url); + +if ( $rss ) { +...display rss feed... +} +else { + echo "An error occured! " . + "Consider donating more $$$ for restoration of services." . + "
Error Message: " . magpie_error(); +} + +DISCUSSION: + +MagpieRSS triggers a warning in a number of circumstances. The 2 most common +circumstances are: if the specified RSS file isn't properly formed (usually +because it includes illegal HTML), or if Magpie can't download the remote RSS +file, and there is no cached version. + +If you don't want your users to see these warnings change your error_reporting +settings to only display ERRORs. Another option is to turn off display_error, +so that WARNINGs, and NOTICEs still go to the error_log but not to the webpages. + +You can do this with: + +ini_set('display_errors', 0); + +See: http://www.php.net/error_reporting, + http://www.php.net/ini_set, + http://www.php.net/manual/en/ref.errorfunc.php + +3. GENERATE A NEW RSS FEED + +PROBLEM: + +Create an RSS feed for other people to use. + +SOLUTION: + +Use Useful Inc's RSSWriter (http://usefulinc.com/rss/rsswriter/) + +DISCUSSION: + +An example of turning a Magpie parsed RSS object back into an RSS file is forth +coming. In the meantime RSSWriter has great documentation. + +4. DISPLAY HEADLINES MORE RECENT THEN X DATE + +PROBLEM: + +You only want to display headlines that were published on, or after a certain +date. + + +SOLUTION: + +require 'rss_utils.inc'; + +# get all headlines published today +$today = getdate(); + +# today, 12AM +$date = mktime(0,0,0,$today['mon'], $today['mday'], $today['year']); + +$rss = fetch_rss($url); + +foreach ( $rss->items as $item ) { + $published = parse_w3cdtf($item['dc']['date']); + if ( $published >= $date ) { + echo "Title: " . $item['title']; + echo "Published: " . date("h:i:s A", $published); + echo "

"; + } +} + +DISCUSSION: + +This recipe only works for RSS 1.0 feeds that include the field. +(which is very good RSS style) + +parse_w3cdtf is defined in rss_utils.inc, and parses RSS style dates into Unix +epoch seconds. + +See: http://www.php.net/manual/en/ref.datetime.php diff --git a/rss/magpie/rss_cache.inc b/rss/magpie/rss_cache.inc new file mode 100644 index 00000000000..fdfa5dd2591 --- /dev/null +++ b/rss/magpie/rss_cache.inc @@ -0,0 +1,184 @@ + + * Version: 0.51 + * License: GPL + * + * The lastest version of MagpieRSS can be obtained from: + * http://magpierss.sourceforge.net + * + * For questions, help, comments, discussion, etc., please join the + * Magpie mailing list: + * http://lists.sourceforge.net/lists/listinfo/magpierss-general + * + */ + +class RSSCache { + var $BASE_CACHE = './cache'; // where the cache files are stored + var $MAX_AGE = 3600; // when are files stale, default one hour + var $ERROR = ""; // accumulate error messages + + function RSSCache ($base='', $age='') { + if ( $base ) { + $this->BASE_CACHE = $base; + } + if ( $age ) { + $this->MAX_AGE = $age; + } + + // attempt to make the cache directory + if ( ! file_exists( $this->BASE_CACHE ) ) { + $status = @mkdir( $this->BASE_CACHE, 0755 ); + + // if make failed + if ( ! $status ) { + $this->error( + "Cache couldn't make dir '" . $this->BASE_CACHE . "'." + ); + } + } + } + +/*=======================================================================*\ + Function: set + Purpose: add an item to the cache, keyed on url + Input: url from wich the rss file was fetched + Output: true on sucess +\*=======================================================================*/ + function set ($url, $rss) { + $this->ERROR = ""; + $cache_file = $this->file_name( $url ); + $fp = @fopen( $cache_file, 'w' ); + + if ( ! $fp ) { + $this->error( + "Cache unable to open file for writing: $cache_file" + ); + return 0; + } + + + $data = $this->serialize( $rss ); + fwrite( $fp, $data ); + fclose( $fp ); + + return $cache_file; + } + +/*=======================================================================*\ + Function: get + Purpose: fetch an item from the cache + Input: url from wich the rss file was fetched + Output: cached object on HIT, false on MISS +\*=======================================================================*/ + function get ($url) { + $this->ERROR = ""; + $cache_file = $this->file_name( $url ); + + if ( ! file_exists( $cache_file ) ) { + $this->debug( + "Cache doesn't contain: $url (cache file: $cache_file)" + ); + return 0; + } + + $fp = @fopen($cache_file, 'r'); + if ( ! $fp ) { + $this->error( + "Failed to open cache file for reading: $cache_file" + ); + return 