From 3ec62cf95ab4436923d4c665fad7aef226cbb822 Mon Sep 17 00:00:00 2001 From: Maryana Rozhankivska Date: Thu, 22 May 2014 17:16:38 +0300 Subject: [PATCH] update to 3.2 version of full-text-rss, issue #694 --- inc/3rdparty/config.php | 104 +- .../content-extractor/ContentExtractor.php | 1455 ++++++----- .../content-extractor/SiteConfig.php | 681 ++--- .../libraries/feedwriter/FeedItem.php | 100 +- .../libraries/feedwriter/FeedWriter.php | 17 +- inc/3rdparty/libraries/html5/TreeBuilder.php | 13 +- .../libraries/humble-http-agent/CookieJar.php | 807 +++--- .../humble-http-agent/HumbleHttpAgent.php | 1589 ++++++------ .../SimplePie_HumbleHttpAgent.php | 157 +- .../language-detect/LanguageDetect.php | 984 +++---- .../libraries/readability/Readability.php | 2274 ++++++++--------- inc/3rdparty/makefulltextfeed.php | 361 +-- inc/3rdparty/makefulltextfeedHelpers.php | 42 +- inc/3rdparty/site_config/index.php | 5 +- inc/3rdparty/site_config/standard/version.txt | 2 +- 15 files changed, 4417 insertions(+), 4174 deletions(-) mode change 100644 => 100755 inc/3rdparty/libraries/feedwriter/FeedItem.php diff --git a/inc/3rdparty/config.php b/inc/3rdparty/config.php index e618117..ec680d8 100755 --- a/inc/3rdparty/config.php +++ b/inc/3rdparty/config.php @@ -19,7 +19,7 @@ if (!isset($options)) $options = new stdClass(); // Enable service // ---------------------- // Set this to false if you want to disable the service. -// If set to false, no feed is produced and users will +// If set to false, no feed is produced and users will // be told that the service is disabled. $options->enabled = true; @@ -43,10 +43,64 @@ $options->default_entries = 5; // ---------------------- // The maximum number of feed items to process when no access key is supplied. // This limits the user-supplied &max=x value. For example, if the user -// asks for 20 items to be processed (&max=20), if max_entries is set to +// asks for 20 items to be processed (&max=20), if max_entries is set to // 10, only 10 will be processed. $options->max_entries = 10; +// Full content +// ---------------------- +// By default Full-Text RSS includes the extracted content in the output. +// You can exclude this from the output by passing '&content=0' in the querystring. +// +// Possible values... +// Always include: true +// Never include: false +// Include unless user overrides (&content=0): 'user' (default) +// +// Note: currently this does not disable full content extraction. It simply omits it +// from the output. +$options->content = 'user'; + +// Excerpts +// ---------------------- +// By default Full-Text RSS does not include excerpts in the output. +// You can enable this by passing '&summary=1' in the querystring. +// This will include a plain text excerpt from the extracted content. +// +// Possible values... +// Always include: true (recommended for new users) +// Never include: false +// Don't include unless user overrides (&summary=1): 'user' (default) +// +// Important: if both content and excerpts are requested, the excerpt will be +// placed in the description element and the full content inside content:encoded. +// If excerpts are not requested, the full content will go inside the description element. +// +// Why are we not returning both excerpts and content by default? +// Mainly for backward compatibility. +// Excerpts should appear in the feed item's description element. Previous versions +// of Full-Text RSS did not return excerpts, so the description element was always +// used for the full content (as recommended by the RSS advisory). When returning both, +// we need somewhere else to place the content (content:encoded). +// Having both enabled should not create any problems for news readers, but it may create +// problems for developers upgrading from one of our earlier versions who may now find +// their applications are returning excerpts instead of the full content they were +// expecting. To avoid such surprises for users who are upgrading Full-Text RSS, +// excerpts must be explicitly requested in the querystring by default. +// +// Why not use a different element name for excerpts? +// According to the RSS advisory: +// "Publishers who employ summaries should store the summary in description and +// the full content in content:encoded, ordering description first within the item. +// On items with no summary, the full content should be stored in description." +// See: http://www.rssboard.org/rss-profile#namespace-elements-content-encoded +// +// For more consistent element naming, we recommend new users set this option to true. +// The full content can still be excluded via the querystring, but the element names +// will not change: when $options->summary = true, the description element will always +// be reserved for the excerpt and content:encoded always for full content. +$options->summary = 'user'; + // Rewrite relative URLs // ---------------------- // With this enabled relative URLs found in the extracted content @@ -67,7 +121,7 @@ $options->exclude_items_on_fail = 'user'; // Enable multi-page support // ------------------------- // If enabled, we will try to follow next page links on multi-page articles. -// Currently this only happens for sites where next_page_link has been defined +// Currently this only happens for sites where next_page_link has been defined // in a site config file. $options->multipage = true; @@ -125,10 +179,10 @@ $options->detect_language = 1; // Registration key // --------------- -// The registration key is optional. It is not required to use Full-Text RSS, -// and does not affect the normal operation of Full-Text RSS. It is currently -// only used on admin pages which help you update site patterns with the -// latest version offered by FiveFilters.org. For these admin-related +// The registration key is optional. It is not required to use Full-Text RSS, +// and does not affect the normal operation of Full-Text RSS. It is currently +// only used on admin pages which help you update site patterns with the +// latest version offered by FiveFilters.org. For these admin-related // tasks to complete, we will require a valid registration key. // If you would like one, you can purchase the latest version of Full-Text RSS // at http://fivefilters.org/content-only/ @@ -144,12 +198,12 @@ $options->registration_key = ''; // ---------------------- // Certain pages/actions, e.g. updating site patterns with our online tool, will require admin credentials. // To use these pages, enter a password here and you'll be prompted for it when you try to access those pages. -// If no password or username is set, pages requiring admin privelages will be inaccessible. +// If no password or username is set, pages requiring admin privelages will be inaccessible. // The default username is 'admin'. // If overriding with an environment variable, separate username and password with a colon, e.g.: // ftr_admin_credentials: admin:my-secret-password // Example: $options->admin_credentials = array('username'=>'admin', 'password'=>'my-secret-password'); -$options->admin_credentials = array('username'=>'admin', 'password'=>'admin'); +$options->admin_credentials = array('username'=>'admin', 'password'=>''); // URLs to allow // ---------------------- @@ -178,12 +232,12 @@ $options->key_required = false; // ---------------------- // By default, when processing feeds, we assume item titles in the feed // have not been truncated. So after processing web pages, the extracted titles -// are not used in the generated feed. If you prefer to have extracted titles in -// the feed you can either set this to false, in which case we will always favour -// extracted titles. Alternatively, if set to 'user' (default) we'll use the +// are not used in the generated feed. If you prefer to have extracted titles in +// the feed you can either set this to false, in which case we will always favour +// extracted titles. Alternatively, if set to 'user' (default) we'll use the // extracted title if you pass '&use_extracted_title' in the querystring. // Possible values: -// * Favour feed titles: true +// * Favour feed titles: true // * Favour extracted titles: false // * Favour feed titles with user override: 'user' (default) // Note: this has no effect when the input URL is to a web page - in these cases @@ -192,17 +246,17 @@ $options->favour_feed_titles = 'user'; // Access keys (password protected access) // ------------------------------------ -// NOTE: You do not need an API key from fivefilters.org to run your own +// NOTE: You do not need an API key from fivefilters.org to run your own // copy of the code. This is here if you'd like to restrict access to // _your_ copy. // Keys let you group users - those with a key and those without - and // restrict access to the service to those without a key. // If you want everyone to access the service in the same way, you can // leave the array below empty and ignore the access key options further down. -// The options further down let you control how the service should behave +// The options further down let you control how the service should behave // in each mode. -// Note: Explicitly including the index number (1 and 2 in the examples below) -// is highly recommended (when generating feeds, we encode the key and +// Note: Explicitly including the index number (1 and 2 in the examples below) +// is highly recommended (when generating feeds, we encode the key and // refer to it by index number and hash). $options->api_keys = array(); // Example: @@ -232,13 +286,13 @@ $options->max_entries_with_key = 10; // filter the resulting HTML for XSS attacks, making it redundant for // Full-Text RSS do the same. Similarly with frameworks/CMS which display // feed content - the content should be treated like any other user-submitted content. -// +// // If you are writing an application yourself which is processing feeds generated by // Full-Text RSS, you can either filter the HTML yourself to remove potential XSS attacks // or enable this option. This might be useful if you are processing our generated // feeds with JavaScript on the client side - although there's client side xss // filtering available too, e.g. https://code.google.com/p/google-caja/wiki/JsHtmlSanitizer -// +// // If enabled, we'll pass retrieved HTML content through htmLawed with // safe flag on and style attributes denied, see // http://www.bioinformatics.org/phplabware/internal_utilities/htmLawed/htmLawed_README.htm#s3.6 @@ -253,8 +307,8 @@ $options->xss_filter = 'user'; // Allowed parsers // ---------------------- // Full-Text RSS attempts to use PHP's libxml extension to process HTML. -// While fast, on some sites it may not always produce good results. -// For these sites, you can specify an alternative HTML parser: +// While fast, on some sites it may not always produce good results. +// For these sites, you can specify an alternative HTML parser: // parser: html5lib // The html5lib parser is bundled with Full-Text RSS. // see http://code.google.com/p/html5lib/ @@ -273,7 +327,7 @@ $options->cors = false; // Use APC user cache? // ---------------------- -// If enabled we will store site config files (when requested +// If enabled we will store site config files (when requested // for the first time) in APC's user cache. Keys prefixed with 'sc.' // This improves performance by reducing disk access. // Note: this has no effect if APC is unavailable on your server. @@ -346,7 +400,7 @@ $options->rewrite_url = array( // Valid actions: // * 'exclude' - exclude this item from the result // * 'link' - create HTML link to the item -$options->content_type_exc = array( +$options->content_type_exc = array( 'application/pdf' => array('action'=>'link', 'name'=>'PDF'), 'image' => array('action'=>'link', 'name'=>'Image'), 'audio' => array('action'=>'link', 'name'=>'Audio'), @@ -375,13 +429,13 @@ $options->cache_cleanup = 100; /// DO NOT CHANGE ANYTHING BELOW THIS /////////// ///////////////////////////////////////////////// -if (!defined('_FF_FTR_VERSION')) define('_FF_FTR_VERSION', '3.1'); +if (!defined('_FF_FTR_VERSION')) define('_FF_FTR_VERSION', '3.2'); if (basename(__FILE__) == 'config.php') { if (file_exists(dirname(__FILE__).'/custom_config.php')) { require_once dirname(__FILE__).'/custom_config.php'; } - + // check for environment variables - often used on cloud platforms // environment variables should be prefixed with 'ftr_', e.g. // ftr_max_entries: 1 diff --git a/inc/3rdparty/libraries/content-extractor/ContentExtractor.php b/inc/3rdparty/libraries/content-extractor/ContentExtractor.php index ddd33bb..21e693e 100644 --- a/inc/3rdparty/libraries/content-extractor/ContentExtractor.php +++ b/inc/3rdparty/libraries/content-extractor/ContentExtractor.php @@ -1,728 +1,727 @@ - true, - 'output-xhtml' => true, - 'logical-emphasis' => true, - 'show-body-only' => false, - 'new-blocklevel-tags' => 'article, aside, footer, header, hgroup, menu, nav, section, details, datagrid', - 'new-inline-tags' => 'mark, time, meter, progress, data', - 'wrap' => 0, - 'drop-empty-paras' => true, - 'drop-proprietary-attributes' => false, - 'enclose-text' => true, - 'enclose-block-text' => true, - 'merge-divs' => true, - 'merge-spans' => true, - 'char-encoding' => 'utf8', - 'hide-comments' => true - ); - protected $html; - protected $config; - protected $title; - protected $author = array(); - protected $language; - protected $date; - protected $body; - protected $success = false; - protected $nextPageUrl; - public $allowedParsers = array('libxml', 'html5lib'); - public $fingerprints = array(); - public $readability; - public $debug = false; - public $debugVerbose = false; - - function __construct($path, $fallback=null) { - SiteConfig::set_config_path($path, $fallback); - } - - protected function debug($msg) { - if ($this->debug) { - $mem = round(memory_get_usage()/1024, 2); - $memPeak = round(memory_get_peak_usage()/1024, 2); - echo '* ',$msg; - if ($this->debugVerbose) echo ' - mem used: ',$mem," (peak: $memPeak)"; - echo "\n"; - ob_flush(); - flush(); - } - } - - public function reset() { - $this->html = null; - $this->readability = null; - $this->config = null; - $this->title = null; - $this->body = null; - $this->author = array(); - $this->language = null; - $this->date = null; - $this->nextPageUrl = null; - $this->success = false; - } - - public function findHostUsingFingerprints($html) { - $this->debug('Checking fingerprints...'); - $head = substr($html, 0, 8000); - foreach ($this->fingerprints as $_fp => $_fphost) { - $lookin = 'html'; - if (is_array($_fphost)) { - if (isset($_fphost['head']) && $_fphost['head']) { - $lookin = 'head'; - } - $_fphost = $_fphost['hostname']; - } - if (strpos($$lookin, $_fp) !== false) { - $this->debug("Found match: $_fphost"); - return $_fphost; - } - } - $this->debug('No fingerprint matches'); - return false; - } - - // returns SiteConfig instance (joined in order: exact match, wildcard, fingerprint, global, default) - public function buildSiteConfig($url, $html='', $add_to_cache=true) { - // extract host name - $host = @parse_url($url, PHP_URL_HOST); - $host = strtolower($host); - if (substr($host, 0, 4) == 'www.') $host = substr($host, 4); - // is merged version already cached? - if (SiteConfig::is_cached("$host.merged")) { - $this->debug("Returning cached and merged site config for $host"); - return SiteConfig::build("$host.merged"); - } - // let's build from site_config/custom/ and standard/ - $config = SiteConfig::build($host); - if ($add_to_cache && $config && !SiteConfig::is_cached("$host")) { - SiteConfig::add_to_cache($host, $config); - } - // if no match, use defaults - if (!$config) $config = new SiteConfig(); - // load fingerprint config? - if ($config->autodetect_on_failure()) { - // check HTML for fingerprints - if (!empty($this->fingerprints) && ($_fphost = $this->findHostUsingFingerprints($html))) { - if ($config_fingerprint = SiteConfig::build($_fphost)) { - $this->debug("Appending site config settings from $_fphost (fingerprint match)"); - $config->append($config_fingerprint); - if ($add_to_cache && !SiteConfig::is_cached($_fphost)) { - //$config_fingerprint->cache_in_apc = true; - SiteConfig::add_to_cache($_fphost, $config_fingerprint); - } - } - } - } - // load global config? - if ($config->autodetect_on_failure()) { - if ($config_global = SiteConfig::build('global', true)) { - $this->debug('Appending site config settings from global.txt'); - $config->append($config_global); - if ($add_to_cache && !SiteConfig::is_cached('global')) { - //$config_global->cache_in_apc = true; - SiteConfig::add_to_cache('global', $config_global); - } - } - } - // store copy of merged config - if ($add_to_cache) { - // do not store in APC if wildcard match - $use_apc = ($host == $config->cache_key); - $config->cache_key = null; - SiteConfig::add_to_cache("$host.merged", $config, $use_apc); - } - return $config; - } - - // returns true on success, false on failure - // $smart_tidy indicates that if tidy is used and no results are produced, we will - // try again without it. Tidy helps us deal with PHP's patchy HTML parsing most of the time - // but it has problems of its own which we try to avoid with this option. - public function process($html, $url, $smart_tidy=true) { - $this->reset(); - $this->config = $this->buildSiteConfig($url, $html); - - // do string replacements - if (!empty($this->config->find_string)) { - if (count($this->config->find_string) == count($this->config->replace_string)) { - $html = str_replace($this->config->find_string, $this->config->replace_string, $html, $_count); - $this->debug("Strings replaced: $_count (find_string and/or replace_string)"); - } else { - $this->debug('Skipped string replacement - incorrect number of find-replace strings in site config'); - } - unset($_count); - } - - // use tidy (if it exists)? - // This fixes problems with some sites which would otherwise - // trouble DOMDocument's HTML parsing. (Although sometimes it - // makes matters worse, which is why you can override it in site config files.) - $tidied = false; - if ($this->config->tidy() && function_exists('tidy_parse_string') && $smart_tidy) { - $this->debug('Using Tidy'); - $tidy = tidy_parse_string($html, self::$tidy_config, 'UTF8'); - if (tidy_clean_repair($tidy)) { - $original_html = $html; - $tidied = true; - $html = $tidy->value; - } - unset($tidy); - } - - // load and parse html - $_parser = $this->config->parser(); - if (!in_array($_parser, $this->allowedParsers)) { - $this->debug("HTML parser $_parser not listed, using libxml instead"); - $_parser = 'libxml'; - } - $this->debug("Attempting to parse HTML with $_parser"); - $this->readability = new Readability($html, $url, $_parser); - - // we use xpath to find elements in the given HTML document - // see http://en.wikipedia.org/wiki/XPath_1.0 - $xpath = new DOMXPath($this->readability->dom); - - // try to get next page link - foreach ($this->config->next_page_link as $pattern) { - $elems = @$xpath->evaluate($pattern, $this->readability->dom); - if (is_string($elems)) { - $this->nextPageUrl = trim($elems); - break; - } elseif ($elems instanceof DOMNodeList && $elems->length > 0) { - foreach ($elems as $item) { - if ($item instanceof DOMElement && $item->hasAttribute('href')) { - $this->nextPageUrl = $item->getAttribute('href'); - break 2; - } elseif ($item instanceof DOMAttr && $item->value) { - $this->nextPageUrl = $item->value; - break 2; - } - } - } - } - - // try to get title - foreach ($this->config->title as $pattern) { - // $this->debug("Trying $pattern"); - $elems = @$xpath->evaluate($pattern, $this->readability->dom); - if (is_string($elems)) { - $this->title = trim($elems); - $this->debug('Title expression evaluated as string: '.$this->title); - $this->debug("...XPath match: $pattern"); - break; - } elseif ($elems instanceof DOMNodeList && $elems->length > 0) { - $this->title = $elems->item(0)->textContent; - $this->debug('Title matched: '.$this->title); - $this->debug("...XPath match: $pattern"); - // remove title from document - try { - $elems->item(0)->parentNode->removeChild($elems->item(0)); - } catch (DOMException $e) { - // do nothing - } - break; - } - } - - // try to get author (if it hasn't already been set) - if (empty($this->author)) { - foreach ($this->config->author as $pattern) { - $elems = @$xpath->evaluate($pattern, $this->readability->dom); - if (is_string($elems)) { - if (trim($elems) != '') { - $this->author[] = trim($elems); - $this->debug('Author expression evaluated as string: '.trim($elems)); - $this->debug("...XPath match: $pattern"); - break; - } - } elseif ($elems instanceof DOMNodeList && $elems->length > 0) { - foreach ($elems as $elem) { - if (!isset($elem->parentNode)) continue; - $this->author[] = trim($elem->textContent); - $this->debug('Author matched: '.trim($elem->textContent)); - } - if (!empty($this->author)) { - $this->debug("...XPath match: $pattern"); - break; - } - } - } - } - - // try to get language - $_lang_xpath = array('//html[@lang]/@lang', '//meta[@name="DC.language"]/@content'); - foreach ($_lang_xpath as $pattern) { - $elems = @$xpath->evaluate($pattern, $this->readability->dom); - if (is_string($elems)) { - if (trim($elems) != '') { - $this->language = trim($elems); - $this->debug('Language matched: '.$this->language); - break; - } - } elseif ($elems instanceof DOMNodeList && $elems->length > 0) { - foreach ($elems as $elem) { - if (!isset($elem->parentNode)) continue; - $this->language = trim($elem->textContent); - $this->debug('Language matched: '.$this->language); - } - if ($this->language) break; - } - } - - // try to get date - foreach ($this->config->date as $pattern) { - $elems = @$xpath->evaluate($pattern, $this->readability->dom); - if (is_string($elems)) { - $this->date = strtotime(trim($elems, "; \t\n\r\0\x0B")); - } elseif ($elems instanceof DOMNodeList && $elems->length > 0) { - $this->date = $elems->item(0)->textContent; - $this->date = strtotime(trim($this->date, "; \t\n\r\0\x0B")); - // remove date from document - // $elems->item(0)->parentNode->removeChild($elems->item(0)); - } - if (!$this->date) { - $this->date = null; - } else { - $this->debug('Date matched: '.date('Y-m-d H:i:s', $this->date)); - $this->debug("...XPath match: $pattern"); - break; - } - } - - // strip elements (using xpath expressions) - foreach ($this->config->strip as $pattern) { - $elems = @$xpath->query($pattern, $this->readability->dom); - // check for matches - if ($elems && $elems->length > 0) { - $this->debug('Stripping '.$elems->length.' elements (strip)'); - for ($i=$elems->length-1; $i >= 0; $i--) { - $elems->item($i)->parentNode->removeChild($elems->item($i)); - } - } - } - - // strip elements (using id and class attribute values) - foreach ($this->config->strip_id_or_class as $string) { - $string = strtr($string, array("'"=>'', '"'=>'')); - $elems = @$xpath->query("//*[contains(@class, '$string') or contains(@id, '$string')]", $this->readability->dom); - // check for matches - if ($elems && $elems->length > 0) { - $this->debug('Stripping '.$elems->length.' elements (strip_id_or_class)'); - for ($i=$elems->length-1; $i >= 0; $i--) { - $elems->item($i)->parentNode->removeChild($elems->item($i)); - } - } - } - - // strip images (using src attribute values) - foreach ($this->config->strip_image_src as $string) { - $string = strtr($string, array("'"=>'', '"'=>'')); - $elems = @$xpath->query("//img[contains(@src, '$string')]", $this->readability->dom); - // check for matches - if ($elems && $elems->length > 0) { - $this->debug('Stripping '.$elems->length.' image elements'); - for ($i=$elems->length-1; $i >= 0; $i--) { - $elems->item($i)->parentNode->removeChild($elems->item($i)); - } - } - } - // strip elements using Readability.com and Instapaper.com ignore class names - // .entry-unrelated and .instapaper_ignore - // See https://www.readability.com/publishers/guidelines/#view-plainGuidelines - // and http://blog.instapaper.com/post/730281947 - $elems = @$xpath->query("//*[contains(concat(' ',normalize-space(@class),' '),' entry-unrelated ') or contains(concat(' ',normalize-space(@class),' '),' instapaper_ignore ')]", $this->readability->dom); - // check for matches - if ($elems && $elems->length > 0) { - $this->debug('Stripping '.$elems->length.' .entry-unrelated,.instapaper_ignore elements'); - for ($i=$elems->length-1; $i >= 0; $i--) { - $elems->item($i)->parentNode->removeChild($elems->item($i)); - } - } - - // strip elements that contain style="display: none;" - $elems = @$xpath->query("//*[contains(@style,'display:none')]", $this->readability->dom); - // check for matches - if ($elems && $elems->length > 0) { - $this->debug('Stripping '.$elems->length.' elements with inline display:none style'); - for ($i=$elems->length-1; $i >= 0; $i--) { - $elems->item($i)->parentNode->removeChild($elems->item($i)); - } - } - - // try to get body - foreach ($this->config->body as $pattern) { - $elems = @$xpath->query($pattern, $this->readability->dom); - // check for matches - if ($elems && $elems->length > 0) { - $this->debug('Body matched'); - $this->debug("...XPath match: $pattern"); - if ($elems->length == 1) { - $this->body = $elems->item(0); - // prune (clean up elements that may not be content) - if ($this->config->prune()) { - $this->debug('...pruning content'); - $this->readability->prepArticle($this->body); - } - break; - } else { - $this->body = $this->readability->dom->createElement('div'); - $this->debug($elems->length.' body elems found'); - foreach ($elems as $elem) { - if (!isset($elem->parentNode)) continue; - $isDescendant = false; - foreach ($this->body->childNodes as $parent) { - if ($this->isDescendant($parent, $elem)) { - $isDescendant = true; - break; - } - } - if ($isDescendant) { - $this->debug('...element is child of another body element, skipping.'); - } else { - // prune (clean up elements that may not be content) - if ($this->config->prune()) { - $this->debug('Pruning content'); - $this->readability->prepArticle($elem); - } - $this->debug('...element added to body'); - $this->body->appendChild($elem); - } - } - if ($this->body->hasChildNodes()) break; - } - } - } - - // auto detect? - $detect_title = $detect_body = $detect_author = $detect_date = false; - // detect title? - if (!isset($this->title)) { - if (empty($this->config->title) || $this->config->autodetect_on_failure()) { - $detect_title = true; - } - } - // detect body? - if (!isset($this->body)) { - if (empty($this->config->body) || $this->config->autodetect_on_failure()) { - $detect_body = true; - } - } - // detect author? - if (empty($this->author)) { - if (empty($this->config->author) || $this->config->autodetect_on_failure()) { - $detect_author = true; - } - } - // detect date? - if (!isset($this->date)) { - if (empty($this->config->date) || $this->config->autodetect_on_failure()) { - $detect_date = true; - } - } - - // check for hNews - if ($detect_title || $detect_body) { - // check for hentry - $elems = @$xpath->query("//*[contains(concat(' ',normalize-space(@class),' '),' hentry ')]", $this->readability->dom); - if ($elems && $elems->length > 0) { - $this->debug('hNews: found hentry'); - $hentry = $elems->item(0); - - if ($detect_title) { - // check for entry-title - $elems = @$xpath->query(".//*[contains(concat(' ',normalize-space(@class),' '),' entry-title ')]", $hentry); - if ($elems && $elems->length > 0) { - $this->title = $elems->item(0)->textContent; - $this->debug('hNews: found entry-title: '.$this->title); - // remove title from document - $elems->item(0)->parentNode->removeChild($elems->item(0)); - $detect_title = false; - } - } - - if ($detect_date) { - // check for time element with pubdate attribute - $elems = @$xpath->query(".//time[@pubdate] | .//abbr[contains(concat(' ',normalize-space(@class),' '),' published ')]", $hentry); - if ($elems && $elems->length > 0) { - $this->date = strtotime(trim($elems->item(0)->textContent)); - // remove date from document - //$elems->item(0)->parentNode->removeChild($elems->item(0)); - if ($this->date) { - $this->debug('hNews: found publication date: '.date('Y-m-d H:i:s', $this->date)); - $detect_date = false; - } else { - $this->date = null; - } - } - } - - if ($detect_author) { - // check for time element with pubdate attribute - $elems = @$xpath->query(".//*[contains(concat(' ',normalize-space(@class),' '),' vcard ') and (contains(concat(' ',normalize-space(@class),' '),' author ') or contains(concat(' ',normalize-space(@class),' '),' byline '))]", $hentry); - if ($elems && $elems->length > 0) { - $author = $elems->item(0); - $fn = @$xpath->query(".//*[contains(concat(' ',normalize-space(@class),' '),' fn ')]", $author); - if ($fn && $fn->length > 0) { - foreach ($fn as $_fn) { - if (trim($_fn->textContent) != '') { - $this->author[] = trim($_fn->textContent); - $this->debug('hNews: found author: '.trim($_fn->textContent)); - } - } - } else { - if (trim($author->textContent) != '') { - $this->author[] = trim($author->textContent); - $this->debug('hNews: found author: '.trim($author->textContent)); - } - } - $detect_author = empty($this->author); - } - } - - // check for entry-content. - // according to hAtom spec, if there are multiple elements marked entry-content, - // we include all of these in the order they appear - see http://microformats.org/wiki/hatom#Entry_Content - if ($detect_body) { - $elems = @$xpath->query(".//*[contains(concat(' ',normalize-space(@class),' '),' entry-content ')]", $hentry); - if ($elems && $elems->length > 0) { - $this->debug('hNews: found entry-content'); - if ($elems->length == 1) { - // what if it's empty? (some sites misuse hNews - place their content outside an empty entry-content element) - $e = $elems->item(0); - if (($e->tagName == 'img') || (trim($e->textContent) != '')) { - $this->body = $elems->item(0); - // prune (clean up elements that may not be content) - if ($this->config->prune()) { - $this->debug('Pruning content'); - $this->readability->prepArticle($this->body); - } - $detect_body = false; - } else { - $this->debug('hNews: skipping entry-content - appears not to contain content'); - } - unset($e); - } else { - $this->body = $this->readability->dom->createElement('div'); - $this->debug($elems->length.' entry-content elems found'); - foreach ($elems as $elem) { - if (!isset($elem->parentNode)) continue; - $isDescendant = false; - foreach ($this->body->childNodes as $parent) { - if ($this->isDescendant($parent, $elem)) { - $isDescendant = true; - break; - } - } - if ($isDescendant) { - $this->debug('Element is child of another body element, skipping.'); - } else { - // prune (clean up elements that may not be content) - if ($this->config->prune()) { - $this->debug('Pruning content'); - $this->readability->prepArticle($elem); - } - $this->debug('Element added to body'); - $this->body->appendChild($elem); - } - } - $detect_body = false; - } - } - } - } - } - - // check for elements marked with instapaper_title - if ($detect_title) { - // check for instapaper_title - $elems = @$xpath->query("//*[contains(concat(' ',normalize-space(@class),' '),' instapaper_title ')]", $this->readability->dom); - if ($elems && $elems->length > 0) { - $this->title = $elems->item(0)->textContent; - $this->debug('Title found (.instapaper_title): '.$this->title); - // remove title from document - $elems->item(0)->parentNode->removeChild($elems->item(0)); - $detect_title = false; - } - } - // check for elements marked with instapaper_body - if ($detect_body) { - $elems = @$xpath->query("//*[contains(concat(' ',normalize-space(@class),' '),' instapaper_body ')]", $this->readability->dom); - if ($elems && $elems->length > 0) { - $this->debug('body found (.instapaper_body)'); - $this->body = $elems->item(0); - // prune (clean up elements that may not be content) - if ($this->config->prune()) { - $this->debug('Pruning content'); - $this->readability->prepArticle($this->body); - } - $detect_body = false; - } - } - - // Find author in rel="author" marked element - // We only use this if there's exactly one. - // If there's more than one, it could indicate more than - // one author, but it could also indicate that we're processing - // a page listing different articles with different authors. - if ($detect_author) { - $elems = @$xpath->query("//a[contains(concat(' ',normalize-space(@rel),' '),' author ')]", $this->readability->dom); - if ($elems && $elems->length == 1) { - $author = trim($elems->item(0)->textContent); - if ($author != '') { - $this->debug("Author found (rel=\"author\"): $author"); - $this->author[] = $author; - $detect_author = false; - } - } - } - - // Find date in pubdate marked time element - // For the same reason given above, we only use this - // if there's exactly one element. - if ($detect_date) { - $elems = @$xpath->query("//time[@pubdate]", $this->readability->dom); - if ($elems && $elems->length == 1) { - $this->date = strtotime(trim($elems->item(0)->textContent)); - // remove date from document - //$elems->item(0)->parentNode->removeChild($elems->item(0)); - if ($this->date) { - $this->debug('Date found (pubdate marked time element): '.date('Y-m-d H:i:s', $this->date)); - $detect_date = false; - } else { - $this->date = null; - } - } - } - - // still missing title or body, so we detect using Readability - if ($detect_title || $detect_body) { - $this->debug('Using Readability'); - // clone body if we're only using Readability for title (otherwise it may interfere with body element) - if (isset($this->body)) $this->body = $this->body->cloneNode(true); - $success = $this->readability->init(); - } - if ($detect_title) { - $this->debug('Detecting title'); - $this->title = $this->readability->getTitle()->textContent; - } - if ($detect_body && $success) { - $this->debug('Detecting body'); - $this->body = $this->readability->getContent(); - if ($this->body->childNodes->length == 1 && $this->body->firstChild->nodeType === XML_ELEMENT_NODE) { - $this->body = $this->body->firstChild; - } - // prune (clean up elements that may not be content) - if ($this->config->prune()) { - $this->debug('Pruning content'); - $this->readability->prepArticle($this->body); - } - } - if (isset($this->body)) { - // remove scripts - $this->readability->removeScripts($this->body); - // remove any h1-h6 elements that appear as first thing in the body - // and which match our title - if (isset($this->title) && ($this->title != '')) { - $firstChild = $this->body->firstChild; - while ($firstChild->nodeType && ($firstChild->nodeType !== XML_ELEMENT_NODE)) { - $firstChild = $firstChild->nextSibling; - } - if (($firstChild->nodeType === XML_ELEMENT_NODE) - && in_array(strtolower($firstChild->tagName), array('h1', 'h2', 'h3', 'h4', 'h5', 'h6')) - && (strtolower(trim($firstChild->textContent)) == strtolower(trim($this->title)))) { - $this->body->removeChild($firstChild); - } - } - // prevent self-closing iframes - $elems = $this->body->getElementsByTagName('iframe'); - for ($i = $elems->length-1; $i >= 0; $i--) { - $e = $elems->item($i); - if (!$e->hasChildNodes()) { - $e->appendChild($this->body->ownerDocument->createTextNode('[embedded content]')); - } - } - // remove image lazy loading - WordPress plugin http://wordpress.org/extend/plugins/lazy-load/ - // the plugin replaces the src attribute to point to a 1x1 gif and puts the original src - // inside the data-lazy-src attribute. It also places the original image inside a noscript element - // next to the amended one. - $elems = @$xpath->query("//img[@data-lazy-src]", $this->body); - for ($i = $elems->length-1; $i >= 0; $i--) { - $e = $elems->item($i); - // let's see if we can grab image from noscript - if ($e->nextSibling !== null && $e->nextSibling->nodeName === 'noscript') { - $_new_elem = $e->ownerDocument->createDocumentFragment(); - @$_new_elem->appendXML($e->nextSibling->innerHTML); - $e->nextSibling->parentNode->replaceChild($_new_elem, $e->nextSibling); - $e->parentNode->removeChild($e); - } else { - // Use data-lazy-src as src value - $e->setAttribute('src', $e->getAttribute('data-lazy-src')); - $e->removeAttribute('data-lazy-src'); - } - } - - $this->success = true; - } - - // if we've had no success and we've used tidy, there's a chance - // that tidy has messed up. So let's try again without tidy... - if (!$this->success && $tidied && $smart_tidy) { - $this->debug('Trying again without tidy'); - $this->process($original_html, $url, false); - } - - return $this->success; - } - - private function isDescendant(DOMElement $parent, DOMElement $child) { - $node = $child->parentNode; - while ($node != null) { - if ($node->isSameNode($parent)) return true; - $node = $node->parentNode; - } - return false; - } - - public function getContent() { - return $this->body; - } - - public function getTitle() { - return $this->title; - } - - public function getAuthors() { - return $this->author; - } - - public function getLanguage() { - return $this->language; - } - - public function getDate() { - return $this->date; - } - - public function getSiteConfig() { - return $this->config; - } - - public function getNextPageUrl() { - return $this->nextPageUrl; - } -} -?> \ No newline at end of file + true, + 'output-xhtml' => true, + 'logical-emphasis' => true, + 'show-body-only' => false, + 'new-blocklevel-tags' => 'article, aside, footer, header, hgroup, menu, nav, section, details, datagrid', + 'new-inline-tags' => 'mark, time, meter, progress, data', + 'wrap' => 0, + 'drop-empty-paras' => true, + 'drop-proprietary-attributes' => false, + 'enclose-text' => true, + 'enclose-block-text' => true, + 'merge-divs' => true, + 'merge-spans' => true, + 'char-encoding' => 'utf8', + 'hide-comments' => true + ); + protected $html; + protected $config; + protected $title; + protected $author = array(); + protected $language; + protected $date; + protected $body; + protected $success = false; + protected $nextPageUrl; + public $allowedParsers = array('libxml', 'html5lib'); + public $fingerprints = array(); + public $readability; + public $debug = false; + public $debugVerbose = false; + + function __construct($path, $fallback=null) { + SiteConfig::set_config_path($path, $fallback); + } + + protected function debug($msg) { + if ($this->debug) { + $mem = round(memory_get_usage()/1024, 2); + $memPeak = round(memory_get_peak_usage()/1024, 2); + echo '* ',$msg; + if ($this->debugVerbose) echo ' - mem used: ',$mem," (peak: $memPeak)"; + echo "\n"; + ob_flush(); + flush(); + } + } + + public function reset() { + $this->html = null; + $this->readability = null; + $this->config = null; + $this->title = null; + $this->body = null; + $this->author = array(); + $this->language = null; + $this->date = null; + $this->nextPageUrl = null; + $this->success = false; + } + + public function findHostUsingFingerprints($html) { + $this->debug('Checking fingerprints...'); + $head = substr($html, 0, 8000); + foreach ($this->fingerprints as $_fp => $_fphost) { + $lookin = 'html'; + if (is_array($_fphost)) { + if (isset($_fphost['head']) && $_fphost['head']) { + $lookin = 'head'; + } + $_fphost = $_fphost['hostname']; + } + if (strpos($$lookin, $_fp) !== false) { + $this->debug("Found match: $_fphost"); + return $_fphost; + } + } + $this->debug('No fingerprint matches'); + return false; + } + + // returns SiteConfig instance (joined in order: exact match, wildcard, fingerprint, global, default) + public function buildSiteConfig($url, $html='', $add_to_cache=true) { + // extract host name + $host = @parse_url($url, PHP_URL_HOST); + $host = strtolower($host); + if (substr($host, 0, 4) == 'www.') $host = substr($host, 4); + // is merged version already cached? + if (SiteConfig::is_cached("$host.merged")) { + $this->debug("Returning cached and merged site config for $host"); + return SiteConfig::build("$host.merged"); + } + // let's build from site_config/custom/ and standard/ + $config = SiteConfig::build($host); + if ($add_to_cache && $config && !SiteConfig::is_cached("$host")) { + SiteConfig::add_to_cache($host, $config); + } + // if no match, use defaults + if (!$config) $config = new SiteConfig(); + // load fingerprint config? + if ($config->autodetect_on_failure()) { + // check HTML for fingerprints + if (!empty($this->fingerprints) && ($_fphost = $this->findHostUsingFingerprints($html))) { + if ($config_fingerprint = SiteConfig::build($_fphost)) { + $this->debug("Appending site config settings from $_fphost (fingerprint match)"); + $config->append($config_fingerprint); + if ($add_to_cache && !SiteConfig::is_cached($_fphost)) { + //$config_fingerprint->cache_in_apc = true; + SiteConfig::add_to_cache($_fphost, $config_fingerprint); + } + } + } + } + // load global config? + if ($config->autodetect_on_failure()) { + if ($config_global = SiteConfig::build('global', true)) { + $this->debug('Appending site config settings from global.txt'); + $config->append($config_global); + if ($add_to_cache && !SiteConfig::is_cached('global')) { + //$config_global->cache_in_apc = true; + SiteConfig::add_to_cache('global', $config_global); + } + } + } + // store copy of merged config + if ($add_to_cache) { + // do not store in APC if wildcard match + $use_apc = ($host == $config->cache_key); + $config->cache_key = null; + SiteConfig::add_to_cache("$host.merged", $config, $use_apc); + } + return $config; + } + + // returns true on success, false on failure + // $smart_tidy indicates that if tidy is used and no results are produced, we will + // try again without it. Tidy helps us deal with PHP's patchy HTML parsing most of the time + // but it has problems of its own which we try to avoid with this option. + public function process($html, $url, $smart_tidy=true) { + $this->reset(); + $this->config = $this->buildSiteConfig($url, $html); + + // do string replacements + if (!empty($this->config->find_string)) { + if (count($this->config->find_string) == count($this->config->replace_string)) { + $html = str_replace($this->config->find_string, $this->config->replace_string, $html, $_count); + $this->debug("Strings replaced: $_count (find_string and/or replace_string)"); + } else { + $this->debug('Skipped string replacement - incorrect number of find-replace strings in site config'); + } + unset($_count); + } + + // use tidy (if it exists)? + // This fixes problems with some sites which would otherwise + // trouble DOMDocument's HTML parsing. (Although sometimes it + // makes matters worse, which is why you can override it in site config files.) + $tidied = false; + if ($this->config->tidy() && function_exists('tidy_parse_string') && $smart_tidy) { + $this->debug('Using Tidy'); + $tidy = tidy_parse_string($html, self::$tidy_config, 'UTF8'); + if (tidy_clean_repair($tidy)) { + $original_html = $html; + $tidied = true; + $html = $tidy->value; + } + unset($tidy); + } + + // load and parse html + $_parser = $this->config->parser(); + if (!in_array($_parser, $this->allowedParsers)) { + $this->debug("HTML parser $_parser not listed, using libxml instead"); + $_parser = 'libxml'; + } + $this->debug("Attempting to parse HTML with $_parser"); + $this->readability = new Readability($html, $url, $_parser); + + // we use xpath to find elements in the given HTML document + // see http://en.wikipedia.org/wiki/XPath_1.0 + $xpath = new DOMXPath($this->readability->dom); + + // try to get next page link + foreach ($this->config->next_page_link as $pattern) { + $elems = @$xpath->evaluate($pattern, $this->readability->dom); + if (is_string($elems)) { + $this->nextPageUrl = trim($elems); + break; + } elseif ($elems instanceof DOMNodeList && $elems->length > 0) { + foreach ($elems as $item) { + if ($item instanceof DOMElement && $item->hasAttribute('href')) { + $this->nextPageUrl = $item->getAttribute('href'); + break 2; + } elseif ($item instanceof DOMAttr && $item->value) { + $this->nextPageUrl = $item->value; + break 2; + } + } + } + } + + // try to get title + foreach ($this->config->title as $pattern) { + // $this->debug("Trying $pattern"); + $elems = @$xpath->evaluate($pattern, $this->readability->dom); + if (is_string($elems)) { + $this->title = trim($elems); + $this->debug('Title expression evaluated as string: '.$this->title); + $this->debug("...XPath match: $pattern"); + break; + } elseif ($elems instanceof DOMNodeList && $elems->length > 0) { + $this->title = $elems->item(0)->textContent; + $this->debug('Title matched: '.$this->title); + $this->debug("...XPath match: $pattern"); + // remove title from document + try { + @$elems->item(0)->parentNode->removeChild($elems->item(0)); + } catch (DOMException $e) { + // do nothing + } + break; + } + } + + // try to get author (if it hasn't already been set) + if (empty($this->author)) { + foreach ($this->config->author as $pattern) { + $elems = @$xpath->evaluate($pattern, $this->readability->dom); + if (is_string($elems)) { + if (trim($elems) != '') { + $this->author[] = trim($elems); + $this->debug('Author expression evaluated as string: '.trim($elems)); + $this->debug("...XPath match: $pattern"); + break; + } + } elseif ($elems instanceof DOMNodeList && $elems->length > 0) { + foreach ($elems as $elem) { + if (!isset($elem->parentNode)) continue; + $this->author[] = trim($elem->textContent); + $this->debug('Author matched: '.trim($elem->textContent)); + } + if (!empty($this->author)) { + $this->debug("...XPath match: $pattern"); + break; + } + } + } + } + + // try to get language + $_lang_xpath = array('//html[@lang]/@lang', '//meta[@name="DC.language"]/@content'); + foreach ($_lang_xpath as $pattern) { + $elems = @$xpath->evaluate($pattern, $this->readability->dom); + if (is_string($elems)) { + if (trim($elems) != '') { + $this->language = trim($elems); + $this->debug('Language matched: '.$this->language); + break; + } + } elseif ($elems instanceof DOMNodeList && $elems->length > 0) { + foreach ($elems as $elem) { + if (!isset($elem->parentNode)) continue; + $this->language = trim($elem->textContent); + $this->debug('Language matched: '.$this->language); + } + if ($this->language) break; + } + } + + // try to get date + foreach ($this->config->date as $pattern) { + $elems = @$xpath->evaluate($pattern, $this->readability->dom); + if (is_string($elems)) { + $this->date = strtotime(trim($elems, "; \t\n\r\0\x0B")); + } elseif ($elems instanceof DOMNodeList && $elems->length > 0) { + $this->date = $elems->item(0)->textContent; + $this->date = strtotime(trim($this->date, "; \t\n\r\0\x0B")); + // remove date from document + // $elems->item(0)->parentNode->removeChild($elems->item(0)); + } + if (!$this->date) { + $this->date = null; + } else { + $this->debug('Date matched: '.date('Y-m-d H:i:s', $this->date)); + $this->debug("...XPath match: $pattern"); + break; + } + } + + // strip elements (using xpath expressions) + foreach ($this->config->strip as $pattern) { + $elems = @$xpath->query($pattern, $this->readability->dom); + // check for matches + if ($elems && $elems->length > 0) { + $this->debug('Stripping '.$elems->length.' elements (strip)'); + for ($i=$elems->length-1; $i >= 0; $i--) { + $elems->item($i)->parentNode->removeChild($elems->item($i)); + } + } + } + + // strip elements (using id and class attribute values) + foreach ($this->config->strip_id_or_class as $string) { + $string = strtr($string, array("'"=>'', '"'=>'')); + $elems = @$xpath->query("//*[contains(@class, '$string') or contains(@id, '$string')]", $this->readability->dom); + // check for matches + if ($elems && $elems->length > 0) { + $this->debug('Stripping '.$elems->length.' elements (strip_id_or_class)'); + for ($i=$elems->length-1; $i >= 0; $i--) { + $elems->item($i)->parentNode->removeChild($elems->item($i)); + } + } + } + + // strip images (using src attribute values) + foreach ($this->config->strip_image_src as $string) { + $string = strtr($string, array("'"=>'', '"'=>'')); + $elems = @$xpath->query("//img[contains(@src, '$string')]", $this->readability->dom); + // check for matches + if ($elems && $elems->length > 0) { + $this->debug('Stripping '.$elems->length.' image elements'); + for ($i=$elems->length-1; $i >= 0; $i--) { + $elems->item($i)->parentNode->removeChild($elems->item($i)); + } + } + } + // strip elements using Readability.com and Instapaper.com ignore class names + // .entry-unrelated and .instapaper_ignore + // See https://www.readability.com/publishers/guidelines/#view-plainGuidelines + // and http://blog.instapaper.com/post/730281947 + $elems = @$xpath->query("//*[contains(concat(' ',normalize-space(@class),' '),' entry-unrelated ') or contains(concat(' ',normalize-space(@class),' '),' instapaper_ignore ')]", $this->readability->dom); + // check for matches + if ($elems && $elems->length > 0) { + $this->debug('Stripping '.$elems->length.' .entry-unrelated,.instapaper_ignore elements'); + for ($i=$elems->length-1; $i >= 0; $i--) { + $elems->item($i)->parentNode->removeChild($elems->item($i)); + } + } + + // strip elements that contain style="display: none;" + $elems = @$xpath->query("//*[contains(@style,'display:none')]", $this->readability->dom); + // check for matches + if ($elems && $elems->length > 0) { + $this->debug('Stripping '.$elems->length.' elements with inline display:none style'); + for ($i=$elems->length-1; $i >= 0; $i--) { + $elems->item($i)->parentNode->removeChild($elems->item($i)); + } + } + + // try to get body + foreach ($this->config->body as $pattern) { + $elems = @$xpath->query($pattern, $this->readability->dom); + // check for matches + if ($elems && $elems->length > 0) { + $this->debug('Body matched'); + $this->debug("...XPath match: $pattern"); + if ($elems->length == 1) { + $this->body = $elems->item(0); + // prune (clean up elements that may not be content) + if ($this->config->prune()) { + $this->debug('...pruning content'); + $this->readability->prepArticle($this->body); + } + break; + } else { + $this->body = $this->readability->dom->createElement('div'); + $this->debug($elems->length.' body elems found'); + foreach ($elems as $elem) { + if (!isset($elem->parentNode)) continue; + $isDescendant = false; + foreach ($this->body->childNodes as $parent) { + if ($this->isDescendant($parent, $elem)) { + $isDescendant = true; + break; + } + } + if ($isDescendant) { + $this->debug('...element is child of another body element, skipping.'); + } else { + // prune (clean up elements that may not be content) + if ($this->config->prune()) { + $this->debug('Pruning content'); + $this->readability->prepArticle($elem); + } + $this->debug('...element added to body'); + $this->body->appendChild($elem); + } + } + if ($this->body->hasChildNodes()) break; + } + } + } + + // auto detect? + $detect_title = $detect_body = $detect_author = $detect_date = false; + // detect title? + if (!isset($this->title)) { + if (empty($this->config->title) || $this->config->autodetect_on_failure()) { + $detect_title = true; + } + } + // detect body? + if (!isset($this->body)) { + if (empty($this->config->body) || $this->config->autodetect_on_failure()) { + $detect_body = true; + } + } + // detect author? + if (empty($this->author)) { + if (empty($this->config->author) || $this->config->autodetect_on_failure()) { + $detect_author = true; + } + } + // detect date? + if (!isset($this->date)) { + if (empty($this->config->date) || $this->config->autodetect_on_failure()) { + $detect_date = true; + } + } + + // check for hNews + if ($detect_title || $detect_body) { + // check for hentry + $elems = @$xpath->query("//*[contains(concat(' ',normalize-space(@class),' '),' hentry ')]", $this->readability->dom); + if ($elems && $elems->length > 0) { + $this->debug('hNews: found hentry'); + $hentry = $elems->item(0); + + if ($detect_title) { + // check for entry-title + $elems = @$xpath->query(".//*[contains(concat(' ',normalize-space(@class),' '),' entry-title ')]", $hentry); + if ($elems && $elems->length > 0) { + $this->title = $elems->item(0)->textContent; + $this->debug('hNews: found entry-title: '.$this->title); + // remove title from document + $elems->item(0)->parentNode->removeChild($elems->item(0)); + $detect_title = false; + } + } + + if ($detect_date) { + // check for time element with pubdate attribute + $elems = @$xpath->query(".//time[@pubdate] | .//abbr[contains(concat(' ',normalize-space(@class),' '),' published ')]", $hentry); + if ($elems && $elems->length > 0) { + $this->date = strtotime(trim($elems->item(0)->textContent)); + // remove date from document + //$elems->item(0)->parentNode->removeChild($elems->item(0)); + if ($this->date) { + $this->debug('hNews: found publication date: '.date('Y-m-d H:i:s', $this->date)); + $detect_date = false; + } else { + $this->date = null; + } + } + } + + if ($detect_author) { + // check for time element with pubdate attribute + $elems = @$xpath->query(".//*[contains(concat(' ',normalize-space(@class),' '),' vcard ') and (contains(concat(' ',normalize-space(@class),' '),' author ') or contains(concat(' ',normalize-space(@class),' '),' byline '))]", $hentry); + if ($elems && $elems->length > 0) { + $author = $elems->item(0); + $fn = @$xpath->query(".//*[contains(concat(' ',normalize-space(@class),' '),' fn ')]", $author); + if ($fn && $fn->length > 0) { + foreach ($fn as $_fn) { + if (trim($_fn->textContent) != '') { + $this->author[] = trim($_fn->textContent); + $this->debug('hNews: found author: '.trim($_fn->textContent)); + } + } + } else { + if (trim($author->textContent) != '') { + $this->author[] = trim($author->textContent); + $this->debug('hNews: found author: '.trim($author->textContent)); + } + } + $detect_author = empty($this->author); + } + } + + // check for entry-content. + // according to hAtom spec, if there are multiple elements marked entry-content, + // we include all of these in the order they appear - see http://microformats.org/wiki/hatom#Entry_Content + if ($detect_body) { + $elems = @$xpath->query(".//*[contains(concat(' ',normalize-space(@class),' '),' entry-content ')]", $hentry); + if ($elems && $elems->length > 0) { + $this->debug('hNews: found entry-content'); + if ($elems->length == 1) { + // what if it's empty? (some sites misuse hNews - place their content outside an empty entry-content element) + $e = $elems->item(0); + if (($e->tagName == 'img') || (trim($e->textContent) != '')) { + $this->body = $elems->item(0); + // prune (clean up elements that may not be content) + if ($this->config->prune()) { + $this->debug('Pruning content'); + $this->readability->prepArticle($this->body); + } + $detect_body = false; + } else { + $this->debug('hNews: skipping entry-content - appears not to contain content'); + } + unset($e); + } else { + $this->body = $this->readability->dom->createElement('div'); + $this->debug($elems->length.' entry-content elems found'); + foreach ($elems as $elem) { + if (!isset($elem->parentNode)) continue; + $isDescendant = false; + foreach ($this->body->childNodes as $parent) { + if ($this->isDescendant($parent, $elem)) { + $isDescendant = true; + break; + } + } + if ($isDescendant) { + $this->debug('Element is child of another body element, skipping.'); + } else { + // prune (clean up elements that may not be content) + if ($this->config->prune()) { + $this->debug('Pruning content'); + $this->readability->prepArticle($elem); + } + $this->debug('Element added to body'); + $this->body->appendChild($elem); + } + } + $detect_body = false; + } + } + } + } + } + + // check for elements marked with instapaper_title + if ($detect_title) { + // check for instapaper_title + $elems = @$xpath->query("//*[contains(concat(' ',normalize-space(@class),' '),' instapaper_title ')]", $this->readability->dom); + if ($elems && $elems->length > 0) { + $this->title = $elems->item(0)->textContent; + $this->debug('Title found (.instapaper_title): '.$this->title); + // remove title from document + $elems->item(0)->parentNode->removeChild($elems->item(0)); + $detect_title = false; + } + } + // check for elements marked with instapaper_body + if ($detect_body) { + $elems = @$xpath->query("//*[contains(concat(' ',normalize-space(@class),' '),' instapaper_body ')]", $this->readability->dom); + if ($elems && $elems->length > 0) { + $this->debug('body found (.instapaper_body)'); + $this->body = $elems->item(0); + // prune (clean up elements that may not be content) + if ($this->config->prune()) { + $this->debug('Pruning content'); + $this->readability->prepArticle($this->body); + } + $detect_body = false; + } + } + + // Find author in rel="author" marked element + // We only use this if there's exactly one. + // If there's more than one, it could indicate more than + // one author, but it could also indicate that we're processing + // a page listing different articles with different authors. + if ($detect_author) { + $elems = @$xpath->query("//a[contains(concat(' ',normalize-space(@rel),' '),' author ')]", $this->readability->dom); + if ($elems && $elems->length == 1) { + $author = trim($elems->item(0)->textContent); + if ($author != '') { + $this->debug("Author found (rel=\"author\"): $author"); + $this->author[] = $author; + $detect_author = false; + } + } + } + + // Find date in pubdate marked time element + // For the same reason given above, we only use this + // if there's exactly one element. + if ($detect_date) { + $elems = @$xpath->query("//time[@pubdate]", $this->readability->dom); + if ($elems && $elems->length == 1) { + $this->date = strtotime(trim($elems->item(0)->textContent)); + // remove date from document + //$elems->item(0)->parentNode->removeChild($elems->item(0)); + if ($this->date) { + $this->debug('Date found (pubdate marked time element): '.date('Y-m-d H:i:s', $this->date)); + $detect_date = false; + } else { + $this->date = null; + } + } + } + + // still missing title or body, so we detect using Readability + if ($detect_title || $detect_body) { + $this->debug('Using Readability'); + // clone body if we're only using Readability for title (otherwise it may interfere with body element) + if (isset($this->body)) $this->body = $this->body->cloneNode(true); + $success = $this->readability->init(); + } + if ($detect_title) { + $this->debug('Detecting title'); + $this->title = $this->readability->getTitle()->textContent; + } + if ($detect_body && $success) { + $this->debug('Detecting body'); + $this->body = $this->readability->getContent(); + if ($this->body->childNodes->length == 1 && $this->body->firstChild->nodeType === XML_ELEMENT_NODE) { + $this->body = $this->body->firstChild; + } + // prune (clean up elements that may not be content) + if ($this->config->prune()) { + $this->debug('Pruning content'); + $this->readability->prepArticle($this->body); + } + } + if (isset($this->body)) { + // remove scripts + $this->readability->removeScripts($this->body); + // remove any h1-h6 elements that appear as first thing in the body + // and which match our title + if (isset($this->title) && ($this->title != '')) { + $firstChild = $this->body->firstChild; + while ($firstChild->nodeType && ($firstChild->nodeType !== XML_ELEMENT_NODE)) { + $firstChild = $firstChild->nextSibling; + } + if (($firstChild->nodeType === XML_ELEMENT_NODE) + && in_array(strtolower($firstChild->tagName), array('h1', 'h2', 'h3', 'h4', 'h5', 'h6')) + && (strtolower(trim($firstChild->textContent)) == strtolower(trim($this->title)))) { + $this->body->removeChild($firstChild); + } + } + // prevent self-closing iframes + $elems = $this->body->getElementsByTagName('iframe'); + for ($i = $elems->length-1; $i >= 0; $i--) { + $e = $elems->item($i); + if (!$e->hasChildNodes()) { + $e->appendChild($this->body->ownerDocument->createTextNode('[embedded content]')); + } + } + // remove image lazy loading - WordPress plugin http://wordpress.org/extend/plugins/lazy-load/ + // the plugin replaces the src attribute to point to a 1x1 gif and puts the original src + // inside the data-lazy-src attribute. It also places the original image inside a noscript element + // next to the amended one. + $elems = @$xpath->query("//img[@data-lazy-src]", $this->body); + for ($i = $elems->length-1; $i >= 0; $i--) { + $e = $elems->item($i); + // let's see if we can grab image from noscript + if ($e->nextSibling !== null && $e->nextSibling->nodeName === 'noscript') { + $_new_elem = $e->ownerDocument->createDocumentFragment(); + @$_new_elem->appendXML($e->nextSibling->innerHTML); + $e->nextSibling->parentNode->replaceChild($_new_elem, $e->nextSibling); + $e->parentNode->removeChild($e); + } else { + // Use data-lazy-src as src value + $e->setAttribute('src', $e->getAttribute('data-lazy-src')); + $e->removeAttribute('data-lazy-src'); + } + } + + $this->success = true; + } + + // if we've had no success and we've used tidy, there's a chance + // that tidy has messed up. So let's try again without tidy... + if (!$this->success && $tidied && $smart_tidy) { + $this->debug('Trying again without tidy'); + $this->process($original_html, $url, false); + } + + return $this->success; + } + + private function isDescendant(DOMElement $parent, DOMElement $child) { + $node = $child->parentNode; + while ($node != null) { + if ($node->isSameNode($parent)) return true; + $node = $node->parentNode; + } + return false; + } + + public function getContent() { + return $this->body; + } + + public function getTitle() { + return $this->title; + } + + public function getAuthors() { + return $this->author; + } + + public function getLanguage() { + return $this->language; + } + + public function getDate() { + return $this->date; + } + + public function getSiteConfig() { + return $this->config; + } + + public function getNextPageUrl() { + return $this->nextPageUrl; + } +} \ No newline at end of file diff --git a/inc/3rdparty/libraries/content-extractor/SiteConfig.php b/inc/3rdparty/libraries/content-extractor/SiteConfig.php index c5e300d..1f6a760 100644 --- a/inc/3rdparty/libraries/content-extractor/SiteConfig.php +++ b/inc/3rdparty/libraries/content-extractor/SiteConfig.php @@ -1,338 +1,343 @@ -tidy)) ? $this->tidy : $this->default_tidy; - return $this->tidy; - } - - // return bool or null - public function prune($use_default=true) { - if ($use_default) return (isset($this->prune)) ? $this->prune : $this->default_prune; - return $this->prune; - } - - // return string or null - public function parser($use_default=true) { - if ($use_default) return (isset($this->parser)) ? $this->parser : $this->default_parser; - return $this->parser; - } - - // return bool or null - public function autodetect_on_failure($use_default=true) { - if ($use_default) return (isset($this->autodetect_on_failure)) ? $this->autodetect_on_failure : $this->default_autodetect_on_failure; - return $this->autodetect_on_failure; - } - - public static function set_config_path($path, $fallback=null) { - self::$config_path = $path; - self::$config_path_fallback = $fallback; - } - - public static function add_to_cache($key, SiteConfig $config, $use_apc=true) { - $key = strtolower($key); - if (substr($key, 0, 4) == 'www.') $key = substr($key, 4); - if ($config->cache_key) $key = $config->cache_key; - self::$config_cache[$key] = $config; - if (self::$apc && $use_apc) { - self::debug("Adding site config to APC cache with key sc.$key"); - apc_add("sc.$key", $config); - } - self::debug("Cached site config with key $key"); - } - - public static function is_cached($key) { - $key = strtolower($key); - if (substr($key, 0, 4) == 'www.') $key = substr($key, 4); - if (array_key_exists($key, self::$config_cache)) { - return true; - } elseif (self::$apc && (bool)apc_fetch("sc.$key")) { - return true; - } - return false; - } - - public function append(SiteConfig $newconfig) { - // check for commands where we accept multiple statements (no test_url) - foreach (array('title', 'body', 'author', 'date', 'strip', 'strip_id_or_class', 'strip_image_src', 'single_page_link', 'single_page_link_in_feed', 'next_page_link', 'http_header', 'find_string', 'replace_string') as $var) { - // append array elements for this config variable from $newconfig to this config - //$this->$var = $this->$var + $newconfig->$var; - $this->$var = array_unique(array_merge($this->$var, $newconfig->$var)); - } - // check for single statement commands - // we do not overwrite existing non null values - foreach (array('tidy', 'prune', 'parser', 'autodetect_on_failure') as $var) { - if ($this->$var === null) $this->$var = $newconfig->$var; - } - } - - // returns SiteConfig instance if an appropriate one is found, false otherwise - // if $exact_host_match is true, we will not look for wildcard config matches - // by default if host is 'test.example.org' we will look for and load '.example.org.txt' if it exists - public static function build($host, $exact_host_match=false) { - $host = strtolower($host); - if (substr($host, 0, 4) == 'www.') $host = substr($host, 4); - if (!$host || (strlen($host) > 200) || !preg_match(self::HOSTNAME_REGEX, ltrim($host, '.'))) return false; - // check for site configuration - $try = array($host); - // should we look for wildcard matches - if (!$exact_host_match) { - $split = explode('.', $host); - if (count($split) > 1) { - array_shift($split); - $try[] = '.'.implode('.', $split); - } - } - - // look for site config file in primary folder - self::debug(". looking for site config for $host in primary folder"); - foreach ($try as $h) { - if (array_key_exists($h, self::$config_cache)) { - self::debug("... site config for $h already loaded in this request"); - return self::$config_cache[$h]; - } elseif (self::$apc && ($sconfig = apc_fetch("sc.$h"))) { - self::debug("... site config for $h in APC cache"); - return $sconfig; - } elseif (file_exists(self::$config_path."/$h.txt")) { - self::debug("... found site config ($h.txt)"); - $file_primary = self::$config_path."/$h.txt"; - $matched_name = $h; - break; - } - } - - // if we found site config, process it - if (isset($file_primary)) { - $config_lines = file($file_primary, FILE_IGNORE_NEW_LINES | FILE_SKIP_EMPTY_LINES); - if (!$config_lines || !is_array($config_lines)) return false; - $config = self::build_from_array($config_lines); - // if APC caching is available and enabled, mark this for cache - //$config->cache_in_apc = true; - $config->cache_key = $matched_name; - - // if autodetec on failure is off (on by default) we do not need to look - // in secondary folder - if (!$config->autodetect_on_failure()) { - self::debug('... autodetect on failure is disabled (no other site config files will be loaded)'); - return $config; - } - } - - // look for site config file in secondary folder - if (isset(self::$config_path_fallback)) { - self::debug(". looking for site config for $host in secondary folder"); - foreach ($try as $h) { - if (file_exists(self::$config_path_fallback."/$h.txt")) { - self::debug("... found site config in secondary folder ($h.txt)"); - $file_secondary = self::$config_path_fallback."/$h.txt"; - $matched_name = $h; - break; - } - } - if (!isset($file_secondary)) { - self::debug("... no site config match in secondary folder"); - } - } - - // return false if no config file found - if (!isset($file_primary) && !isset($file_secondary)) { - self::debug("... no site config match for $host"); - return false; - } - - // return primary config if secondary not found - if (!isset($file_secondary) && isset($config)) { - return $config; - } - - // process secondary config file - $config_lines = file($file_secondary, FILE_IGNORE_NEW_LINES | FILE_SKIP_EMPTY_LINES); - if (!$config_lines || !is_array($config_lines)) { - // failed to process secondary - if (isset($config)) { - // return primary config - return $config; - } else { - return false; - } - } - - // merge with primary and return - if (isset($config)) { - self::debug('. merging config files'); - $config->append(self::build_from_array($config_lines)); - return $config; - } else { - // return just secondary - $config = self::build_from_array($config_lines); - // if APC caching is available and enabled, mark this for cache - //$config->cache_in_apc = true; - $config->cache_key = $matched_name; - return $config; - } - } - - public static function build_from_array(array $lines) { - $config = new SiteConfig(); - foreach ($lines as $line) { - $line = trim($line); - - // skip comments, empty lines - if ($line == '' || $line[0] == '#') continue; - - // get command - $command = explode(':', $line, 2); - // if there's no colon ':', skip this line - if (count($command) != 2) continue; - $val = trim($command[1]); - $command = trim($command[0]); - if ($command == '' || $val == '') continue; - - // check for commands where we accept multiple statements - if (in_array($command, array('title', 'body', 'author', 'date', 'strip', 'strip_id_or_class', 'strip_image_src', 'single_page_link', 'single_page_link_in_feed', 'next_page_link', 'http_header', 'test_url', 'find_string', 'replace_string'))) { - array_push($config->$command, $val); - // check for single statement commands that evaluate to true or false - } elseif (in_array($command, array('tidy', 'prune', 'autodetect_on_failure'))) { - $config->$command = ($val == 'yes'); - // check for single statement commands stored as strings - } elseif (in_array($command, array('parser'))) { - $config->$command = $val; - // check for replace_string(find): replace - } elseif ((substr($command, -1) == ')') && preg_match('!^([a-z0-9_]+)\((.*?)\)$!i', $command, $match)) { - if (in_array($match[1], array('replace_string'))) { - $command = $match[1]; - array_push($config->find_string, $match[2]); - array_push($config->$command, $val); - } - } - } - return $config; - } -} -?> \ No newline at end of file +tidy)) ? $this->tidy : $this->default_tidy; + return $this->tidy; + } + + // return bool or null + public function prune($use_default=true) { + if ($use_default) return (isset($this->prune)) ? $this->prune : $this->default_prune; + return $this->prune; + } + + // return string or null + public function parser($use_default=true) { + if ($use_default) return (isset($this->parser)) ? $this->parser : $this->default_parser; + return $this->parser; + } + + // return bool or null + public function autodetect_on_failure($use_default=true) { + if ($use_default) return (isset($this->autodetect_on_failure)) ? $this->autodetect_on_failure : $this->default_autodetect_on_failure; + return $this->autodetect_on_failure; + } + + public static function set_config_path($path, $fallback=null) { + self::$config_path = $path; + self::$config_path_fallback = $fallback; + } + + public static function add_to_cache($key, SiteConfig $config, $use_apc=true) { + $key = strtolower($key); + if (substr($key, 0, 4) == 'www.') $key = substr($key, 4); + if ($config->cache_key) $key = $config->cache_key; + self::$config_cache[$key] = $config; + if (self::$apc && $use_apc) { + self::debug("Adding site config to APC cache with key sc.$key"); + apc_add("sc.$key", $config); + } + self::debug("Cached site config with key $key"); + } + + public static function is_cached($key) { + $key = strtolower($key); + if (substr($key, 0, 4) == 'www.') $key = substr($key, 4); + if (array_key_exists($key, self::$config_cache)) { + return true; + } elseif (self::$apc && (bool)apc_fetch("sc.$key")) { + return true; + } + return false; + } + + public function append(SiteConfig $newconfig) { + // check for commands where we accept multiple statements (no test_url) + foreach (array('title', 'body', 'author', 'date', 'strip', 'strip_id_or_class', 'strip_image_src', 'single_page_link', 'single_page_link_in_feed', 'next_page_link', 'http_header') as $var) { + // append array elements for this config variable from $newconfig to this config + //$this->$var = $this->$var + $newconfig->$var; + $this->$var = array_unique(array_merge($this->$var, $newconfig->$var)); + } + // check for single statement commands + // we do not overwrite existing non null values + foreach (array('tidy', 'prune', 'parser', 'autodetect_on_failure') as $var) { + if ($this->$var === null) $this->$var = $newconfig->$var; + } + // treat find_string and replace_string separately (don't apply array_unique) (thanks fabrizio!) + foreach (array('find_string', 'replace_string') as $var) { + // append array elements for this config variable from $newconfig to this config + //$this->$var = $this->$var + $newconfig->$var; + $this->$var = array_merge($this->$var, $newconfig->$var); + } + } + + // returns SiteConfig instance if an appropriate one is found, false otherwise + // if $exact_host_match is true, we will not look for wildcard config matches + // by default if host is 'test.example.org' we will look for and load '.example.org.txt' if it exists + public static function build($host, $exact_host_match=false) { + $host = strtolower($host); + if (substr($host, 0, 4) == 'www.') $host = substr($host, 4); + if (!