\r
// set include path\r
set_include_path(realpath(dirname(__FILE__).'/libraries').PATH_SEPARATOR.get_include_path());\r
-// Autoloading of classes allows us to include files only when they're\r
-// needed. If we've got a cached copy, for example, only Zend_Cache is loaded.\r
-function autoload($class_name) {\r
- static $dir = null;\r
- if ($dir === null) $dir = dirname(__FILE__).'/libraries/';\r
- static $mapping = array(\r
- // Include FeedCreator for RSS/Atom creation\r
- 'FeedWriter' => 'feedwriter/FeedWriter.php',\r
- 'FeedItem' => 'feedwriter/FeedItem.php',\r
- // Include ContentExtractor and Readability for identifying and extracting content from URLs\r
- 'ContentExtractor' => 'content-extractor/ContentExtractor.php',\r
- 'SiteConfig' => 'content-extractor/SiteConfig.php',\r
- 'Readability' => 'readability/Readability.php',\r
- // Include Humble HTTP Agent to allow parallel requests and response caching\r
- 'HumbleHttpAgent' => 'humble-http-agent/HumbleHttpAgent.php',\r
- 'SimplePie_HumbleHttpAgent' => 'humble-http-agent/SimplePie_HumbleHttpAgent.php',\r
- 'CookieJar' => 'humble-http-agent/CookieJar.php',\r
- // Include Zend Cache to improve performance (cache results)\r
- 'Zend_Cache' => 'Zend/Cache.php',\r
- // Language detect\r
- 'Text_LanguageDetect' => 'language-detect/LanguageDetect.php',\r
- // HTML5 Lib\r
- 'HTML5_Parser' => 'html5/Parser.php',\r
- // htmLawed - used if XSS filter is enabled (xss_filter)\r
- 'htmLawed' => 'htmLawed/htmLawed.php'\r
- );\r
- if (isset($mapping[$class_name])) {\r
- debug("** Loading class $class_name ({$mapping[$class_name]})");\r
- require $dir.$mapping[$class_name];\r
- return true;\r
- } else {\r
- return false;\r
- }\r
-}\r
-spl_autoload_register('autoload');\r
-require dirname(__FILE__).'/libraries/simplepie/autoloader.php';\r
+\r
+require_once dirname(__FILE__).'/makefulltextfeedHelpers.php';\r
\r
////////////////////////////////\r
// Load config file\r
//////////////////////////////////\r
// Set up HTTP agent\r
//////////////////////////////////\r
+global $http;\r
$http = new HumbleHttpAgent();\r
$http->debug = $debug_mode;\r
$http->userAgentMap = $options->user_agents;\r
$isDummyFeed = true;\r
unset($feed, $result);\r
// create single item dummy feed object\r
- class DummySingleItemFeed {\r
- public $item;\r
- function __construct($url) { $this->item = new DummySingleItem($url); }\r
- public function get_title() { return ''; }\r
- public function get_description() { return 'Content extracted from '.$this->item->url; }\r
- public function get_link() { return $this->item->url; }\r
- public function get_language() { return false; }\r
- public function get_image_url() { return false; }\r
- public function get_items($start=0, $max=1) { return array(0=>$this->item); }\r
- }\r
- class DummySingleItem {\r
- public $url;\r
- function __construct($url) { $this->url = $url; }\r
- public function get_permalink() { return $this->url; }\r
- public function get_title() { return null; }\r
- public function get_date($format='') { return false; }\r
- public function get_author($key=0) { return null; }\r
- public function get_authors() { return null; }\r
- public function get_description() { return ''; }\r