0; + } + + $data = fread( $fp, filesize($cache_file) ); + $rss = $this->unserialize( $data ); + + return $rss; + } + +/*=======================================================================*\ + Function: check_cache + Purpose: check a url for membership in the cache + and whether the object is older then MAX_AGE (ie. STALE) + Input: url from wich the rss file was fetched + Output: cached object on HIT, false on MISS +\*=======================================================================*/ + function check_cache ( $url ) { + $this->ERROR = ""; + $filename = $this->file_name( $url ); + + if ( file_exists( $filename ) ) { + // find how long ago the file was added to the cache + // and whether that is longer then MAX_AGE + $mtime = filemtime( $filename ); + $age = time() - $mtime; + if ( $this->MAX_AGE > $age ) { + // object exists and is current + return 'HIT'; + } + else { + // object exists but is old + return 'STALE'; + } + } + else { + // object does not exist + return 'MISS'; + } + } + +/*=======================================================================*\ + Function: serialize +\*=======================================================================*/ + function serialize ( $rss ) { + return serialize( $rss ); + } + +/*=======================================================================*\ + Function: unserialize +\*=======================================================================*/ + function unserialize ( $data ) { + return unserialize( $data ); + } + +/*=======================================================================*\ + Function: file_name + Purpose: map url to location in cache + Input: url from wich the rss file was fetched + Output: a file name +\*=======================================================================*/ + function file_name ($url) { + $filename = md5( $url ); + return join( DIRECTORY_SEPARATOR, array( $this->BASE_CACHE, $filename ) ); + } + +/*=======================================================================*\ + Function: error + Purpose: register error +\*=======================================================================*/ + function error ($errormsg, $lvl=E_USER_WARNING) { + // append PHP's error message if track_errors enabled + if ( isset($php_errormsg) ) { + $errormsg .= " ($php_errormsg)"; + } + $this->ERROR = $errormsg; + if ( MAGPIE_DEBUG ) { + trigger_error( $errormsg, $lvl); + } + else { + error_log( $errormsg, 0); + } + } + + function debug ($debugmsg, $lvl=E_USER_NOTICE) { + if ( MAGPIE_DEBUG ) { + $this->error("MagpieRSS [debug] $debugmsg", $lvl); + } + } + +} + +?> diff --git a/rss/magpie/rss_fetch.inc b/rss/magpie/rss_fetch.inc new file mode 100644 index 00000000000..c41945cc730 --- /dev/null +++ b/rss/magpie/rss_fetch.inc @@ -0,0 +1,459 @@ + + * License: GPL + * + * The lastest version of MagpieRSS can be obtained from: + * http://magpierss.sourceforge.net + * + * For questions, help, comments, discussion, etc., please join the + * Magpie mailing list: + * magpierss-general@lists.sourceforge.net + * + */ + +// Setup MAGPIE_DIR for use on hosts that don't include +// the current path in include_path. +// with thanks to rajiv and smarty +if (!defined('DIR_SEP')) { + define('DIR_SEP', DIRECTORY_SEPARATOR); +} + +if (!defined('MAGPIE_DIR')) { + define('MAGPIE_DIR', dirname(__FILE__) . DIR_SEP); +} + +require_once( MAGPIE_DIR . 'rss_parse.inc' ); +require_once( MAGPIE_DIR . 'rss_cache.inc' ); + +// for including 3rd party libraries +define('MAGPIE_EXTLIB', MAGPIE_DIR . 'extlib' . DIR_SEP); +require_once( MAGPIE_EXTLIB . 'Snoopy.class.inc'); + + +/* + * CONSTANTS - redefine these in your script to change the + * behaviour of fetch_rss() currently, most options effect the cache + * + * MAGPIE_CACHE_ON - Should Magpie cache parsed RSS objects? + * For me a built in cache was essential to creating a "PHP-like" + * feel to Magpie, see rss_cache.inc for rationale + * + * + * MAGPIE_CACHE_DIR - Where should Magpie cache parsed RSS objects? + * This should be a location that the webserver can write to. If this + * directory does not already exist Mapie will try to be smart and create + * it. This will often fail for permissions reasons. + * + * + * MAGPIE_CACHE_AGE - How long to store cached RSS objects? In seconds. + * + * + * MAGPIE_CACHE_FRESH_ONLY - If remote fetch fails, throw error + * instead of returning stale object? + * + * MAGPIE_DEBUG - Display debugging notices? + * +*/ + + +/*=======================================================================*\ + Function: fetch_rss: + Purpose: return RSS object for the give url + maintain the cache + Input: url of RSS file + Output: parsed RSS object (see rss_parse.inc) + + NOTES ON CACHEING: + If caching is on (MAGPIE_CACHE_ON) fetch_rss will first check the cache. + + NOTES ON RETRIEVING REMOTE FILES: + If conditional gets are on (MAGPIE_CONDITIONAL_GET_ON) fetch_rss will + return a cached object, and touch the cache object upon recieving a + 304. + + NOTES ON FAILED REQUESTS: + If there is an HTTP error while fetching an RSS object, the cached + version will be return, if it exists (and if MAGPIE_CACHE_FRESH_ONLY is off) +\*=======================================================================*/ + +define('MAGPIE_VERSION', '0.7'); + +$MAGPIE_ERROR = ""; + +function fetch_rss ($url) { + // initialize constants + init(); + + if ( !isset($url) ) { + do_error("fetch_rss called without a url"); + return false; + } + + // if cache is disabled + if ( !MAGPIE_CACHE_ON ) { + // fetch file, and parse it + $resp = _fetch_remote_file( $url ); + if ( is_success( $resp->status ) ) { + return _response_to_rss( $resp ); + } + else { + do_error("Failed to fetch $url and cache is off"); + return false; + } + } + // else cache is ON + else { + // Flow + // 1. check cache + // 2. if there is a hit, make sure its fresh + // 3. if cached obj fails freshness check, fetch remote + // 4. if remote fails, return stale object, or error + + $cache = new RSSCache( MAGPIE_CACHE_DIR, MAGPIE_CACHE_AGE ); + + if (MAGPIE_DEBUG and $cache->ERROR) { + do_debug($cache->ERROR, E_USER_WARNING); + } + + + $cache_status = 0; // response of check_cache + $request_headers = array(); // HTTP headers to send with fetch + $rss = 0; // parsed RSS object + $errormsg = 0; // errors, if any + + // store parsed XML by desired output encoding + // as character munging happens at parse time + $cache_key = $url . MAGPIE_OUTPUT_ENCODING; + + if (!$cache->ERROR) { + // return cache HIT, MISS, or STALE + $cache_status = $cache->check_cache( $cache_key); + } + + // if object cached, and cache is fresh, return cached obj + if ( $cache_status == 'HIT' ) { + $rss = $cache->get( $cache_key ); + if ( isset($rss) and $rss ) { + $rss->from_cache = 1; + if ( MAGPIE_DEBUG > 1) { + do_debug("MagpieRSS: Cache HIT", E_USER_NOTICE); + } + return $rss; + } + } + + // else attempt a conditional get + + // setup headers + if ( $cache_status == 'STALE' ) { + $rss = $cache->get( $url ); + if ( $rss->etag and $rss->last_modified ) { + $request_headers['If-None-Match'] = $rss->etag; + $request_headers['If-Last-Modified'] = $rss->last_modified; + } + } + + $resp = _fetch_remote_file( $url, $request_headers ); + + if (isset($resp) and $resp) { + if ($resp->status == '304' ) { + // we have the most current copy + if ( MAGPIE_DEBUG > 1) { + do_debug("Got 304 for $url"); + } + // reset cache on 304 (at minutillo insistent prodding) + $cache->set($cache_key, $rss); + return $rss; + } + elseif ( is_success( $resp->status ) ) { + $rss = _response_to_rss( $resp ); + if ( $rss ) { + if (MAGPIE_DEBUG > 1) { + do_debug("Fetch successful"); + } + // add object to cache + $cache->set( $cache_key, $rss ); + return $rss; + } + } + else { + $errormsg = "Failed to fetch $url. "; + if ( $resp->status == '-100' ) { + $errormsg .= "(Request timed out after " . MAGPIE_FETCH_TIME_OUT . " seconds)"; + } + elseif ( $resp->error ) { + # compensate for Snoopy's annoying habbit to tacking + # on '\n' + $http_error = substr($resp->error, 0, -2); + $errormsg .= "(HTTP Error: $http_error)"; + } + else { + $errormsg .= "(HTTP Response: " . $resp->response_code .')'; + } + } + } + else { + $errormsg = "Unable to retrieve RSS file for unknown reasons."; + } + + // else fetch failed + + // attempt to return cached object + if ($rss) { + if ( MAGPIE_DEBUG ) { + do_debug("Returning STALE object for $url"); + } + return $rss; + } + + // else we totally failed + do_error( $errormsg ); + + return false; + + } // end if ( !MAGPIE_CACHE_ON ) { +} // end fetch_rss() + +/*=======================================================================*\ + Function: error + Purpose: set MAGPIE_ERROR, and trigger error +\*=======================================================================*/ +//Daryl Hawes note: had to rename this function from error to do_error for moodle +// due to conflict with existing moodle function name +function do_error ($errormsg, $lvl=E_USER_WARNING) { + global $MAGPIE_ERROR; + + // append PHP's error message if track_errors enabled + if ( isset($php_errormsg) ) { + $errormsg .