$host || (strlen($host) > 200) || !preg_match(self::HOSTNAME_REGEX, ltrim($host, '.'))) return false; + // check for site configuration + $try = array($host); + // should we look for wildcard matches + if (!$exact_host_match) { + $split = explode('.', $host); + if (count($split) > 1) { + array_shift($split); + $try[] = '.'.implode('.', $split); + } + } + + // look for site config file in primary folder + self::debug(". looking for site config for $host in primary folder"); + foreach ($try as $h) { + if (array_key_exists($h, self::$config_cache)) { + self::debug("... site config for $h already loaded in this request"); + return self::$config_cache[$h]; + } elseif (self::$apc && ($sconfig = apc_fetch("sc.$h"))) { + self::debug("... site config for $h in APC cache"); + return $sconfig; + } elseif (file_exists(self::$config_path."/$h.txt")) { + self::debug("... found site config ($h.txt)"); + $file_primary = self::$config_path."/$h.txt"; + $matched_name = $h; + break; + } + } + + // if we found site config, process it + if (isset($file_primary)) { + $config_lines = file($file_primary, FILE_IGNORE_NEW_LINES | FILE_SKIP_EMPTY_LINES); + if (!$config_lines || !is_array($config_lines)) return false; + $config = self::build_from_array($config_lines); + // if APC caching is available and enabled, mark this for cache + //$config->cache_in_apc = true; + $config->cache_key = $matched_name; + + // if autodetec on failure is off (on by default) we do not need to look + // in secondary folder + if (!$config->autodetect_on_failure()) { + self::debug('... autodetect on failure is disabled (no other site config files will be loaded)'); + return $config; + } + } + + // look for site config file in secondary folder + if (isset(self::$config_path_fallback)) { + self::debug(". looking for site config for $host in secondary folder"); + foreach ($try as $h) { + if (file_exists(self::$config_path_fallback."/$h.txt")) { + self::debug("... found site config in secondary folder ($h.txt)"); + $file_secondary = self::$config_path_fallback."/$h.txt"; + $matched_name = $h; + break; + } + } + if (!isset($file_secondary)) { + self::debug("... no site config match in secondary folder"); + } + } + + // return false if no config file found + if (!isset($file_primary) && !isset($file_secondary)) { + self::debug("... no site config match for $host"); + return false; + } + + // return primary config if secondary not found + if (!isset($file_secondary) && isset($config)) { + return $config; + } + + // process secondary config file + $config_lines = file($file_secondary, FILE_IGNORE_NEW_LINES | FILE_SKIP_EMPTY_LINES); + if (!$config_lines || !is_array($config_lines)) { + // failed to process secondary + if (isset($config)) { + // return primary config + return $config; + } else { + return false; + } + } + + // merge with primary and return + if (isset($config)) { + self::debug('. merging config files'); + $config->append(self::build_from_array($config_lines)); + return $config; + } else { + // return just secondary + $config = self::build_from_array($config_lines); + // if APC caching is available and enabled, mark this for cache + //$config->cache_in_apc = true; + $config->cache_key = $matched_name; + return $config; + } + } + + public static function build_from_array(array $lines) { + $config = new SiteConfig(); + foreach ($lines as $line) { + $line = trim($line); + + // skip comments, empty lines + if ($line == '' || $line[0] == '#') continue; + + // get command + $command = explode(':', $line, 2); + // if there's no colon ':', skip this line + if (count($command) != 2) continue; + $val = trim($command[1]); + $command = trim($command[0]); + if ($command == '' || $val == '') continue; + + // check for commands where we accept multiple statements + if (in_array($command, array('title', 'body', 'author', 'date', 'strip', 'strip_id_or_class', 'strip_image_src', 'single_page_link', 'single_page_link_in_feed', 'next_page_link', 'http_header', 'test_url', 'find_string', 'replace_string'))) { + array_push($config->$command, $val); + // check for single statement commands that evaluate to true or false + } elseif (in_array($command, array('tidy', 'prune', 'autodetect_on_failure'))) { + $config->$command = ($val == 'yes'); + // check for single statement commands stored as strings + } elseif (in_array($command, array('parser'))) { + $config->$command = $val; + // check for replace_string(find): replace + } elseif ((substr($command, -1) == ')') && preg_match('!^([a-z0-9_]+)\((.*?)\)$!i', $command, $match)) { + if (in_array($match[1], array('replace_string'))) { + $command = $match[1]; + array_push($config->find_string, $match[2]); + array_push($config->$command, $val); + } + } + } + return $config; + } +} \ No newline at end of file diff --git a/inc/3rdparty/libraries/feedwriter/FeedItem.php b/inc/3rdparty/libraries/feedwriter/FeedItem.php old mode 100644 new mode 100755 index 54a56f2..4078659 --- a/inc/3rdparty/libraries/feedwriter/FeedItem.php +++ b/inc/3rdparty/libraries/feedwriter/FeedItem.php @@ -1,7 +1,7 @@ version = $version; } /** * Set element (overwrites existing elements with $elementName) - * + * * @access public * @param srting The tag name of an element * @param srting The content of tag @@ -38,11 +38,11 @@ unset($this->elements[$elementName]); } $this->addElement($elementName, $content, $attributes); - } - + } + /** * Add an element to elements array - * + * * @access public * @param srting The tag name of an element * @param srting The content of tag @@ -61,11 +61,11 @@ $this->elements[$elementName][$i]['content'] = $content; $this->elements[$elementName][$i]['attributes'] = $attributes; } - + /** - * Set multiple feed elements from an array. + * Set multiple feed elements from an array. * Elements which have attributes cannot be added by this method - * + * * @access public * @param array array of elements in 'tagName' => 'tagContent' format. * @return void @@ -73,15 +73,15 @@ public function addElementArray($elementArray) { if(! is_array($elementArray)) return; - foreach ($elementArray as $elementName => $content) + foreach ($elementArray as $elementName => $content) { $this->addElement($elementName, $content); } } - + /** * Return the collection of elements in this feed item - * + * * @access public * @return array */ @@ -89,68 +89,74 @@ { return $this->elements; } - + // Wrapper functions ------------------------------------------------------ - + /** * Set the 'dscription' element of feed item - * + * * @access public * @param string The content of 'description' element * @return void */ - public function setDescription($description) + public function setDescription($description) { - $this->setElement('description', $description); + $tag = ($this->version == ATOM)? 'summary' : 'description'; + $this->setElement($tag, $description); } - + /** * @desc Set the 'title' element of feed item * @access public * @param string The content of 'title' element * @return void */ - public function setTitle($title) + public function setTitle($title) { - $this->setElement('title', $title); + $this->setElement('title', $title); } - + /** * Set the 'date' element of feed item - * + * * @access public * @param string The content of 'date' element * @return void */ - public function setDate($date) + public function setDate($date) { if(! is_numeric($date)) { $date = strtotime($date); } - - if($this->version == RSS2) + + if($this->version == ATOM) { - $tag = 'pubDate'; - $value = date(DATE_RSS, $date); + $tag = 'updated'; + $value = date(DATE_ATOM, $date); } - else + elseif($this->version == RSS2) { - $tag = 'dc:date'; - $value = date("Y-m-d", $date); + $tag = 'pubDate'; + $value = date(DATE_RSS, $date); } - - $this->setElement($tag, $value); + else + { + $tag = 'dc:date'; + $value = date("Y-m-d", $date); + } + + $this->setElement($tag, $value); } - + /** * Set the 'link' element of feed item - * + * * @access public * @param string The content of 'link' element * @return void */ - public function setLink($link) + public function setLink($link) { if($this->version == RSS2 || $this->version == RSS1) { @@ -161,27 +167,27 @@ { $this->setElement('link','',array('href'=>$link)); $this->setElement('id', FeedWriter::uuid($link,'urn:uuid:')); - } - + } + } /** * Set the 'source' element of feed item - * + * * @access public * @param string The content of 'source' element * @return void */ - public function setSource($link) + public function setSource($link) { $attributes = array('url'=>$link); $this->setElement('source', "wallabag",$attributes); } - + /** * Set the 'encloser' element of feed item * For RSS 2.0 only - * + * * @access public * @param string The url attribute of encloser tag * @param string The length attribute of encloser tag @@ -193,6 +199,6 @@ $attributes = array('url'=>$url, 'length'=>$length, 'type'=>$type); $this->setElement('enclosure','',$attributes); } - + } // end of class FeedItem ?> \ No newline at end of file diff --git a/inc/3rdparty/libraries/feedwriter/FeedWriter.php b/inc/3rdparty/libraries/feedwriter/FeedWriter.php index d708e99..7775569 100755 --- a/inc/3rdparty/libraries/feedwriter/FeedWriter.php +++ b/inc/3rdparty/libraries/feedwriter/FeedWriter.php @@ -97,15 +97,12 @@ define('JSONP', 3, true); header('X-content-type-options: nosniff'); } elseif ($this->version == JSON) { header('Content-type: application/json; charset=UTF-8'); + $this->json = new stdClass(); } elseif ($this->version == JSONP) { header('Content-type: application/javascript; charset=UTF-8'); + $this->json = new stdClass(); } } - - if ($this->version == JSON || $this->version == JSONP) { - $this->json = new stdClass(); - } - $this->printHead(); $this->printChannels(); @@ -116,6 +113,11 @@ define('JSONP', 3, true); } } + public function &getItems() + { + return $this->items; + } + /** * Create a new FeedItem. * @@ -199,7 +201,8 @@ define('JSONP', 3, true); */ public function setDescription($description) { - $this->setChannelElement('description', $description); + $tag = ($this->version == ATOM)? 'subtitle' : 'description'; + $this->setChannelElement($tag, $desciption); } /** @@ -244,7 +247,7 @@ define('JSONP', 3, true); { $out = ''."\n"; if ($this->xsl) $out .= 'xsl).'"?>' . PHP_EOL; - $out .= '' . PHP_EOL; + $out .= '' . PHP_EOL; echo $out; } elseif ($this->version == JSON || $this->version == JSONP) diff --git a/inc/3rdparty/libraries/html5/TreeBuilder.php b/inc/3rdparty/libraries/html5/TreeBuilder.php index 2f5244f..c4a48b2 100644 --- a/inc/3rdparty/libraries/html5/TreeBuilder.php +++ b/inc/3rdparty/libraries/html5/TreeBuilder.php @@ -134,6 +134,7 @@ class HTML5_TreeBuilder { // Namespaces for foreign content const NS_HTML = null; // to prevent DOM from requiring NS on everything + const NS_XHTML = 'http://www.w3.org/1999/xhtml'; const NS_MATHML = 'http://www.w3.org/1998/Math/MathML'; const NS_SVG = 'http://www.w3.org/2000/svg'; const NS_XLINK = 'http://www.w3.org/1999/xlink'; @@ -3157,11 +3158,19 @@ class HTML5_TreeBuilder { } private function insertElement($token, $append = true) { - $el = $this->dom->createElementNS(self::NS_HTML, $token['name']); + //$el = $this->dom->createElementNS(self::NS_HTML, $token['name']); + $namespaceURI = strpos($token['name'], ':') ? self::NS_XHTML : self::NS_HTML; + $el = $this->dom->createElementNS($namespaceURI, $token['name']); if (!empty($token['attr'])) { foreach($token['attr'] as $attr) { - if(!$el->hasAttribute($attr['name'])) { + + // mike@macgirvin.com 2011-11-17, check attribute name for + // validity (ignoring extenders and combiners) as illegal chars in names + // causes everything to abort + + $valid = preg_match('/^[a-zA-Z\_\:]([\-a-zA-Z0-9\_\:\.]+$)/',$attr['name']); + if($attr['name'] && (!$el->hasAttribute($attr['name'])) && ($valid)) { $el->setAttribute($attr['name'], $attr['value']); } } diff --git a/inc/3rdparty/libraries/humble-http-agent/CookieJar.php b/inc/3rdparty/libraries/humble-http-agent/CookieJar.php index 83e94f1..e4d5f49 100644 --- a/inc/3rdparty/libraries/humble-http-agent/CookieJar.php +++ b/inc/3rdparty/libraries/humble-http-agent/CookieJar.php @@ -1,404 +1,403 @@ - - * - * This class should be used to handle cookies (storing cookies from HTTP response messages, and - * sending out cookies in HTTP request messages). This has been adapted for FiveFilters.org - * from the original version used in HTTP Navigator. See http://www.keyvan.net/code/http-navigator/ - * - * This class is mainly based on Cookies.pm from the libwww-perl collection . - * Unlike Cookies.pm, this class only supports the Netscape cookie spec, not RFC 2965. - * - * @version 0.5 - * @date 2011-03-15 - * @see http://php.net/HttpRequestPool - * @author Keyvan Minoukadeh - * @copyright 2011 Keyvan Minoukadeh - * @license http://www.gnu.org/licenses/agpl-3.0.html AGPL v3 - */ - -class CookieJar -{ - /** - * Cookies - array containing all cookies. - * - * 
-    * Cookies are stored like this:
-    *   [domain][path][name] = array
-    * where array is:
-    *   0 => value, 1 => secure, 2 => expires
-    *
- * @var array - * @access private - */ - public $cookies = array(); - public $debug = false; - - /** - * Constructor - */ - function __construct() { - } - - protected function debug($msg, $file=null, $line=null) { - if ($this->debug) { - $mem = round(memory_get_usage()/1024, 2); - $memPeak = round(memory_get_peak_usage()/1024, 2); - echo '* ',$msg; - if (isset($file, $line)) echo " ($file line $line)"; - echo ' - mem used: ',$mem," (peak: $memPeak)\n"; - ob_flush(); - flush(); - } - } - - /** - * Get matching cookies - * - * Only use this method if you cannot use add_cookie_header(), for example, if you want to use - * this cookie jar class without using the request class. - * - * @param array $param associative array containing 'domain', 'path', 'secure' keys - * @return string - * @see add_cookie_header() - */ - public function getMatchingCookies($url) - { - if (($parts = @parse_url($url)) && isset($parts['scheme'], $parts['host'], $parts['path'])) { - $param['domain'] = $parts['host']; - $param['path'] = $parts['path']; - $param['secure'] = (strtolower($parts['scheme']) == 'https'); - unset($parts); - } else { - return false; - } - // RFC 2965 notes: - // If multiple cookies satisfy the criteria above, they are ordered in - // the Cookie header such that those with more specific Path attributes - // precede those with less specific. Ordering with respect to other - // attributes (e.g., Domain) is unspecified. - $domain = $param['domain']; - if (strpos($domain, '.') === false) $domain .= '.local'; - $request_path = $param['path']; - if ($request_path == '') $request_path = '/'; - $request_secure = $param['secure']; - $now = time(); - $matched_cookies = array(); - // domain - find matching domains - $this->debug('Finding matching domains for '.$domain, __FILE__, __LINE__); - while (strpos($domain, '.') !== false) { - if (isset($this->cookies[$domain])) { - $this->debug(' domain match found: '.$domain); - $cookies =& $this->cookies[$domain]; - } else { - $domain = $this->_reduce_domain($domain); - continue; - } - // paths - find matching paths starting from most specific - $this->debug(' - Finding matching paths for '.$request_path); - $paths = array_keys($cookies); - usort($paths, array($this, '_cmp_length')); - foreach ($paths as $path) { - // continue to next cookie if request path does not path-match cookie path - if (!$this->_path_match($request_path, $path)) continue; - // loop through cookie names - $this->debug(' path match found: '.$path); - foreach ($cookies[$path] as $name => $values) { - // if this cookie is secure but request isn't, continue to next cookie - if ($values[1] && !$request_secure) continue; - // if cookie is not a session cookie and has expired, continue to next cookie - if (is_int($values[2]) && ($values[2] < $now)) continue; - // cookie matches request - $this->debug(' cookie match: '.$name.'='.$values[0]); - $matched_cookies[] = $name.'='.$values[0]; - } - } - $domain = $this->_reduce_domain($domain); - } - // return cookies - return implode('; ', $matched_cookies); - } - - /** - * Parse Set-Cookie values. - * - * Only use this method if you cannot use extract_cookies(), for example, if you want to use - * this cookie jar class without using the response class. - * - * @param array $set_cookies array holding 1 or more "Set-Cookie" header values - * @param array $param associative array containing 'host', 'path' keys - * @return void - * @see extract_cookies() - */ - public function storeCookies($url, $set_cookies) - { - if (count($set_cookies) == 0) return; - $param = @parse_url($url); - if (!is_array($param) || !isset($param['host'])) return; - $request_host = $param['host']; - if (strpos($request_host, '.') === false) $request_host .= '.local'; - $request_path = @$param['path']; - if ($request_path == '') $request_path = '/'; - // - // loop through set-cookie headers - // - foreach ($set_cookies as $set_cookie) { - $this->debug('Parsing: '.$set_cookie); - // temporary cookie store (before adding to jar) - $tmp_cookie = array(); - $param = explode(';', $set_cookie); - // loop through params - for ($x=0; $x$key, 'value'=>$val); - continue; - } - $key = strtolower($key); - if (in_array($key, array('expires', 'path', 'domain', 'secure'))) { - $tmp_cookie[$key] = $val; - } - } - // - // set cookie - // - // check domain - if (isset($tmp_cookie['domain']) && ($tmp_cookie['domain'] != $request_host) && - ($tmp_cookie['domain'] != ".$request_host")) { - $domain = $tmp_cookie['domain']; - if ((strpos($domain, '.') === false) && ($domain != 'local')) { - $this->debug(' - domain "'.$domain.'" has no dot and is not a local domain'); - continue; - } - if (preg_match('/\.[0-9]+$/', $domain)) { - $this->debug(' - domain "'.$domain.'" appears to be an ip address'); - continue; - } - if (substr($domain, 0, 1) != '.') $domain = ".$domain"; - if (!$this->_domain_match($request_host, $domain)) { - $this->debug(' - request host "'.$request_host.'" does not domain-match "'.$domain.'"'); - continue; - } - } else { - // if domain is not specified in the set-cookie header, domain will default to - // the request host - $domain = $request_host; - } - // check path - if (isset($tmp_cookie['path']) && ($tmp_cookie['path'] != '')) { - $path = urldecode($tmp_cookie['path']); - if (!$this->_path_match($request_path, $path)) { - $this->debug(' - request path "'.$request_path.'" does not path-match "'.$path.'"'); - continue; - } - } else { - $path = $request_path; - $path = substr($path, 0, strrpos($path, '/')); - if ($path == '') $path = '/'; - } - // check if secure - $secure = (isset($tmp_cookie['secure'])) ? true : false; - // check expiry - if (isset($tmp_cookie['expires'])) { - if (($expires = strtotime($tmp_cookie['expires'])) < 0) { - $expires = null; - } - } else { - $expires = null; - } - // set cookie - $this->set_cookie($domain, $path, $tmp_cookie['name'], $tmp_cookie['value'], $secure, $expires); - } - } - - // return array of set-cookie values extracted from HTTP response headers (string $h) - public function extractCookies($h) { - $x = 0; - $lines = 0; - $headers = array(); - $last_match = false; - $h = explode("\n", $h); - foreach ($h as $line) { - $line = rtrim($line); - $lines++; - - $trimmed_line = trim($line); - if (isset($line_last)) { - // check if we have \r\n\r\n (indicating the end of headers) - // some servers will not use CRLF (\r\n), so we make CR (\r) optional. - // if (preg_match('/\015?\012\015?\012/', $line_last.$line)) { - // break; - // } - // As an alternative, we can check if the current trimmed line is empty - if ($trimmed_line == '') { - break; - } - - // check for continuation line... - // RFC 2616 Section 2.2 "Basic Rules": - // HTTP/1.1 header field values can be folded onto multiple lines if the - // continuation line begins with a space or horizontal tab. All linear - // white space, including folding, has the same semantics as SP. A - // recipient MAY replace any linear white space with a single SP before - // interpreting the field value or forwarding the message downstream. - if ($last_match && preg_match('/^\s+(.*)/', $line, $match)) { - // append to previous header value - $headers[$x-1] .= ' '.rtrim($match[1]); - continue; - } - } - $line_last = $line; - - // split header name and value - if (preg_match('/^Set-Cookie\s*:\s*(.*)/i', $line, $match)) { - $headers[$x++] = rtrim($match[1]); - $last_match = true; - } else { - $last_match = false; - } - } - return $headers; - } - - /** - * Set Cookie - * @param string $domain - * @param string $path - * @param string $name cookie name - * @param string $value cookie value - * @param bool $secure - * @param int $expires expiry time (null if session cookie, <= 0 will delete cookie) - * @return void - */ - function set_cookie($domain, $path, $name, $value, $secure=false, $expires=null) - { - if ($domain == '') return; - if ($path == '') return; - if ($name == '') return; - // check if cookie needs to go - if (isset($expires) && ($expires <= 0)) { - if (isset($this->cookies[$domain][$path][$name])) unset($this->cookies[$domain][$path][$name]); - return; - } - if ($value == '') return; - $this->cookies[$domain][$path][$name] = array($value, $secure, $expires); - return; - } - - /** - * Clear cookies - [domain [,path [,name]]] - call method with no arguments to clear all cookies. - * @param string $domain - * @param string $path - * @param string $name - * @return void - */ - function clear($domain=null, $path=null, $name=null) - { - if (!isset($domain)) { - $this->cookies = array(); - } elseif (!isset($path)) { - if (isset($this->cookies[$domain])) unset($this->cookies[$domain]); - } elseif (!isset($name)) { - if (isset($this->cookies[$domain][$path])) unset($this->cookies[$domain][$path]); - } elseif (isset($name)) { - if (isset($this->cookies[$domain][$path][$name])) unset($this->cookies[$domain][$path][$name]); - } - } - - /** - * Compare string length - used for sorting - * @access private - * @return int - */ - function _cmp_length($a, $b) - { - $la = strlen($a); $lb = strlen($b); - if ($la == $lb) return 0; - return ($la > $lb) ? -1 : 1; - } - - /** - * Reduce domain - * @param string $domain - * @return string - * @access private - */ - function _reduce_domain($domain) - { - if ($domain == '') return ''; - if (substr($domain, 0, 1) == '.') return substr($domain, 1); - return substr($domain, strpos($domain, '.')); - } - - /** - * Path match - check if path1 path-matches path2 - * - * From RFC 2965: - * For two strings that represent paths, P1 and P2, P1 path-matches P2 - * if P2 is a prefix of P1 (including the case where P1 and P2 string- - * compare equal). Thus, the string /tec/waldo path-matches /tec. - * @param string $path1 - * @param string $path2 - * @return bool - * @access private - */ - function _path_match($path1, $path2) - { - return (substr($path1, 0, strlen($path2)) == $path2); - } - - /** - * Domain match - check if domain1 domain-matches domain2 - * - * A few extracts from RFC 2965: - * - A Set-Cookie2 from request-host y.x.foo.com for Domain=.foo.com - * would be rejected, because H is y.x and contains a dot. - * - * - A Set-Cookie2 from request-host x.foo.com for Domain=.foo.com - * would be accepted. - * - * - A Set-Cookie2 with Domain=.com or Domain=.com., will always be - * rejected, because there is no embedded dot. - * - * - A Set-Cookie2 from request-host example for Domain=.local will - * be accepted, because the effective host name for the request- - * host is example.local, and example.local domain-matches .local. - * - * I'm ignoring the first point for now (must check to see how other browsers handle - * this rule for Set-Cookie headers) - * - * @param string $domain1 - * @param string $domain2 - * @return bool - * @access private - */ - function _domain_match($domain1, $domain2) - { - $domain1 = strtolower($domain1); - $domain2 = strtolower($domain2); - while (strpos($domain1, '.') !== false) { - if ($domain1 == $domain2) return true; - $domain1 = $this->_reduce_domain($domain1); - continue; - } - return false; - } -} -?> \ No newline at end of file + + * + * This class should be used to handle cookies (storing cookies from HTTP response messages, and + * sending out cookies in HTTP request messages). This has been adapted for FiveFilters.org + * from the original version used in HTTP Navigator. See http://www.keyvan.net/code/http-navigator/ + * + * This class is mainly based on Cookies.pm from the libwww-perl collection . + * Unlike Cookies.pm, this class only supports the Netscape cookie spec, not RFC 2965. + * + * @version 0.5 + * @date 2011-03-15 + * @see http://php.net/HttpRequestPool + * @author Keyvan Minoukadeh + * @copyright 2011 Keyvan Minoukadeh + * @license http://www.gnu.org/licenses/agpl-3.0.html AGPL v3 + */ + +class CookieJar +{ + /** + * Cookies - array containing all cookies. + * + * 
+    * Cookies are stored like this:
+    *   [domain][path][name] = array
+    * where array is:
+    *   0 => value, 1 => secure, 2 => expires
+    *
+ * @var array + * @access private + */ + public $cookies = array(); + public $debug = false; + + /** + * Constructor + */ + function __construct() { + } + + protected function debug($msg, $file=null, $line=null) { + if ($this->debug) { + $mem = round(memory_get_usage()/1024, 2); + $memPeak = round(memory_get_peak_usage()/1024, 2); + echo '* ',$msg; + if (isset($file, $line)) echo " ($file line $line)"; + echo ' - mem used: ',$mem," (peak: $memPeak)\n"; + ob_flush(); + flush(); + } + } + + /** + * Get matching cookies + * + * Only use this method if you cannot use add_cookie_header(), for example, if you want to use + * this cookie jar class without using the request class. + * + * @param array $param associative array containing 'domain', 'path', 'secure' keys + * @return string + * @see add_cookie_header() + */ + public function getMatchingCookies($url) + { + if (($parts = @parse_url($url)) && isset($parts['scheme'], $parts['host'], $parts['path'])) { + $param['domain'] = $parts['host']; + $param['path'] = $parts['path']; + $param['secure'] = (strtolower($parts['scheme']) == 'https'); + unset($parts); + } else { + return false; + } + // RFC 2965 notes: + // If multiple cookies satisfy the criteria above, they are ordered in + // the Cookie header such that those with more specific Path attributes + // precede those with less specific. Ordering with respect to other + // attributes (e.g., Domain) is unspecified. + $domain = $param['domain']; + if (strpos($domain, '.') === false) $domain .= '.local'; + $request_path = $param['path']; + if ($request_path == '') $request_path = '/'; + $request_secure = $param['secure']; + $now = time(); + $matched_cookies = array(); + // domain - find matching domains + $this->debug('Finding matching domains for '.$domain, __FILE__, __LINE__); + while (strpos($domain, '.') !== false) { + if (isset($this->cookies[$domain])) { + $this->debug(' domain match found: '.$domain); + $cookies =& $this->cookies[$domain]; + } else { + $domain = $this->_reduce_domain($domain); + continue; + } + // paths - find matching paths starting from most specific + $this->debug(' - Finding matching paths for '.$request_path); + $paths = array_keys($cookies); + usort($paths, array($this, '_cmp_length')); + foreach ($paths as $path) { + // continue to next cookie if request path does not path-match cookie path + if (!$this->_path_match($request_path, $path)) continue; + // loop through cookie names + $this->debug(' path match found: '.$path); + foreach ($cookies[$path] as $name => $values) { + // if this cookie is secure but request isn't, continue to next cookie + if ($values[1] && !$request_secure) continue; + // if cookie is not a session cookie and has expired, continue to next cookie + if (is_int($values[2]) && ($values[2] < $now)) continue; + // cookie matches request + $this->debug(' cookie match: '.$name.'='.$values[0]); + $matched_cookies[] = $name.'='.$values[0]; + } + } + $domain = $this->_reduce_domain($domain); + } + // return cookies + return implode('; ', $matched_cookies); + } + + /** + * Parse Set-Cookie values. + * + * Only use this method if you cannot use extract_cookies(), for example, if you want to use + * this cookie jar class without using the response class. + * + * @param array $set_cookies array holding 1 or more "Set-Cookie" header values + * @param array $param associative array containing 'host', 'path' keys + * @return void + * @see extract_cookies() + */ + public function storeCookies($url, $set_cookies) + { + if (count($set_cookies) == 0) return; + $param = @parse_url($url); + if (!is_array($param) || !isset($param['host'])) return; + $request_host = $param['host']; + if (strpos($request_host, '.') === false) $request_host .= '.local'; + $request_path = @$param['path']; + if ($request_path == '') $request_path = '/'; + // + // loop through set-cookie headers + // + foreach ($set_cookies as $set_cookie) { + $this->debug('Parsing: '.$set_cookie); + // temporary cookie store (before adding to jar) + $tmp_cookie = array(); + $param = explode(';', $set_cookie); + // loop through params + for ($x=0; $x$key, 'value'=>$val); + continue; + } + $key = strtolower($key); + if (in_array($key, array('expires', 'path', 'domain', 'secure'))) { + $tmp_cookie[$key] = $val; + } + } + // + // set cookie + // + // check domain + if (isset($tmp_cookie['domain']) && ($tmp_cookie['domain'] != $request_host) && + ($tmp_cookie['domain'] != ".$request_host")) { + $domain = $tmp_cookie['domain']; + if ((strpos($domain, '.') === false) && ($domain != 'local')) { + $this->debug(' - domain "'.$domain.'" has no dot and is not a local domain'); + continue; + } + if (preg_match('/\.[0-9]+$/', $domain)) { + $this->debug(' - domain "'.$domain.'" appears to be an ip address'); + continue; + } + if (substr($domain, 0, 1) != '.') $domain = ".$domain"; + if (!$this->_domain_match($request_host, $domain)) { + $this->debug(' - request host "'.$request_host.'" does not domain-match "'.$domain.'"'); + continue; + } + } else { + // if domain is not specified in the set-cookie header, domain will default to + // the request host + $domain = $request_host; + } + // check path + if (isset($tmp_cookie['path']) && ($tmp_cookie['path'] != '')) { + $path = urldecode($tmp_cookie['path']); + if (!$this->_path_match($request_path, $path)) { + $this->debug(' - request path "'.$request_path.'" does not path-match "'.$path.'"'); + continue; + } + } else { + $path = $request_path; + $path = substr($path, 0, strrpos($path, '/')); + if ($path == '') $path = '/'; + } + // check if secure + $secure = (isset($tmp_cookie['secure'])) ? true : false; + // check expiry + if (isset($tmp_cookie['expires'])) { + if (($expires = strtotime($tmp_cookie['expires'])) < 0) { + $expires = null; + } + } else { + $expires = null; + } + // set cookie + $this->set_cookie($domain, $path, $tmp_cookie['name'], $tmp_cookie['value'], $secure, $expires); + } + } + + // return array of set-cookie values extracted from HTTP response headers (string $h) + public function extractCookies($h) { + $x = 0; + $lines = 0; + $headers = array(); + $last_match = false; + $h = explode("\n", $h); + foreach ($h as $line) { + $line = rtrim($line); + $lines++; + + $trimmed_line = trim($line); + if (isset($line_last)) { + // check if we have \r\n\r\n (indicating the end of headers) + // some servers will not use CRLF (\r\n), so we make CR (\r) optional. + // if (preg_match('/\015?\012\015?\012/', $line_last.$line)) { + // break; + // } + // As an alternative, we can check if the current trimmed line is empty + if ($trimmed_line == '') { + break; + } + + // check for continuation line... + // RFC 2616 Section 2.2 "Basic Rules": + // HTTP/1.1 header field values can be folded onto multiple lines if the + // continuation line begins with a space or horizontal tab. All linear + // white space, including folding, has the same semantics as SP. A + // recipient MAY replace any linear white space with a single SP before + // interpreting the field value or forwarding the message downstream. + if ($last_match && preg_match('/^\s+(.*)/', $line, $match)) { + // append to previous header value + $headers[$x-1] .= ' '.rtrim($match[1]); + continue; + } + } + $line_last = $line; + + // split header name and value + if (preg_match('/^Set-Cookie\s*:\s*(.*)/i', $line, $match)) { + $headers[$x++] = rtrim($match[1]); + $last_match = true; + } else { + $last_match = false; + } + } + return $headers; + } + + /** + * Set Cookie + * @param string $domain + * @param string $path + * @param string $name cookie name + * @param string $value cookie value + * @param bool $secure + * @param int $expires expiry time (null if session cookie, <= 0 will delete cookie) + * @return void + */ + function set_cookie($domain, $path, $name, $value, $secure=false, $expires=null) + { + if ($domain == '') return; + if ($path == '') return; + if ($name == '') return; + // check if cookie needs to go + if (isset($expires) && ($expires <= 0)) { + if (isset($this->cookies[$domain][$path][$name])) unset($this->cookies[$domain][$path][$name]); + return; + } + if ($value == '') return; + $this->cookies[$domain][$path][$name] = array($value, $secure, $expires); + return; + } + + /** + * Clear cookies - [domain [,path [,name]]] - call method with no arguments to clear all cookies. + * @param string $domain + * @param string $path + * @param string $name + * @return void + */ + function clear($domain=null, $path=null, $name=null) + { + if (!isset($domain)) { + $this->cookies = array(); + } elseif (!isset($path)) { + if (isset($this->cookies[$domain])) unset($this->cookies[$domain]); + } elseif (!isset($name)) { + if (isset($this->cookies[$domain][$path])) unset($this->cookies[$domain][$path]); + } elseif (isset($name)) { + if (isset($this->cookies[$domain][$path][$name])) unset($this->cookies[$domain][$path][$name]); + } + } + + /** + * Compare string length - used for sorting + * @access private + * @return int + */ + function _cmp_length($a, $b) + { + $la = strlen($a); $lb = strlen($b); + if ($la == $lb) return 0; + return ($la > $lb) ? -1 : 1; + } + + /** + * Reduce domain + * @param string $domain + * @return string + * @access private + */ + function _reduce_domain($domain) + { + if ($domain == '') return ''; + if (substr($domain, 0, 1) == '.') return substr($domain, 1); + return substr($domain, strpos($domain, '.')); + } + + /** + * Path match - check if path1 path-matches path2 + * + * From RFC 2965: + * For two strings that represent paths, P1 and P2, P1 path-matches P2 + * if P2 is a prefix of P1 (including the case where P1 and P2 string- + * compare equal). Thus, the string /tec/waldo path-matches /tec. + * @param string $path1 + * @param string $path2 + * @return bool + * @access private + */ + function _path_match($path1, $path2) + { + return (substr($path1, 0, strlen($path2)) == $path2); + } + + /** + * Domain match - check if domain1 domain-matches domain2 + * + * A few extracts from RFC 2965: + * - A Set-Cookie2 from request-host y.x.foo.com for Domain=.foo.com + * would be rejected, because H is y.x and contains a dot. + * + * - A Set-Cookie2 from request-host x.foo.com for Domain=.foo.com + * would be accepted. + * + * - A Set-Cookie2 with Domain=.com or Domain=.com., will always be + * rejected, because there is no embedded dot. + * + * - A Set-Cookie2 from request-host example for Domain=.local will + * be accepted, because the effective host name for the request- + * host is example.local, and example.local domain-matches .local. + * + * I'm ignoring the first point for now (must check to see how other browsers handle + * this rule for Set-Cookie headers) + * + * @param string $domain1 + * @param string $domain2 + * @return bool + * @access private + */ + function _domain_match($domain1, $domain2) + { + $domain1 = strtolower($domain1); + $domain2 = strtolower($domain2); + while (strpos($domain1, '.') !== false) { + if ($domain1 == $domain2) return true; + $domain1 = $this->_reduce_domain($domain1); + continue; + } + return false; + } +} \ No newline at end of file diff --git a/inc/3rdparty/libraries/humble-http-agent/HumbleHttpAgent.php b/inc/3rdparty/libraries/humble-http-agent/HumbleHttpAgent.php index e4f1b3b..963f0c0 100644 --- a/inc/3rdparty/libraries/humble-http-agent/HumbleHttpAgent.php +++ b/inc/3rdparty/libraries/humble-http-agent/HumbleHttpAgent.php @@ -1,779 +1,810 @@ -userAgentDefault = self::UA_BROWSER; - $this->referer = self::REF_GOOGLE; - // set the request method - if (in_array($method, array(1,2,4))) { - $this->method = $method; - } else { - if (class_exists('HttpRequestPool')) { - $this->method = self::METHOD_REQUEST_POOL; - } elseif (function_exists('curl_multi_init')) { - $this->method = self::METHOD_CURL_MULTI; - } else { - $this->method = self::METHOD_FILE_GET_CONTENTS; - } - } - if ($this->method == self::METHOD_CURL_MULTI) { - require_once(dirname(__FILE__).'/RollingCurl.php'); - } - // create cookie jar - $this->cookieJar = new CookieJar(); - // set request options (redirect must be 0) - $this->requestOptions = array( - 'timeout' => 15, - 'redirect' => 0 // we handle redirects manually so we can rewrite the new hashbang URLs that are creeping up over the web - // TODO: test onprogress? - ); - if (is_array($requestOptions)) { - $this->requestOptions = array_merge($this->requestOptions, $requestOptions); - } - $this->httpContext = array( - 'http' => array( - 'ignore_errors' => true, - 'timeout' => $this->requestOptions['timeout'], - 'max_redirects' => $this->requestOptions['redirect'], - 'header' => "Accept: */*\r\n" - ) - ); - } - - protected function debug($msg) { - if ($this->debug) { - $mem = round(memory_get_usage()/1024, 2); - $memPeak = round(memory_get_peak_usage()/1024, 2); - echo '* ',$msg; - if ($this->debugVerbose) echo ' - mem used: ',$mem," (peak: $memPeak)"; - echo "\n"; - ob_flush(); - flush(); - } - } - - protected function getUserAgent($url, $asArray=false) { - $host = @parse_url($url, PHP_URL_HOST); - if (strtolower(substr($host, 0, 4)) == 'www.') { - $host = substr($host, 4); - } - if ($host) { - $try = array($host); - $split = explode('.', $host); - if (count($split) > 1) { - array_shift($split); - $try[] = '.'.implode('.', $split); - } - foreach ($try as $h) { - if (isset($this->userAgentMap[$h])) { - $ua = $this->userAgentMap[$h]; - break; - } - } - } - if (!isset($ua)) $ua = $this->userAgentDefault; - if ($asArray) { - return array('User-Agent' => $ua); - } else { - return 'User-Agent: '.$ua; - } - } - - public function rewriteHashbangFragment($url) { - // return $url if there's no '#!' - if (strpos($url, '#!') === false) return $url; - // split $url and rewrite - // TODO: is SimplePie_IRI included? - $iri = new SimplePie_IRI($url); - $fragment = substr($iri->fragment, 1); // strip '!' - $iri->fragment = null; - if (isset($iri->query)) { - parse_str($iri->query, $query); - } else { - $query = array(); - } - $query['_escaped_fragment_'] = (string)$fragment; - $iri->query = str_replace('%2F', '/', http_build_query($query)); // needed for some sites - return $iri->get_iri(); - } - - public function getUglyURL($url, $html) { - if ($html == '') return false; - $found = false; - foreach ($this->ajaxTriggers as $string) { - if (stripos($html, $string)) { - $found = true; - break; - } - } - if (!$found) return false; - $iri = new SimplePie_IRI($url); - if (isset($iri->query)) { - parse_str($iri->query, $query); - } else { - $query = array(); - } - $query['_escaped_fragment_'] = ''; - $iri->query = str_replace('%2F', '/', http_build_query($query)); // needed for some sites - return $iri->get_iri(); - } - - public function removeFragment($url) { - $pos = strpos($url, '#'); - if ($pos === false) { - return $url; - } else { - return substr($url, 0, $pos); - } - } - - public function rewriteUrls($url) { - foreach ($this->rewriteUrls as $find => $action) { - if (strpos($url, $find) !== false) { - if (is_array($action)) { - return strtr($url, $action); - } - } - } - return $url; - } - - public function enableDebug($bool=true) { - $this->debug = (bool)$bool; - } - - public function minimiseMemoryUse($bool = true) { - $this->minimiseMemoryUse = $bool; - } - - public function setMaxParallelRequests($max) { - $this->maxParallelRequests = $max; - } - - public function validateUrl($url) { - $url = filter_var($url, FILTER_SANITIZE_URL); - $test = filter_var($url, FILTER_VALIDATE_URL, FILTER_FLAG_SCHEME_REQUIRED); - // deal with bug http://bugs.php.net/51192 (present in PHP 5.2.13 and PHP 5.3.2) - if ($test === false) { - $test = filter_var(strtr($url, '-', '_'), FILTER_VALIDATE_URL, FILTER_FLAG_SCHEME_REQUIRED); - } - if ($test !== false && $test !== null && preg_match('!^https?://!', $url)) { - return $url; - } else { - return false; - } - } - - public function fetchAll(array $urls) { - $this->fetchAllOnce($urls, $isRedirect=false); - $redirects = 0; - while (!empty($this->redirectQueue) && ++$redirects <= $this->maxRedirects) { - $this->debug("Following redirects #$redirects..."); - $this->fetchAllOnce($this->redirectQueue, $isRedirect=true); - } - } - - // fetch all URLs without following redirects - public function fetchAllOnce(array $urls, $isRedirect=false) { - if (!$isRedirect) $urls = array_unique($urls); - if (empty($urls)) return; - - ////////////////////////////////////////////////////// - // parallel (HttpRequestPool) - if ($this->method == self::METHOD_REQUEST_POOL) { - $this->debug('Starting parallel fetch (HttpRequestPool)'); - try { - while (count($urls) > 0) { - $this->debug('Processing set of '.min($this->maxParallelRequests, count($urls))); - $subset = array_splice($urls, 0, $this->maxParallelRequests); - $pool = new HttpRequestPool(); - foreach ($subset as $orig => $url) { - if (!$isRedirect) $orig = $url; - unset($this->redirectQueue[$orig]); - $this->debug("...$url"); - if (!$isRedirect && isset($this->requests[$url])) { - $this->debug("......in memory"); - /* - } elseif ($this->isCached($url)) { - $this->debug("......is cached"); - if (!$this->minimiseMemoryUse) { - $this->requests[$url] = $this->getCached($url); - } - */ - } else { - $this->debug("......adding to pool"); - $req_url = $this->rewriteUrls($url); - $req_url = ($this->rewriteHashbangFragment) ? $this->rewriteHashbangFragment($req_url) : $req_url; - $req_url = $this->removeFragment($req_url); - if (!empty($this->headerOnlyTypes) && !isset($this->requests[$orig]['wrongGuess']) && $this->possibleUnsupportedType($req_url)) { - $_meth = HttpRequest::METH_HEAD; - } else { - $_meth = HttpRequest::METH_GET; - unset($this->requests[$orig]['wrongGuess']); - } - $httpRequest = new HttpRequest($req_url, $_meth, $this->requestOptions); - // send cookies, if we have any - if ($cookies = $this->cookieJar->getMatchingCookies($req_url)) { - $this->debug("......sending cookies: $cookies"); - $httpRequest->addHeaders(array('Cookie' => $cookies)); - } - //$httpRequest->addHeaders(array('User-Agent' => $this->userAgent)); - $httpRequest->addHeaders($this->getUserAgent($req_url, true)); - // add referer for picky sites - $httpRequest->addheaders(array('Referer' => $this->referer)); - $this->requests[$orig] = array('headers'=>null, 'body'=>null, 'httpRequest'=>$httpRequest); - $this->requests[$orig]['original_url'] = $orig; - $pool->attach($httpRequest); - } - } - // did we get anything into the pool? - if (count($pool) > 0) { - $this->debug('Sending request...'); - try { - $pool->send(); - } catch (HttpRequestPoolException $e) { - // do nothing - } - $this->debug('Received responses'); - foreach($subset as $orig => $url) { - if (!$isRedirect) $orig = $url; - $request = $this->requests[$orig]['httpRequest']; - //$this->requests[$orig]['headers'] = $this->headersToString($request->getResponseHeader()); - // getResponseHeader() doesn't return status line, so, for consistency... - $this->requests[$orig]['headers'] = substr($request->getRawResponseMessage(), 0, $request->getResponseInfo('header_size')); - // check content type - // TODO: use getResponseHeader('content-type') or getResponseInfo() - if ($this->headerOnlyType($this->requests[$orig]['headers'])) { - $this->requests[$orig]['body'] = ''; - $_header_only_type = true; - $this->debug('Header only type returned'); - } else { - $this->requests[$orig]['body'] = $request->getResponseBody(); - $_header_only_type = false; - } - $this->requests[$orig]['effective_url'] = $request->getResponseInfo('effective_url'); - $this->requests[$orig]['status_code'] = $status_code = $request->getResponseCode(); - // is redirect? - if ((in_array($status_code, array(300, 301, 302, 303, 307)) || $status_code > 307 && $status_code < 400) && $request->getResponseHeader('location')) { - $redirectURL = $request->getResponseHeader('location'); - if (!preg_match('!^https?://!i', $redirectURL)) { - $redirectURL = SimplePie_Misc::absolutize_url($redirectURL, $url); - } - if ($this->validateURL($redirectURL)) { - $this->debug('Redirect detected. Valid URL: '.$redirectURL); - // store any cookies - $cookies = $request->getResponseHeader('set-cookie'); - if ($cookies && !is_array($cookies)) $cookies = array($cookies); - if ($cookies) $this->cookieJar->storeCookies($url, $cookies); - $this->redirectQueue[$orig] = $redirectURL; - } else { - $this->debug('Redirect detected. Invalid URL: '.$redirectURL); - } - } elseif (!$_header_only_type && $request->getMethod() === HttpRequest::METH_HEAD) { - // the response content-type did not match our 'header only' types, - // but we'd issues a HEAD request because we assumed it would. So - // let's queue a proper GET request for this item... - $this->debug('Wrong guess at content-type, queing GET request'); - $this->requests[$orig]['wrongGuess'] = true; - $this->redirectQueue[$orig] = $this->requests[$orig]['effective_url']; - } elseif (strpos($this->requests[$orig]['effective_url'], '_escaped_fragment_') === false) { - // check for - // for AJAX sites, e.g. Blogger with its dynamic views templates. - // Based on Google's spec: https://developers.google.com/webmasters/ajax-crawling/docs/specification - if (isset($this->requests[$orig]['body'])) { - $redirectURL = $this->getUglyURL($this->requests[$orig]['effective_url'], substr($this->requests[$orig]['body'], 0, 4000)); - if ($redirectURL) { - $this->debug('AJAX trigger (meta name="fragment" content="!") found. Queueing '.$redirectURL); - $this->redirectQueue[$orig] = $redirectURL; - } - } - } - //die($url.' -multi- '.$request->getResponseInfo('effective_url')); - $pool->detach($request); - unset($this->requests[$orig]['httpRequest'], $request); - /* - if ($this->minimiseMemoryUse) { - if ($this->cache($url)) { - unset($this->requests[$url]); - } - } - */ - } - } - } - } catch (HttpException $e) { - $this->debug($e); - return false; - } - } - - ////////////////////////////////////////////////////////// - // parallel (curl_multi_*) - elseif ($this->method == self::METHOD_CURL_MULTI) { - $this->debug('Starting parallel fetch (curl_multi_*)'); - while (count($urls) > 0) { - $this->debug('Processing set of '.min($this->maxParallelRequests, count($urls))); - $subset = array_splice($urls, 0, $this->maxParallelRequests); - $pool = new RollingCurl(array($this, 'handleCurlResponse')); - $pool->window_size = count($subset); - - foreach ($subset as $orig => $url) { - if (!$isRedirect) $orig = $url; - unset($this->redirectQueue[$orig]); - $this->debug("...$url"); - if (!$isRedirect && isset($this->requests[$url])) { - $this->debug("......in memory"); - /* - } elseif ($this->isCached($url)) { - $this->debug("......is cached"); - if (!$this->minimiseMemoryUse) { - $this->requests[$url] = $this->getCached($url); - } - */ - } else { - $this->debug("......adding to pool"); - $req_url = $this->rewriteUrls($url); - $req_url = ($this->rewriteHashbangFragment) ? $this->rewriteHashbangFragment($req_url) : $req_url; - $req_url = $this->removeFragment($req_url); - if (!empty($this->headerOnlyTypes) && !isset($this->requests[$orig]['wrongGuess']) && $this->possibleUnsupportedType($req_url)) { - $_meth = 'HEAD'; - } else { - $_meth = 'GET'; - unset($this->requests[$orig]['wrongGuess']); - } - $headers = array(); - //$headers[] = 'User-Agent: '.$this->userAgent; - $headers[] = $this->getUserAgent($req_url); - // add referer for picky sites - $headers[] = 'Referer: '.$this->referer; - // send cookies, if we have any - if ($cookies = $this->cookieJar->getMatchingCookies($req_url)) { - $this->debug("......sending cookies: $cookies"); - $headers[] = 'Cookie: '.$cookies; - } - $httpRequest = new RollingCurlRequest($req_url, $_meth, null, $headers, array( - CURLOPT_CONNECTTIMEOUT => $this->requestOptions['timeout'], - CURLOPT_TIMEOUT => $this->requestOptions['timeout'] - )); - $httpRequest->set_original_url($orig); - $this->requests[$orig] = array('headers'=>null, 'body'=>null, 'httpRequest'=>$httpRequest); - $this->requests[$orig]['original_url'] = $orig; // TODO: is this needed anymore? - $pool->add($httpRequest); - } - } - // did we get anything into the pool? - if (count($pool) > 0) { - $this->debug('Sending request...'); - $pool->execute(); // this will call handleCurlResponse() and populate $this->requests[$orig] - $this->debug('Received responses'); - foreach($subset as $orig => $url) { - if (!$isRedirect) $orig = $url; - // $this->requests[$orig]['headers'] - // $this->requests[$orig]['body'] - // $this->requests[$orig]['effective_url'] - // check content type - if ($this->headerOnlyType($this->requests[$orig]['headers'])) { - $this->requests[$orig]['body'] = ''; - $_header_only_type = true; - $this->debug('Header only type returned'); - } else { - $_header_only_type = false; - } - $status_code = $this->requests[$orig]['status_code']; - if ((in_array($status_code, array(300, 301, 302, 303, 307)) || $status_code > 307 && $status_code < 400) && isset($this->requests[$orig]['location'])) { - $redirectURL = $this->requests[$orig]['location']; - if (!preg_match('!^https?://!i', $redirectURL)) { - $redirectURL = SimplePie_Misc::absolutize_url($redirectURL, $url); - } - if ($this->validateURL($redirectURL)) { - $this->debug('Redirect detected. Valid URL: '.$redirectURL); - // store any cookies - $cookies = $this->cookieJar->extractCookies($this->requests[$orig]['headers']); - if (!empty($cookies)) $this->cookieJar->storeCookies($url, $cookies); - $this->redirectQueue[$orig] = $redirectURL; - } else { - $this->debug('Redirect detected. Invalid URL: '.$redirectURL); - } - } elseif (!$_header_only_type && $this->requests[$orig]['method'] == 'HEAD') { - // the response content-type did not match our 'header only' types, - // but we'd issues a HEAD request because we assumed it would. So - // let's queue a proper GET request for this item... - $this->debug('Wrong guess at content-type, queing GET request'); - $this->requests[$orig]['wrongGuess'] = true; - $this->redirectQueue[$orig] = $this->requests[$orig]['effective_url']; - } elseif (strpos($this->requests[$orig]['effective_url'], '_escaped_fragment_') === false) { - // check for - // for AJAX sites, e.g. Blogger with its dynamic views templates. - // Based on Google's spec: https://developers.google.com/webmasters/ajax-crawling/docs/specification - if (isset($this->requests[$orig]['body'])) { - $redirectURL = $this->getUglyURL($this->requests[$orig]['effective_url'], substr($this->requests[$orig]['body'], 0, 4000)); - if ($redirectURL) { - $this->debug('AJAX trigger (meta name="fragment" content="!") found. Queueing '.$redirectURL); - $this->redirectQueue[$orig] = $redirectURL; - } - } - } - // die($url.' -multi- '.$request->getResponseInfo('effective_url')); - unset($this->requests[$orig]['httpRequest'], $this->requests[$orig]['method']); - } - } - } - } - - ////////////////////////////////////////////////////// - // sequential (file_get_contents) - else { - $this->debug('Starting sequential fetch (file_get_contents)'); - $this->debug('Processing set of '.count($urls)); - foreach ($urls as $orig => $url) { - if (!$isRedirect) $orig = $url; - unset($this->redirectQueue[$orig]); - $this->debug("...$url"); - if (!$isRedirect && isset($this->requests[$url])) { - $this->debug("......in memory"); - /* - } elseif ($this->isCached($url)) { - $this->debug("......is cached"); - if (!$this->minimiseMemoryUse) { - $this->requests[$url] = $this->getCached($url); - } - */ - } else { - $this->debug("Sending request for $url"); - $this->requests[$orig]['original_url'] = $orig; - $req_url = $this->rewriteUrls($url); - $req_url = ($this->rewriteHashbangFragment) ? $this->rewriteHashbangFragment($req_url) : $req_url; - $req_url = $this->removeFragment($req_url); - // send cookies, if we have any - $httpContext = $this->httpContext; - $httpContext['http']['header'] .= $this->getUserAgent($req_url)."\r\n"; - // add referer for picky sites - $httpContext['http']['header'] .= 'Referer: '.$this->referer."\r\n"; - if ($cookies = $this->cookieJar->getMatchingCookies($req_url)) { - $this->debug("......sending cookies: $cookies"); - $httpContext['http']['header'] .= 'Cookie: '.$cookies."\r\n"; - } - if (false !== ($html = @file_get_contents($req_url, false, stream_context_create($httpContext)))) { - $this->debug('Received response'); - // get status code - if (!isset($http_response_header[0]) || !preg_match('!^HTTP/\d+\.\d+\s+(\d+)!', trim($http_response_header[0]), $match)) { - $this->debug('Error: no status code found'); - // TODO: handle error - no status code - } else { - $this->requests[$orig]['headers'] = $this->headersToString($http_response_header, false); - // check content type - if ($this->headerOnlyType($this->requests[$orig]['headers'])) { - $this->requests[$orig]['body'] = ''; - } else { - $this->requests[$orig]['body'] = $html; - } - $this->requests[$orig]['effective_url'] = $req_url; - $this->requests[$orig]['status_code'] = $status_code = (int)$match[1]; - unset($match); - // handle redirect - if (preg_match('/^Location:(.*?)$/mi', $this->requests[$orig]['headers'], $match)) { - $this->requests[$orig]['location'] = trim($match[1]); - } - if ((in_array($status_code, array(300, 301, 302, 303, 307)) || $status_code > 307 && $status_code < 400) && isset($this->requests[$orig]['location'])) { - $redirectURL = $this->requests[$orig]['location']; - if (!preg_match('!^https?://!i', $redirectURL)) { - $redirectURL = SimplePie_Misc::absolutize_url($redirectURL, $url); - } - if ($this->validateURL($redirectURL)) { - $this->debug('Redirect detected. Valid URL: '.$redirectURL); - // store any cookies - $cookies = $this->cookieJar->extractCookies($this->requests[$orig]['headers']); - if (!empty($cookies)) $this->cookieJar->storeCookies($url, $cookies); - $this->redirectQueue[$orig] = $redirectURL; - } else { - $this->debug('Redirect detected. Invalid URL: '.$redirectURL); - } - } elseif (strpos($this->requests[$orig]['effective_url'], '_escaped_fragment_') === false) { - // check for - // for AJAX sites, e.g. Blogger with its dynamic views templates. - // Based on Google's spec: https://developers.google.com/webmasters/ajax-crawling/docs/specification - if (isset($this->requests[$orig]['body'])) { - $redirectURL = $this->getUglyURL($this->requests[$orig]['effective_url'], substr($this->requests[$orig]['body'], 0, 4000)); - if ($redirectURL) { - $this->debug('AJAX trigger (meta name="fragment" content="!") found. Queueing '.$redirectURL); - $this->redirectQueue[$orig] = $redirectURL; - } - } - } - } - } else { - $this->debug('Error retrieving URL'); - //print_r($req_url); - //print_r($http_response_header); - //print_r($html); - - // TODO: handle error - failed to retrieve URL - } - } - } - } - } - - public function handleCurlResponse($response, $info, $request) { - $orig = $request->url_original; - $this->requests[$orig]['headers'] = substr($response, 0, $info['header_size']); - $this->requests[$orig]['body'] = substr($response, $info['header_size']); - $this->requests[$orig]['method'] = $request->method; - $this->requests[$orig]['effective_url'] = $info['url']; - $this->requests[$orig]['status_code'] = (int)$info['http_code']; - if (preg_match('/^Location:(.*?)$/mi', $this->requests[$orig]['headers'], $match)) { - $this->requests[$orig]['location'] = trim($match[1]); - } - } - - protected function headersToString(array $headers, $associative=true) { - if (!$associative) { - return implode("\n", $headers); - } else { - $str = ''; - foreach ($headers as $key => $val) { - if (is_array($val)) { - foreach ($val as $v) $str .= "$key: $v\n"; - } else { - $str .= "$key: $val\n"; - } - } - return rtrim($str); - } - } - - public function get($url, $remove=false, $gzdecode=true) { - $url = "$url"; - if (isset($this->requests[$url]) && isset($this->requests[$url]['body'])) { - $this->debug("URL already fetched - in memory ($url, effective: {$this->requests[$url]['effective_url']})"); - $response = $this->requests[$url]; - /* - } elseif ($this->isCached($url)) { - $this->debug("URL already fetched - in disk cache ($url)"); - $response = $this->getCached($url); - $this->requests[$url] = $response; - */ - } else { - $this->debug("Fetching URL ($url)"); - $this->fetchAll(array($url)); - if (isset($this->requests[$url]) && isset($this->requests[$url]['body'])) { - $response = $this->requests[$url]; - } else { - $this->debug("Request failed"); - $response = false; - } - } - /* - if ($this->minimiseMemoryUse && $response) { - $this->cache($url); - unset($this->requests[$url]); - } - */ - if ($remove && $response) unset($this->requests[$url]); - if ($gzdecode && stripos($response['headers'], 'Content-Encoding: gzip')) { - if ($html = gzdecode($response['body'])) { - $response['body'] = $html; - } - } - return $response; - } - - public function parallelSupport() { - return class_exists('HttpRequestPool') || function_exists('curl_multi_init'); - } - - private function headerOnlyType($headers) { - if (preg_match('!^Content-Type:\s*(([a-z-]+)/([^;\r\n ]+))!im', $headers, $match)) { - // look for full mime type (e.g. image/jpeg) or just type (e.g. image) - $match[1] = strtolower(trim($match[1])); - $match[2] = strtolower(trim($match[2])); - foreach (array($match[1], $match[2]) as $mime) { - if (in_array($mime, $this->headerOnlyTypes)) return true; - } - } - return false; - } - - private function possibleUnsupportedType($url) { - $path = @parse_url($url, PHP_URL_PATH); - if ($path && strpos($path, '.') !== false) { - $ext = strtolower(trim(pathinfo($path, PATHINFO_EXTENSION))); - return in_array($ext, $this->headerOnlyClues); - } - return false; - } -} - -// gzdecode from http://www.php.net/manual/en/function.gzdecode.php#82930 -if (!function_exists('gzdecode')) { - function gzdecode($data,&$filename='',&$error='',$maxlength=null) - { - $len = strlen($data); - if ($len < 18 || strcmp(substr($data,0,2),"\x1f\x8b")) { - $error = "Not in GZIP format."; - return null; // Not GZIP format (See RFC 1952) - } - $method = ord(substr($data,2,1)); // Compression method - $flags = ord(substr($data,3,1)); // Flags - if ($flags & 31 != $flags) { - $error = "Reserved bits not allowed."; - return null; - } - // NOTE: $mtime may be negative (PHP integer limitations) - $mtime = unpack("V", substr($data,4,4)); - $mtime = $mtime[1]; - $xfl = substr($data,8,1); - $os = substr($data,8,1); - $headerlen = 10; - $extralen = 0; - $extra = ""; - if ($flags & 4) { - // 2-byte length prefixed EXTRA data in header - if ($len - $headerlen - 2 < 8) { - return false; // invalid - } - $extralen = unpack("v",substr($data,8,2)); - $extralen = $extralen[1]; - if ($len - $headerlen - 2 - $extralen < 8) { - return false; // invalid - } - $extra = substr($data,10,$extralen); - $headerlen += 2 + $extralen; - } - $filenamelen = 0; - $filename = ""; - if ($flags & 8) { - // C-style string - if ($len - $headerlen - 1 < 8) { - return false; // invalid - } - $filenamelen = strpos(substr($data,$headerlen),chr(0)); - if ($filenamelen === false || $len - $headerlen - $filenamelen - 1 < 8) { - return false; // invalid - } - $filename = substr($data,$headerlen,$filenamelen); - $headerlen += $filenamelen + 1; - } - $commentlen = 0; - $comment = ""; - if ($flags & 16) { - // C-style string COMMENT data in header - if ($len - $headerlen - 1 < 8) { - return false; // invalid - } - $commentlen = strpos(substr($data,$headerlen),chr(0)); - if ($commentlen === false || $len - $headerlen - $commentlen - 1 < 8) { - return false; // Invalid header format - } - $comment = substr($data,$headerlen,$commentlen); - $headerlen += $commentlen + 1; - } - $headercrc = ""; - if ($flags & 2) { - // 2-bytes (lowest order) of CRC32 on header present - if ($len - $headerlen - 2 < 8) { - return false; // invalid - } - $calccrc = crc32(substr($data,0,$headerlen)) & 0xffff; - $headercrc = unpack("v", substr($data,$headerlen,2)); - $headercrc = $headercrc[1]; - if ($headercrc != $calccrc) { - $error = "Header checksum failed."; - return false; // Bad header CRC - } - $headerlen += 2; - } - // GZIP FOOTER - $datacrc = unpack("V",substr($data,-8,4)); - $datacrc = sprintf('%u',$datacrc[1] & 0xFFFFFFFF); - $isize = unpack("V",substr($data,-4)); - $isize = $isize[1]; - // decompression: - $bodylen = $len-$headerlen-8; - if ($bodylen < 1) { - // IMPLEMENTATION BUG! - return null; - } - $body = substr($data,$headerlen,$bodylen); - $data = ""; - if ($bodylen > 0) { - switch ($method) { - case 8: - // Currently the only supported compression method: - $data = gzinflate($body,$maxlength); - break; - default: - $error = "Unknown compression method."; - return false; - } - } // zero-byte body content is allowed - // Verifiy CRC32 - $crc = sprintf("%u",crc32($data)); - $crcOK = $crc == $datacrc; - $lenOK = $isize == strlen($data); - if (!$lenOK || !$crcOK) { - $error = ( $lenOK ? '' : 'Length check FAILED. ') . ( $crcOK ? '' : 'Checksum FAILED.'); - return false; - } - return $data; - } -} -?> \ No newline at end of file +userAgentDefault = self::UA_BROWSER; + $this->referer = self::REF_GOOGLE; + // set the request method + if (in_array($method, array(1,2,4))) { + $this->method = $method; + } else { + if (class_exists('HttpRequestPool')) { + $this->method = self::METHOD_REQUEST_POOL; + } elseif (function_exists('curl_multi_init')) { + $this->method = self::METHOD_CURL_MULTI; + } else { + $this->method = self::METHOD_FILE_GET_CONTENTS; + } + } + if ($this->method == self::METHOD_CURL_MULTI) { + require_once(dirname(__FILE__).'/RollingCurl.php'); + } + // create cookie jar + $this->cookieJar = new CookieJar(); + // set request options (redirect must be 0) + $this->requestOptions = array( + 'timeout' => 15, + 'connecttimeout' => 15, + 'dns_cache_timeout' => 300, + 'redirect' => 0 // we handle redirects manually so we can rewrite the new hashbang URLs that are creeping up over the web + // TODO: test onprogress? + ); + if (is_array($requestOptions)) { + $this->requestOptions = array_merge($this->requestOptions, $requestOptions); + } + $this->httpContext = array( + 'http' => array( + 'ignore_errors' => true, + 'timeout' => $this->requestOptions['timeout'], + 'max_redirects' => $this->requestOptions['redirect'], + 'header' => "Accept: */*\r\n" + ) + ); + } + + protected function debug($msg) { + if ($this->debug) { + $mem = round(memory_get_usage()/1024, 2); + $memPeak = round(memory_get_peak_usage()/1024, 2); + echo '* ',$msg; + if ($this->debugVerbose) echo ' - mem used: ',$mem," (peak: $memPeak)"; + echo "\n"; + ob_flush(); + flush(); + } + } + + protected function getUserAgent($url, $asArray=false) { + $host = @parse_url($url, PHP_URL_HOST); + if (strtolower(substr($host, 0, 4)) == 'www.') { + $host = substr($host, 4); + } + if ($host) { + $try = array($host); + $split = explode('.', $host); + if (count($split) > 1) { + array_shift($split); + $try[] = '.'.implode('.', $split); + } + foreach ($try as $h) { + if (isset($this->userAgentMap[$h])) { + $ua = $this->userAgentMap[$h]; + break; + } + } + } + if (!isset($ua)) $ua = $this->userAgentDefault; + if ($asArray) { + return array('User-Agent' => $ua); + } else { + return 'User-Agent: '.$ua; + } + } + + public function rewriteHashbangFragment($url) { + // return $url if there's no '#!' + if (strpos($url, '#!') === false) return $url; + // split $url and rewrite + // TODO: is SimplePie_IRI included? + $iri = new SimplePie_IRI($url); + $fragment = substr($iri->fragment, 1); // strip '!' + $iri->fragment = null; + if (isset($iri->query)) { + parse_str($iri->query, $query); + } else { + $query = array(); + } + $query['_escaped_fragment_'] = (string)$fragment; + $iri->query = str_replace('%2F', '/', http_build_query($query)); // needed for some sites + return $iri->get_iri(); + } + + public function getRedirectURLfromHTML($url, $html) { + $redirect_url = $this->getMetaRefreshURL($url, $html); + if (!$redirect_url) { + $redirect_url = $this->getUglyURL($url, $html); + } + return $redirect_url; + } + + public function getMetaRefreshURL($url, $html) { + if ($html == '') return false; + // + if (!preg_match('!]+)["\']*>!i', $html, $match)) { + return false; + } + $redirect_url = $match[1]; + if (preg_match('!^https?://!i', $redirect_url)) { + // already absolute + $this->debug('Meta refresh redirect found (http-equiv="refresh"), new URL: '.$redirect_url); + return $redirect_url; + } + // absolutize redirect URL + $base = new SimplePie_IRI($url); + // remove '//' in URL path (causes URLs not to resolve properly) + if (isset($base->path)) $base->path = preg_replace('!//+!', '/', $base->path); + if ($absolute = SimplePie_IRI::absolutize($base, $redirect_url)) { + $this->debug('Meta refresh redirect found (http-equiv="refresh"), new URL: '.$absolute); + return $absolute; + } + return false; + } + + public function getUglyURL($url, $html) { + if ($html == '') return false; + $found = false; + foreach ($this->ajaxTriggers as $string) { + if (stripos($html, $string)) { + $found = true; + break; + } + } + if (!$found) return false; + $iri = new SimplePie_IRI($url); + if (isset($iri->query)) { + parse_str($iri->query, $query); + } else { + $query = array(); + } + $query['_escaped_fragment_'] = ''; + $iri->query = str_replace('%2F', '/', http_build_query($query)); // needed for some sites + $ugly_url = $iri->get_iri(); + $this->debug('AJAX trigger (meta name="fragment" content="!") found, new URL: '.$ugly_url); + return $ugly_url; + } + + public function removeFragment($url) { + $pos = strpos($url, '#'); + if ($pos === false) { + return $url; + } else { + return substr($url, 0, $pos); + } + } + + public function rewriteUrls($url) { + foreach ($this->rewriteUrls as $find => $action) { + if (strpos($url, $find) !== false) { + if (is_array($action)) { + return strtr($url, $action); + } + } + } + return $url; + } + + public function enableDebug($bool=true) { + $this->debug = (bool)$bool; + } + + public function minimiseMemoryUse($bool = true) { + $this->minimiseMemoryUse = $bool; + } + + public function setMaxParallelRequests($max) { + $this->maxParallelRequests = $max; + } + + public function validateUrl($url) { + $url = filter_var($url, FILTER_SANITIZE_URL); + $test = filter_var($url, FILTER_VALIDATE_URL, FILTER_FLAG_SCHEME_REQUIRED); + // deal with bug http://bugs.php.net/51192 (present in PHP 5.2.13 and PHP 5.3.2) + if ($test === false) { + $test = filter_var(strtr($url, '-', '_'), FILTER_VALIDATE_URL, FILTER_FLAG_SCHEME_REQUIRED); + } + if ($test !== false && $test !== null && preg_match('!^https?://!', $url)) { + return $url; + } else { + return false; + } + } + + public function fetchAll(array $urls) { + $this->fetchAllOnce($urls, $isRedirect=false); + $redirects = 0; + while (!empty($this->redirectQueue) && ++$redirects <= $this->maxRedirects) { + $this->debug("Following redirects #$redirects..."); + $this->fetchAllOnce($this->redirectQueue, $isRedirect=true); + } + } + + // fetch all URLs without following redirects + public function fetchAllOnce(array $urls, $isRedirect=false) { + if (!$isRedirect) $urls = array_unique($urls); + if (empty($urls)) return; + + ////////////////////////////////////////////////////// + // parallel (HttpRequestPool) + if ($this->method == self::METHOD_REQUEST_POOL) { + $this->debug('Starting parallel fetch (HttpRequestPool)'); + try { + while (count($urls) > 0) { + $this->debug('Processing set of '.min($this->maxParallelRequests, count($urls))); + $subset = array_splice($urls, 0, $this->maxParallelRequests); + $pool = new HttpRequestPool(); + foreach ($subset as $orig => $url) { + if (!$isRedirect) $orig = $url; + unset($this->redirectQueue[$orig]); + $this->debug("...$url"); + if (!$isRedirect && isset($this->requests[$url])) { + $this->debug("......in memory"); + /* + } elseif ($this->isCached($url)) { + $this->debug("......is cached"); + if (!$this->minimiseMemoryUse) { + $this->requests[$url] = $this->getCached($url); + } + */ + } else { + $this->debug("......adding to pool"); + $req_url = $this->rewriteUrls($url); + $req_url = ($this->rewriteHashbangFragment) ? $this->rewriteHashbangFragment($req_url) : $req_url; + $req_url = $this->removeFragment($req_url); + if (!empty($this->headerOnlyTypes) && !isset($this->requests[$orig]['wrongGuess']) && $this->possibleUnsupportedType($req_url)) { + $_meth = HttpRequest::METH_HEAD; + } else { + $_meth = HttpRequest::METH_GET; + unset($this->requests[$orig]['wrongGuess']); + } + $httpRequest = new HttpRequest($req_url, $_meth, $this->requestOptions); + // send cookies, if we have any + if ($cookies = $this->cookieJar->getMatchingCookies($req_url)) { + $this->debug("......sending cookies: $cookies"); + $httpRequest->addHeaders(array('Cookie' => $cookies)); + } + //$httpRequest->addHeaders(array('User-Agent' => $this->userAgent)); + $httpRequest->addHeaders($this->getUserAgent($req_url, true)); + // add referer for picky sites + $httpRequest->addheaders(array('Referer' => $this->referer)); + $this->requests[$orig] = array('headers'=>null, 'body'=>null, 'httpRequest'=>$httpRequest); + $this->requests[$orig]['original_url'] = $orig; + $pool->attach($httpRequest); + } + } + // did we get anything into the pool? + if (count($pool) > 0) { + $this->debug('Sending request...'); + try { + $pool->send(); + } catch (HttpRequestPoolException $e) { + // do nothing + } + $this->debug('Received responses'); + foreach($subset as $orig => $url) { + if (!$isRedirect) $orig = $url; + $request = $this->requests[$orig]['httpRequest']; + //$this->requests[$orig]['headers'] = $this->headersToString($request->getResponseHeader()); + // getResponseHeader() doesn't return status line, so, for consistency... + $this->requests[$orig]['headers'] = substr($request->getRawResponseMessage(), 0, $request->getResponseInfo('header_size')); + // check content type + // TODO: use getResponseHeader('content-type') or getResponseInfo() + if ($this->headerOnlyType($this->requests[$orig]['headers'])) { + $this->requests[$orig]['body'] = ''; + $_header_only_type = true; + $this->debug('Header only type returned'); + } else { + $this->requests[$orig]['body'] = $request->getResponseBody(); + $_header_only_type = false; + } + $this->requests[$orig]['effective_url'] = $request->getResponseInfo('effective_url'); + $this->requests[$orig]['status_code'] = $status_code = $request->getResponseCode(); + // is redirect? + if ((in_array($status_code, array(300, 301, 302, 303, 307)) || $status_code > 307 && $status_code < 400) && $request->getResponseHeader('location')) { + $redirectURL = $request->getResponseHeader('location'); + if (!preg_match('!