- public function get_enclosure($key=0, $prefer=null) { return null; }\r
- public function get_enclosures() { return null; }\r
- public function get_categories() { return null; }\r
- }\r
$feed = new DummySingleItemFeed($url);\r
}\r
\r
// add effective URL (URL after redirects)\r
if (isset($effective_url)) {\r
//TODO: ensure $effective_url is valid witout - sometimes it causes problems, e.g.\r
- //http://www.siasat.pk/forum/showthread.php?108883-Pakistan-Chowk-by-Rana-Mubashir-\96-25th-March-2012-Special-Program-from-Liari-(Karachi)\r
+ //http://www.siasat.pk/forum/showthread.php?108883-Pakistan-Chowk-by-Rana-Mubashir-�-25th-March-2012-Special-Program-from-Liari-(Karachi)\r
//temporary measure: use utf8_encode()\r
$newitem->addElement('dc:identifier', remove_url_cruft(utf8_encode($effective_url)));\r
} else {\r
}\r
if ($add_to_cache) {\r
ob_start();\r
- $output->genarateFeed();\r
+ $output->genarateFeed(false);\r
$output = ob_get_contents();\r
ob_end_clean();\r
if ($html_only && $item_count == 0) {\r
}\r
echo $output;\r
} else {\r
- $output->genarateFeed();\r
+ $output->genarateFeed(false);\r
}\r
if ($callback) echo ');';\r
}\r
\r
-///////////////////////////////\r
-// HELPER FUNCTIONS\r
-///////////////////////////////\r
-\r
-function url_allowed($url) {\r
- global $options;\r
- if (!empty($options->allowed_urls)) {\r
- $allowed = false;\r
- foreach ($options->allowed_urls as $allowurl) {\r
- if (stristr($url, $allowurl) !== false) {\r
- $allowed = true;\r
- break;\r
- }\r
- }\r
- if (!$allowed) return false;\r
- } else {\r
- foreach ($options->blocked_urls as $blockurl) {\r
- if (stristr($url, $blockurl) !== false) {\r
- return false;\r
- }\r
- }\r
- }\r
- return true;\r
-}\r
-\r
-//////////////////////////////////////////////\r
-// Convert $html to UTF8\r
-// (uses HTTP headers and HTML to find encoding)\r
-// adapted from http://stackoverflow.com/questions/910793/php-detect-encoding-and-make-everything-utf-8\r
-//////////////////////////////////////////////\r
-function convert_to_utf8($html, $header=null)\r
-{\r
- $encoding = null;\r
- if ($html || $header) {\r
- if (is_array($header)) $header = implode("\n", $header);\r
- if (!$header || !preg_match_all('/^Content-Type:\s+([^;]+)(?:;\s*charset=["\']?([^;"\'\n]*))?/im', $header, $match, PREG_SET_ORDER)) {\r
- // error parsing the response\r
- debug('Could not find Content-Type header in HTTP response');\r
- } else {\r
- $match = end($match); // get last matched element (in case of redirects)\r
- if (isset($match[2])) $encoding = trim($match[2], "\"' \r\n\0\x0B\t");\r
- }\r
- // TODO: check to see if encoding is supported (can we convert it?)\r
- // If it's not, result will be empty string.\r
- // For now we'll check for invalid encoding types returned by some sites, e.g. 'none'\r
- // Problem URL: http://facta.co.jp/blog/archives/20111026001026.html\r
- if (!$encoding || $encoding == 'none') {\r
- // search for encoding in HTML - only look at the first 50000 characters\r
- // Why 50000? See, for example, http://www.lemonde.fr/festival-de-cannes/article/2012/05/23/deux-cretes-en-goguette-sur-la-croisette_1705732_766360.html\r
- // TODO: improve this so it looks at smaller chunks first\r
- $html_head = substr($html, 0, 50000);\r