= " ($php_errormsg)"; + } + if ( $errormsg ) { + $errormsg = "MagpieRSS: $errormsg"; + $MAGPIE_ERROR = $errormsg; + trigger_error( $errormsg, $lvl); + } +} + +//Daryl Hawes note: renamed this function from debug to do_debug for moodle +function do_debug ($debugmsg, $lvl=E_USER_NOTICE) { + trigger_error("MagpieRSS [debug] $debugmsg", $lvl); +} + +/*=======================================================================*\ + Function: magpie_error + Purpose: accessor for the magpie error variable +\*=======================================================================*/ +function magpie_error ($errormsg="") { + global $MAGPIE_ERROR; + + if ( isset($errormsg) and $errormsg ) { + $MAGPIE_ERROR = $errormsg; + } + + return $MAGPIE_ERROR; +} + +/*=======================================================================*\ + Function: _fetch_remote_file + Purpose: retrieve an arbitrary remote file + Input: url of the remote file + headers to send along with the request (optional) + Output: an HTTP response object (see Snoopy.class.inc) +\*=======================================================================*/ +function _fetch_remote_file ($url, $headers = "" ) { + // Snoopy is an HTTP client in PHP + $client = new Snoopy(); + $client->agent = MAGPIE_USER_AGENT; + $client->read_timeout = MAGPIE_FETCH_TIME_OUT; + $client->use_gzip = MAGPIE_USE_GZIP; + if (is_array($headers) ) { + $client->rawheaders = $headers; + } + + @$client->fetch($url); + return $client; + +} + +/*=======================================================================*\ + Function: _response_to_rss + Purpose: parse an HTTP response object into an RSS object + Input: an HTTP response object (see Snoopy) + Output: parsed RSS object (see rss_parse) +\*=======================================================================*/ +function _response_to_rss ($resp) { + $rss = new MagpieRSS( $resp->results, MAGPIE_OUTPUT_ENCODING, MAGPIE_INPUT_ENCODING, MAGPIE_DETECT_ENCODING ); + + // if RSS parsed successfully + if ( $rss and !$rss->ERROR) { + + // find Etag, and Last-Modified + foreach($resp->headers as $h) { + // 2003-03-02 - Nicola Asuni (www.tecnick.com) - fixed bug "Undefined offset: 1" + if (strpos($h, ": ")) { + list($field, $val) = explode(": ", $h, 2); + } + else { + $field = $h; + $val = ""; + } + + if ( $field == 'ETag' ) { + $rss->etag = $val; + } + + if ( $field == 'Last-Modified' ) { + $rss->last_modified = $val; + } + } + + return $rss; + } // else construct error message + else { + $errormsg = "Failed to parse RSS file."; + + if ($rss) { + $errormsg .= " (" . $rss->ERROR . ")"; + } + do_error($errormsg); + + return false; + } // end if ($rss and !$rss->error) +} + +/*=======================================================================*\ + Function: init + Purpose: setup constants with default values + check for user overrides +\*=======================================================================*/ +function init () { + if ( defined('MAGPIE_INITALIZED') ) { + return; + } + else { + define('MAGPIE_INITALIZED', true); + } + + if ( !defined('MAGPIE_CACHE_ON') ) { + define('MAGPIE_CACHE_ON', true); + } + + if ( !defined('MAGPIE_CACHE_DIR') ) { + define('MAGPIE_CACHE_DIR', './cache'); + } + + if ( !defined('MAGPIE_CACHE_AGE') ) { + define('MAGPIE_CACHE_AGE', 60*60); // one hour + } + + if ( !defined('MAGPIE_CACHE_FRESH_ONLY') ) { + define('MAGPIE_CACHE_FRESH_ONLY', false); + } + + if ( !defined('MAGPIE_OUTPUT_ENCODING') ) { + define('MAGPIE_OUTPUT_ENCODING', 'ISO-8859-1'); + } + + if ( !defined('MAGPIE_INPUT_ENCODING') ) { + define('MAGPIE_INPUT_ENCODING', null); + } + + if ( !defined('MAGPIE_DETECT_ENCODING') ) { + define('MAGPIE_DETECT_ENCODING', true); + } + + if ( !defined('MAGPIE_DEBUG') ) { + define('MAGPIE_DEBUG', 0); + } + + if ( !defined('MAGPIE_USER_AGENT') ) { + $ua = 'MagpieRSS/'. MAGPIE_VERSION . ' (+http://magpierss.sf.net'; + + if ( MAGPIE_CACHE_ON ) { + $ua = $ua . ')'; + } + else { + $ua = $ua . '; No cache)'; + } + + define('MAGPIE_USER_AGENT', $ua); + } + + if ( !defined('MAGPIE_FETCH_TIME_OUT') ) { + define('MAGPIE_FETCH_TIME_OUT', 5); // 5 second timeout + } + + // use gzip encoding to fetch rss files if supported? + if ( !defined('MAGPIE_USE_GZIP') ) { + define('MAGPIE_USE_GZIP', true); + } +} + +// NOTE: the following code should really be in Snoopy, or at least +// somewhere other then rss_fetch! + +/*=======================================================================*\ + HTTP STATUS CODE PREDICATES + These functions attempt to classify an HTTP status code + based on RFC 2616 and RFC 2518. + + All of them take an HTTP status code as input, and return true or false + + All this code is adapted from LWP's HTTP::Status. +\*=======================================================================*/ + + +/*=======================================================================*\ + Function: is_info + Purpose: return true if Informational status code +\*=======================================================================*/ +function is_info ($sc) { + return $sc >= 100 && $sc < 200; +} + +/*=======================================================================*\ + Function: is_success + Purpose: return true if Successful status code +\*=======================================================================*/ +function is_success ($sc) { + return $sc >= 200 && $sc < 300; +} + +/*=======================================================================*\ + Function: is_redirect + Purpose: return true if Redirection status code +\*=======================================================================*/ +function is_redirect ($sc) { + return $sc >= 300 && $sc < 400; +} + +/*=======================================================================*\ + Function: is_error + Purpose: return true if Error status code +\*=======================================================================*/ +function is_error ($sc) { + return $sc >= 400 && $sc < 600; +} + +/*=======================================================================*\ + Function: is_client_error + Purpose: return true if Error status code, and its a client error +\*=======================================================================*/ +function is_client_error ($sc) { + return $sc >= 400 && $sc < 500; +} + +/*=======================================================================*\ + Function: is_client_error + Purpose: return true if Error status code, and its a server error +\*=======================================================================*/ +function is_server_error ($sc) { + return $sc >= 500 && $sc < 600; +} + +?> diff --git a/rss/magpie/rss_parse.inc b/rss/magpie/rss_parse.inc new file mode 100644 index 00000000000..af4020ebfb4 --- /dev/null +++ b/rss/magpie/rss_parse.inc @@ -0,0 +1,589 @@ + +* @version 0.7a +* @license GPL +* +*/ + +define('RSS', 'RSS'); +define('ATOM', 'Atom'); + +require_once (MAGPIE_DIR . 'rss_utils.inc'); + +/** +* Hybrid parser, and object, takes RSS as a string and returns a simple object. +* +* see: rss_fetch.inc for a simpler interface with integrated caching support +* +*/ +class MagpieRSS { + var $parser; + + var $current_item = array(); // item currently being parsed + var $items = array(); // collection of parsed items + var $channel = array(); // hash of channel fields + var $textinput = array(); + var $image = array(); + var $feed_type; + var $feed_version; + var $encoding = ''; // output encoding of parsed rss + + var $_source_encoding = ''; // only set if we have to parse xml prolog + + var $ERROR = ""; + var $WARNING = ""; + + // define some constants + + var $_CONTENT_CONSTRUCTS = array('content', 'summary', 'info', 'title', 'tagline', 'copyright'); + var $_KNOWN_ENCODINGS = array('UTF-8', 'US-ASCII', 'ISO-8859-1'); + + // parser variables, useless if you're not a parser, treat as private + var $stack = array(); // parser stack + var $inchannel = false; + var $initem = false; + var $incontent = false; // if in Atom field + var $intextinput = false; + var $inimage = false; + var $current_field = ''; + var $current_namespace = false; + + + /** + * Set up XML parser, parse source, and return populated RSS object.. + * + * @param string $source string containing the RSS to be parsed + * + * NOTE: Probably a good idea to leave the encoding options alone unless + * you know what you're doing as PHP's character set support is + * a little weird. + * + * NOTE: A lot of this is unnecessary but harmless with PHP5 + * + * + * @param string $output_encoding output the parsed RSS in this character + * set defaults to ISO-8859-1 as this is PHP's + * default. + * + * NOTE: might be changed to UTF-8 in future + * versions. + * + * @param string $input_encoding the character set of the incoming RSS source. + * Leave blank and Magpie will try to figure it + * out. + * + * + * @param bool $detect_encoding if false Magpie won't attempt to detect + * source encoding. (caveat emptor) + * + */ + function