^https?://!i', $redirectURL)) { + $redirectURL = SimplePie_Misc::absolutize_url($redirectURL, $url); + } + if ($this->validateURL($redirectURL)) { + $this->debug('Redirect detected. Valid URL: '.$redirectURL); + // store any cookies + $cookies = $request->getResponseHeader('set-cookie'); + if ($cookies && !is_array($cookies)) $cookies = array($cookies); + if ($cookies) $this->cookieJar->storeCookies($url, $cookies); + $this->redirectQueue[$orig] = $redirectURL; + } else { + $this->debug('Redirect detected. Invalid URL: '.$redirectURL); + } + } elseif (!$_header_only_type && $request->getMethod() === HttpRequest::METH_HEAD) { + // the response content-type did not match our 'header only' types, + // but we'd issues a HEAD request because we assumed it would. So + // let's queue a proper GET request for this item... + $this->debug('Wrong guess at content-type, queing GET request'); + $this->requests[$orig]['wrongGuess'] = true; + $this->redirectQueue[$orig] = $this->requests[$orig]['effective_url']; + } elseif (strpos($this->requests[$orig]['effective_url'], '_escaped_fragment_') === false) { + // check for + // for AJAX sites, e.g. Blogger with its dynamic views templates. + // Based on Google's spec: https://developers.google.com/webmasters/ajax-crawling/docs/specification + if (isset($this->requests[$orig]['body'])) { + $redirectURL = $this->getRedirectURLfromHTML($this->requests[$orig]['effective_url'], substr($this->requests[$orig]['body'], 0, 4000)); + if ($redirectURL) { + $this->redirectQueue[$orig] = $redirectURL; + } + } + } + //die($url.' -multi- '.$request->getResponseInfo('effective_url')); + $pool->detach($request); + unset($this->requests[$orig]['httpRequest'], $request); + /* + if ($this->minimiseMemoryUse) { + if ($this->cache($url)) { + unset($this->requests[$url]); + } + } + */ + } + } + } + } catch (HttpException $e) { + $this->debug($e); + return false; + } + } + + ////////////////////////////////////////////////////////// + // parallel (curl_multi_*) + elseif ($this->method == self::METHOD_CURL_MULTI) { + $this->debug('Starting parallel fetch (curl_multi_*)'); + while (count($urls) > 0) { + $this->debug('Processing set of '.min($this->maxParallelRequests, count($urls))); + $subset = array_splice($urls, 0, $this->maxParallelRequests); + $pool = new RollingCurl(array($this, 'handleCurlResponse')); + $pool->window_size = count($subset); + + foreach ($subset as $orig => $url) { + if (!$isRedirect) $orig = $url; + unset($this->redirectQueue[$orig]); + $this->debug("...$url"); + if (!$isRedirect && isset($this->requests[$url])) { + $this->debug("......in memory"); + /* + } elseif ($this->isCached($url)) { + $this->debug("......is cached"); + if (!$this->minimiseMemoryUse) { + $this->requests[$url] = $this->getCached($url); + } + */ + } else { + $this->debug("......adding to pool"); + $req_url = $this->rewriteUrls($url); + $req_url = ($this->rewriteHashbangFragment) ? $this->rewriteHashbangFragment($req_url) : $req_url; + $req_url = $this->removeFragment($req_url); + if (!empty($this->headerOnlyTypes) && !isset($this->requests[$orig]['wrongGuess']) && $this->possibleUnsupportedType($req_url)) { + $_meth = 'HEAD'; + } else { + $_meth = 'GET'; + unset($this->requests[$orig]['wrongGuess']); + } + $headers = array(); + //$headers[] = 'User-Agent: '.$this->userAgent; + $headers[] = $this->getUserAgent($req_url); + // add referer for picky sites + $headers[] = 'Referer: '.$this->referer; + // send cookies, if we have any + if ($cookies = $this->cookieJar->getMatchingCookies($req_url)) { + $this->debug("......sending cookies: $cookies"); + $headers[] = 'Cookie: '.$cookies; + } + $httpRequest = new RollingCurlRequest($req_url, $_meth, null, $headers, array( + CURLOPT_CONNECTTIMEOUT => $this->requestOptions['timeout'], + CURLOPT_TIMEOUT => $this->requestOptions['timeout'] + )); + $httpRequest->set_original_url($orig); + $this->requests[$orig] = array('headers'=>null, 'body'=>null, 'httpRequest'=>$httpRequest); + $this->requests[$orig]['original_url'] = $orig; // TODO: is this needed anymore? + $pool->add($httpRequest); + } + } + // did we get anything into the pool? + if (count($pool) > 0) { + $this->debug('Sending request...'); + $pool->execute(); // this will call handleCurlResponse() and populate $this->requests[$orig] + $this->debug('Received responses'); + foreach($subset as $orig => $url) { + if (!$isRedirect) $orig = $url; + // $this->requests[$orig]['headers'] + // $this->requests[$orig]['body'] + // $this->requests[$orig]['effective_url'] + // check content type + if ($this->headerOnlyType($this->requests[$orig]['headers'])) { + $this->requests[$orig]['body'] = ''; + $_header_only_type = true; + $this->debug('Header only type returned'); + } else { + $_header_only_type = false; + } + $status_code = $this->requests[$orig]['status_code']; + if ((in_array($status_code, array(300, 301, 302, 303, 307)) || $status_code > 307 && $status_code < 400) && isset($this->requests[$orig]['location'])) { + $redirectURL = $this->requests[$orig]['location']; + if (!preg_match('!^https?://!i', $redirectURL)) { + $redirectURL = SimplePie_Misc::absolutize_url($redirectURL, $url); + } + if ($this->validateURL($redirectURL)) { + $this->debug('Redirect detected. Valid URL: '.$redirectURL); + // store any cookies + $cookies = $this->cookieJar->extractCookies($this->requests[$orig]['headers']); + if (!empty($cookies)) $this->cookieJar->storeCookies($url, $cookies); + $this->redirectQueue[$orig] = $redirectURL; + } else { + $this->debug('Redirect detected. Invalid URL: '.$redirectURL); + } + } elseif (!$_header_only_type && $this->requests[$orig]['method'] == 'HEAD') { + // the response content-type did not match our 'header only' types, + // but we'd issues a HEAD request because we assumed it would. So + // let's queue a proper GET request for this item... + $this->debug('Wrong guess at content-type, queing GET request'); + $this->requests[$orig]['wrongGuess'] = true; + $this->redirectQueue[$orig] = $this->requests[$orig]['effective_url']; + } elseif (strpos($this->requests[$orig]['effective_url'], '_escaped_fragment_') === false) { + // check for + // for AJAX sites, e.g. Blogger with its dynamic views templates. + // Based on Google's spec: https://developers.google.com/webmasters/ajax-crawling/docs/specification + if (isset($this->requests[$orig]['body'])) { + $redirectURL = $this->getRedirectURLfromHTML($this->requests[$orig]['effective_url'], substr($this->requests[$orig]['body'], 0, 4000)); + if ($redirectURL) { + $this->redirectQueue[$orig] = $redirectURL; + } + } + } + // die($url.' -multi- '.$request->getResponseInfo('effective_url')); + unset($this->requests[$orig]['httpRequest'], $this->requests[$orig]['method']); + } + } + } + } + + ////////////////////////////////////////////////////// + // sequential (file_get_contents) + else { + $this->debug('Starting sequential fetch (file_get_contents)'); + $this->debug('Processing set of '.count($urls)); + foreach ($urls as $orig => $url) { + if (!$isRedirect) $orig = $url; + unset($this->redirectQueue[$orig]); + $this->debug("...$url"); + if (!$isRedirect && isset($this->requests[$url])) { + $this->debug("......in memory"); + /* + } elseif ($this->isCached($url)) { + $this->debug("......is cached"); + if (!$this->minimiseMemoryUse) { + $this->requests[$url] = $this->getCached($url); + } + */ + } else { + $this->debug("Sending request for $url"); + $this->requests[$orig]['original_url'] = $orig; + $req_url = $this->rewriteUrls($url); + $req_url = ($this->rewriteHashbangFragment) ? $this->rewriteHashbangFragment($req_url) : $req_url; + $req_url = $this->removeFragment($req_url); + // send cookies, if we have any + $httpContext = $this->httpContext; + $httpContext['http']['header'] .= $this->getUserAgent($req_url)."\r\n"; + // add referer for picky sites + $httpContext['http']['header'] .= 'Referer: '.$this->referer."\r\n"; + if ($cookies = $this->cookieJar->getMatchingCookies($req_url)) { + $this->debug("......sending cookies: $cookies"); + $httpContext['http']['header'] .= 'Cookie: '.$cookies."\r\n"; + } + if (false !== ($html = @file_get_contents($req_url, false, stream_context_create($httpContext)))) { + $this->debug('Received response'); + // get status code + if (!isset($http_response_header[0]) || !preg_match('!^HTTP/\d+\.\d+\s+(\d+)!', trim($http_response_header[0]), $match)) { + $this->debug('Error: no status code found'); + // TODO: handle error - no status code + } else { + $this->requests[$orig]['headers'] = $this->headersToString($http_response_header, false); + // check content type + if ($this->headerOnlyType($this->requests[$orig]['headers'])) { + $this->requests[$orig]['body'] = ''; + } else { + $this->requests[$orig]['body'] = $html; + } + $this->requests[$orig]['effective_url'] = $req_url; + $this->requests[$orig]['status_code'] = $status_code = (int)$match[1]; + unset($match); + // handle redirect + if (preg_match('/^Location:(.*?)$/mi', $this->requests[$orig]['headers'], $match)) { + $this->requests[$orig]['location'] = trim($match[1]); + } + if ((in_array($status_code, array(300, 301, 302, 303, 307)) || $status_code > 307 && $status_code < 400) && isset($this->requests[$orig]['location'])) { + $redirectURL = $this->requests[$orig]['location']; + if (!preg_match('!^https?://!i', $redirectURL)) { + $redirectURL = SimplePie_Misc::absolutize_url($redirectURL, $url); + } + if ($this->validateURL($redirectURL)) { + $this->debug('Redirect detected. Valid URL: '.$redirectURL); + // store any cookies + $cookies = $this->cookieJar->extractCookies($this->requests[$orig]['headers']); + if (!empty($cookies)) $this->cookieJar->storeCookies($url, $cookies); + $this->redirectQueue[$orig] = $redirectURL; + } else { + $this->debug('Redirect detected. Invalid URL: '.$redirectURL); + } + } elseif (strpos($this->requests[$orig]['effective_url'], '_escaped_fragment_') === false) { + // check for + // for AJAX sites, e.g. Blogger with its dynamic views templates. + // Based on Google's spec: https://developers.google.com/webmasters/ajax-crawling/docs/specification + if (isset($this->requests[$orig]['body'])) { + $redirectURL = $this->getRedirectURLfromHTML($this->requests[$orig]['effective_url'], substr($this->requests[$orig]['body'], 0, 4000)); + if ($redirectURL) { + $this->redirectQueue[$orig] = $redirectURL; + } + } + } + } + } else { + $this->debug('Error retrieving URL'); + //print_r($req_url); + //print_r($http_response_header); + //print_r($html); + + // TODO: handle error - failed to retrieve URL + } + } + } + } + } + + public function handleCurlResponse($response, $info, $request) { + $orig = $request->url_original; + $this->requests[$orig]['headers'] = substr($response, 0, $info['header_size']); + $this->requests[$orig]['body'] = substr($response, $info['header_size']); + $this->requests[$orig]['method'] = $request->method; + $this->requests[$orig]['effective_url'] = $info['url']; + $this->requests[$orig]['status_code'] = (int)$info['http_code']; + if (preg_match('/^Location:(.*?)$/mi', $this->requests[$orig]['headers'], $match)) { + $this->requests[$orig]['location'] = trim($match[1]); + } + } + + protected function headersToString(array $headers, $associative=true) { + if (!$associative) { + return implode("\n", $headers); + } else { + $str = ''; + foreach ($headers as $key => $val) { + if (is_array($val)) { + foreach ($val as $v) $str .= "$key: $v\n"; + } else { + $str .= "$key: $val\n"; + } + } + return rtrim($str); + } + } + + public function get($url, $remove=false, $gzdecode=true) { + $url = "$url"; + if (isset($this->requests[$url]) && isset($this->requests[$url]['body'])) { + $this->debug("URL already fetched - in memory ($url, effective: {$this->requests[$url]['effective_url']})"); + $response = $this->requests[$url]; + /* + } elseif ($this->isCached($url)) { + $this->debug("URL already fetched - in disk cache ($url)"); + $response = $this->getCached($url); + $this->requests[$url] = $response; + */ + } else { + $this->debug("Fetching URL ($url)"); + $this->fetchAll(array($url)); + if (isset($this->requests[$url]) && isset($this->requests[$url]['body'])) { + $response = $this->requests[$url]; + } else { + $this->debug("Request failed"); + $response = false; + } + } + /* + if ($this->minimiseMemoryUse && $response) { + $this->cache($url); + unset($this->requests[$url]); + } + */ + if ($remove && $response) unset($this->requests[$url]); + if ($gzdecode && stripos($response['headers'], 'Content-Encoding: gzip')) { + if ($html = gzdecode($response['body'])) { + $response['body'] = $html; + } + } + return $response; + } + + public function parallelSupport() { + return class_exists('HttpRequestPool') || function_exists('curl_multi_init'); + } + + private function headerOnlyType($headers) { + if (preg_match('!^Content-Type:\s*(([a-z-]+)/([^;\r\n ]+))!im', $headers, $match)) { + // look for full mime type (e.g. image/jpeg) or just type (e.g. image) + $match[1] = strtolower(trim($match[1])); + $match[2] = strtolower(trim($match[2])); + foreach (array($match[1], $match[2]) as $mime) { + if (in_array($mime, $this->headerOnlyTypes)) return true; + } + } + return false; + } + + private function possibleUnsupportedType($url) { + $path = @parse_url($url, PHP_URL_PATH); + if ($path && strpos($path, '.') !== false) { + $ext = strtolower(trim(pathinfo($path, PATHINFO_EXTENSION))); + return in_array($ext, $this->headerOnlyClues); + } + return false; + } +} + +// gzdecode from http://www.php.net/manual/en/function.gzdecode.php#82930 +if (!function_exists('gzdecode')) { + function gzdecode($data,&$filename='',&$error='',$maxlength=null) + { + $len = strlen($data); + if ($len < 18 || strcmp(substr($data,0,2),"\x1f\x8b")) { + $error = "Not in GZIP format."; + return null; // Not GZIP format (See RFC 1952) + } + $method = ord(substr($data,2,1)); // Compression method + $flags = ord(substr($data,3,1)); // Flags + if ($flags & 31 != $flags) { + $error = "Reserved bits not allowed."; + return null; + } + // NOTE: $mtime may be negative (PHP integer limitations) + $mtime = unpack("V", substr($data,4,4)); + $mtime = $mtime[1]; + $xfl = substr($data,8,1); + $os = substr($data,8,1); + $headerlen = 10; + $extralen = 0; + $extra = ""; + if ($flags & 4) { + // 2-byte length prefixed EXTRA data in header + if ($len - $headerlen - 2 < 8) { + return false; // invalid + } + $extralen = unpack("v",substr($data,8,2)); + $extralen = $extralen[1]; + if ($len - $headerlen - 2 - $extralen < 8) { + return false; // invalid + } + $extra = substr($data,10,$extralen); + $headerlen += 2 + $extralen; + } + $filenamelen = 0; + $filename = ""; + if ($flags & 8) { + // C-style string + if ($len - $headerlen - 1 < 8) { + return false; // invalid + } + $filenamelen = strpos(substr($data,$headerlen),chr(0)); + if ($filenamelen === false || $len - $headerlen - $filenamelen - 1 < 8) { + return false; // invalid + } + $filename = substr($data,$headerlen,$filenamelen); + $headerlen += $filenamelen + 1; + } + $commentlen = 0; + $comment = ""; + if ($flags & 16) { + // C-style string COMMENT data in header + if ($len - $headerlen - 1 < 8) { + return false; // invalid + } + $commentlen = strpos(substr($data,$headerlen),chr(0)); + if ($commentlen === false || $len - $headerlen - $commentlen - 1 < 8) { + return false; // Invalid header format + } + $comment = substr($data,$headerlen,$commentlen); + $headerlen += $commentlen + 1; + } + $headercrc = ""; + if ($flags & 2) { + // 2-bytes (lowest order) of CRC32 on header present + if ($len - $headerlen - 2 < 8) { + return false; // invalid + } + $calccrc = crc32(substr($data,0,$headerlen)) & 0xffff; + $headercrc = unpack("v", substr($data,$headerlen,2)); + $headercrc = $headercrc[1]; + if ($headercrc != $calccrc) { + $error = "Header checksum failed."; + return false; // Bad header CRC + } + $headerlen += 2; + } + // GZIP FOOTER + $datacrc = unpack("V",substr($data,-8,4)); + $datacrc = sprintf('%u',$datacrc[1] & 0xFFFFFFFF); + $isize = unpack("V",substr($data,-4)); + $isize = $isize[1]; + // decompression: + $bodylen = $len-$headerlen-8; + if ($bodylen < 1) { + // IMPLEMENTATION BUG! + return null; + } + $body = substr($data,$headerlen,$bodylen); + $data = ""; + if ($bodylen > 0) { + switch ($method) { + case 8: + // Currently the only supported compression method: + $data = gzinflate($body,$maxlength); + break; + default: + $error = "Unknown compression method."; + return false; + } + } // zero-byte body content is allowed + // Verifiy CRC32 + $crc = sprintf("%u",crc32($data)); + $crcOK = $crc == $datacrc; + $lenOK = $isize == strlen($data); + if (!$lenOK || !$crcOK) { + $error = ( $lenOK ? '' : 'Length check FAILED. ') . ( $crcOK ? '' : 'Checksum FAILED.'); + return false; + } + return $data; + } +} \ No newline at end of file diff --git a/inc/3rdparty/libraries/humble-http-agent/SimplePie_HumbleHttpAgent.php b/inc/3rdparty/libraries/humble-http-agent/SimplePie_HumbleHttpAgent.php index ecd46d5..c524a1e 100644 --- a/inc/3rdparty/libraries/humble-http-agent/SimplePie_HumbleHttpAgent.php +++ b/inc/3rdparty/libraries/humble-http-agent/SimplePie_HumbleHttpAgent.php @@ -1,79 +1,78 @@ -encode($parsed['authority']), $parsed['path'], $parsed['query'], $parsed['fragment']); - } - $this->url = $url; - $this->useragent = $useragent; - if (preg_match('/^http(s)?:\/\//i', $url)) - { - if (!is_array($headers)) - { - $headers = array(); - } - $this->method = SIMPLEPIE_FILE_SOURCE_REMOTE | SIMPLEPIE_FILE_SOURCE_CURL; - $headers2 = array(); - foreach ($headers as $key => $value) { - $headers2[] = "$key: $value"; - } - //TODO: allow for HTTP headers - // curl_setopt($fp, CURLOPT_HTTPHEADER, $headers2); - - $response = self::$agent->get($url); - - if ($response === false || !isset($response['status_code'])) { - $this->error = 'failed to fetch URL'; - $this->success = false; - } else { - // The extra lines at the end are there to satisfy SimplePie's HTTP parser. - // The class expects a full HTTP message, whereas we're giving it only - // headers - the new lines indicate the start of the body. - $parser = new SimplePie_HTTP_Parser($response['headers']."\r\n\r\n"); - if ($parser->parse()) { - $this->headers = $parser->headers; - //$this->body = $parser->body; - $this->body = $response['body']; - $this->status_code = $parser->status_code; - } - } - } - else - { - $this->error = 'invalid URL'; - $this->success = false; - } - } -} -?> \ No newline at end of file +encode($parsed['authority']), $parsed['path'], $parsed['query'], $parsed['fragment']); + } + $this->url = $url; + $this->useragent = $useragent; + if (preg_match('/^http(s)?:\/\//i', $url)) + { + if (!is_array($headers)) + { + $headers = array(); + } + $this->method = SIMPLEPIE_FILE_SOURCE_REMOTE | SIMPLEPIE_FILE_SOURCE_CURL; + $headers2 = array(); + foreach ($headers as $key => $value) { + $headers2[] = "$key: $value"; + } + //TODO: allow for HTTP headers + // curl_setopt($fp, CURLOPT_HTTPHEADER, $headers2); + + $response = self::$agent->get($url); + + if ($response === false || !isset($response['status_code'])) { + $this->error = 'failed to fetch URL'; + $this->success = false; + } else { + // The extra lines at the end are there to satisfy SimplePie's HTTP parser. + // The class expects a full HTTP message, whereas we're giving it only + // headers - the new lines indicate the start of the body. + $parser = new SimplePie_HTTP_Parser($response['headers']."\r\n\r\n"); + if ($parser->parse()) { + $this->headers = $parser->headers; + //$this->body = $parser->body; + $this->body = $response['body']; + $this->status_code = $parser->status_code; + } + } + } + else + { + $this->error = 'invalid URL'; + $this->success = false; + } + } +} \ No newline at end of file diff --git a/inc/3rdparty/libraries/language-detect/LanguageDetect.php b/inc/3rdparty/libraries/language-detect/LanguageDetect.php index 09b1154..382d869 100644 --- a/inc/3rdparty/libraries/language-detect/LanguageDetect.php +++ b/inc/3rdparty/libraries/language-detect/LanguageDetect.php @@ -6,23 +6,24 @@ * Attempts to detect the language of a sample of text by correlating ranked * 3-gram frequencies to a table of 3-gram frequencies of known languages. * - * Implements a version of a technique originally proposed by Cavnar & Trenkle - * (1994): "N-Gram-Based Text Categorization" + * Implements a version of a technique originally proposed by Cavnar & Trenkle + * (1994): "N-Gram-Based Text Categorization" * - * PHP versions 4 and 5 + * PHP version 5 * - * @category Text - * @package Text_LanguageDetect - * @author Nicholas Pisarro - * @copyright 2005-2006 Nicholas Pisarro - * @license http://www.debian.org/misc/bsd.license BSD - * @version CVS: $Id: LanguageDetect.php,v 1.20 2008/07/01 02:09:15 taak Exp $ - * @link http://pear.php.net/package/Text_LanguageDetect/ - * @link http://langdetect.blogspot.com/ + * @category Text + * @package Text_LanguageDetect + * @author Nicholas Pisarro + * @copyright 2005-2006 Nicholas Pisarro + * @license http://www.debian.org/misc/bsd.license BSD + * @version SVN: $Id: LanguageDetect.php 322353 2012-01-16 08:41:43Z cweiske $ + * @link http://pear.php.net/package/Text_LanguageDetect/ + * @link http://langdetect.blogspot.com/ */ -//require_once 'PEAR.php'; -require_once 'Parser.php'; +require_once 'LanguageDetect/Exception.php'; +require_once 'LanguageDetect/Parser.php'; +require_once 'LanguageDetect/ISO639.php'; /** * Language detection class @@ -41,9 +42,10 @@ require_once 'Parser.php'; * * echo "Supported languages:\n"; * - * $langs = $l->getLanguages(); - * if (PEAR::isError($langs)) { - * die($langs->getMessage()); + * try { + * $langs = $l->getLanguages(); + * } catch (Text_LanguageDetect_Exception $e) { + * die($e->getMessage()); * } * * sort($langs); @@ -54,38 +56,38 @@ require_once 'Parser.php'; * } * * - * @category Text - * @package Text_LanguageDetect - * @author Nicholas Pisarro - * @copyright 2005 Nicholas Pisarro - * @license http://www.debian.org/misc/bsd.license BSD - * @version Release: @package_version@ - * @todo allow users to generate their own language models + * @category Text + * @package Text_LanguageDetect + * @author Nicholas Pisarro + * @copyright 2005 Nicholas Pisarro + * @license http://www.debian.org/misc/bsd.license BSD + * @version Release: @package_version@ + * @link http://pear.php.net/package/Text_LanguageDetect/ + * @todo allow users to generate their own language models */ - class Text_LanguageDetect { - /** + /** * The filename that stores the trigram data for the detector * - * If this value starts with a slash (/) or a dot (.) the value of + * If this value starts with a slash (/) or a dot (.) the value of * $this->_data_dir will be ignored - * + * * @var string * @access private */ - var $_db_filename = './lang.dat'; + var $_db_filename = 'lang.dat'; /** * The filename that stores the unicode block definitions * - * If this value starts with a slash (/) or a dot (.) the value of + * If this value starts with a slash (/) or a dot (.) the value of * $this->_data_dir will be ignored - * + * * @var string * @access private */ - var $_unicode_db_filename = './unicode_blocks.dat'; + var $_unicode_db_filename = 'unicode_blocks.dat'; /** * The data directory @@ -99,11 +101,8 @@ class Text_LanguageDetect /** * The trigram data for comparison - * - * Will be loaded on start from $this->_db_filename * - * May be set to a PEAR_Error object if there is an error during its - * initialization + * Will be loaded on start from $this->_db_filename * * @var array * @access private @@ -120,7 +119,7 @@ class Text_LanguageDetect /** * The size of the trigram data arrays - * + * * @var int * @access private */ @@ -140,7 +139,7 @@ class Text_LanguageDetect /** * Whether or not to simulate perl's Language::Guess exactly - * + * * @access private * @var bool * @see setPerlCompatible() @@ -164,19 +163,25 @@ class Text_LanguageDetect */ var $_clusters; + /** + * Which type of "language names" are accepted and returned: + * + * 0 - language name ("english") + * 2 - 2-letter ISO 639-1 code ("en") + * 3 - 3-letter ISO 639-2 code ("eng") + */ + var $_name_mode = 0; + /** * Constructor * * Will attempt to load the language database. If it fails, you will get - * a PEAR_Error object returned when you try to use detect() - * + * an exception. */ - function Text_LanguageDetect($db=null, $unicode_db=null) + function __construct() { - if (isset($db)) $this->_db_filename = $db; - if (isset($unicode_db)) $this->_unicode_db_filename = $unicode_db; - $data = $this->_readdb($this->_db_filename); + $this->_checkTrigram($data['trigram']); $this->_lang_db = $data['trigram']; if (isset($data['trigram-unicodemap'])) { @@ -186,29 +191,32 @@ class Text_LanguageDetect // Not yet implemented: if (isset($data['trigram-clusters'])) { $this->_clusters = $data['trigram-clusters']; - } + } } /** * Returns the path to the location of the database * - * @access private - * @return string expected path to the language model database + * @param string $fname File name to load + * + * @return string expected path to the language model database + * @access private */ function _get_data_loc($fname) { - return $fname; + return dirname(__FILE__).'/'.$fname; } /** * Loads the language trigram database from filename * * Trigram datbase should be a serialize()'d array - * - * @access private - * @param string $fname the filename where the data is stored - * @return array the language model data - * @throws PEAR_Error + * + * @param string $fname the filename where the data is stored + * + * @return array the language model data + * @throws Text_LanguageDetect_Exception + * @access private */ function _readdb($fname) { @@ -217,79 +225,74 @@ class Text_LanguageDetect // input check if (!file_exists($fname)) { - throw new Exception('Language database does not exist.'); + throw new Text_LanguageDetect_Exception( + 'Language database does not exist: ' . $fname, + Text_LanguageDetect_Exception::DB_NOT_FOUND + ); } elseif (!is_readable($fname)) { - throw new Exception('Language database is not readable.'); + throw new Text_LanguageDetect_Exception( + 'Language database is not readable: ' . $fname, + Text_LanguageDetect_Exception::DB_NOT_READABLE + ); } - if (function_exists('file_get_contents')) { - return unserialize(file_get_contents($fname)); - } else { - // if you don't have file_get_contents(), - // then this is the next fastest way - ob_start(); - readfile($fname); - $contents = ob_get_contents(); - ob_end_clean(); - return unserialize($contents); - } + return unserialize(file_get_contents($fname)); } /** * Checks if this object is ready to detect languages - * - * @access private - * @param mixed &$err error object to be returned by reference, if any - * @return bool true if no errors + * + * @param array $trigram Trigram data from database + * + * @return void + * @access private */ - function _setup_ok(&$err) + function _checkTrigram($trigram) { - if (!is_array($this->_lang_db)) { + if (!is_array($trigram)) { if (ini_get('magic_quotes_runtime')) { - throw new Exception('Error loading database. Try turning magic_quotes_runtime off.'); - } else { - throw new Exception('Language database is not an array.'); + throw new Text_LanguageDetect_Exception( + 'Error loading database. Try turning magic_quotes_runtime off.', + Text_LanguageDetect_Exception::MAGIC_QUOTES + ); } - return false; - - } elseif (empty($this->_lang_db)) { - throw new Exception('Language database has no elements.'); - return false; - - } else { - return true; + throw new Text_LanguageDetect_Exception( + 'Language database is not an array.', + Text_LanguageDetect_Exception::DB_NOT_ARRAY + ); + } elseif (empty($trigram)) { + throw new Text_LanguageDetect_Exception( + 'Language database has no elements.', + Text_LanguageDetect_Exception::DB_EMPTY + ); } } /** * Omits languages * - * Pass this function the name of or an array of names of + * Pass this function the name of or an array of names of * languages that you don't want considered * - * If you're only expecting a limited set of languages, this can greatly + * If you're only expecting a limited set of languages, this can greatly * speed up processing * - * @access public - * @param mixed $omit_list language name or array of names to omit - * @param bool $include_only if true will include (rather than - * exclude) only those in the list - * @return int number of languages successfully deleted - * @throws PEAR_Error + * @param mixed $omit_list language name or array of names to omit + * @param bool $include_only if true will include (rather than + * exclude) only those in the list + * + * @return int number of languages successfully deleted + * @throws Text_LanguageDetect_Exception */ - function omitLanguages($omit_list, $include_only = false) + public function omitLanguages($omit_list, $include_only = false) { - - // setup check - if (!$this->_setup_ok($err)) { - return $err; - } - $deleted = 0; - // deleting the given languages + $omit_list = $this->_convertFromNameMode($omit_list); + if (!$include_only) { + // deleting the given languages if (!is_array($omit_list)) { $omit_list = strtolower($omit_list); // case desensitize if (isset($this->_lang_db[$omit_list])) { @@ -301,12 +304,12 @@ class Text_LanguageDetect if (isset($this->_lang_db[$omit_lang])) { unset($this->_lang_db[$omit_lang]); $deleted++; - } + } } } - // deleting all except the given languages } else { + // deleting all except the given languages if (!is_array($omit_list)) { $omit_list = array($omit_list); } @@ -327,7 +330,7 @@ class Text_LanguageDetect // reset the cluster cache if the number of languages changes // this will then have to be recalculated if (isset($this->_clusters) && $deleted > 0) { - unset($this->_clusters); + $this->_clusters = null; } return $deleted; @@ -339,49 +342,40 @@ class Text_LanguageDetect * * @access public * @return int the number of languages - * @throws PEAR_Error + * @throws Text_LanguageDetect_Exception */ function getLanguageCount() { - if (!$this->_setup_ok($err)) { - return $err; - } else { - return count($this->_lang_db); - } + return count($this->_lang_db); } /** - * Returns true if a given language exists + * Checks if the language with the given name exists in the database * - * If passed an array of names, will return true only if all exist + * @param mixed $lang Language name or array of language names * - * @access public - * @param mixed $lang language name or array of language names - * @return bool true if language model exists - * @throws PEAR_Error + * @return bool true if language model exists */ - function languageExists($lang) + public function languageExists($lang) { - if (!$this->_setup_ok($err)) { - return $err; - } else { - // string - if (is_string($lang)) { - return isset($this->_lang_db[strtolower($lang)]); + $lang = $this->_convertFromNameMode($lang); - // array - } elseif (is_array($lang)) { - foreach ($lang as $test_lang) { - if (!isset($this->_lang_db[strtolower($test_lang)])) { - return false; - } + if (is_string($lang)) { + return isset($this->_lang_db[strtolower($lang)]); + + } elseif (is_array($lang)) { + foreach ($lang as $test_lang) { + if (!isset($this->_lang_db[strtolower($test_lang)])) { + return false; } - return true; - - // other (error) - } else { - throw new Exception('Unknown type passed to languageExists()'); } + return true; + + } else { + throw new Text_LanguageDetect_Exception( + 'Unsupported parameter type passed to languageExists()', + Text_LanguageDetect_Exception::PARAM_TYPE + ); } } @@ -389,25 +383,24 @@ class Text_LanguageDetect * Returns the list of detectable languages * * @access public - * @return array the names of the languages known to this object - * @throws PEAR_Error + * @return array the names of the languages known to this object<<<<<<< + * @throws Text_LanguageDetect_Exception */ function getLanguages() { - if (!$this->_setup_ok($err)) { - return $err; - } else { - return array_keys($this->_lang_db); - } + return $this->_convertToNameMode( + array_keys($this->_lang_db) + ); } /** * Make this object behave like Language::Guess - * - * @access public - * @param bool $setting false to turn off perl compatibility + * + * @param bool $setting false to turn off perl compatibility + * + * @return void */ - function setPerlCompatible($setting = true) + public function setPerlCompatible($setting = true) { if (is_bool($setting)) { // input check $this->_perl_compatible = $setting; @@ -421,6 +414,21 @@ class Text_LanguageDetect } + /** + * Sets the way how language names are accepted and returned. + * + * @param integer $name_mode One of the following modes: + * 0 - language name ("english") + * 2 - 2-letter ISO 639-1 code ("en") + * 3 - 3-letter ISO 639-2 code ("eng") + * + * @return void + */ + function setNameMode($name_mode) + { + $this->_name_mode = $name_mode; + } + /** * Whether to use unicode block ranges in detection * @@ -429,10 +437,11 @@ class Text_LanguageDetect * in languages that use latin scripts. In other cases it should speed up * detection noticeably. * - * @access public - * @param bool $setting false to turn off + * @param bool $setting false to turn off + * + * @return void */ - function useUnicodeBlocks($setting = true) + public function useUnicodeBlocks($setting = true) { if (is_bool($setting)) { $this->_use_unicode_narrowing = $setting; @@ -442,15 +451,15 @@ class Text_LanguageDetect /** * Converts a piece of text into trigrams * - * Superceded by the Text_LanguageDetect_Parser class + * @param string $text text to convert * - * @access private - * @param string $text text to convert - * @return array array of trigram frequencies + * @return array array of trigram frequencies + * @access private + * @deprecated Superceded by the Text_LanguageDetect_Parser class */ function _trigram($text) { - $s = new Text_LanguageDetect_Parser($text, $this->_db_filename, $this->_unicode_db_filename); + $s = new Text_LanguageDetect_Parser($text); $s->prepareTrigram(); $s->prepareUnicode(false); $s->setPadStart(!$this->_perl_compatible); @@ -463,11 +472,12 @@ class Text_LanguageDetect * * Thresholds (cuts off) the list at $this->_threshold * - * @access protected - * @param array $arr array of trgram - * @return array ranks of trigrams + * @param array $arr array of trigram + * + * @return array ranks of trigrams + * @access protected */ - function _arr_rank(&$arr) + function _arr_rank($arr) { // sorts alphabetically first as a standard way of breaking rank ties @@ -494,14 +504,17 @@ class Text_LanguageDetect /** * Sorts an array by value breaking ties alphabetically - * - * @access private - * @param array &$arr the array to sort + * + * @param array &$arr the array to sort + * + * @return void + * @access private */ function _bub_sort(&$arr) { // should do the same as this perl statement: - // sort { $trigrams{$b} == $trigrams{$a} ? $a cmp $b : $trigrams{$b} <=> $trigrams{$a} } + // sort { $trigrams{$b} == $trigrams{$a} + // ? $a cmp $b : $trigrams{$b} <=> $trigrams{$a} } // needs to sort by both key and value at once // using the key to break ties for the value @@ -528,13 +541,14 @@ class Text_LanguageDetect /** * Sort function used by bubble sort * - * Callback function for usort(). + * Callback function for usort(). * - * @access private - * @param array first param passed by usort() - * @param array second param passed by usort() - * @return int 1 if $a is greater, -1 if not - * @see _bub_sort() + * @param array $a first param passed by usort() + * @param array $b second param passed by usort() + * + * @return int 1 if $a is greater, -1 if not + * @see _bub_sort() + * @access private */ function _sort_func($a, $b) { @@ -542,12 +556,12 @@ class Text_LanguageDetect list($a_key, $a_value) = $a; list($b_key, $b_value) = $b; - // if the values are the same, break ties using the key if ($a_value == $b_value) { + // if the values are the same, break ties using the key return strcmp($a_key, $b_key); - // if not, just sort normally } else { + // if not, just sort normally if ($a_value > $b_value) { return -1; } else { @@ -559,23 +573,24 @@ class Text_LanguageDetect } /** - * Calculates a linear rank-order distance statistic between two sets of + * Calculates a linear rank-order distance statistic between two sets of * ranked trigrams * - * Sums the differences in rank for each trigram. If the trigram does not + * Sums the differences in rank for each trigram. If the trigram does not * appear in both, consider it a difference of $this->_threshold. * * This distance measure was proposed by Cavnar & Trenkle (1994). Despite * its simplicity it has been shown to be highly accurate for language * identification tasks. * - * @access private - * @param array $arr1 the reference set of trigram ranks - * @param array $arr2 the target set of trigram ranks - * @return int the sum of the differences between the ranks of - * the two trigram sets + * @param array $arr1 the reference set of trigram ranks + * @param array $arr2 the target set of trigram ranks + * + * @return int the sum of the differences between the ranks of + * the two trigram sets + * @access private */ - function _distance(&$arr1, &$arr2) + function _distance($arr1, $arr2) { $sumdist = 0; @@ -598,14 +613,15 @@ class Text_LanguageDetect /** * Normalizes the score returned by _distance() - * + * * Different if perl compatible or not * - * @access private - * @param int $score the score from _distance() - * @param int $base_count the number of trigrams being considered - * @return float the normalized score - * @see _distance() + * @param int $score the score from _distance() + * @param int $base_count the number of trigrams being considered + * + * @return float the normalized score + * @see _distance() + * @access private */ function _normalize_score($score, $base_count = null) { @@ -630,29 +646,24 @@ class Text_LanguageDetect * * If perl compatible, the score is 300-0, 0 being most similar. * Otherwise, it's 0-1 with 1 being most similar. - * + * * The $sample text should be at least a few sentences in length; * should be ascii-7 or utf8 encoded, if another and the mbstring extension * is present it will try to detect and convert. However, experience has - * shown that mb_detect_encoding() *does not work very well* with at least + * shown that mb_detect_encoding() *does not work very well* with at least * some types of encoding. * - * @access public - * @param string $sample a sample of text to compare. - * @param int $limit if specified, return an array of the most likely - * $limit languages and their scores. - * @return mixed sorted array of language scores, blank array if no - * useable text was found, or PEAR_Error if error - * with the object setup - * @see _distance() - * @throws PEAR_Error + * @param string $sample a sample of text to compare. + * @param int $limit if specified, return an array of the most likely + * $limit languages and their scores. + * + * @return mixed sorted array of language scores, blank array if no + * useable text was found + * @see _distance() + * @throws Text_LanguageDetect_Exception */ - function detect($sample, $limit = 0) + public function detect($sample, $limit = 0) { - if (!$this->_setup_ok($err)) { - return $err; - } - // input check if (!Text_LanguageDetect_Parser::validateString($sample)) { return array(); @@ -660,36 +671,27 @@ class Text_LanguageDetect // check char encoding // (only if mbstring extension is compiled and PHP > 4.0.6) - if (function_exists('mb_detect_encoding') - && function_exists('mb_convert_encoding')) { - + if (function_exists('mb_detect_encoding') + && function_exists('mb_convert_encoding') + ) { // mb_detect_encoding isn't very reliable, to say the least - // detection should still work with a sufficient sample of ascii characters + // detection should still work with a sufficient sample + // of ascii characters $encoding = mb_detect_encoding($sample); // mb_detect_encoding() will return FALSE if detection fails // don't attempt conversion if that's the case - if ($encoding != 'ASCII' && $encoding != 'UTF-8' && $encoding !== false) { - - if (function_exists('mb_list_encodings')) { - - // verify the encoding exists in mb_list_encodings - if (in_array($encoding, mb_list_encodings())) { - $sample = mb_convert_encoding($sample, 'UTF-8', $encoding); - } - - // if the previous condition failed: - // somehow we detected an encoding that also we don't support - - } else { - // php 4 doesnt have mb_list_encodings() - // so attempt with error suppression - $sample = @mb_convert_encoding($sample, 'UTF-8', $encoding); + if ($encoding != 'ASCII' && $encoding != 'UTF-8' + && $encoding !== false + ) { + // verify the encoding exists in mb_list_encodings + if (in_array($encoding, mb_list_encodings())) { + $sample = mb_convert_encoding($sample, 'UTF-8', $encoding); } } } - $sample_obj = new Text_LanguageDetect_Parser($sample, $this->_db_filename, $this->_unicode_db_filename); + $sample_obj = new Text_LanguageDetect_Parser($sample); $sample_obj->prepareTrigram(); if ($this->_use_unicode_narrowing) { $sample_obj->prepareUnicode(); @@ -713,7 +715,10 @@ class Text_LanguageDetect if (is_array($blocks)) { $present_blocks = array_keys($blocks); } else { - throw new Exception('Error during block detection'); + throw new Text_LanguageDetect_Exception( + 'Error during block detection', + Text_LanguageDetect_Exception::BLOCK_DETECTION + ); } $possible_langs = array(); @@ -731,30 +736,30 @@ class Text_LanguageDetect } // could also try an intersect operation rather than a union - // in other words, choose languages whose trigrams contain + // in other words, choose languages whose trigrams contain // ALL of the unicode blocks found in this sample // would improve speed but would be completely thrown off by an // unexpected character, like an umlaut appearing in english text $possible_langs = array_intersect( - array_keys($this->_lang_db), - array_unique($possible_langs) + array_keys($this->_lang_db), + array_unique($possible_langs) ); - // needs to intersect it with the keys of _lang_db in case + // needs to intersect it with the keys of _lang_db in case // languages have been omitted - // or just try 'em all } else { + // or just try 'em all $possible_langs = array_keys($this->_lang_db); } foreach ($possible_langs as $lang) { - $scores[$lang] = - $this->_normalize_score( - $this->_distance($this->_lang_db[$lang], $trigram_freqs), - $trigram_count); + $scores[$lang] = $this->_normalize_score( + $this->_distance($this->_lang_db[$lang], $trigram_freqs), + $trigram_count + ); } unset($sample_obj); @@ -772,7 +777,6 @@ class Text_LanguageDetect $limited_scores = array(); $i = 0; - foreach ($scores as $key => $value) { if ($i++ >= $limit) { break; @@ -781,9 +785,9 @@ class Text_LanguageDetect $limited_scores[$key] = $value; } - return $limited_scores; + return $this->_convertToNameMode($limited_scores, true); } else { - return $scores; + return $this->_convertToNameMode($scores, true); } } @@ -791,35 +795,33 @@ class Text_LanguageDetect * Returns only the most similar language to the text sample * * Calls $this->detect() and returns only the top result - * - * @access public - * @param string $sample text to detect the language of - * @return string the name of the most likely language - * or null if no language is similar - * @see detect() - * @throws PEAR_Error + * + * @param string $sample text to detect the language of + * + * @return string the name of the most likely language + * or null if no language is similar + * @see detect() + * @throws Text_LanguageDetect_Exception */ - function detectSimple($sample) + public function detectSimple($sample) { $scores = $this->detect($sample, 1); // if top language has the maximum possible score, // then the top score will have been picked at random - if ( !is_array($scores) - || empty($scores) - || current($scores) == $this->_max_score) { - + if (!is_array($scores) || empty($scores) + || current($scores) == $this->_max_score + ) { return null; - } else { - return ucfirst(key($scores)); + return key($scores); } } /** * Returns an array containing the most similar language and a confidence * rating - * + * * Confidence is a simple measure calculated from the similarity score * minus the similarity score from the next most similar language * divided by the highest possible score. Languages that have closely @@ -827,46 +829,43 @@ class Text_LanguageDetect * confidence scores. * * The similarity score answers the question "How likely is the text the - * returned language regardless of the other languages considered?" The + * returned language regardless of the other languages considered?" The * confidence score is one way of answering the question "how likely is the * text the detected language relative to the rest of the language model * set?" * * To see how similar languages are a priori, see languageSimilarity() - * - * @access public - * @param string $sample text for which language will be detected - * @return array most similar language, score and confidence rating - * or null if no language is similar - * @see detect() - * @throws PEAR_Error + * + * @param string $sample text for which language will be detected + * + * @return array most similar language, score and confidence rating + * or null if no language is similar + * @see detect() + * @throws Text_LanguageDetect_Exception */ - function detectConfidence($sample) + public function detectConfidence($sample) { $scores = $this->detect($sample, 2); - // if most similar language has the max score, it + // if most similar language has the max score, it // will have been picked at random - if ( !is_array($scores) - || empty($scores) - || current($scores) == $this->_max_score) { - + if (!is_array($scores) || empty($scores) + || current($scores) == $this->_max_score + ) { return null; } - $arr['language'] = ucfirst(key($scores)); + $arr['language'] = key($scores); $arr['similarity'] = current($scores); if (next($scores) !== false) { // if false then no next element // the goal is to return a higher value if the distance between // the similarity of the first score and the second score is high if ($this->_perl_compatible) { - - $arr['confidence'] = - (current($scores) - $arr['similarity']) / $this->_max_score; + $arr['confidence'] = (current($scores) - $arr['similarity']) + / $this->_max_score; } else { - $arr['confidence'] = $arr['similarity'] - current($scores); } @@ -882,32 +881,26 @@ class Text_LanguageDetect * Returns the distribution of unicode blocks in a given utf8 string * * For the block name of a single char, use unicodeBlockName() - * - * @access public - * @param string $str input string. Must be ascii or utf8 - * @param bool $skip_symbols if true, skip ascii digits, symbols and - * non-printing characters. Includes spaces, - * newlines and common punctutation characters. + * + * @param string $str input string. Must be ascii or utf8 + * @param bool $skip_symbols if true, skip ascii digits, symbols and + * non-printing characters. Includes spaces, + * newlines and common punctutation characters. + * * @return array - * @throws PEAR_Error + * @throws Text_LanguageDetect_Exception */ - function detectUnicodeBlocks($str, $skip_symbols) + public function detectUnicodeBlocks($str, $skip_symbols) { - // input check - if (!is_bool($skip_symbols)) { - throw new Exception('Second parameter must be boolean'); - } + $skip_symbols = (bool)$skip_symbols; + $str = (string)$str; - if (!is_string($str)) { - throw new Exception('First parameter was not a string'); - } - - $sample_obj = new Text_LanguageDetect_Parser($str, $this->_db_filename, $this->_unicode_db_filename); + $sample_obj = new Text_LanguageDetect_Parser($str); $sample_obj->prepareUnicode(); $sample_obj->prepareTrigram(false); $sample_obj->setUnicodeSkipSymbols($skip_symbols); $sample_obj->analyze(); - $blocks =& $sample_obj->getUnicodeBlocks(); + $blocks = $sample_obj->getUnicodeBlocks(); unset($sample_obj); return $blocks; } @@ -915,38 +908,37 @@ class Text_LanguageDetect /** * Returns the block name for a given unicode value * - * If passed a string, will assume it is being passed a UTF8-formatted + * If passed a string, will assume it is being passed a UTF8-formatted * character and will automatically convert. Otherwise it will assume it * is being passed a numeric unicode value. * * Make sure input is of the correct type! * - * @access public * @param mixed $unicode unicode value or utf8 char + * * @return mixed the block name string or false if not found - * @throws PEAR_Error + * @throws Text_LanguageDetect_Exception */ - function unicodeBlockName($unicode) { + public function unicodeBlockName($unicode) + { if (is_string($unicode)) { // assume it is being passed a utf8 char, so convert it - - // input check - if ($this->utf8strlen($unicode) > 1) { - throw new Exception('Pass this function only a single char'); + if (self::utf8strlen($unicode) > 1) { + throw new Text_LanguageDetect_Exception( + 'Pass a single char only to this method', + Text_LanguageDetect_Exception::PARAM_TYPE + ); } - $unicode = $this->_utf8char2unicode($unicode); - if ($unicode == -1) { - throw new Exception('Malformatted char'); - } - - // input check } elseif (!is_int($unicode)) { - throw new Exception('Input must be of type string or int.'); + throw new Text_LanguageDetect_Exception( + 'Input must be of type string or int.', + Text_LanguageDetect_Exception::PARAM_TYPE + ); } - $blocks =& $this->_read_unicode_block_db(); + $blocks = $this->_read_unicode_block_db(); $result = $this->_unicode_block_name($unicode, $blocks); @@ -964,14 +956,17 @@ class Text_LanguageDetect * the public interface for this function, which does input checks which * this function omits for speed. * - * @access protected - * @param int $unicode the unicode value - * @param array &$blocks the block database - * @param int $block_count the number of defined blocks in the database - * @see unicodeBlockName() + * @param int $unicode the unicode value + * @param array $blocks the block database + * @param int $block_count the number of defined blocks in the database + * + * @return mixed Block name, -1 if it failed + * @see unicodeBlockName() + * @access protected */ - function _unicode_block_name($unicode, &$blocks, $block_count = -1) { - // for a reference, see + function _unicode_block_name($unicode, $blocks, $block_count = -1) + { + // for a reference, see // http://www.unicode.org/Public/UNIDATA/Blocks.txt // assume that ascii characters are the most common @@ -994,35 +989,36 @@ class Text_LanguageDetect while ($low <= $high) { $mid = floor(($low + $high) / 2); - // if it's lower than the lower bound if ($unicode < $blocks[$mid][0]) { + // if it's lower than the lower bound $high = $mid - 1; - // if it's higher than the upper bound } elseif ($unicode > $blocks[$mid][1]) { + // if it's higher than the upper bound $low = $mid + 1; - // found it } else { + // found it return $blocks[$mid]; } } - // failed to find the block + // failed to find the block return -1; - // todo: differentiate when it's out of range or when it falls + // todo: differentiate when it's out of range or when it falls // into an unassigned range? } /** * Brings up the unicode block database * - * @access protected * @return array the database of unicode block definitions - * @throws PEAR_Error + * @throws Text_LanguageDetect_Exception + * @access protected */ - function &_read_unicode_block_db() { + function _read_unicode_block_db() + { // since the unicode definitions are always going to be the same, // might as well share the memory for the db with all other instances // of this class @@ -1037,29 +1033,27 @@ class Text_LanguageDetect /** * Calculate the similarities between the language models - * + * * Use this function to see how similar languages are to each other. * * If passed 2 language names, will return just those languages compared. * If passed 1 language name, will return that language compared to * all others. - * If passed none, will return an array of every language model compared + * If passed none, will return an array of every language model compared * to every other one. * - * @access public - * @param string $lang1 the name of the first language to be compared - * @param string $lang2 the name of the second language to be compared - * @return array scores of every language compared - * or the score of just the provided languages - * or null if one of the supplied languages does not exist - * @throws PEAR_Error + * @param string $lang1 the name of the first language to be compared + * @param string $lang2 the name of the second language to be compared + * + * @return array scores of every language compared + * or the score of just the provided languages + * or null if one of the supplied languages does not exist + * @throws Text_LanguageDetect_Exception */ - function languageSimilarity($lang1 = null, $lang2 = null) + public function languageSimilarity($lang1 = null, $lang2 = null) { - if (!$this->_setup_ok($err)) { - return $err; - } - + $lang1 = $this->_convertFromNameMode($lang1); + $lang2 = $this->_convertFromNameMode($lang2); if ($lang1 != null) { $lang1 = strtolower($lang1); @@ -1069,12 +1063,8 @@ class Text_LanguageDetect } if ($lang2 != null) { - - // can't only set the second param - if ($lang1 == null) { - return null; - // check if language model exists - } elseif (!isset($this->_lang_db[$lang2])) { + if (!isset($this->_lang_db[$lang2])) { + // check if language model exists return null; } @@ -1088,14 +1078,15 @@ class Text_LanguageDetect ) ); - - // compare just $lang1 to all languages } else { + // compare just $lang1 to all languages $return_arr = array(); foreach ($this->_lang_db as $key => $value) { - if ($key != $lang1) { // don't compare a language to itself + if ($key != $lang1) { + // don't compare a language to itself $return_arr[$key] = $this->_normalize_score( - $this->_distance($this->_lang_db[$lang1], $value)); + $this->_distance($this->_lang_db[$lang1], $value) + ); } } asort($return_arr); @@ -1104,30 +1095,27 @@ class Text_LanguageDetect } - // compare all languages to each other } else { + // compare all languages to each other $return_arr = array(); foreach (array_keys($this->_lang_db) as $lang1) { foreach (array_keys($this->_lang_db) as $lang2) { - // skip comparing languages to themselves - if ($lang1 != $lang2) { - - // don't re-calculate what's already been done + if ($lang1 != $lang2) { + if (isset($return_arr[$lang2][$lang1])) { + // don't re-calculate what's already been done + $return_arr[$lang1][$lang2] + = $return_arr[$lang2][$lang1]; - $return_arr[$lang1][$lang2] = - $return_arr[$lang2][$lang1]; - - // calculate } else { - - $return_arr[$lang1][$lang2] = - $this->_normalize_score( - $this->_distance( - $this->_lang_db[$lang1], - $this->_lang_db[$lang2] - ) + // calculate + $return_arr[$lang1][$lang2] + = $this->_normalize_score( + $this->_distance( + $this->_lang_db[$lang1], + $this->_lang_db[$lang2] + ) ); } @@ -1150,20 +1138,14 @@ class Text_LanguageDetect * * @access public * @return array language cluster data - * @throws PEAR_Error + * @throws Text_LanguageDetect_Exception * @see languageSimilarity() - * @deprecated this function will eventually be removed and placed into + * @deprecated this function will eventually be removed and placed into * the model generation class */ function clusterLanguages() { // todo: set the maximum number of clusters - - // setup check - if (!$this->_setup_ok($err)) { - return $err; - } - // return cached result, if any if (isset($this->_clusters)) { return $this->_clusters; @@ -1177,7 +1159,10 @@ class Text_LanguageDetect foreach ($langs as $lang) { if (!isset($this->_lang_db[$lang])) { - throw new Exception("missing $lang!\n"); + throw new Text_LanguageDetect_Exception( + "missing $lang!", + Text_LanguageDetect_Exception::UNKNOWN_LANGUAGE + ); } } @@ -1186,7 +1171,9 @@ class Text_LanguageDetect $langs[$lang1] = $lang1; unset($langs[$old_key]); } - + + $result_data = $really_map = array(); + $i = 0; while (count($langs) > 2 && $i++ < 200) { $highest_score = -1; @@ -1194,18 +1181,22 @@ class Text_LanguageDetect $highest_key2 = ''; foreach ($langs as $lang1) { foreach ($langs as $lang2) { - if ( $lang1 != $lang2 - && $arr[$lang1][$lang2] > $highest_score) { + if ($lang1 != $lang2 + && $arr[$lang1][$lang2] > $highest_score + ) { $highest_score = $arr[$lang1][$lang2]; $highest_key1 = $lang1; $highest_key2 = $lang2; } } } - + if (!$highest_key1) { // should not ever happen - throw new Exception("no highest key? (step: $i)"); + throw new Text_LanguageDetect_Exception( + "no highest key? (step: $i)", + Text_LanguageDetect_Exception::NO_HIGHEST_KEY + ); } if ($highest_score == 0) { @@ -1217,7 +1208,7 @@ class Text_LanguageDetect $sum1 = array_sum($arr[$highest_key1]); $sum2 = array_sum($arr[$highest_key2]); - // use the score for the one that is most similar to the rest of + // use the score for the one that is most similar to the rest of // the field as the score for the group // todo: could try averaging or "centroid" method instead // seems like that might make more sense @@ -1248,7 +1239,7 @@ class Text_LanguageDetect $really_lang = $replaceme; while (isset($really_map[$really_lang])) { $really_lang = $really_map[$really_lang]; - } + } $really_map[$newkey] = $really_lang; @@ -1259,8 +1250,8 @@ class Text_LanguageDetect $arr[$key1][$newkey] = $arr[$key1][$key2]; unset($arr[$key1][$key2]); // replacing $arr[$key1][$key2] with $arr[$key1][$newkey] - } - + } + if ($key1 == $replaceme) { $arr[$newkey][$key2] = $arr[$key1][$key2]; unset($arr[$key1][$key2]); @@ -1273,7 +1264,7 @@ class Text_LanguageDetect } } } - + unset($langs[$highest_key1]); unset($langs[$highest_key2]); @@ -1293,7 +1284,7 @@ class Text_LanguageDetect } $return_val = array( - 'open_forks' => $langs, + 'open_forks' => $langs, // the top level of clusters // clusters that are mutually exclusive // or specified by a specific maximum @@ -1323,11 +1314,11 @@ class Text_LanguageDetect * use, and it may disappear or its functionality may change in future * releases without notice. * - * This compares the sample text to top the top level of clusters. If the + * This compares the sample text to top the top level of clusters. If the * sample is similar to the cluster it will drop down and compare it to the * languages in the cluster, and so on until it hits a leaf node. * - * this should find the language in considerably fewer compares + * this should find the language in considerably fewer compares * (the equivalent of a binary search), however clusterLanguages() is costly * and the loss of accuracy from this technique is significant. * @@ -1337,15 +1328,14 @@ class Text_LanguageDetect * was very large, however in such cases some method of Bayesian inference * might be more helpful. * - * @see clusterLanguages() - * @access public - * @param string $str input string - * @return array language scores (only those compared) - * @throws PEAR_Error + * @param string $str input string + * + * @return array language scores (only those compared) + * @throws Text_LanguageDetect_Exception + * @see clusterLanguages() */ - function clusteredSearch($str) + public function clusteredSearch($str) { - // input check if (!Text_LanguageDetect_Parser::validateString($str)) { return array(); @@ -1359,7 +1349,7 @@ class Text_LanguageDetect $dendogram_data = $result['fork_data']; $dendogram_alias = $result['name_map']; - $sample_obj = new Text_LanguageDetect_Parser($str, $this->_db_filename, $this->_unicode_db_filename); + $sample_obj = new Text_LanguageDetect_Parser($str); $sample_obj->prepareTrigram(); $sample_obj->setPadStart(!$this->_perl_compatible); $sample_obj->analyze(); @@ -1372,7 +1362,7 @@ class Text_LanguageDetect } $i = 0; // counts the number of steps - + foreach ($dendogram_start as $lang) { if (isset($dendogram_alias[$lang])) { $lang_key = $dendogram_alias[$lang]; @@ -1382,7 +1372,8 @@ class Text_LanguageDetect $scores[$lang] = $this->_normalize_score( $this->_distance($this->_lang_db[$lang_key], $sample_result), - $sample_count); + $sample_count + ); $i++; } @@ -1411,7 +1402,8 @@ class Text_LanguageDetect $scores[$lang] = $this->_normalize_score( $this->_distance($this->_lang_db[$lang_key], $sample_result), - $sample_count); + $sample_count + ); //todo: does not need to do same comparison again } @@ -1428,8 +1420,8 @@ class Text_LanguageDetect $diff = $scores[$cur_key] - $scores[$loser_key]; - // $cur_key ({$dendogram_alias[$cur_key]}) wins - // over $loser_key ({$dendogram_alias[$loser_key]}) + // $cur_key ({$dendogram_alias[$cur_key]}) wins + // over $loser_key ({$dendogram_alias[$loser_key]}) // with a difference of $diff } @@ -1439,9 +1431,9 @@ class Text_LanguageDetect // which paths the algorithm decided to take along the tree // but sometimes the last item is only the second highest - if ( ($this->_perl_compatible && (end($scores) > prev($scores))) - || (!$this->_perl_compatible && (end($scores) < prev($scores)))) { - + if (($this->_perl_compatible && (end($scores) > prev($scores))) + || (!$this->_perl_compatible && (end($scores) < prev($scores))) + ) { $real_last_score = current($scores); $real_last_key = key($scores); @@ -1449,7 +1441,7 @@ class Text_LanguageDetect unset($scores[$real_last_key]); $scores[$real_last_key] = $real_last_score; } - + if (!$this->_perl_compatible) { $scores = array_reverse($scores, true); @@ -1464,12 +1456,11 @@ class Text_LanguageDetect * * Returns the numbers of characters (not bytes) in a utf8 string * - * @static - * @access public - * @param string $str string to get the length of - * @return int number of chars + * @param string $str string to get the length of + * + * @return int number of chars */ - function utf8strlen($str) + public static function utf8strlen($str) { // utf8_decode() will convert unknown chars to '?', which is actually // ideal for counting. @@ -1482,53 +1473,45 @@ class Text_LanguageDetect /** * Returns the unicode value of a utf8 char * - * @access protected - * @param string $char a utf8 (possibly multi-byte) char - * @return int unicode value or -1 if malformatted + * @param string $char a utf8 (possibly multi-byte) char + * + * @return int unicode value + * @access protected + * @link http://en.wikipedia.org/wiki/UTF-8 */ - function _utf8char2unicode($char) { - + function _utf8char2unicode($char) + { // strlen() here will actually get the binary length of a single char switch (strlen($char)) { + case 1: + // normal ASCII-7 byte + // 0xxxxxxx --> 0xxxxxxx + return ord($char{0}); - // for a reference, see http://en.wikipedia.org/wiki/UTF-8 + case 2: + // 2 byte unicode + // 110zzzzx 10xxxxxx --> 00000zzz zxxxxxxx + $z = (ord($char{0}) & 0x000001F) << 6; + $x = (ord($char{1}) & 0x0000003F); + return ($z | $x); - case 1: - // normal ASCII-7 byte - // 0xxxxxxx --> 0xxxxxxx - return ord($char{0}); + case 3: + // 3 byte unicode + // 1110zzzz 10zxxxxx 10xxxxxx --> zzzzzxxx xxxxxxxx + $z = (ord($char{0}) & 0x0000000F) << 12; + $x1 = (ord($char{1}) & 0x0000003F) << 6; + $x2 = (ord($char{2}) & 0x0000003F); + return ($z | $x1 | $x2); - case 2: - // 2 byte unicode - // 110zzzzx 10xxxxxx --> 00000zzz zxxxxxxx - $z = (ord($char{0}) & 0x000001F) << 6; - $x = (ord($char{1}) & 0x0000003F); - - return ($z | $x); - - case 3: - // 3 byte unicode - // 1110zzzz 10zxxxxx 10xxxxxx --> zzzzzxxx xxxxxxxx - $z = (ord($char{0}) & 0x0000000F) << 12; - $x1 = (ord($char{1}) & 0x0000003F) << 6; - $x2 = (ord($char{2}) & 0x0000003F); - - return ($z | $x1 | $x2); - - case 4: - // 4 byte unicode - // 11110zzz 10zzxxxx 10xxxxxx 10xxxxxx --> - // 000zzzzz xxxxxxxx xxxxxxxx - $z1 = (ord($char{0}) & 0x00000007) << 18; - $z2 = (ord($char{1}) & 0x0000003F) << 12; - $x1 = (ord($char{2}) & 0x0000003F) << 6; - $x2 = (ord($char{3}) & 0x0000003F); - - return ($z1 | $z2 | $x1 | $x2); - - default: - // error: malformatted char? - return -1; + case 4: + // 4 byte unicode + // 11110zzz 10zzxxxx 10xxxxxx 10xxxxxx --> + // 000zzzzz xxxxxxxx xxxxxxxx + $z1 = (ord($char{0}) & 0x00000007) << 18; + $z2 = (ord($char{1}) & 0x0000003F) << 12; + $x1 = (ord($char{2}) & 0x0000003F) << 6; + $x2 = (ord($char{3}) & 0x0000003F); + return ($z1 | $z2 | $x1 | $x2); } } @@ -1536,18 +1519,18 @@ class Text_LanguageDetect * utf8-safe fast character iterator * * Will get the next character starting from $counter, which will then be - * incremented. If a multi-byte char the bytes will be concatenated and + * incremented. If a multi-byte char the bytes will be concatenated and * $counter will be incremeted by the number of bytes in the char. * - * @access private - * @param string &$str the string being iterated over - * @param int &$counter the iterator, will increment by reference - * @param bool $special_convert whether to do special conversions - * @return char the next (possibly multi-byte) char from $counter + * @param string $str the string being iterated over + * @param int &$counter the iterator, will increment by reference + * @param bool $special_convert whether to do special conversions + * + * @return char the next (possibly multi-byte) char from $counter + * @access private */ - function _next_char(&$str, &$counter, $special_convert = false) + static function _next_char($str, &$counter, $special_convert = false) { - $char = $str{$counter++}; $ord = ord($char); @@ -1556,7 +1539,6 @@ class Text_LanguageDetect // normal ascii one byte char if ($ord <= 127) { - // special conversions needed for this package // (that only apply to regular ascii characters) // lower case, and convert all non-alphanumeric characters @@ -1571,8 +1553,8 @@ class Text_LanguageDetect return $char; - // multi-byte chars } elseif ($ord >> 5 == 6) { // two-byte char + // multi-byte chars $nextchar = $str{$counter++}; // get next byte // lower-casing of non-ascii characters is still incomplete @@ -1582,27 +1564,27 @@ class Text_LanguageDetect if ($ord == 195) { $nextord = ord($nextchar); $nextord_adj = $nextord + 64; - // for a reference, see + // for a reference, see // http://www.ramsch.org/martin/uni/fmi-hp/iso8859-1.html // À - Þ but not × - if ( $nextord_adj >= 192 - && $nextord_adj <= 222 - && $nextord_adj != 215) { - - $nextchar = chr($nextord + 32); + if ($nextord_adj >= 192 + && $nextord_adj <= 222 + && $nextord_adj != 215 + ) { + $nextchar = chr($nextord + 32); } - // lower case cyrillic alphabet } elseif ($ord == 208) { + // lower case cyrillic alphabet $nextord = ord($nextchar); // if A - Pe if ($nextord >= 144 && $nextord <= 159) { // lower case $nextchar = chr($nextord + 32); - // if Er - Ya } elseif ($nextord >= 160 && $nextord <= 175) { + // if Er - Ya // lower case $char = chr(209); // == $ord++ $nextchar = chr($nextord - 32); @@ -1611,12 +1593,11 @@ class Text_LanguageDetect } // tag on next byte - return $char . $nextchar; - + return $char . $nextchar; } elseif ($ord >> 4 == 14) { // three-byte char - + // tag on next 2 bytes - return $char . $str{$counter++} . $str{$counter++}; + return $char . $str{$counter++} . $str{$counter++}; } elseif ($ord >> 3 == 30) { // four-byte char @@ -1628,8 +1609,85 @@ class Text_LanguageDetect } } + /** + * Converts an $language input parameter from the configured mode + * to the language name that is used internally. + * + * Works for strings and arrays. + * + * @param string|array $lang A language description ("english"/"en"/"eng") + * @param boolean $convertKey If $lang is an array, setting $key + * converts the keys to the language name. + * + * @return string|array Language name + */ + function _convertFromNameMode($lang, $convertKey = false) + { + if ($this->_name_mode == 0) { + return $lang; + } + + if ($this->_name_mode == 2) { + $method = 'code2ToName'; + } else { + $method = 'code3ToName'; + } + + if (is_string($lang)) { + return (string)Text_LanguageDetect_ISO639::$method($lang); + } + + $newlang = array(); + foreach ($lang as $key => $val) { + if ($convertKey) { + $newkey = (string)Text_LanguageDetect_ISO639::$method($key); + $newlang[$newkey] = $val; + } else { + $newlang[$key] = (string)Text_LanguageDetect_ISO639::$method($val); + } + } + return $newlang; + } + + /** + * Converts an $language output parameter from the language name that is + * used internally to the configured mode. + * + * Works for strings and arrays. + * + * @param string|array $lang A language description ("english"/"en"/"eng") + * @param boolean $convertKey If $lang is an array, setting $key + * converts the keys to the language name. + * + * @return string|array Language name + */ + function _convertToNameMode($lang, $convertKey = false) + { + if ($this->_name_mode == 0) { + return $lang; + } + + if ($this->_name_mode == 2) { + $method = 'nameToCode2'; + } else { + $method = 'nameToCode3'; + } + + if (is_string($lang)) { + return Text_LanguageDetect_ISO639::$method($lang); + } + + $newlang = array(); + foreach ($lang as $key => $val) { + if ($convertKey) { + $newkey = Text_LanguageDetect_ISO639::$method($key); + $newlang[$newkey] = $val; + } else { + $newlang[$key] = Text_LanguageDetect_ISO639::$method($val); + } + } + return $newlang; + } } -/* vim: set expandtab tabstop=4 shiftwidth=4 softtabstop=4: */ - -?> +/* vim: set expandtab tabstop=4 shiftwidth=4 softtabstop=4: */ \ No newline at end of file diff --git a/inc/3rdparty/libraries/readability/Readability.php b/inc/3rdparty/libraries/readability/Readability.php index 2e8991c..d0f09d7 100644 --- a/inc/3rdparty/libraries/readability/Readability.php +++ b/inc/3rdparty/libraries/readability/Readability.php @@ -1,1138 +1,1138 @@ -init(); -echo $r->articleContent->innerHTML; -*/ - -class Readability -{ - public $version = '1.7.1-without-multi-page'; - public $convertLinksToFootnotes = false; - public $revertForcedParagraphElements = true; - public $articleTitle; - public $articleContent; - public $dom; - public $url = null; // optional - URL where HTML was retrieved - public $debug = false; - public $lightClean = true; // preserves more content (experimental) added 2012-09-19 - protected $body = null; // - protected $bodyCache = null; // Cache the body HTML in case we need to re-use it later - protected $flags = 7; // 1 | 2 | 4; // Start with all flags set. - protected $success = false; // indicates whether we were able to extract or not - - /** - * All of the regular expressions in use within readability. - * Defined up here so we don't instantiate them repeatedly in loops. - **/ - public $regexps = array( - 'unlikelyCandidates' => '/combx|comment|community|disqus|extra|foot|header|menu|remark|rss|shoutbox|sidebar|sponsor|ad-break|agegate|pagination|pager|popup/i', - 'okMaybeItsACandidate' => '/and|article|body|column|main|shadow/i', - 'positive' => '/article|body|content|entry|hentry|main|page|attachment|pagination|post|text|blog|story/i', - 'negative' => '/combx|comment|com-|contact|foot|footer|_nav|footnote|masthead|media|meta|outbrain|promo|related|scroll|shoutbox|sidebar|sponsor|shopping|tags|tool|widget/i', - 'divToPElements' => '/<(a|blockquote|dl|div|img|ol|p|pre|table|ul)/i', - 'replaceBrs' => '/(]*>[ \n\r\t]*){2,}/i', - 'replaceFonts' => '/<(\/?)font[^>]*>/i', - // 'trimRe' => '/^\s+|\s+$/g', // PHP has trim() - 'normalize' => '/\s{2,}/', - 'killBreaks' => '/((\s| ?)*){1,}/', - 'video' => '!//(player\.|www\.)?(youtube|vimeo|viddler)\.com!i', - 'skipFootnoteLink' => '/^\s*(\[?[a-z0-9]{1,2}\]?|^|edit|citation needed)\s*$/i' - ); - - /* constants */ - const FLAG_STRIP_UNLIKELYS = 1; - const FLAG_WEIGHT_CLASSES = 2; - const FLAG_CLEAN_CONDITIONALLY = 4; - - /** - * Create instance of Readability - * @param string UTF-8 encoded string - * @param string (optional) URL associated with HTML (used for footnotes) - * @param string which parser to use for turning raw HTML into a DOMDocument (either 'libxml' or 'html5lib') - */ - function __construct($html, $url=null, $parser='libxml') - { - $this->url = $url; - /* Turn all double br's into p's */ - $html = preg_replace($this->regexps['replaceBrs'], '