- if (preg_match('/^<\?xml\s+version=(?:"[^"]*"|\'[^\']*\')\s+encoding=("[^"]*"|\'[^\']*\')/s', $html_head, $match)) {\r
- $encoding = trim($match[1], '"\'');\r
- } elseif (preg_match('/<meta\s+http-equiv=["\']?Content-Type["\']? content=["\'][^;]+;\s*charset=["\']?([^;"\'>]+)/i', $html_head, $match)) {\r
- $encoding = trim($match[1]);\r
- } elseif (preg_match_all('/<meta\s+([^>]+)>/i', $html_head, $match)) {\r
- foreach ($match[1] as $_test) {\r
- if (preg_match('/charset=["\']?([^"\']+)/i', $_test, $_m)) {\r
- $encoding = trim($_m[1]);\r
- break;\r
- }\r
- }\r
- }\r
- }\r
- if (isset($encoding)) $encoding = trim($encoding);\r
- // trim is important here!\r
- if (!$encoding || (strtolower($encoding) == 'iso-8859-1')) {\r
- // replace MS Word smart qutoes\r
- $trans = array();\r
- $trans[chr(130)] = '‚'; // Single Low-9 Quotation Mark\r
- $trans[chr(131)] = 'ƒ'; // Latin Small Letter F With Hook\r
- $trans[chr(132)] = '„'; // Double Low-9 Quotation Mark\r
- $trans[chr(133)] = '…'; // Horizontal Ellipsis\r
- $trans[chr(134)] = '†'; // Dagger\r
- $trans[chr(135)] = '‡'; // Double Dagger\r
- $trans[chr(136)] = 'ˆ'; // Modifier Letter Circumflex Accent\r
- $trans[chr(137)] = '‰'; // Per Mille Sign\r
- $trans[chr(138)] = 'Š'; // Latin Capital Letter S With Caron\r
- $trans[chr(139)] = '‹'; // Single Left-Pointing Angle Quotation Mark\r
- $trans[chr(140)] = 'Œ'; // Latin Capital Ligature OE\r
- $trans[chr(145)] = '‘'; // Left Single Quotation Mark\r
- $trans[chr(146)] = '’'; // Right Single Quotation Mark\r
- $trans[chr(147)] = '“'; // Left Double Quotation Mark\r
- $trans[chr(148)] = '”'; // Right Double Quotation Mark\r
- $trans[chr(149)] = '•'; // Bullet\r
- $trans[chr(150)] = '–'; // En Dash\r
- $trans[chr(151)] = '—'; // Em Dash\r
- $trans[chr(152)] = '˜'; // Small Tilde\r
- $trans[chr(153)] = '™'; // Trade Mark Sign\r
- $trans[chr(154)] = 'š'; // Latin Small Letter S With Caron\r
- $trans[chr(155)] = '›'; // Single Right-Pointing Angle Quotation Mark\r
- $trans[chr(156)] = 'œ'; // Latin Small Ligature OE\r
- $trans[chr(159)] = 'Ÿ'; // Latin Capital Letter Y With Diaeresis\r
- $html = strtr($html, $trans);\r
- }\r
- if (!$encoding) {\r
- debug('No character encoding found, so treating as UTF-8');\r
- $encoding = 'utf-8';\r
- } else {\r
- debug('Character encoding: '.$encoding);\r
- if (strtolower($encoding) != 'utf-8') {\r
- debug('Converting to UTF-8');\r
- $html = SimplePie_Misc::change_encoding($html, $encoding, 'utf-8');\r
- /*\r
- if (function_exists('iconv')) {\r
- // iconv appears to handle certain character encodings better than mb_convert_encoding\r
- $html = iconv($encoding, 'utf-8', $html);\r
- } else {\r
- $html = mb_convert_encoding($html, 'utf-8', $encoding);\r
- }\r
- */\r
- }\r
- }\r
- }\r
- return $html;\r
-}\r
-\r
-function makeAbsolute($base, $elem) {\r
- $base = new SimplePie_IRI($base);\r
- // remove '//' in URL path (used to prevent URLs from resolving properly)\r
- // TODO: check if this is still the case\r
- if (isset($base->path)) $base->path = preg_replace('!//+!', '/', $base->path);\r
- foreach(array('a'=>'href', 'img'=>'src') as $tag => $attr) {\r
- $elems = $elem->getElementsByTagName($tag);\r
- for ($i = $elems->length-1; $i >= 0; $i--) {\r