MagpieRSS ($source, $output_encoding='ISO-8859-1', + $input_encoding=null, $detect_encoding=true) + { + # if PHP xml isn't compiled in, die + # + if (!function_exists('xml_parser_create')) { + $this->error( "Failed to load PHP's XML Extension. " . + "http://www.php.net/manual/en/ref.xml.php", + E_USER_ERROR ); + } + + list($parser, $source) = $this->create_parser($source, + $output_encoding, $input_encoding, $detect_encoding); + + + if (!is_resource($parser)) { + $this->error( "Failed to create an instance of PHP's XML parser. " . + "http://www.php.net/manual/en/ref.xml.php", + E_USER_ERROR ); + } + + + $this->parser = $parser; + + # pass in parser, and a reference to this object + # setup handlers + # + xml_set_object( $this->parser, $this ); + xml_set_element_handler($this->parser, + 'feed_start_element', 'feed_end_element' ); + + xml_set_character_data_handler( $this->parser, 'feed_cdata' ); + + $status = xml_parse( $this->parser, $source ); + + if (! $status ) { + $errorcode = xml_get_error_code( $this->parser ); + if ( $errorcode != XML_ERROR_NONE ) { + $xml_error = xml_error_string( $errorcode ); + $error_line = xml_get_current_line_number($this->parser); + $error_col = xml_get_current_column_number($this->parser); + $errormsg = "$xml_error at line $error_line, column $error_col"; + + $this->error( $errormsg ); + } + } + + xml_parser_free( $this->parser ); + + $this->normalize(); + } + + function feed_start_element($p, $element, &$attrs) { + $el = $element = strtolower($element); + $attrs = array_change_key_case($attrs, CASE_LOWER); + + // check for a namespace, and split if found + $ns = false; + if ( strpos( $element, ':' ) ) { + list($ns, $el) = split( ':', $element, 2); + } + if ( $ns and $ns != 'rdf' ) { + $this->current_namespace = $ns; + } + + # if feed type isn't set, then this is first element of feed + # identify feed from root element + # + if (!isset($this->feed_type) ) { + if ( $el == 'rdf' ) { + $this->feed_type = RSS; + $this->feed_version = '1.0'; + } + elseif ( $el == 'rss' ) { + $this->feed_type = RSS; + $this->feed_version = $attrs['version']; + } + elseif ( $el == 'feed' ) { + $this->feed_type = ATOM; + $this->feed_version = $attrs['version']; + $this->inchannel = true; + } + return; + } + + if ( $el == 'channel' ) + { + $this->inchannel = true; + } + elseif ($el == 'item' or $el == 'entry' ) + { + $this->initem = true; + if ( isset($attrs['rdf:about']) ) { + $this->current_item['about'] = $attrs['rdf:about']; + } + } + + // if we're in the default namespace of an RSS feed, + // record textinput or image fields + elseif ( + $this->feed_type == RSS and + $this->current_namespace == '' and + $el == 'textinput' ) + { + $this->intextinput = true; + } + + elseif ( + $this->feed_type == RSS and + $this->current_namespace == '' and + $el == 'image' ) + { + $this->inimage = true; + } + + # handle atom content constructs + elseif ( $this->feed_type == ATOM and in_array($el, $this->_CONTENT_CONSTRUCTS) ) + { + // avoid clashing w/ RSS mod_content + if ($el == 'content' ) { + $el = 'atom_content'; + } + + $this->incontent = $el; + + + } + + // if inside an Atom content construct (e.g. content or summary) field treat tags as text + elseif ($this->feed_type == ATOM and $this->incontent ) + { + // if tags are inlined, then flatten + $attrs_str = join(' ', + array_map('map_attrs', + array_keys($attrs), + array_values($attrs) ) ); + + $this->append_content( "<$element $attrs_str>" ); + + array_unshift( $this->stack, $el ); + } + + // Atom support many links per containging element. + // Magpie treats link elements of type rel='alternate' + // as being equivalent to RSS's simple link element. + // + elseif ($this->feed_type == ATOM and $el == 'link' ) + { + if ( isset($attrs['rel']) and $attrs['rel'] == 'alternate' ) + { + $link_el = 'link'; + } + else { + $link_el = 'link_' . $attrs['rel']; + } + + $this->append($link_el, $attrs['href']); + } + // set stack[0] to current element + else { + array_unshift($this->stack, $el); + } + } + + + + function feed_cdata ($p, $text) { + + if ($this->feed_type == ATOM and $this->incontent) + { + $this->append_content( $text ); + } + else { + $current_el = join('_', array_reverse($this->stack)); + $this->append($current_el, $text); + } + } + + function feed_end_element ($p, $el) { + $el = strtolower($el); + + if ( $el == 'item' or $el == 'entry' ) + { + $this->items[] = $this->current_item; + $this->current_item = array(); + $this->initem = false; + } + elseif ($this->feed_type == RSS and $this->current_namespace == '' and $el == 'textinput' ) + { + $this->intextinput = false; + } + elseif ($this->feed_type == RSS and $this->current_namespace == '' and $el == 'image' ) + { + $this->inimage = false; + } + elseif ($this->feed_type == ATOM and in_array($el, $this->_CONTENT_CONSTRUCTS) ) + { + $this->incontent = false; + } + elseif ($el == 'channel' or $el == 'feed' ) + { + $this->inchannel = false; + } + elseif ($this->feed_type == ATOM and $this->incontent ) { + // balance tags properly + // note: i don't think this is actually neccessary + if ( $this->stack[0] == $el ) + { + $this->append_content(""); + } + else { + $this->append_content("<$el />"); + } + + array_shift( $this->stack ); + } + else { + array_shift( $this->stack ); + } + + $this->current_namespace = false; + } + + function concat (&$str1, $str2="") { + if (!isset($str1) ) { + $str1=""; + } + $str1 .= $str2; + } + + + + function append_content($text) { + if ( $this->initem ) { + $this->concat( $this->current_item[ $this->incontent ], $text ); + } + elseif ( $this->inchannel ) { + $this->concat( $this->channel[ $this->incontent ], $text ); + } + } + + // smart append - field and namespace aware + function append($el, $text) { + if (!$el) { + return; + } + if ( $this->current_namespace ) + { + if ( $this->initem ) { + $this->concat( + $this->current_item[ $this->current_namespace ][ $el ], $text); + } + elseif ($this->inchannel) { + $this->concat( + $this->channel[ $this->current_namespace][ $el ], $text ); + } + elseif ($this->intextinput) { + $this->concat( + $this->textinput[ $this->current_namespace][ $el ], $text ); + } + elseif ($this->inimage) { + $this->concat( + $this->image[ $this->current_namespace ][ $el ], $text ); + } + } + else { + if ( $this->initem ) { + $this->concat( + $this->current_item[ $el ], $text); + } + elseif ($this->intextinput) { + $this->concat( + $this->textinput[ $el ], $text ); + } + elseif ($this->inimage) { + $this->concat( + $this->image[ $el ], $text ); + } + elseif ($this->inchannel) { + $this->concat( + $this->channel[ $el ], $text ); + } + + } + } + + function normalize () { + // if atom populate rss fields + if ( $this->is_atom() ) { + $this->channel['description'] = $this->channel['tagline']; + for ( $i = 0; $i < count($this->items); $i++) { + $item = $this->items[$i]; + if ( isset($item['summary']) ) + $item['description'] = $item['summary']; + if ( isset($item['atom_content'])) + $item['content']['encoded'] = $item['atom_content']; + + $atom_date = (isset($item['issued']) ) ? $item['issued'] : $item['modified']; + if ( $atom_date ) { + $epoch = @parse_w3cdtf($item['modified']); + if ($epoch and $epoch > 0) { + $item['date_timestamp'] = $epoch; + } + } + + $this->items[$i] = $item; + } + } + elseif ( $this->is_rss() ) { + $this->channel['tagline'] = $this->channel['description']; + for ( $i = 0; $i < count($this->items); $i++) { + $item = $this->items[$i]; + if ( isset($item['description'])) + $item['summary'] = $item['description']; + if ( isset($item['content']['encoded'] ) ) + $item['atom_content'] = $item['content']['encoded']; + + if ( $this->is_rss() == '1.0' and isset($item['dc']['date']) ) { + $epoch = @parse_w3cdtf($item['dc']['date']); + if ($epoch and $epoch > 0) { + $item['date_timestamp'] = $epoch; + } + } + elseif ( isset($item['pubdate']) ) { + $epoch = @strtotime($item['pubdate']); + if ($epoch > 0) { + $item['date_timestamp'] = $epoch; + } + } + + $this->items[$i] = $item; + } + } + } + + + function is_rss () { + if ( $this->feed_type == RSS ) { + return $this->feed_version; + } + else { + return false; + } + } + + function is_atom() { + if ( $this->feed_type == ATOM ) { + return $this->feed_version; + } + else { + return false; + } + } + + /** + * return XML parser, and possibly re-encoded source + * + */ + function create_parser($source, $out_enc, $in_enc, $detect) { + if ( substr(phpversion(),0,1) == 5) { + $parser = $this->php5_create_parser($in_enc, $detect); + } + else { + list($parser, $source) = $this->php4_create_parser($source, $in_enc, $detect); + } + if ($out_enc) { + $this->encoding = $out_enc; + xml_parser_set_option($parser, XML_OPTION_TARGET_ENCODING, $out_enc); + } + + return array($parser, $source); + } + + /** + * Instantiate an XML parser under PHP5 + * + * PHP5 will do a fine job of detecting input encoding + * if passed an empty string as the encoding. + * + * All hail libxml2! + * + */ + function php5_create_parser($in_enc, $detect) { + // by default php5 does a fine job of detecting input encodings + if(!