', $html); - $html = preg_replace($this->regexps['replaceFonts'], '<$1span>', $html); - $html = mb_convert_encoding($html, 'HTML-ENTITIES', "UTF-8"); - if (trim($html) == '') $html = ''; - if ($parser=='html5lib' && ($this->dom = HTML5_Parser::parse($html))) { - // all good - } else { - $this->dom = new DOMDocument(); - $this->dom->preserveWhiteSpace = false; - @$this->dom->loadHTML($html); - } - $this->dom->registerNodeClass('DOMElement', 'JSLikeHTMLElement'); - } - - /** - * Get article title element - * @return DOMElement - */ - public function getTitle() { - return $this->articleTitle; - } - - /** - * Get article content element - * @return DOMElement - */ - public function getContent() { - return $this->articleContent; - } - - /** - * Runs readability. - * - * Workflow: - * 1. Prep the document by removing script tags, css, etc. - * 2. Build readability's DOM tree. - * 3. Grab the article content from the current dom tree. - * 4. Replace the current DOM tree with the new one. - * 5. Read peacefully. - * - * @return boolean true if we found content, false otherwise - **/ - public function init() - { - if (!isset($this->dom->documentElement)) return false; - $this->removeScripts($this->dom); - //die($this->getInnerHTML($this->dom->documentElement)); - - // Assume successful outcome - $this->success = true; - - $bodyElems = $this->dom->getElementsByTagName('body'); - if ($bodyElems->length > 0) { - if ($this->bodyCache == null) { - $this->bodyCache = $bodyElems->item(0)->innerHTML; - } - if ($this->body == null) { - $this->body = $bodyElems->item(0); - } - } - - $this->prepDocument(); - - //die($this->dom->documentElement->parentNode->nodeType); - //$this->setInnerHTML($this->dom->documentElement, $this->getInnerHTML($this->dom->documentElement)); - //die($this->getInnerHTML($this->dom->documentElement)); - - /* Build readability's DOM tree */ - $overlay = $this->dom->createElement('div'); - $innerDiv = $this->dom->createElement('div'); - $articleTitle = $this->getArticleTitle(); - $articleContent = $this->grabArticle(); - - if (!$articleContent) { - $this->success = false; - $articleContent = $this->dom->createElement('div'); - $articleContent->setAttribute('id', 'readability-content'); - $articleContent->innerHTML = '

Sorry, Readability was unable to parse this page for content.

'; - } - - $overlay->setAttribute('id', 'readOverlay'); - $innerDiv->setAttribute('id', 'readInner'); - - /* Glue the structure of our document together. */ - $innerDiv->appendChild($articleTitle); - $innerDiv->appendChild($articleContent); - $overlay->appendChild($innerDiv); - - /* Clear the old HTML, insert the new content. */ - $this->body->innerHTML = ''; - $this->body->appendChild($overlay); - //document.body.insertBefore(overlay, document.body.firstChild); - $this->body->removeAttribute('style'); - - $this->postProcessContent($articleContent); - - // Set title and content instance variables - $this->articleTitle = $articleTitle; - $this->articleContent = $articleContent; - - return $this->success; - } - - /** - * Debug - */ - protected function dbg($msg) { - if ($this->debug) echo '* ',$msg, "\n"; - } - - /** - * Run any post-process modifications to article content as necessary. - * - * @param DOMElement - * @return void - */ - public function postProcessContent($articleContent) { - if ($this->convertLinksToFootnotes && !preg_match('/wikipedia\.org/', @$this->url)) { - $this->addFootnotes($articleContent); - } - } - - /** - * Get the article title as an H1. - * - * @return DOMElement - */ - protected function getArticleTitle() { - $curTitle = ''; - $origTitle = ''; - - try { - $curTitle = $origTitle = $this->getInnerText($this->dom->getElementsByTagName('title')->item(0)); - } catch(Exception $e) {} - - if (preg_match('/ [\|\-] /', $curTitle)) - { - $curTitle = preg_replace('/(.*)[\|\-] .*/i', '$1', $origTitle); - - if (count(explode(' ', $curTitle)) < 3) { - $curTitle = preg_replace('/[^\|\-]*[\|\-](.*)/i', '$1', $origTitle); - } - } - else if (strpos($curTitle, ': ') !== false) - { - $curTitle = preg_replace('/.*:(.*)/i', '$1', $origTitle); - - if (count(explode(' ', $curTitle)) < 3) { - $curTitle = preg_replace('/[^:]*[:](.*)/i','$1', $origTitle); - } - } - else if(strlen($curTitle) > 150 || strlen($curTitle) < 15) - { - $hOnes = $this->dom->getElementsByTagName('h1'); - if($hOnes->length == 1) - { - $curTitle = $this->getInnerText($hOnes->item(0)); - } - } - - $curTitle = trim($curTitle); - - if (count(explode(' ', $curTitle)) <= 4) { - $curTitle = $origTitle; - } - - $articleTitle = $this->dom->createElement('h1'); - $articleTitle->innerHTML = $curTitle; - - return $articleTitle; - } - - /** - * Prepare the HTML document for readability to scrape it. - * This includes things like stripping javascript, CSS, and handling terrible markup. - * - * @return void - **/ - protected function prepDocument() { - /** - * In some cases a body element can't be found (if the HTML is totally hosed for example) - * so we create a new body node and append it to the document. - */ - if ($this->body == null) - { - $this->body = $this->dom->createElement('body'); - $this->dom->documentElement->appendChild($this->body); - } - $this->body->setAttribute('id', 'readabilityBody'); - - /* Remove all style tags in head */ - $styleTags = $this->dom->getElementsByTagName('style'); - for ($i = $styleTags->length-1; $i >= 0; $i--) - { - $styleTags->item($i)->parentNode->removeChild($styleTags->item($i)); - } - - /* Turn all double br's into p's */ - /* Note, this is pretty costly as far as processing goes. Maybe optimize later. */ - //document.body.innerHTML = document.body.innerHTML.replace(readability.regexps.replaceBrs, '