$detect && $in_enc) { + return xml_parser_create($in_enc); + } + else { + return xml_parser_create(''); + } + } + + /** + * Instaniate an XML parser under PHP4 + * + * Unfortunately PHP4's support for character encodings + * and especially XML and character encodings sucks. As + * long as the documents you parse only contain characters + * from the ISO-8859-1 character set (a superset of ASCII, + * and a subset of UTF-8) you're fine. However once you + * step out of that comfy little world things get mad, bad, + * and dangerous to know. + * + * The following code is based on SJM's work with FoF + * @see http://minutillo.com/steve/weblog/2004/6/17/php-xml-and-character-encodings-a-tale-of-sadness-rage-and-data-loss + * + */ + function php4_create_parser($source, $in_enc, $detect) { + if ( !$detect ) { + return array(xml_parser_create($in_enc), $source); + } + + if (!$in_enc) { + if (preg_match('//m', $source, $m)) { + $in_enc = strtoupper($m[1]); + $this->source_encoding = $in_enc; + } + else { + $in_enc = 'UTF-8'; + } + } + + if ($this->known_encoding($in_enc)) { + return array(xml_parser_create($in_enc), $source); + } + + // the dectected encoding is not one of the simple encodings PHP knows + + // attempt to use the iconv extension to + // cast the XML to a known encoding + // @see http://php.net/iconv + + if (function_exists('iconv')) { + $encoded_source = iconv($in_enc,'UTF-8', $source); + if ($encoded_source) { + return array(xml_parser_create('UTF-8'), $encoded_source); + } + } + + // iconv didn't work, try mb_convert_encoding + // @see http://php.net/mbstring + if(function_exists('mb_convert_encoding')) { + $encoded_source = mb_convert_encoding($source, 'UTF-8', $in_enc ); + if ($encoded_source) { + return array(xml_parser_create('UTF-8'), $encoded_source); + } + } + + // else + $this->error("Feed is in an unsupported character encoding. ($in_enc) " . + "You may see strange artifacts, and mangled characters.", + E_USER_NOTICE); + + return array(xml_parser_create(), $source); + } + + function known_encoding($enc) { + $enc = strtoupper($enc); + if ( in_array($enc, $this->_KNOWN_ENCODINGS) ) { + return $enc; + } + else { + return false; + } + } + + function error ($errormsg, $lvl=E_USER_WARNING) { + // append PHP's error message if track_errors enabled + if ( $php_errormsg ) { + $errormsg .= " ($php_errormsg)"; + } + if ( MAGPIE_DEBUG ) { + trigger_error( $errormsg, $lvl); + } + else { + error_log( $errormsg, 0); + } + + $notices = E_USER_NOTICE|E_NOTICE; + if ( $lvl&$notices ) { + $this->WARNING = $errormsg; + } else { + $this->ERROR = $errormsg; + } + } + + +} // end class RSS + +function map_attrs($k, $v) { + return "$k=\"$v\""; +} + + +?> diff --git a/rss/magpie/rss_utils.inc b/rss/magpie/rss_utils.inc new file mode 100644 index 00000000000..2a29e72a969 --- /dev/null +++ b/rss/magpie/rss_utils.inc @@ -0,0 +1,67 @@ + + * Version: 0.51 + * License: GPL + * + * The lastest version of MagpieRSS can be obtained from: + * http://magpierss.sourceforge.net + * + * For questions, help, comments, discussion, etc., please join the + * Magpie mailing list: + * magpierss-general@lists.sourceforge.net + */ + + +/*======================================================================*\ + Function: parse_w3cdtf + Purpose: parse a W3CDTF date into unix epoch + + NOTE: http://www.w3.org/TR/NOTE-datetime +\*======================================================================*/ + +function parse_w3cdtf ( $date_str ) { + + # regex to match wc3dtf + $pat = "/(\d{4})-(\d{2})-(\d{2})T(\d{2}):(\d{2})(:(\d{2}))?(?:([-+])(\d{2}):?(\d{2})|(Z))?/"; + + if ( preg_match( $pat, $date_str, $match ) ) { + list( $year, $month, $day, $hours, $minutes, $seconds) = + array( $match[1], $match[2], $match[3], $match[4], $match[5], $match[6]); + + # calc epoch for current date assuming GMT + $epoch = gmmktime( $hours, $minutes, $seconds, $month, $day, $year); + + $offset = 0; + if ( $match[10] == 'Z' ) { + # zulu time, aka GMT + } + else { + list( $tz_mod, $tz_hour, $tz_min ) = + array( $match[8], $match[9], $match[10]); + + # zero out the variables + if ( ! $tz_hour ) { $tz_hour = 0; } + if ( ! $tz_min ) { $tz_min = 0; } + + $offset_secs = (($tz_hour*60)+$tz_min)*60; + + # is timezone ahead of GMT? then subtract offset + # + if ( $tz_mod == '+' ) { + $offset_secs = $offset_secs * -1; + } + + $offset = $offset_secs; + } + $epoch = $epoch + $offset; + return $epoch; + } + else { + return -1; + } +} + +?>