').replace(readability.regexps.replaceFonts, '<$1span>'); - // We do this in the constructor for PHP as that's when we have raw HTML - before parsing it into a DOM tree. - // Manipulating innerHTML as it's done in JS is not possible in PHP. - } - - /** - * For easier reading, convert this document to have footnotes at the bottom rather than inline links. - * @see http://www.roughtype.com/archives/2010/05/experiments_in.php - * - * @return void - **/ - public function addFootnotes($articleContent) { - $footnotesWrapper = $this->dom->createElement('div'); - $footnotesWrapper->setAttribute('id', 'readability-footnotes'); - $footnotesWrapper->innerHTML = '

References

'; - - $articleFootnotes = $this->dom->createElement('ol'); - $articleFootnotes->setAttribute('id', 'readability-footnotes-list'); - $footnotesWrapper->appendChild($articleFootnotes); - - $articleLinks = $articleContent->getElementsByTagName('a'); - - $linkCount = 0; - for ($i = 0; $i < $articleLinks->length; $i++) - { - $articleLink = $articleLinks->item($i); - $footnoteLink = $articleLink->cloneNode(true); - $refLink = $this->dom->createElement('a'); - $footnote = $this->dom->createElement('li'); - $linkDomain = @parse_url($footnoteLink->getAttribute('href'), PHP_URL_HOST); - if (!$linkDomain && isset($this->url)) $linkDomain = @parse_url($this->url, PHP_URL_HOST); - //linkDomain = footnoteLink.host ? footnoteLink.host : document.location.host, - $linkText = $this->getInnerText($articleLink); - - if ((strpos($articleLink->getAttribute('class'), 'readability-DoNotFootnote') !== false) || preg_match($this->regexps['skipFootnoteLink'], $linkText)) { - continue; - } - - $linkCount++; - - /** Add a superscript reference after the article link */ - $refLink->setAttribute('href', '#readabilityFootnoteLink-' . $linkCount); - $refLink->innerHTML = '[' . $linkCount . ']'; - $refLink->setAttribute('class', 'readability-DoNotFootnote'); - $refLink->setAttribute('style', 'color: inherit;'); - - //TODO: does this work or should we use DOMNode.isSameNode()? - if ($articleLink->parentNode->lastChild == $articleLink) { - $articleLink->parentNode->appendChild($refLink); - } else { - $articleLink->parentNode->insertBefore($refLink, $articleLink->nextSibling); - } - - $articleLink->setAttribute('style', 'color: inherit; text-decoration: none;'); - $articleLink->setAttribute('name', 'readabilityLink-' . $linkCount); - - $footnote->innerHTML = '^ '; - - $footnoteLink->innerHTML = ($footnoteLink->getAttribute('title') != '' ? $footnoteLink->getAttribute('title') : $linkText); - $footnoteLink->setAttribute('name', 'readabilityFootnoteLink-' . $linkCount); - - $footnote->appendChild($footnoteLink); - if ($linkDomain) $footnote->innerHTML = $footnote->innerHTML . ' (' . $linkDomain . ')'; - - $articleFootnotes->appendChild($footnote); - } - - if ($linkCount > 0) { - $articleContent->appendChild($footnotesWrapper); - } - } - - /** - * Reverts P elements with class 'readability-styled' - * to text nodes - which is what they were before. - * - * @param DOMElement - * @return void - */ - function revertReadabilityStyledElements($articleContent) { - $xpath = new DOMXPath($articleContent->ownerDocument); - $elems = $xpath->query('.//p[@class="readability-styled"]', $articleContent); - //$elems = $articleContent->getElementsByTagName('p'); - for ($i = $elems->length-1; $i >= 0; $i--) { - $e = $elems->item($i); - $e->parentNode->replaceChild($articleContent->ownerDocument->createTextNode($e->textContent), $e); - //if ($e->hasAttribute('class') && $e->getAttribute('class') == 'readability-styled') { - // $e->parentNode->replaceChild($this->dom->createTextNode($e->textContent), $e); - //} - } - } - - /** - * Prepare the article node for display. Clean out any inline styles, - * iframes, forms, strip extraneous

tags, etc. - * - * @param DOMElement - * @return void - */ - function prepArticle($articleContent) { - $this->cleanStyles($articleContent); - $this->killBreaks($articleContent); - if ($this->revertForcedParagraphElements) { - $this->revertReadabilityStyledElements($articleContent); - } - - /* Clean out junk from the article content */ - $this->cleanConditionally($articleContent, 'form'); - $this->clean($articleContent, 'object'); - $this->clean($articleContent, 'h1'); - - /** - * If there is only one h2, they are probably using it - * as a header and not a subheader, so remove it since we already have a header. - ***/ - if (!$this->lightClean && ($articleContent->getElementsByTagName('h2')->length == 1)) { - $this->clean($articleContent, 'h2'); - } - $this->clean($articleContent, 'iframe'); - - $this->cleanHeaders($articleContent); - - /* Do these last as the previous stuff may have removed junk that will affect these */ - $this->cleanConditionally($articleContent, 'table'); - $this->cleanConditionally($articleContent, 'ul'); - $this->cleanConditionally($articleContent, 'div'); - - /* Remove extra paragraphs */ - $articleParagraphs = $articleContent->getElementsByTagName('p'); - for ($i = $articleParagraphs->length-1; $i >= 0; $i--) - { - $imgCount = $articleParagraphs->item($i)->getElementsByTagName('img')->length; - $embedCount = $articleParagraphs->item($i)->getElementsByTagName('embed')->length; - $objectCount = $articleParagraphs->item($i)->getElementsByTagName('object')->length; - $iframeCount = $articleParagraphs->item($i)->getElementsByTagName('iframe')->length; - - if ($imgCount === 0 && $embedCount === 0 && $objectCount === 0 && $iframeCount === 0 && $this->getInnerText($articleParagraphs->item($i), false) == '') - { - $articleParagraphs->item($i)->parentNode->removeChild($articleParagraphs->item($i)); - } - } - - try { - $articleContent->innerHTML = preg_replace('/]*>\s*

innerHTML); - //articleContent.innerHTML = articleContent.innerHTML.replace(/]*>\s*

dbg("Cleaning innerHTML of breaks failed. This is an IE strict-block-elements bug. Ignoring.: " . $e); - } - } - - /** - * Initialize a node with the readability object. Also checks the - * className/id for special names to add to its score. - * - * @param Element - * @return void - **/ - protected function initializeNode($node) { - $readability = $this->dom->createAttribute('readability'); - $readability->value = 0; // this is our contentScore - $node->setAttributeNode($readability); - - switch (strtoupper($node->tagName)) { // unsure if strtoupper is needed, but using it just in case - case 'DIV': - $readability->value += 5; - break; - - case 'PRE': - case 'TD': - case 'BLOCKQUOTE': - $readability->value += 3; - break; - - case 'ADDRESS': - case 'OL': - case 'UL': - case 'DL': - case 'DD': - case 'DT': - case 'LI': - case 'FORM': - $readability->value -= 3; - break; - - case 'H1': - case 'H2': - case 'H3': - case 'H4': - case 'H5': - case 'H6': - case 'TH': - $readability->value -= 5; - break; - } - $readability->value += $this->getClassWeight($node); - } - - /*** - * grabArticle - Using a variety of metrics (content score, classname, element types), find the content that is - * most likely to be the stuff a user wants to read. Then return it wrapped up in a div. - * - * @return DOMElement - **/ - protected function grabArticle($page=null) { - $stripUnlikelyCandidates = $this->flagIsActive(self::FLAG_STRIP_UNLIKELYS); - if (!$page) $page = $this->dom; - $allElements = $page->getElementsByTagName('*'); - /** - * First, node prepping. Trash nodes that look cruddy (like ones with the class name "comment", etc), and turn divs - * into P tags where they have been used inappropriately (as in, where they contain no other block level elements.) - * - * Note: Assignment from index for performance. See http://www.peachpit.com/articles/article.aspx?p=31567&seqNum=5 - * TODO: Shouldn't this be a reverse traversal? - **/ - $node = null; - $nodesToScore = array(); - for ($nodeIndex = 0; ($node = $allElements->item($nodeIndex)); $nodeIndex++) { - //for ($nodeIndex=$targetList->length-1; $nodeIndex >= 0; $nodeIndex--) { - //$node = $targetList->item($nodeIndex); - $tagName = strtoupper($node->tagName); - /* Remove unlikely candidates */ - if ($stripUnlikelyCandidates) { - $unlikelyMatchString = $node->getAttribute('class') . $node->getAttribute('id'); - if ( - preg_match($this->regexps['unlikelyCandidates'], $unlikelyMatchString) && - !preg_match($this->regexps['okMaybeItsACandidate'], $unlikelyMatchString) && - $tagName != 'BODY' - ) - { - $this->dbg('Removing unlikely candidate - ' . $unlikelyMatchString); - //$nodesToRemove[] = $node; - $node->parentNode->removeChild($node); - $nodeIndex--; - continue; - } - } - - if ($tagName == 'P' || $tagName == 'TD' || $tagName == 'PRE') { - $nodesToScore[] = $node; - } - - /* Turn all divs that don't have children block level elements into p's */ - if ($tagName == 'DIV') { - if (!preg_match($this->regexps['divToPElements'], $node->innerHTML)) { - //$this->dbg('Altering div to p'); - $newNode = $this->dom->createElement('p'); - try { - $newNode->innerHTML = $node->innerHTML; - //$nodesToReplace[] = array('new'=>$newNode, 'old'=>$node); - $node->parentNode->replaceChild($newNode, $node); - $nodeIndex--; - $nodesToScore[] = $node; // or $newNode? - } - catch(Exception $e) { - $this->dbg('Could not alter div to p, reverting back to div.: ' . $e); - } - } - else - { - /* EXPERIMENTAL */ - // TODO: change these p elements back to text nodes after processing - for ($i = 0, $il = $node->childNodes->length; $i < $il; $i++) { - $childNode = $node->childNodes->item($i); - if ($childNode->nodeType == 3) { // XML_TEXT_NODE - //$this->dbg('replacing text node with a p tag with the same content.'); - $p = $this->dom->createElement('p'); - $p->innerHTML = $childNode->nodeValue; - $p->setAttribute('style', 'display: inline;'); - $p->setAttribute('class', 'readability-styled'); - $childNode->parentNode->replaceChild($p, $childNode); - } - } - } - } - } - - /** - * Loop through all paragraphs, and assign a score to them based on how content-y they look. - * Then add their score to their parent node. - * - * A score is determined by things like number of commas, class names, etc. Maybe eventually link density. - **/ - $candidates = array(); - for ($pt=0; $pt < count($nodesToScore); $pt++) { - $parentNode = $nodesToScore[$pt]->parentNode; - // $grandParentNode = $parentNode ? $parentNode->parentNode : null; - $grandParentNode = !$parentNode ? null : (($parentNode->parentNode instanceof DOMElement) ? $parentNode->parentNode : null); - $innerText = $this->getInnerText($nodesToScore[$pt]); - - if (!$parentNode || !isset($parentNode->tagName)) { - continue; - } - - /* If this paragraph is less than 25 characters, don't even count it. */ - if(strlen($innerText) < 25) { - continue; - } - - /* Initialize readability data for the parent. */ - if (!$parentNode->hasAttribute('readability')) - { - $this->initializeNode($parentNode); - $candidates[] = $parentNode; - } - - /* Initialize readability data for the grandparent. */ - if ($grandParentNode && !$grandParentNode->hasAttribute('readability') && isset($grandParentNode->tagName)) - { - $this->initializeNode($grandParentNode); - $candidates[] = $grandParentNode; - } - - $contentScore = 0; - - /* Add a point for the paragraph itself as a base. */ - $contentScore++; - - /* Add points for any commas within this paragraph */ - $contentScore += count(explode(',', $innerText)); - - /* For every 100 characters in this paragraph, add another point. Up to 3 points. */ - $contentScore += min(floor(strlen($innerText) / 100), 3); - - /* Add the score to the parent. The grandparent gets half. */ - $parentNode->getAttributeNode('readability')->value += $contentScore; - - if ($grandParentNode) { - $grandParentNode->getAttributeNode('readability')->value += $contentScore/2; - } - } - - /** - * After we've calculated scores, loop through all of the possible candidate nodes we found - * and find the one with the highest score. - **/ - $topCandidate = null; - for ($c=0, $cl=count($candidates); $c < $cl; $c++) - { - /** - * Scale the final candidates score based on link density. Good content should have a - * relatively small link density (5% or less) and be mostly unaffected by this operation. - **/ - $readability = $candidates[$c]->getAttributeNode('readability'); - $readability->value = $readability->value * (1-$this->getLinkDensity($candidates[$c])); - - $this->dbg('Candidate: ' . $candidates[$c]->tagName . ' (' . $candidates[$c]->getAttribute('class') . ':' . $candidates[$c]->getAttribute('id') . ') with score ' . $readability->value); - - if (!$topCandidate || $readability->value > (int)$topCandidate->getAttribute('readability')) { - $topCandidate = $candidates[$c]; - } - } - - /** - * If we still have no top candidate, just use the body as a last resort. - * We also have to copy the body node so it is something we can modify. - **/ - if ($topCandidate === null || strtoupper($topCandidate->tagName) == 'BODY') - { - $topCandidate = $this->dom->createElement('div'); - if ($page instanceof DOMDocument) { - if (!isset($page->documentElement)) { - // we don't have a body either? what a mess! :) - } else { - $topCandidate->innerHTML = $page->documentElement->innerHTML; - $page->documentElement->innerHTML = ''; - $page->documentElement->appendChild($topCandidate); - } - } else { - $topCandidate->innerHTML = $page->innerHTML; - $page->innerHTML = ''; - $page->appendChild($topCandidate); - } - $this->initializeNode($topCandidate); - } - - /** - * Now that we have the top candidate, look through its siblings for content that might also be related. - * Things like preambles, content split by ads that we removed, etc. - **/ - $articleContent = $this->dom->createElement('div'); - $articleContent->setAttribute('id', 'readability-content'); - $siblingScoreThreshold = max(10, ((int)$topCandidate->getAttribute('readability')) * 0.2); - $siblingNodes = $topCandidate->parentNode->childNodes; - if (!isset($siblingNodes)) { - $siblingNodes = new stdClass; - $siblingNodes->length = 0; - } - - for ($s=0, $sl=$siblingNodes->length; $s < $sl; $s++) - { - $siblingNode = $siblingNodes->item($s); - $append = false; - - $this->dbg('Looking at sibling node: ' . $siblingNode->nodeName . (($siblingNode->nodeType === XML_ELEMENT_NODE && $siblingNode->hasAttribute('readability')) ? (' with score ' . $siblingNode->getAttribute('readability')) : '')); - - //dbg('Sibling has score ' . ($siblingNode->readability ? siblingNode.readability.contentScore : 'Unknown')); - - if ($siblingNode === $topCandidate) - // or if ($siblingNode->isSameNode($topCandidate)) - { - $append = true; - } - - $contentBonus = 0; - /* Give a bonus if sibling nodes and top candidates have the example same classname */ - if ($siblingNode->nodeType === XML_ELEMENT_NODE && $siblingNode->getAttribute('class') == $topCandidate->getAttribute('class') && $topCandidate->getAttribute('class') != '') { - $contentBonus += ((int)$topCandidate->getAttribute('readability')) * 0.2; - } - - if ($siblingNode->nodeType === XML_ELEMENT_NODE && $siblingNode->hasAttribute('readability') && (((int)$siblingNode->getAttribute('readability')) + $contentBonus) >= $siblingScoreThreshold) - { - $append = true; - } - - if (strtoupper($siblingNode->nodeName) == 'P') { - $linkDensity = $this->getLinkDensity($siblingNode); - $nodeContent = $this->getInnerText($siblingNode); - $nodeLength = strlen($nodeContent); - - if ($nodeLength > 80 && $linkDensity < 0.25) - { - $append = true; - } - else if ($nodeLength < 80 && $linkDensity === 0 && preg_match('/\.( |$)/', $nodeContent)) - { - $append = true; - } - } - - if ($append) - { - $this->dbg('Appending node: ' . $siblingNode->nodeName); - - $nodeToAppend = null; - $sibNodeName = strtoupper($siblingNode->nodeName); - if ($sibNodeName != 'DIV' && $sibNodeName != 'P') { - /* We have a node that isn't a common block level element, like a form or td tag. Turn it into a div so it doesn't get filtered out later by accident. */ - - $this->dbg('Altering siblingNode of ' . $sibNodeName . ' to div.'); - $nodeToAppend = $this->dom->createElement('div'); - try { - $nodeToAppend->setAttribute('id', $siblingNode->getAttribute('id')); - $nodeToAppend->innerHTML = $siblingNode->innerHTML; - } - catch(Exception $e) - { - $this->dbg('Could not alter siblingNode to div, reverting back to original.'); - $nodeToAppend = $siblingNode; - $s--; - $sl--; - } - } else { - $nodeToAppend = $siblingNode; - $s--; - $sl--; - } - - /* To ensure a node does not interfere with readability styles, remove its classnames */ - $nodeToAppend->removeAttribute('class'); - - /* Append sibling and subtract from our list because it removes the node when you append to another node */ - $articleContent->appendChild($nodeToAppend); - } - } - - /** - * So we have all of the content that we need. Now we clean it up for presentation. - **/ - $this->prepArticle($articleContent); - - /** - * Now that we've gone through the full algorithm, check to see if we got any meaningful content. - * If we didn't, we may need to re-run grabArticle with different flags set. This gives us a higher - * likelihood of finding the content, and the sieve approach gives us a higher likelihood of - * finding the -right- content. - **/ - if (strlen($this->getInnerText($articleContent, false)) < 250) - { - // TODO: find out why element disappears sometimes, e.g. for this URL http://www.businessinsider.com/6-hedge-fund-etfs-for-average-investors-2011-7 - // in the meantime, we check and create an empty element if it's not there. - if (!isset($this->body->childNodes)) $this->body = $this->dom->createElement('body'); - $this->body->innerHTML = $this->bodyCache; - - if ($this->flagIsActive(self::FLAG_STRIP_UNLIKELYS)) { - $this->removeFlag(self::FLAG_STRIP_UNLIKELYS); - return $this->grabArticle($this->body); - } - else if ($this->flagIsActive(self::FLAG_WEIGHT_CLASSES)) { - $this->removeFlag(self::FLAG_WEIGHT_CLASSES); - return $this->grabArticle($this->body); - } - else if ($this->flagIsActive(self::FLAG_CLEAN_CONDITIONALLY)) { - $this->removeFlag(self::FLAG_CLEAN_CONDITIONALLY); - return $this->grabArticle($this->body); - } - else { - return false; - } - } - return $articleContent; - } - - /** - * Remove script tags from document - * - * @param DOMElement - * @return void - */ - public function removeScripts($doc) { - $scripts = $doc->getElementsByTagName('script'); - for($i = $scripts->length-1; $i >= 0; $i--) - { - $scripts->item($i)->parentNode->removeChild($scripts->item($i)); - } - } - - /** - * Get the inner text of a node. - * This also strips out any excess whitespace to be found. - * - * @param DOMElement $ - * @param boolean $normalizeSpaces (default: true) - * @return string - **/ - public function getInnerText($e, $normalizeSpaces=true) { - $textContent = ''; - - if (!isset($e->textContent) || $e->textContent == '') { - return ''; - } - - $textContent = trim($e->textContent); - - if ($normalizeSpaces) { - return preg_replace($this->regexps['normalize'], ' ', $textContent); - } else { - return $textContent; - } - } - - /** - * Get the number of times a string $s appears in the node $e. - * - * @param DOMElement $e - * @param string - what to count. Default is "," - * @return number (integer) - **/ - public function getCharCount($e, $s=',') { - return substr_count($this->getInnerText($e), $s); - } - - /** - * Remove the style attribute on every $e and under. - * - * @param DOMElement $e - * @return void - */ - public function cleanStyles($e) { - if (!is_object($e)) return; - $elems = $e->getElementsByTagName('*'); - foreach ($elems as $elem) { - $elem->removeAttribute('style'); - } - } - - /** - * Get the density of links as a percentage of the content - * This is the amount of text that is inside a link divided by the total text in the node. - * - * @param DOMElement $e - * @return number (float) - */ - public function getLinkDensity($e) { - $links = $e->getElementsByTagName('a'); - $textLength = strlen($this->getInnerText($e)); - $linkLength = 0; - for ($i=0, $il=$links->length; $i < $il; $i++) - { - $linkLength += strlen($this->getInnerText($links->item($i))); - } - if ($textLength > 0) { - return $linkLength / $textLength; - } else { - return 0; - } - } - - /** - * Get an elements class/id weight. Uses regular expressions to tell if this - * element looks good or bad. - * - * @param DOMElement $e - * @return number (Integer) - */ - public function getClassWeight($e) { - if(!$this->flagIsActive(self::FLAG_WEIGHT_CLASSES)) { - return 0; - } - - $weight = 0; - - /* Look for a special classname */ - if ($e->hasAttribute('class') && $e->getAttribute('class') != '') - { - if (preg_match($this->regexps['negative'], $e->getAttribute('class'))) { - $weight -= 25; - } - if (preg_match($this->regexps['positive'], $e->getAttribute('class'))) { - $weight += 25; - } - } - - /* Look for a special ID */ - if ($e->hasAttribute('id') && $e->getAttribute('id') != '') - { - if (preg_match($this->regexps['negative'], $e->getAttribute('id'))) { - $weight -= 25; - } - if (preg_match($this->regexps['positive'], $e->getAttribute('id'))) { - $weight += 25; - } - } - return $weight; - } - - /** - * Remove extraneous break tags from a node. - * - * @param DOMElement $node - * @return void - */ - public function killBreaks($node) { - $html = $node->innerHTML; - $html = preg_replace($this->regexps['killBreaks'], '
', $html); - $node->innerHTML = $html; - } - - /** - * Clean a node of all elements of type "tag". - * (Unless it's a youtube/vimeo video. People love movies.) - * - * Updated 2012-09-18 to preserve youtube/vimeo iframes - * - * @param DOMElement $e - * @param string $tag - * @return void - */ - public function clean($e, $tag) { - $targetList = $e->getElementsByTagName($tag); - $isEmbed = ($tag == 'iframe' || $tag == 'object' || $tag == 'embed'); - - for ($y=$targetList->length-1; $y >= 0; $y--) { - /* Allow youtube and vimeo videos through as people usually want to see those. */ - if ($isEmbed) { - $attributeValues = ''; - for ($i=0, $il=$targetList->item($y)->attributes->length; $i < $il; $i++) { - $attributeValues .= $targetList->item($y)->attributes->item($i)->value . '|'; // DOMAttr? (TODO: test) - } - - /* First, check the elements attributes to see if any of them contain youtube or vimeo */ - if (preg_match($this->regexps['video'], $attributeValues)) { - continue; - } - - /* Then check the elements inside this element for the same. */ - if (preg_match($this->regexps['video'], $targetList->item($y)->innerHTML)) { - continue; - } - } - $targetList->item($y)->parentNode->removeChild($targetList->item($y)); - } - } - - /** - * Clean an element of all tags of type "tag" if they look fishy. - * "Fishy" is an algorithm based on content length, classnames, - * link density, number of images & embeds, etc. - * - * @param DOMElement $e - * @param string $tag - * @return void - */ - public function cleanConditionally($e, $tag) { - if (!$this->flagIsActive(self::FLAG_CLEAN_CONDITIONALLY)) { - return; - } - - $tagsList = $e->getElementsByTagName($tag); - $curTagsLength = $tagsList->length; - - /** - * Gather counts for other typical elements embedded within. - * Traverse backwards so we can remove nodes at the same time without effecting the traversal. - * - * TODO: Consider taking into account original contentScore here. - */ - for ($i=$curTagsLength-1; $i >= 0; $i--) { - $weight = $this->getClassWeight($tagsList->item($i)); - $contentScore = ($tagsList->item($i)->hasAttribute('readability')) ? (int)$tagsList->item($i)->getAttribute('readability') : 0; - - $this->dbg('Cleaning Conditionally ' . $tagsList->item($i)->tagName . ' (' . $tagsList->item($i)->getAttribute('class') . ':' . $tagsList->item($i)->getAttribute('id') . ')' . (($tagsList->item($i)->hasAttribute('readability')) ? (' with score ' . $tagsList->item($i)->getAttribute('readability')) : '')); - - if ($weight + $contentScore < 0) { - $tagsList->item($i)->parentNode->removeChild($tagsList->item($i)); - } - else if ( $this->getCharCount($tagsList->item($i), ',') < 10) { - /** - * If there are not very many commas, and the number of - * non-paragraph elements is more than paragraphs or other ominous signs, remove the element. - **/ - $p = $tagsList->item($i)->getElementsByTagName('p')->length; - $img = $tagsList->item($i)->getElementsByTagName('img')->length; - $li = $tagsList->item($i)->getElementsByTagName('li')->length-100; - $input = $tagsList->item($i)->getElementsByTagName('input')->length; - $a = $tagsList->item($i)->getElementsByTagName('a')->length; - - $embedCount = 0; - $embeds = $tagsList->item($i)->getElementsByTagName('embed'); - for ($ei=0, $il=$embeds->length; $ei < $il; $ei++) { - if (preg_match($this->regexps['video'], $embeds->item($ei)->getAttribute('src'))) { - $embedCount++; - } - } - $embeds = $tagsList->item($i)->getElementsByTagName('iframe'); - for ($ei=0, $il=$embeds->length; $ei < $il; $ei++) { - if (preg_match($this->regexps['video'], $embeds->item($ei)->getAttribute('src'))) { - $embedCount++; - } - } - - $linkDensity = $this->getLinkDensity($tagsList->item($i)); - $contentLength = strlen($this->getInnerText($tagsList->item($i))); - $toRemove = false; - - if ($this->lightClean) { - $this->dbg('Light clean...'); - if ( ($img > $p) && ($img > 4) ) { - $this->dbg(' more than 4 images and more image elements than paragraph elements'); - $toRemove = true; - } else if ($li > $p && $tag != 'ul' && $tag != 'ol') { - $this->dbg(' too many

  • elements, and parent is not
      or
        '); - $toRemove = true; - } else if ( $input > floor($p/3) ) { - $this->dbg(' too many elements'); - $toRemove = true; - } else if ($contentLength < 25 && ($embedCount === 0 && ($img === 0 || $img > 2))) { - $this->dbg(' content length less than 25 chars, 0 embeds and either 0 images or more than 2 images'); - $toRemove = true; - } else if($weight < 25 && $linkDensity > 0.2) { - $this->dbg(' weight smaller than 25 and link density above 0.2'); - $toRemove = true; - } else if($a > 2 && ($weight >= 25 && $linkDensity > 0.5)) { - $this->dbg(' more than 2 links and weight above 25 but link density greater than 0.5'); - $toRemove = true; - } else if($embedCount > 3) { - $this->dbg(' more than 3 embeds'); - $toRemove = true; - } - } else { - $this->dbg('Standard clean...'); - if ( $img > $p ) { - $this->dbg(' more image elements than paragraph elements'); - $toRemove = true; - } else if ($li > $p && $tag != 'ul' && $tag != 'ol') { - $this->dbg(' too many
      1. elements, and parent is not
          or
            '); - $toRemove = true; - } else if ( $input > floor($p/3) ) { - $this->dbg(' too many elements'); - $toRemove = true; - } else if ($contentLength < 25 && ($img === 0 || $img > 2) ) { - $this->dbg(' content length less than 25 chars and 0 images, or more than 2 images'); - $toRemove = true; - } else if($weight < 25 && $linkDensity > 0.2) { - $this->dbg(' weight smaller than 25 and link density above 0.2'); - $toRemove = true; - } else if($weight >= 25 && $linkDensity > 0.5) { - $this->dbg(' weight above 25 but link density greater than 0.5'); - $toRemove = true; - } else if(($embedCount == 1 && $contentLength < 75) || $embedCount > 1) { - $this->dbg(' 1 embed and content length smaller than 75 chars, or more than one embed'); - $toRemove = true; - } - } - - if ($toRemove) { - //$this->dbg('Removing: '.$tagsList->item($i)->innerHTML); - $tagsList->item($i)->parentNode->removeChild($tagsList->item($i)); - } - } - } - } - - /** - * Clean out spurious headers from an Element. Checks things like classnames and link density. - * - * @param DOMElement $e - * @return void - */ - public function cleanHeaders($e) { - for ($headerIndex = 1; $headerIndex < 3; $headerIndex++) { - $headers = $e->getElementsByTagName('h' . $headerIndex); - for ($i=$headers->length-1; $i >=0; $i--) { - if ($this->getClassWeight($headers->item($i)) < 0 || $this->getLinkDensity($headers->item($i)) > 0.33) { - $headers->item($i)->parentNode->removeChild($headers->item($i)); - } - } - } - } - - public function flagIsActive($flag) { - return ($this->flags & $flag) > 0; - } - - public function addFlag($flag) { - $this->flags = $this->flags | $flag; - } - - public function removeFlag($flag) { - $this->flags = $this->flags & ~$flag; - } -} +init(); +echo $r->articleContent->innerHTML; +*/ + +class Readability +{ + public $version = '1.7.1-without-multi-page'; + public $convertLinksToFootnotes = false; + public $revertForcedParagraphElements = true; + public $articleTitle; + public $articleContent; + public $dom; + public $url = null; // optional - URL where HTML was retrieved + public $debug = false; + public $lightClean = true; // preserves more content (experimental) added 2012-09-19 + protected $body = null; // + protected $bodyCache = null; // Cache the body HTML in case we need to re-use it later + protected $flags = 7; // 1 | 2 | 4; // Start with all flags set. + protected $success = false; // indicates whether we were able to extract or not + + /** + * All of the regular expressions in use within readability. + * Defined up here so we don't instantiate them repeatedly in loops. + **/ + public $regexps = array( + 'unlikelyCandidates' => '/combx|comment|community|disqus|extra|foot|header|menu|remark|rss|shoutbox|sidebar|sponsor|ad-break|agegate|pagination|pager|popup/i', + 'okMaybeItsACandidate' => '/and|article|body|column|main|shadow/i', + 'positive' => '/article|body|content|entry|hentry|main|page|attachment|pagination|post|text|blog|story/i', + 'negative' => '/combx|comment|com-|contact|foot|footer|_nav|footnote|masthead|media|meta|outbrain|promo|related|scroll|shoutbox|sidebar|sponsor|shopping|tags|tool|widget/i', + 'divToPElements' => '/<(a|blockquote|dl|div|img|ol|p|pre|table|ul)/i', + 'replaceBrs' => '/(]*>[ \n\r\t]*){2,}/i', + 'replaceFonts' => '/<(\/?)font[^>]*>/i', + // 'trimRe' => '/^\s+|\s+$/g', // PHP has trim() + 'normalize' => '/\s{2,}/', + 'killBreaks' => '/((\s| ?)*){1,}/', + 'video' => '!//(player\.|www\.)?(youtube|vimeo|viddler)\.com!i', + 'skipFootnoteLink' => '/^\s*(\[?[a-z0-9]{1,2}\]?|^|edit|citation needed)\s*$/i' + ); + + /* constants */ + const FLAG_STRIP_UNLIKELYS = 1; + const FLAG_WEIGHT_CLASSES = 2; + const FLAG_CLEAN_CONDITIONALLY = 4; + + /** + * Create instance of Readability + * @param string UTF-8 encoded string + * @param string (optional) URL associated with HTML (used for footnotes) + * @param string which parser to use for turning raw HTML into a DOMDocument (either 'libxml' or 'html5lib') + */ + function __construct($html, $url=null, $parser='libxml') + { + $this->url = $url; + /* Turn all double br's into p's */ + $html = preg_replace($this->regexps['replaceBrs'], '

            ', $html); + $html = preg_replace($this->regexps['replaceFonts'], '<$1span>', $html); + $html = mb_convert_encoding($html, 'HTML-ENTITIES', "UTF-8"); + if (trim($html) == '') $html = ''; + if ($parser=='html5lib' && ($this->dom = HTML5_Parser::parse($html))) { + // all good + } else { + $this->dom = new DOMDocument(); + $this->dom->preserveWhiteSpace = false; + @$this->dom->loadHTML($html); + } + $this->dom->registerNodeClass('DOMElement', 'JSLikeHTMLElement'); + } + + /** + * Get article title element + * @return DOMElement + */ + public function getTitle() { + return $this->articleTitle; + } + + /** + * Get article content element + * @return DOMElement + */ + public function getContent() { + return $this->articleContent; + } + + /** + * Runs readability. + * + * Workflow: + * 1. Prep the document by removing script tags, css, etc. + * 2. Build readability's DOM tree. + * 3. Grab the article content from the current dom tree. + * 4. Replace the current DOM tree with the new one. + * 5. Read peacefully. + * + * @return boolean true if we found content, false otherwise + **/ + public function init() + { + if (!isset($this->dom->documentElement)) return false; + $this->removeScripts($this->dom); + //die($this->getInnerHTML($this->dom->documentElement)); + + // Assume successful outcome + $this->success = true; + + $bodyElems = $this->dom->getElementsByTagName('body'); + if ($bodyElems->length > 0) { + if ($this->bodyCache == null) { + $this->bodyCache = $bodyElems->item(0)->innerHTML; + } + if ($this->body == null) { + $this->body = $bodyElems->item(0); + } + } + + $this->prepDocument(); + + //die($this->dom->documentElement->parentNode->nodeType); + //$this->setInnerHTML($this->dom->documentElement, $this->getInnerHTML($this->dom->documentElement)); + //die($this->getInnerHTML($this->dom->documentElement)); + + /* Build readability's DOM tree */ + $overlay = $this->dom->createElement('div'); + $innerDiv = $this->dom->createElement('div'); + $articleTitle = $this->getArticleTitle(); + $articleContent = $this->grabArticle(); + + if (!$articleContent) { + $this->success = false; + $articleContent = $this->dom->createElement('div'); + $articleContent->setAttribute('id', 'readability-content'); + $articleContent->innerHTML = '

            Sorry, Readability was unable to parse this page for content.

            '; + } + + $overlay->setAttribute('id', 'readOverlay'); + $innerDiv->setAttribute('id', 'readInner'); + + /* Glue the structure of our document together. */ + $innerDiv->appendChild($articleTitle); + $innerDiv->appendChild($articleContent); + $overlay->appendChild($innerDiv); + + /* Clear the old HTML, insert the new content. */ + $this->body->innerHTML = ''; + $this->body->appendChild($overlay); + //document.body.insertBefore(overlay, document.body.firstChild); + $this->body->removeAttribute('style'); + + $this->postProcessContent($articleContent); + + // Set title and content instance variables + $this->articleTitle = $articleTitle; + $this->articleContent = $articleContent; + + return $this->success; + } + + /** + * Debug + */ + protected function dbg($msg) { + if ($this->debug) echo '* ',$msg, "\n"; + } + + /** + * Run any post-process modifications to article content as necessary. + * + * @param DOMElement + * @return void + */ + public function postProcessContent($articleContent) { + if ($this->convertLinksToFootnotes && !preg_match('/wikipedia\.org/', @$this->url)) { + $this->addFootnotes($articleContent); + } + } + + /** + * Get the article title as an H1. + * + * @return DOMElement + */ + protected function getArticleTitle() { + $curTitle = ''; + $origTitle = ''; + + try { + $curTitle = $origTitle = $this->getInnerText($this->dom->getElementsByTagName('title')->item(0)); + } catch(Exception $e) {} + + if (preg_match('/ [\|\-] /', $curTitle)) + { + $curTitle = preg_replace('/(.*)[\|\-] .*/i', '$1', $origTitle); + + if (count(explode(' ', $curTitle)) < 3) { + $curTitle = preg_replace('/[^\|\-]*[\|\-](.*)/i', '$1', $origTitle); + } + } + else if (strpos($curTitle, ': ') !== false) + { + $curTitle = preg_replace('/.*:(.*)/i', '$1', $origTitle); + + if (count(explode(' ', $curTitle)) < 3) { + $curTitle = preg_replace('/[^:]*[:](.*)/i','$1', $origTitle); + } + } + else if(strlen($curTitle) > 150 || strlen($curTitle) < 15) + { + $hOnes = $this->dom->getElementsByTagName('h1'); + if($hOnes->length == 1) + { + $curTitle = $this->getInnerText($hOnes->item(0)); + } + } + + $curTitle = trim($curTitle); + + if (count(explode(' ', $curTitle)) <= 4) { + $curTitle = $origTitle; + } + + $articleTitle = $this->dom->createElement('h1'); + $articleTitle->innerHTML = $curTitle; + + return $articleTitle; + } + + /** + * Prepare the HTML document for readability to scrape it. + * This includes things like stripping javascript, CSS, and handling terrible markup. + * + * @return void + **/ + protected function prepDocument() { + /** + * In some cases a body element can't be found (if the HTML is totally hosed for example) + * so we create a new body node and append it to the document. + */ + if ($this->body == null) + { + $this->body = $this->dom->createElement('body'); + $this->dom->documentElement->appendChild($this->body); + } + $this->body->setAttribute('id', 'readabilityBody'); + + /* Remove all style tags in head */ + $styleTags = $this->dom->getElementsByTagName('style'); + for ($i = $styleTags->length-1; $i >= 0; $i--) + { + $styleTags->item($i)->parentNode->removeChild($styleTags->item($i)); + } + + /* Turn all double br's into p's */ + /* Note, this is pretty costly as far as processing goes. Maybe optimize later. */ + //document.body.innerHTML = document.body.innerHTML.replace(readability.regexps.replaceBrs, '

            ').replace(readability.regexps.replaceFonts, '<$1span>'); + // We do this in the constructor for PHP as that's when we have raw HTML - before parsing it into a DOM tree. + // Manipulating innerHTML as it's done in JS is not possible in PHP. + } + + /** + * For easier reading, convert this document to have footnotes at the bottom rather than inline links. + * @see http://www.roughtype.com/archives/2010/05/experiments_in.php + * + * @return void + **/ + public function addFootnotes($articleContent) { + $footnotesWrapper = $this->dom->createElement('div'); + $footnotesWrapper->setAttribute('id', 'readability-footnotes'); + $footnotesWrapper->innerHTML = '

            References

            '; + + $articleFootnotes = $this->dom->createElement('ol'); + $articleFootnotes->setAttribute('id', 'readability-footnotes-list'); + $footnotesWrapper->appendChild($articleFootnotes); + + $articleLinks = $articleContent->getElementsByTagName('a'); + + $linkCount = 0; + for ($i = 0; $i < $articleLinks->length; $i++) + { + $articleLink = $articleLinks->item($i); + $footnoteLink = $articleLink->cloneNode(true); + $refLink = $this->dom->createElement('a'); + $footnote = $this->dom->createElement('li'); + $linkDomain = @parse_url($footnoteLink->getAttribute('href'), PHP_URL_HOST); + if (!$linkDomain && isset($this->url)) $linkDomain = @parse_url($this->url, PHP_URL_HOST); + //linkDomain = footnoteLink.host ? footnoteLink.host : document.location.host, + $linkText = $this->getInnerText($articleLink); + + if ((strpos($articleLink->getAttribute('class'), 'readability-DoNotFootnote') !== false) || preg_match($this->regexps['skipFootnoteLink'], $linkText)) { + continue; + } + + $linkCount++; + + /** Add a superscript reference after the article link */ + $refLink->setAttribute('href', '#readabilityFootnoteLink-' . $linkCount); + $refLink->innerHTML = '[' . $linkCount . ']'; + $refLink->setAttribute('class', 'readability-DoNotFootnote'); + $refLink->setAttribute('style', 'color: inherit;'); + + //TODO: does this work or should we use DOMNode.isSameNode()? + if ($articleLink->parentNode->lastChild == $articleLink) { + $articleLink->parentNode->appendChild($refLink); + } else { + $articleLink->parentNode->insertBefore($refLink, $articleLink->nextSibling); + } + + $articleLink->setAttribute('style', 'color: inherit; text-decoration: none;'); + $articleLink->setAttribute('name', 'readabilityLink-' . $linkCount); + + $footnote->innerHTML = '^ '; + + $footnoteLink->innerHTML = ($footnoteLink->getAttribute('title') != '' ? $footnoteLink->getAttribute('title') : $linkText); + $footnoteLink->setAttribute('name', 'readabilityFootnoteLink-' . $linkCount); + + $footnote->appendChild($footnoteLink); + if ($linkDomain) $footnote->innerHTML = $footnote->innerHTML . ' (' . $linkDomain . ')'; + + $articleFootnotes->appendChild($footnote); + } + + if ($linkCount > 0) { + $articleContent->appendChild($footnotesWrapper); + } + } + + /** + * Reverts P elements with class 'readability-styled' + * to text nodes - which is what they were before. + * + * @param DOMElement + * @return void + */ + function revertReadabilityStyledElements($articleContent) { + $xpath = new DOMXPath($articleContent->ownerDocument); + $elems = $xpath->query('.//p[@class="readability-styled"]', $articleContent); + //$elems = $articleContent->getElementsByTagName('p'); + for ($i = $elems->length-1; $i >= 0; $i--) { + $e = $elems->item($i); + $e->parentNode->replaceChild($articleContent->ownerDocument->createTextNode($e->textContent), $e); + //if ($e->hasAttribute('class') && $e->getAttribute('class') == 'readability-styled') { + // $e->parentNode->replaceChild($this->dom->createTextNode($e->textContent), $e); + //} + } + } + + /** + * Prepare the article node for display. Clean out any inline styles, + * iframes, forms, strip extraneous

            tags, etc. + * + * @param DOMElement + * @return void + */ + function prepArticle($articleContent) { + $this->cleanStyles($articleContent); + $this->killBreaks($articleContent); + if ($this->revertForcedParagraphElements) { + $this->revertReadabilityStyledElements($articleContent); + } + + /* Clean out junk from the article content */ + $this->cleanConditionally($articleContent, 'form'); + $this->clean($articleContent, 'object'); + $this->clean($articleContent, 'h1'); + + /** + * If there is only one h2, they are probably using it + * as a header and not a subheader, so remove it since we already have a header. + ***/ + if (!$this->lightClean && ($articleContent->getElementsByTagName('h2')->length == 1)) { + $this->clean($articleContent, 'h2'); + } + $this->clean($articleContent, 'iframe'); + + $this->cleanHeaders($articleContent); + + /* Do these last as the previous stuff may have removed junk that will affect these */ + $this->cleanConditionally($articleContent, 'table'); + $this->cleanConditionally($articleContent, 'ul'); + $this->cleanConditionally($articleContent, 'div'); + + /* Remove extra paragraphs */ + $articleParagraphs = $articleContent->getElementsByTagName('p'); + for ($i = $articleParagraphs->length-1; $i >= 0; $i--) + { + $imgCount = $articleParagraphs->item($i)->getElementsByTagName('img')->length; + $embedCount = $articleParagraphs->item($i)->getElementsByTagName('embed')->length; + $objectCount = $articleParagraphs->item($i)->getElementsByTagName('object')->length; + $iframeCount = $articleParagraphs->item($i)->getElementsByTagName('iframe')->length; + + if ($imgCount === 0 && $embedCount === 0 && $objectCount === 0 && $iframeCount === 0 && $this->getInnerText($articleParagraphs->item($i), false) == '') + { + $articleParagraphs->item($i)->parentNode->removeChild($articleParagraphs->item($i)); + } + } + + try { + $articleContent->innerHTML = preg_replace('/]*>\s*

            innerHTML); + //articleContent.innerHTML = articleContent.innerHTML.replace(/]*>\s*

            dbg("Cleaning innerHTML of breaks failed. This is an IE strict-block-elements bug. Ignoring.: " . $e); + } + } + + /** + * Initialize a node with the readability object. Also checks the + * className/id for special names to add to its score. + * + * @param Element + * @return void + **/ + protected function initializeNode($node) { + $readability = $this->dom->createAttribute('readability'); + $readability->value = 0; // this is our contentScore + $node->setAttributeNode($readability); + + switch (strtoupper($node->tagName)) { // unsure if strtoupper is needed, but using it just in case + case 'DIV': + $readability->value += 5; + break; + + case 'PRE': + case 'TD': + case 'BLOCKQUOTE': + $readability->value += 3; + break; + + case 'ADDRESS': + case 'OL': + case 'UL': + case 'DL': + case 'DD': + case 'DT': + case 'LI': + case 'FORM': + $readability->value -= 3; + break; + + case 'H1': + case 'H2': + case 'H3': + case 'H4': + case 'H5': + case 'H6': + case 'TH': + $readability->value -= 5; + break; + } + $readability->value += $this->getClassWeight($node); + } + + /*** + * grabArticle - Using a variety of metrics (content score, classname, element types), find the content that is + * most likely to be the stuff a user wants to read. Then return it wrapped up in a div. + * + * @return DOMElement + **/ + protected function grabArticle($page=null) { + $stripUnlikelyCandidates = $this->flagIsActive(self::FLAG_STRIP_UNLIKELYS); + if (!$page) $page = $this->dom; + $allElements = $page->getElementsByTagName('*'); + /** + * First, node prepping. Trash nodes that look cruddy (like ones with the class name "comment", etc), and turn divs + * into P tags where they have been used inappropriately (as in, where they contain no other block level elements.) + * + * Note: Assignment from index for performance. See http://www.peachpit.com/articles/article.aspx?p=31567&seqNum=5 + * TODO: Shouldn't this be a reverse traversal? + **/ + $node = null; + $nodesToScore = array(); + for ($nodeIndex = 0; ($node = $allElements->item($nodeIndex)); $nodeIndex++) { + //for ($nodeIndex=$targetList->length-1; $nodeIndex >= 0; $nodeIndex--) { + //$node = $targetList->item($nodeIndex); + $tagName = strtoupper($node->tagName); + /* Remove unlikely candidates */ + if ($stripUnlikelyCandidates) { + $unlikelyMatchString = $node->getAttribute('class') . $node->getAttribute('id'); + if ( + preg_match($this->regexps['unlikelyCandidates'], $unlikelyMatchString) && + !preg_match($this->regexps['okMaybeItsACandidate'], $unlikelyMatchString) && + $tagName != 'BODY' + ) + { + $this->dbg('Removing unlikely candidate - ' . $unlikelyMatchString); + //$nodesToRemove[] = $node; + $node->parentNode->removeChild($node); + $nodeIndex--; + continue; + } + } + + if ($tagName == 'P' || $tagName == 'TD' || $tagName == 'PRE') { + $nodesToScore[] = $node; + } + + /* Turn all divs that don't have children block level elements into p's */ + if ($tagName == 'DIV') { + if (!preg_match($this->regexps['divToPElements'], $node->innerHTML)) { + //$this->dbg('Altering div to p'); + $newNode = $this->dom->createElement('p'); + try { + $newNode->innerHTML = $node->innerHTML; + //$nodesToReplace[] = array('new'=>$newNode, 'old'=>$node); + $node->parentNode->replaceChild($newNode, $node); + $nodeIndex--; + $nodesToScore[] = $node; // or $newNode? + } + catch(Exception $e) { + $this->dbg('Could not alter div to p, reverting back to div.: ' . $e); + } + } + else + { + /* EXPERIMENTAL */ + // TODO: change these p elements back to text nodes after processing + for ($i = 0, $il = $node->childNodes->length; $i < $il; $i++) { + $childNode = $node->childNodes->item($i); + if ($childNode->nodeType == 3) { // XML_TEXT_NODE + //$this->dbg('replacing text node with a p tag with the same content.'); + $p = $this->dom->createElement('p'); + $p->innerHTML = $childNode->nodeValue; + $p->setAttribute('style', 'display: inline;'); + $p->setAttribute('class', 'readability-styled'); + $childNode->parentNode->replaceChild($p, $childNode); + } + } + } + } + } + + /** + * Loop through all paragraphs, and assign a score to them based on how content-y they look. + * Then add their score to their parent node. + * + * A score is determined by things like number of commas, class names, etc. Maybe eventually link density. + **/ + $candidates = array(); + for ($pt=0; $pt < count($nodesToScore); $pt++) { + $parentNode = $nodesToScore[$pt]->parentNode; + // $grandParentNode = $parentNode ? $parentNode->parentNode : null; + $grandParentNode = !$parentNode ? null : (($parentNode->parentNode instanceof DOMElement) ? $parentNode->parentNode : null); + $innerText = $this->getInnerText($nodesToScore[$pt]); + + if (!$parentNode || !isset($parentNode->tagName)) { + continue; + } + + /* If this paragraph is less than 25 characters, don't even count it. */ + if(strlen($innerText) < 25) { + continue; + } + + /* Initialize readability data for the parent. */ + if (!$parentNode->hasAttribute('readability')) + { + $this->initializeNode($parentNode); + $candidates[] = $parentNode; + } + + /* Initialize readability data for the grandparent. */ + if ($grandParentNode && !$grandParentNode->hasAttribute('readability') && isset($grandParentNode->tagName)) + { + $this->initializeNode($grandParentNode); + $candidates[] = $grandParentNode; + } + + $contentScore = 0; + + /* Add a point for the paragraph itself as a base. */ + $contentScore++; + + /* Add points for any commas within this paragraph */ + $contentScore += count(explode(',', $innerText)); + + /* For every 100 characters in this paragraph, add another point. Up to 3 points. */ + $contentScore += min(floor(strlen($innerText) / 100), 3); + + /* Add the score to the parent. The grandparent gets half. */ + $parentNode->getAttributeNode('readability')->value += $contentScore; + + if ($grandParentNode) { + $grandParentNode->getAttributeNode('readability')->value += $contentScore/2; + } + } + + /** + * After we've calculated scores, loop through all of the possible candidate nodes we found + * and find the one with the highest score. + **/ + $topCandidate = null; + for ($c=0, $cl=count($candidates); $c < $cl; $c++) + { + /** + * Scale the final candidates score based on link density. Good content should have a + * relatively small link density (5% or less) and be mostly unaffected by this operation. + **/ + $readability = $candidates[$c]->getAttributeNode('readability'); + $readability->value = $readability->value * (1-$this->getLinkDensity($candidates[$c])); + + $this->dbg('Candidate: ' . $candidates[$c]->tagName . ' (' . $candidates[$c]->getAttribute('class') . ':' . $candidates[$c]->getAttribute('id') . ') with score ' . $readability->value); + + if (!$topCandidate || $readability->value > (int)$topCandidate->getAttribute('readability')) { + $topCandidate = $candidates[$c]; + } + } + + /** + * If we still have no top candidate, just use the body as a last resort. + * We also have to copy the body node so it is something we can modify. + **/ + if ($topCandidate === null || strtoupper($topCandidate->tagName) == 'BODY') + { + $topCandidate = $this->dom->createElement('div'); + if ($page instanceof DOMDocument) { + if (!isset($page->documentElement)) { + // we don't have a body either? what a mess! :) + } else { + $topCandidate->innerHTML = $page->documentElement->innerHTML; + $page->documentElement->innerHTML = ''; + $page->documentElement->appendChild($topCandidate); + } + } else { + $topCandidate->innerHTML = $page->innerHTML; + $page->innerHTML = ''; + $page->appendChild($topCandidate); + } + $this->initializeNode($topCandidate); + } + + /** + * Now that we have the top candidate, look through its siblings for content that might also be related. + * Things like preambles, content split by ads that we removed, etc. + **/ + $articleContent = $this->dom->createElement('div'); + $articleContent->setAttribute('id', 'readability-content'); + $siblingScoreThreshold = max(10, ((int)$topCandidate->getAttribute('readability')) * 0.2); + $siblingNodes = $topCandidate->parentNode->childNodes; + if (!isset($siblingNodes)) { + $siblingNodes = new stdClass; + $siblingNodes->length = 0; + } + + for ($s=0, $sl=$siblingNodes->length; $s < $sl; $s++) + { + $siblingNode = $siblingNodes->item($s); + $append = false; + + $this->dbg('Looking at sibling node: ' . $siblingNode->nodeName . (($siblingNode->nodeType === XML_ELEMENT_NODE && $siblingNode->hasAttribute('readability')) ? (' with score ' . $siblingNode->getAttribute('readability')) : '')); + + //dbg('Sibling has score ' . ($siblingNode->readability ? siblingNode.readability.contentScore : 'Unknown')); + + if ($siblingNode === $topCandidate) + // or if ($siblingNode->isSameNode($topCandidate)) + { + $append = true; + } + + $contentBonus = 0; + /* Give a bonus if sibling nodes and top candidates have the example same classname */ + if ($siblingNode->nodeType === XML_ELEMENT_NODE && $siblingNode->getAttribute('class') == $topCandidate->getAttribute('class') && $topCandidate->getAttribute('class') != '') { + $contentBonus += ((int)$topCandidate->getAttribute('readability')) * 0.2; + } + + if ($siblingNode->nodeType === XML_ELEMENT_NODE && $siblingNode->hasAttribute('readability') && (((int)$siblingNode->getAttribute('readability')) + $contentBonus) >= $siblingScoreThreshold) + { + $append = true; + } + + if (strtoupper($siblingNode->nodeName) == 'P') { + $linkDensity = $this->getLinkDensity($siblingNode); + $nodeContent = $this->getInnerText($siblingNode); + $nodeLength = strlen($nodeContent); + + if ($nodeLength > 80 && $linkDensity < 0.25) + { + $append = true; + } + else if ($nodeLength < 80 && $linkDensity === 0 && preg_match('/\.( |$)/', $nodeContent)) + { + $append = true; + } + } + + if ($append) + { + $this->dbg('Appending node: ' . $siblingNode->nodeName); + + $nodeToAppend = null; + $sibNodeName = strtoupper($siblingNode->nodeName); + if ($sibNodeName != 'DIV' && $sibNodeName != 'P') { + /* We have a node that isn't a common block level element, like a form or td tag. Turn it into a div so it doesn't get filtered out later by accident. */ + + $this->dbg('Altering siblingNode of ' . $sibNodeName . ' to div.'); + $nodeToAppend = $this->dom->createElement('div'); + try { + $nodeToAppend->setAttribute('id', $siblingNode->getAttribute('id')); + $nodeToAppend->innerHTML = $siblingNode->innerHTML; + } + catch(Exception $e) + { + $this->dbg('Could not alter siblingNode to div, reverting back to original.'); + $nodeToAppend = $siblingNode; + $s--; + $sl--; + } + } else { + $nodeToAppend = $siblingNode; + $s--; + $sl--; + } + + /* To ensure a node does not interfere with readability styles, remove its classnames */ + $nodeToAppend->removeAttribute('class'); + + /* Append sibling and subtract from our list because it removes the node when you append to another node */ + $articleContent->appendChild($nodeToAppend); + } + } + + /** + * So we have all of the content that we need. Now we clean it up for presentation. + **/ + $this->prepArticle($articleContent); + + /** + * Now that we've gone through the full algorithm, check to see if we got any meaningful content. + * If we didn't, we may need to re-run grabArticle with different flags set. This gives us a higher + * likelihood of finding the content, and the sieve approach gives us a higher likelihood of + * finding the -right- content. + **/ + if (strlen($this->getInnerText($articleContent, false)) < 250) + { + // TODO: find out why element disappears sometimes, e.g. for this URL http://www.businessinsider.com/6-hedge-fund-etfs-for-average-investors-2011-7 + // in the meantime, we check and create an empty element if it's not there. + if (!isset($this->body->childNodes)) $this->body = $this->dom->createElement('body'); + $this->body->innerHTML = $this->bodyCache; + + if ($this->flagIsActive(self::FLAG_STRIP_UNLIKELYS)) { + $this->removeFlag(self::FLAG_STRIP_UNLIKELYS); + return $this->grabArticle($this->body); + } + else if ($this->flagIsActive(self::FLAG_WEIGHT_CLASSES)) { + $this->removeFlag(self::FLAG_WEIGHT_CLASSES); + return $this->grabArticle($this->body); + } + else if ($this->flagIsActive(self::FLAG_CLEAN_CONDITIONALLY)) { + $this->removeFlag(self::FLAG_CLEAN_CONDITIONALLY); + return $this->grabArticle($this->body); + } + else { + return false; + } + } + return $articleContent; + } + + /** + * Remove script tags from document + * + * @param DOMElement + * @return void + */ + public function removeScripts($doc) { + $scripts = $doc->getElementsByTagName('script'); + for($i = $scripts->length-1; $i >= 0; $i--) + { + $scripts->item($i)->parentNode->removeChild($scripts->item($i)); + } + } + + /** + * Get the inner text of a node. + * This also strips out any excess whitespace to be found. + * + * @param DOMElement $ + * @param boolean $normalizeSpaces (default: true) + * @return string + **/ + public function getInnerText($e, $normalizeSpaces=true) { + $textContent = ''; + + if (!isset($e->textContent) || $e->textContent == '') { + return ''; + } + + $textContent = trim($e->textContent); + + if ($normalizeSpaces) { + return preg_replace($this->regexps['normalize'], ' ', $textContent); + } else { + return $textContent; + } + } + + /** + * Get the number of times a string $s appears in the node $e. + * + * @param DOMElement $e + * @param string - what to count. Default is "," + * @return number (integer) + **/ + public function getCharCount($e, $s=',') { + return substr_count($this->getInnerText($e), $s); + } + + /** + * Remove the style attribute on every $e and under. + * + * @param DOMElement $e + * @return void + */ + public function cleanStyles($e) { + if (!is_object($e)) return; + $elems = $e->getElementsByTagName('*'); + foreach ($elems as $elem) { + $elem->removeAttribute('style'); + } + } + + /** + * Get the density of links as a percentage of the content + * This is the amount of text that is inside a link divided by the total text in the node. + * + * @param DOMElement $e + * @return number (float) + */ + public function getLinkDensity($e) { + $links = $e->getElementsByTagName('a'); + $textLength = strlen($this->getInnerText($e)); + $linkLength = 0; + for ($i=0, $il=$links->length; $i < $il; $i++) + { + $linkLength += strlen($this->getInnerText($links->item($i))); + } + if ($textLength > 0) { + return $linkLength / $textLength; + } else { + return 0; + } + } + + /** + * Get an elements class/id weight. Uses regular expressions to tell if this + * element looks good or bad. + * + * @param DOMElement $e + * @return number (Integer) + */ + public function getClassWeight($e) { + if(!$this->flagIsActive(self::FLAG_WEIGHT_CLASSES)) { + return 0; + } + + $weight = 0; + + /* Look for a special classname */ + if ($e->hasAttribute('class') && $e->getAttribute('class') != '') + { + if (preg_match($this->regexps['negative'], $e->getAttribute('class'))) { + $weight -= 25; + } + if (preg_match($this->regexps['positive'], $e->getAttribute('class'))) { + $weight += 25; + } + } + + /* Look for a special ID */ + if ($e->hasAttribute('id') && $e->getAttribute('id') != '') + { + if (preg_match($this->regexps['negative'], $e->getAttribute('id'))) { + $weight -= 25; + } + if (preg_match($this->regexps['positive'], $e->getAttribute('id'))) { + $weight += 25; + } + } + return $weight; + } + + /** + * Remove extraneous break tags from a node. + * + * @param DOMElement $node + * @return void + */ + public function killBreaks($node) { + $html = $node->innerHTML; + $html = preg_replace($this->regexps['killBreaks'], '
            ', $html); + $node->innerHTML = $html; + } + + /** + * Clean a node of all elements of type "tag". + * (Unless it's a youtube/vimeo video. People love movies.) + * + * Updated 2012-09-18 to preserve youtube/vimeo iframes + * + * @param DOMElement $e + * @param string $tag + * @return void + */ + public function clean($e, $tag) { + $targetList = $e->getElementsByTagName($tag); + $isEmbed = ($tag == 'iframe' || $tag == 'object' || $tag == 'embed'); + + for ($y=$targetList->length-1; $y >= 0; $y--) { + /* Allow youtube and vimeo videos through as people usually want to see those. */ + if ($isEmbed) { + $attributeValues = ''; + for ($i=0, $il=$targetList->item($y)->attributes->length; $i < $il; $i++) { + $attributeValues .= $targetList->item($y)->attributes->item($i)->value . '|'; // DOMAttr? (TODO: test) + } + + /* First, check the elements attributes to see if any of them contain youtube or vimeo */ + if (preg_match($this->regexps['video'], $attributeValues)) { + continue; + } + + /* Then check the elements inside this element for the same. */ + if (preg_match($this->regexps['video'], $targetList->item($y)->innerHTML)) { + continue; + } + } + $targetList->item($y)->parentNode->removeChild($targetList->item($y)); + } + } + + /** + * Clean an element of all tags of type "tag" if they look fishy. + * "Fishy" is an algorithm based on content length, classnames, + * link density, number of images & embeds, etc. + * + * @param DOMElement $e + * @param string $tag + * @return void + */ + public function cleanConditionally($e, $tag) { + if (!$this->flagIsActive(self::FLAG_CLEAN_CONDITIONALLY)) { + return; + } + + $tagsList = $e->getElementsByTagName($tag); + $curTagsLength = $tagsList->length; + + /** + * Gather counts for other typical elements embedded within. + * Traverse backwards so we can remove nodes at the same time without effecting the traversal. + * + * TODO: Consider taking into account original contentScore here. + */ + for ($i=$curTagsLength-1; $i >= 0; $i--) { + $weight = $this->getClassWeight($tagsList->item($i)); + $contentScore = ($tagsList->item($i)->hasAttribute('readability')) ? (int)$tagsList->item($i)->getAttribute('readability') : 0; + + $this->dbg('Cleaning Conditionally ' . $tagsList->item($i)->tagName . ' (' . $tagsList->item($i)->getAttribute('class') . ':' . $tagsList->item($i)->getAttribute('id') . ')' . (($tagsList->item($i)->hasAttribute('readability')) ? (' with score ' . $tagsList->item($i)->getAttribute('readability')) : '')); + + if ($weight + $contentScore < 0) { + $tagsList->item($i)->parentNode->removeChild($tagsList->item($i)); + } + else if ( $this->getCharCount($tagsList->item($i), ',') < 10) { + /** + * If there are not very many commas, and the number of + * non-paragraph elements is more than paragraphs or other ominous signs, remove the element. + **/ + $p = $tagsList->item($i)->getElementsByTagName('p')->length; + $img = $tagsList->item($i)->getElementsByTagName('img')->length; + $li = $tagsList->item($i)->getElementsByTagName('li')->length-100; + $input = $tagsList->item($i)->getElementsByTagName('input')->length; + $a = $tagsList->item($i)->getElementsByTagName('a')->length; + + $embedCount = 0; + $embeds = $tagsList->item($i)->getElementsByTagName('embed'); + for ($ei=0, $il=$embeds->length; $ei < $il; $ei++) { + if (preg_match($this->regexps['video'], $embeds->item($ei)->getAttribute('src'))) { + $embedCount++; + } + } + $embeds = $tagsList->item($i)->getElementsByTagName('iframe'); + for ($ei=0, $il=$embeds->length; $ei < $il; $ei++) { + if (preg_match($this->regexps['video'], $embeds->item($ei)->getAttribute('src'))) { + $embedCount++; + } + } + + $linkDensity = $this->getLinkDensity($tagsList->item($i)); + $contentLength = strlen($this->getInnerText($tagsList->item($i))); + $toRemove = false; + + if ($this->lightClean) { + $this->dbg('Light clean...'); + if ( ($img > $p) && ($img > 4) ) { + $this->dbg(' more than 4 images and more image elements than paragraph elements'); + $toRemove = true; + } else if ($li > $p && $tag != 'ul' && $tag != 'ol') { + $this->dbg(' too many

          1. elements, and parent is not
              or
                '); + $toRemove = true; + } else if ( $input > floor($p/3) ) { + $this->dbg(' too many elements'); + $toRemove = true; + } else if ($contentLength < 10 && ($embedCount === 0 && ($img === 0 || $img > 2))) { + $this->dbg(' content length less than 10 chars, 0 embeds and either 0 images or more than 2 images'); + $toRemove = true; + } else if($weight < 25 && $linkDensity > 0.2) { + $this->dbg(' weight smaller than 25 and link density above 0.2'); + $toRemove = true; + } else if($a > 2 && ($weight >= 25 && $linkDensity > 0.5)) { + $this->dbg(' more than 2 links and weight above 25 but link density greater than 0.5'); + $toRemove = true; + } else if($embedCount > 3) { + $this->dbg(' more than 3 embeds'); + $toRemove = true; + } + } else { + $this->dbg('Standard clean...'); + if ( $img > $p ) { + $this->dbg(' more image elements than paragraph elements'); + $toRemove = true; + } else if ($li > $p && $tag != 'ul' && $tag != 'ol') { + $this->dbg(' too many
              1. elements, and parent is not
                  or
                    '); + $toRemove = true; + } else if ( $input > floor($p/3) ) { + $this->dbg(' too many elements'); + $toRemove = true; + } else if ($contentLength < 25 && ($img === 0 || $img > 2) ) { + $this->dbg(' content length less than 25 chars and 0 images, or more than 2 images'); + $toRemove = true; + } else if($weight < 25 && $linkDensity > 0.2) { + $this->dbg(' weight smaller than 25 and link density above 0.2'); + $toRemove = true; + } else if($weight >= 25 && $linkDensity > 0.5) { + $this->dbg(' weight above 25 but link density greater than 0.5'); + $toRemove = true; + } else if(($embedCount == 1 && $contentLength < 75) || $embedCount > 1) { + $this->dbg(' 1 embed and content length smaller than 75 chars, or more than one embed'); + $toRemove = true; + } + } + + if ($toRemove) { + //$this->dbg('Removing: '.$tagsList->item($i)->innerHTML); + $tagsList->item($i)->parentNode->removeChild($tagsList->item($i)); + } + } + } + } + + /** + * Clean out spurious headers from an Element. Checks things like classnames and link density. + * + * @param DOMElement $e + * @return void + */ + public function cleanHeaders($e) { + for ($headerIndex = 1; $headerIndex < 3; $headerIndex++) { + $headers = $e->getElementsByTagName('h' . $headerIndex); + for ($i=$headers->length-1; $i >=0; $i--) { + if ($this->getClassWeight($headers->item($i)) < 0 || $this->getLinkDensity($headers->item($i)) > 0.33) { + $headers->item($i)->parentNode->removeChild($headers->item($i)); + } + } + } + } + + public function flagIsActive($flag) { + return ($this->flags & $flag) > 0; + } + + public function addFlag($flag) { + $this->flags = $this->flags | $flag; + } + + public function removeFlag($flag) { + $this->flags = $this->flags & ~$flag; + } +} ?> \ No newline at end of file diff --git a/inc/3rdparty/makefulltextfeed.php b/inc/3rdparty/makefulltextfeed.php index 4faad6d..7a56be8 100755 --- a/inc/3rdparty/makefulltextfeed.php +++ b/inc/3rdparty/makefulltextfeed.php @@ -3,8 +3,8 @@ // Author: Keyvan Minoukadeh // Copyright (c) 2013 Keyvan Minoukadeh // License: AGPLv3 -// Version: 3.1 -// Date: 2013-03-05 +// Version: 3.2 +// Date: 2013-05-13 // More info: http://fivefilters.org/content-only/ // Help: http://help.fivefilters.org @@ -25,12 +25,8 @@ along with this program. If not, see . // Usage // ----- -// Request this file passing it your feed in the querystring: makefulltextfeed.php?url=mysite.org -// The following options can be passed in the querystring: -// * URL: url=[feed or website url] (required, should be URL-encoded - in php: urlencode($url)) -// * URL points to HTML (not feed): html=true (optional, by default it's automatically detected) -// * API key: key=[api key] (optional, refer to config.php) -// * Max entries to process: max=[max number of items] (optional) +// Request this file passing it a web page or feed URL in the querystring: makefulltextfeed.php?url=example.org/article +// For more request parameters, see http://help.fivefilters.org/customer/portal/articles/226660-usage error_reporting(E_ALL ^ E_NOTICE); ini_set("display_errors", 1); @@ -76,8 +72,8 @@ header('X-Robots-Tag: noindex, nofollow'); //////////////////////////////// // Check if service is enabled //////////////////////////////// -if (!$options->enabled) { - die('The full-text RSS service is currently disabled'); +if (!$options->enabled) { + die('The full-text RSS service is currently disabled'); } //////////////////////////////// @@ -121,8 +117,8 @@ $options->smart_cache = $options->smart_cache && function_exists('apc_inc'); //////////////////////////////// // Check for feed URL //////////////////////////////// -if (!isset($_GET['url'])) { - die('No URL supplied'); +if (!isset($_GET['url'])) { + die('No URL supplied'); } $url = trim($_GET['url']); if (strtolower(substr($url, 0, 7)) == 'feed://') { @@ -161,10 +157,12 @@ if (isset($_GET['key']) && ($key_index = array_search($_GET['key'], $options->ap if (isset($_GET['links'])) $redirect .= '&links='.urlencode($_GET['links']); if (isset($_GET['exc'])) $redirect .= '&exc='.urlencode($_GET['exc']); if (isset($_GET['format'])) $redirect .= '&format='.urlencode($_GET['format']); - if (isset($_GET['callback'])) $redirect .= '&callback='.urlencode($_GET['callback']); + if (isset($_GET['callback'])) $redirect .= '&callback='.urlencode($_GET['callback']); if (isset($_GET['l'])) $redirect .= '&l='.urlencode($_GET['l']); if (isset($_GET['xss'])) $redirect .= '&xss'; if (isset($_GET['use_extracted_title'])) $redirect .= '&use_extracted_title'; + if (isset($_GET['content'])) $redirect .= '&content='.urlencode($_GET['content']); + if (isset($_GET['summary'])) $redirect .= '&summary='.urlencode($_GET['summary']); if (isset($_GET['debug'])) $redirect .= '&debug'; if ($debug_mode) { debug('Redirecting to hide access key, follow URL below to continue'); @@ -177,7 +175,7 @@ if (isset($_GET['key']) && ($key_index = array_search($_GET['key'], $options->ap /////////////////////////////////////////////// // Set timezone. -// Prevents warnings, but needs more testing - +// Prevents warnings, but needs more testing - // perhaps if timezone is set in php.ini we // don't need to set it at all... /////////////////////////////////////////////// @@ -199,7 +197,7 @@ if (isset($_GET['key']) && isset($_GET['hash']) && isset($options->api_keys[(int } $key_index = ($valid_key) ? (int)$_GET['key'] : 0; if (!$valid_key && $options->key_required) { - die('A valid key must be supplied'); + die('A valid key must be supplied'); } if (!$valid_key && isset($_GET['key']) && $_GET['key'] != '') { die('The entered key is invalid'); @@ -250,6 +248,28 @@ if ($options->favour_feed_titles == 'user') { $favour_feed_titles = $options->favour_feed_titles; } +/////////////////////////////////////////////// +// Include full content in output? +/////////////////////////////////////////////// +if ($options->content === 'user') { + if (isset($_GET['content']) && $_GET['content'] === '0') { + $options->content = false; + } else { + $options->content = true; + } +} + +/////////////////////////////////////////////// +// Include summaries in output? +/////////////////////////////////////////////// +if ($options->summary === 'user') { + if (isset($_GET['summary']) && $_GET['summary'] === '1') { + $options->summary = true; + } else { + $options->summary = false; + } +} + /////////////////////////////////////////////// // Exclude items if extraction fails /////////////////////////////////////////////// @@ -272,15 +292,6 @@ if ($options->detect_language === 'user') { $detect_language = $options->detect_language; } -if ($detect_language >= 2) { - $language_codes = array('albanian' => 'sq','arabic' => 'ar','azeri' => 'az','bengali' => 'bn','bulgarian' => 'bg', - 'cebuano' => 'ceb', // ISO 639-2 - 'croatian' => 'hr','czech' => 'cs','danish' => 'da','dutch' => 'nl','english' => 'en','estonian' => 'et','farsi' => 'fa','finnish' => 'fi','french' => 'fr','german' => 'de','hausa' => 'ha', - 'hawaiian' => 'haw', // ISO 639-2 - 'hindi' => 'hi','hungarian' => 'hu','icelandic' => 'is','indonesian' => 'id','italian' => 'it','kazakh' => 'kk','kyrgyz' => 'ky','latin' => 'la','latvian' => 'lv','lithuanian' => 'lt','macedonian' => 'mk','mongolian' => 'mn','nepali' => 'ne','norwegian' => 'no','pashto' => 'ps', - 'pidgin' => 'cpe', // ISO 639-2 - 'polish' => 'pl','portuguese' => 'pt','romanian' => 'ro','russian' => 'ru','serbian' => 'sr','slovak' => 'sk','slovene' => 'sl','somali' => 'so','spanish' => 'es','swahili' => 'sw','swedish' => 'sv','tagalog' => 'tl','turkish' => 'tr','ukrainian' => 'uk','urdu' => 'ur','uzbek' => 'uz','vietnamese' => 'vi','welsh' => 'cy'); -} $use_cld = extension_loaded('cld') && (version_compare(PHP_VERSION, '5.3.0') >= 0); ///////////////////////////////////// @@ -330,7 +341,7 @@ if ($options->cors) header('Access-Control-Allow-Origin: *'); ////////////////////////////////// if ($options->caching) { debug('Caching is enabled...'); - $cache_id = md5($max.$url.$valid_key.$links.$favour_feed_titles.$xss_filter.$exclude_on_fail.$format.$detect_language.(int)isset($_GET['pubsub'])); + $cache_id = md5($max.$url.(int)$valid_key.$links.(int)$favour_feed_titles.(int)$options->content.(int)$options->summary.(int)$xss_filter.(int)$exclude_on_fail.$format.$detect_language.(int)isset($_GET['pubsub'])); $check_cache = true; if ($options->apc && $options->smart_cache) { apc_add("cache.$cache_id", 0, 10*60); @@ -468,7 +479,7 @@ if ($img_url = $feed->get_image_url()) { //////////////////////////////////////////// // Loop through feed items //////////////////////////////////////////// -$items = $feed->get_items(0, $max); +$items = $feed->get_items(0, $max); // Request all feed items in parallel (if supported) $urls_sanitized = array(); $urls = array(); @@ -550,24 +561,43 @@ foreach ($items as $key => $item) { $is_single_page = false; if ($single_page_response = getSinglePage($item, $html, $effective_url)) { $is_single_page = true; - $html = $single_page_response['body']; - // remove strange things - $html = str_replace('', '', $html); - $html = convert_to_utf8($html, $single_page_response['headers']); $effective_url = $single_page_response['effective_url']; - debug("Retrieved single-page view from $effective_url"); + // check if action defined for returned Content-Type + $mime_info = get_mime_action_info($single_page_response['headers']); + if (isset($mime_info['action'])) { + if ($mime_info['action'] == 'exclude') { + continue; // skip this feed item entry + } elseif ($mime_info['action'] == 'link') { + if ($mime_info['type'] == 'image') { + $html = "\"{$mime_info['name']}\""; + } else { + $html = "Download {$mime_info['name']}"; + } + $extracted_title = $mime_info['name']; + $do_content_extraction = false; + } + } + if ($do_content_extraction) { + $html = $single_page_response['body']; + // remove strange things + $html = str_replace('', '', $html); + $html = convert_to_utf8($html, $single_page_response['headers']); + debug("Retrieved single-page view from $effective_url"); + } unset($single_page_response); } + } + if ($do_content_extraction) { debug('--------'); debug('Attempting to extract content'); $extract_result = $extractor->process($html, $effective_url); $readability = $extractor->readability; - $content_block = ($extract_result) ? $extractor->getContent() : null; + $content_block = ($extract_result) ? $extractor->getContent() : null; $extracted_title = ($extract_result) ? $extractor->getTitle() : ''; // Deal with multi-page articles //die('Next: '.$extractor->getNextPageUrl()); $is_multi_page = (!$is_single_page && $extract_result && $extractor->getNextPageUrl()); - if ($options->multipage && $is_multi_page) { + if ($options->multipage && $is_multi_page && $options->content) { debug('--------'); debug('Attempting to process multi-page article'); $multi_page_urls = array(); @@ -580,7 +610,7 @@ foreach ($items as $key => $item) { // check it's not what we have already! if (!in_array($next_page_url, $multi_page_urls)) { // it's not, so let's attempt to fetch it - $multi_page_urls[] = $next_page_url; + $multi_page_urls[] = $next_page_url; $_prev_ref = $http->referer; if (($response = $http->get($next_page_url, true)) && $response['status_code'] < 300) { // make sure mime type is not something with a different action associated @@ -605,13 +635,15 @@ foreach ($items as $key => $item) { // did we successfully deal with this multi-page article? if (empty($multi_page_content)) { debug('Failed to extract all parts of multi-page article, so not going to include them'); - $multi_page_content[] = $readability->dom->createElement('p')->innerHTML = 'This article appears to continue on subsequent pages which we could not extract'; + $_page = $readability->dom->createElement('p'); + $_page->innerHTML = 'This article appears to continue on subsequent pages which we could not extract'; + $multi_page_content[] = $_page; } foreach ($multi_page_content as $_page) { $_page = $content_block->ownerDocument->importNode($_page, true); $content_block->appendChild($_page); } - unset($multi_page_urls, $multi_page_content, $page_mime_info, $next_page_url); + unset($multi_page_urls, $multi_page_content, $page_mime_info, $next_page_url, $_page); } } // use extracted title for both feed and item title if we're using single-item dummy feed @@ -658,7 +690,7 @@ foreach ($items as $key => $item) { } else { $html = $content_block->ownerDocument->saveXML($content_block); // essentially outerHTML } - unset($content_block); + //unset($content_block); // post-processing cleanup $html = preg_replace('!

                    [\s\h\v]*

                    !u', '', $html); if ($links == 'remove') { @@ -671,130 +703,155 @@ foreach ($items as $key => $item) { } } - if ($valid_key && isset($_GET['pubsub'])) { // used only on fivefilters.org at the moment - $newitem->addElement('guid', 'http://fivefilters.org/content-only/redirect.php?url='.urlencode($item->get_permalink()), array('isPermaLink'=>'false')); + if ($valid_key && isset($_GET['pubsub'])) { // used only on fivefilters.org at the moment + $newitem->addElement('guid', 'http://fivefilters.org/content-only/redirect.php?url='.urlencode($item->get_permalink()), array('isPermaLink'=>'false')); + } else { + $newitem->addElement('guid', $item->get_permalink(), array('isPermaLink'=>'true')); + } + // filter xss? + if ($xss_filter) { + debug('Filtering HTML to remove XSS'); + $html = htmLawed::hl($html, array('safe'=>1, 'deny_attribute'=>'style', 'comment'=>1, 'cdata'=>1)); + } + + // add content + if ($options->summary === true) { + // get summary + $summary = ''; + if (!$do_content_extraction) { + $summary = $html; } else { - $newitem->addElement('guid', $item->get_permalink(), array('isPermaLink'=>'true')); - } - // filter xss? - if ($xss_filter) { - debug('Filtering HTML to remove XSS'); - $html = htmLawed::hl($html, array('safe'=>1, 'deny_attribute'=>'style', 'comment'=>1, 'cdata'=>1)); - } - $newitem->setDescription($html); - - // set date - if ((int)$item->get_date('U') > 0) { - $newitem->setDate((int)$item->get_date('U')); - } elseif ($extractor->getDate()) { - $newitem->setDate($extractor->getDate()); - } - - // add authors - if ($authors = $item->get_authors()) { - foreach ($authors as $author) { - // for some feeds, SimplePie stores author's name as email, e.g. http://feeds.feedburner.com/nymag/intel - if ($author->get_name() !== null) { - $newitem->addElement('dc:creator', $author->get_name()); - } elseif ($author->get_email() !== null) { - $newitem->addElement('dc:creator', $author->get_email()); + // Try to get first few paragraphs + if (isset($content_block) && ($content_block instanceof DOMElement)) { + $_paras = $content_block->getElementsByTagName('p'); + foreach ($_paras as $_para) { + $summary .= preg_replace("/[\n\r\t ]+/", ' ', $_para->textContent).' '; + if (strlen($summary) > 200) break; } - } - } elseif ($authors = $extractor->getAuthors()) { - //TODO: make sure the list size is reasonable - foreach ($authors as $author) { - // TODO: xpath often selects authors from other articles linked from the page. - // for now choose first item - $newitem->addElement('dc:creator', $author); - break; + } else { + $summary = $html; } } - - // add language - if ($detect_language) { - $language = $extractor->getLanguage(); - if (!$language) $language = $feed->get_language(); - if (($detect_language == 3 || (!$language && $detect_language == 2)) && $text_sample) { - try { - if ($use_cld) { - // Use PHP-CLD extension - $php_cld = 'CLD\detect'; // in quotes to prevent PHP 5.2 parse error - $res = $php_cld($text_sample); - if (is_array($res) && count($res) > 0) { - $language = $res[0]['code']; - } - } else { - //die('what'); - // Use PEAR's Text_LanguageDetect - if (!isset($l)) { - $l = new Text_LanguageDetect('libraries/language-detect/lang.dat', 'libraries/language-detect/unicode_blocks.dat'); - } - $l_result = $l->detect($text_sample, 1); - if (count($l_result) > 0) { - $language = $language_codes[key($l_result)]; - } + unset($_paras, $_para); + $summary = get_excerpt($summary); + $newitem->setDescription($summary); + if ($options->content) $newitem->setElement('content:encoded', $html); + } else { + if ($options->content) $newitem->setDescription($html); + } + + // set date + if ((int)$item->get_date('U') > 0) { + $newitem->setDate((int)$item->get_date('U')); + } elseif ($extractor->getDate()) { + $newitem->setDate($extractor->getDate()); + } + + // add authors + if ($authors = $item->get_authors()) { + foreach ($authors as $author) { + // for some feeds, SimplePie stores author's name as email, e.g. http://feeds.feedburner.com/nymag/intel + if ($author->get_name() !== null) { + $newitem->addElement('dc:creator', $author->get_name()); + } elseif ($author->get_email() !== null) { + $newitem->addElement('dc:creator', $author->get_email()); + } + } + } elseif ($authors = $extractor->getAuthors()) { + //TODO: make sure the list size is reasonable + foreach ($authors as $author) { + // TODO: xpath often selects authors from other articles linked from the page. + // for now choose first item + $newitem->addElement('dc:creator', $author); + break; + } + } + + // add language + if ($detect_language) { + $language = $extractor->getLanguage(); + if (!$language) $language = $feed->get_language(); + if (($detect_language == 3 || (!$language && $detect_language == 2)) && $text_sample) { + try { + if ($use_cld) { + // Use PHP-CLD extension + $php_cld = 'CLD\detect'; // in quotes to prevent PHP 5.2 parse error + $res = $php_cld($text_sample); + if (is_array($res) && count($res) > 0) { + $language = $res[0]['code']; } - } catch (Exception $e) { - //die('error: '.$e); - // do nothing - } - } - if ($language && (strlen($language) < 7)) { - $newitem->addElement('dc:language', $language); - } - } - - // add MIME type (if it appeared in our exclusions lists) - if (isset($mime_info['mime'])) $newitem->addElement('dc:format', $mime_info['mime']); - // add effective URL (URL after redirects) - if (isset($effective_url)) { - //TODO: ensure $effective_url is valid witout - sometimes it causes problems, e.g. - //http://www.siasat.pk/forum/showthread.php?108883-Pakistan-Chowk-by-Rana-Mubashir-�-25th-March-2012-Special-Program-from-Liari-(Karachi) - //temporary measure: use utf8_encode() - $newitem->addElement('dc:identifier', remove_url_cruft(utf8_encode($effective_url))); - } else { - $newitem->addElement('dc:identifier', remove_url_cruft($item->get_permalink())); - } - - // add categories - if ($categories = $item->get_categories()) { - foreach ($categories as $category) { - if ($category->get_label() !== null) { - $newitem->addElement('category', $category->get_label()); - } - } - } - - // check for enclosures - if ($options->keep_enclosures) { - if ($enclosures = $item->get_enclosures()) { - foreach ($enclosures as $enclosure) { - // thumbnails - foreach ((array)$enclosure->get_thumbnails() as $thumbnail) { - $newitem->addElement('media:thumbnail', '', array('url'=>$thumbnail)); + } else { + //die('what'); + // Use PEAR's Text_LanguageDetect + if (!isset($l)) { + $l = new Text_LanguageDetect(); + $l->setNameMode(2); // return ISO 639-1 codes (e.g. "en") + } + $l_result = $l->detect($text_sample, 1); + if (count($l_result) > 0) { + $language = key($l_result); } - if (!$enclosure->get_link()) continue; - $enc = array(); - // Media RSS spec ($enc): http://search.yahoo.com/mrss - // SimplePie methods ($enclosure): http://simplepie.org/wiki/reference/start#methods4 - $enc['url'] = $enclosure->get_link(); - if ($enclosure->get_length()) $enc['fileSize'] = $enclosure->get_length(); - if ($enclosure->get_type()) $enc['type'] = $enclosure->get_type(); - if ($enclosure->get_medium()) $enc['medium'] = $enclosure->get_medium(); - if ($enclosure->get_expression()) $enc['expression'] = $enclosure->get_expression(); - if ($enclosure->get_bitrate()) $enc['bitrate'] = $enclosure->get_bitrate(); - if ($enclosure->get_framerate()) $enc['framerate'] = $enclosure->get_framerate(); - if ($enclosure->get_sampling_rate()) $enc['samplingrate'] = $enclosure->get_sampling_rate(); - if ($enclosure->get_channels()) $enc['channels'] = $enclosure->get_channels(); - if ($enclosure->get_duration()) $enc['duration'] = $enclosure->get_duration(); - if ($enclosure->get_height()) $enc['height'] = $enclosure->get_height(); - if ($enclosure->get_width()) $enc['width'] = $enclosure->get_width(); - if ($enclosure->get_language()) $enc['lang'] = $enclosure->get_language(); - $newitem->addElement('media:content', '', $enc); } + } catch (Exception $e) { + //die('error: '.$e); + // do nothing } } - /* } */ + if ($language && (strlen($language) < 7)) { + $newitem->addElement('dc:language', $language); + } + } + + // add MIME type (if it appeared in our exclusions lists) + if (isset($mime_info['mime'])) $newitem->addElement('dc:format', $mime_info['mime']); + // add effective URL (URL after redirects) + if (isset($effective_url)) { + //TODO: ensure $effective_url is valid witout - sometimes it causes problems, e.g. + //http://www.siasat.pk/forum/showthread.php?108883-Pakistan-Chowk-by-Rana-Mubashir-�-25th-March-2012-Special-Program-from-Liari-(Karachi) + //temporary measure: use utf8_encode() + $newitem->addElement('dc:identifier', remove_url_cruft(utf8_encode($effective_url))); + } else { + $newitem->addElement('dc:identifier', remove_url_cruft($item->get_permalink())); + } + + // add categories + if ($categories = $item->get_categories()) { + foreach ($categories as $category) { + if ($category->get_label() !== null) { + $newitem->addElement('category', $category->get_label()); + } + } + } + + // check for enclosures + if ($options->keep_enclosures) { + if ($enclosures = $item->get_enclosures()) { + foreach ($enclosures as $enclosure) { + // thumbnails + foreach ((array)$enclosure->get_thumbnails() as $thumbnail) { + $newitem->addElement('media:thumbnail', '', array('url'=>$thumbnail)); + } + if (!$enclosure->get_link()) continue; + $enc = array(); + // Media RSS spec ($enc): http://search.yahoo.com/mrss + // SimplePie methods ($enclosure): http://simplepie.org/wiki/reference/start#methods4 + $enc['url'] = $enclosure->get_link(); + if ($enclosure->get_length()) $enc['fileSize'] = $enclosure->get_length(); + if ($enclosure->get_type()) $enc['type'] = $enclosure->get_type(); + if ($enclosure->get_medium()) $enc['medium'] = $enclosure->get_medium(); + if ($enclosure->get_expression()) $enc['expression'] = $enclosure->get_expression(); + if ($enclosure->get_bitrate()) $enc['bitrate'] = $enclosure->get_bitrate(); + if ($enclosure->get_framerate()) $enc['framerate'] = $enclosure->get_framerate(); + if ($enclosure->get_sampling_rate()) $enc['samplingrate'] = $enclosure->get_sampling_rate(); + if ($enclosure->get_channels()) $enc['channels'] = $enclosure->get_channels(); + if ($enclosure->get_duration()) $enc['duration'] = $enclosure->get_duration(); + if ($enclosure->get_height()) $enc['height'] = $enclosure->get_height(); + if ($enclosure->get_width()) $enc['width'] = $enclosure->get_width(); + if ($enclosure->get_language()) $enc['lang'] = $enclosure->get_language(); + $newitem->addElement('media:content', '', $enc); + } + } + } $output->addItem($newitem); unset($html); $item_count++; diff --git a/inc/3rdparty/makefulltextfeedHelpers.php b/inc/3rdparty/makefulltextfeedHelpers.php index 1c11b8f..4e98537 100755 --- a/inc/3rdparty/makefulltextfeedHelpers.php +++ b/inc/3rdparty/makefulltextfeedHelpers.php @@ -66,6 +66,38 @@ class DummySingleItem { // HELPER FUNCTIONS /////////////////////////////// +// Adapted from WordPress +// http://core.trac.wordpress.org/browser/tags/3.5.1/wp-includes/formatting.php#L2173 +function get_excerpt($text, $num_words=55, $more=null) { + if (null === $more) $more = '…'; + $text = strip_tags($text); + //TODO: Check if word count is based on single characters (East Asian characters) + /* + if (1==2) { + $text = trim(preg_replace("/[\n\r\t ]+/", ' ', $text), ' '); + preg_match_all('/./u', $text, $words_array); + $words_array = array_slice($words_array[0], 0, $num_words + 1); + $sep = ''; + } else { + $words_array = preg_split("/[\n\r\t ]+/", $text, $num_words + 1, PREG_SPLIT_NO_EMPTY); + $sep = ' '; + } + */ + $words_array = preg_split("/[\n\r\t ]+/", $text, $num_words + 1, PREG_SPLIT_NO_EMPTY); + $sep = ' '; + if (count($words_array) > $num_words) { + array_pop($words_array); + $text = implode($sep, $words_array); + $text = $text.$more; + } else { + $text = implode($sep, $words_array); + } + // trim whitespace at beginning or end of string + // See: http://stackoverflow.com/questions/4166896/trim-unicode-whitespace-in-php-5-2 + $text = preg_replace('/^[\pZ\pC]+|[\pZ\pC]+$/u', '', $text); + return $text; +} + function url_allowed($url) { global $options; if (!empty($options->allowed_urls)) { @@ -165,14 +197,6 @@ function convert_to_utf8($html, $header=null) if (strtolower($encoding) != 'utf-8') { debug('Converting to UTF-8'); $html = SimplePie_Misc::change_encoding($html, $encoding, 'utf-8'); - /* - if (function_exists('iconv')) { - // iconv appears to handle certain character encodings better than mb_convert_encoding - $html = iconv($encoding, 'utf-8', $html); - } else { - $html = mb_convert_encoding($html, 'utf-8', $encoding); - } - */ } } } @@ -196,7 +220,7 @@ function makeAbsolute($base, $elem) { } function makeAbsoluteAttr($base, $e, $attr) { if ($e->hasAttribute($attr)) { - // Trim leading and trailing white space. I don't really like this but + // Trim leading and trailing white space. I don't really like this but // unfortunately it does appear on some sites. e.g. $url = trim(str_replace('%20', ' ', $e->getAttribute($attr))); $url = str_replace(' ', '%20', $url); diff --git a/inc/3rdparty/site_config/index.php b/inc/3rdparty/site_config/index.php index a1b767f..76ca8b3 100644 --- a/inc/3rdparty/site_config/index.php +++ b/inc/3rdparty/site_config/index.php @@ -1,3 +1,2 @@ - \ No newline at end of file +