- $e = $elems->item($i);\r
- //$e->parentNode->replaceChild($articleContent->ownerDocument->createTextNode($e->textContent), $e);\r
- makeAbsoluteAttr($base, $e, $attr);\r
- }\r
- if (strtolower($elem->tagName) == $tag) makeAbsoluteAttr($base, $elem, $attr);\r
- }\r
-}\r
-function makeAbsoluteAttr($base, $e, $attr) {\r
- if ($e->hasAttribute($attr)) {\r
- // Trim leading and trailing white space. I don't really like this but \r
- // unfortunately it does appear on some sites. e.g. <img src=" /path/to/image.jpg" />\r
- $url = trim(str_replace('%20', ' ', $e->getAttribute($attr)));\r
- $url = str_replace(' ', '%20', $url);\r
- if (!preg_match('!https?://!i', $url)) {\r
- if ($absolute = SimplePie_IRI::absolutize($base, $url)) {\r
- $e->setAttribute($attr, $absolute);\r
- }\r
- }\r
- }\r
-}\r
-function makeAbsoluteStr($base, $url) {\r
- $base = new SimplePie_IRI($base);\r
- // remove '//' in URL path (causes URLs not to resolve properly)\r
- if (isset($base->path)) $base->path = preg_replace('!//+!', '/', $base->path);\r
- if (preg_match('!^https?://!i', $url)) {\r
- // already absolute\r
- return $url;\r
- } else {\r
- if ($absolute = SimplePie_IRI::absolutize($base, $url)) {\r
- return $absolute;\r
- }\r
- return false;\r
- }\r
-}\r
-// returns single page response, or false if not found\r
-function getSinglePage($item, $html, $url) {\r
- global $http, $extractor;\r
- debug('Looking for site config files to see if single page link exists');\r
- $site_config = $extractor->buildSiteConfig($url, $html);\r
- $splink = null;\r
- if (!empty($site_config->single_page_link)) {\r
- $splink = $site_config->single_page_link;\r
- } elseif (!empty($site_config->single_page_link_in_feed)) {\r
- // single page link xpath is targeted at feed\r
- $splink = $site_config->single_page_link_in_feed;\r
- // so let's replace HTML with feed item description\r
- $html = $item->get_description();\r
- }\r
- if (isset($splink)) {\r
- // Build DOM tree from HTML\r
- $readability = new Readability($html, $url);\r
- $xpath = new DOMXPath($readability->dom);\r
- // Loop through single_page_link xpath expressions\r
- $single_page_url = null;\r
- foreach ($splink as $pattern) {\r
- $elems = @$xpath->evaluate($pattern, $readability->dom);\r
- if (is_string($elems)) {\r
- $single_page_url = trim($elems);\r
- break;\r
- } elseif ($elems instanceof DOMNodeList && $elems->length > 0) {\r
- foreach ($elems as $item) {\r
- if ($item instanceof DOMElement && $item->hasAttribute('href')) {\r
- $single_page_url = $item->getAttribute('href');\r
- break 2;\r
- } elseif ($item instanceof DOMAttr && $item->value) {\r
- $single_page_url = $item->value;\r
- break 2;\r
- }\r
- }\r
- }\r
- }\r
- // If we've got URL, resolve against $url\r
- if (isset($single_page_url) && ($single_page_url = makeAbsoluteStr($url, $single_page_url))) {\r
- // check it's not what we have already!\r
- if ($single_page_url != $url) {\r
- // it's not, so let's try to fetch it...\r
- $_prev_ref = $http->referer;\r
- $http->referer = $single_page_url;\r
- if (($response = $http->get($single_page_url, true)) && $response['status_code'] < 300) {\r
- $http->referer = $_prev_ref;\r
- return $response;\r
- }\r
- $http->referer = $_prev_ref;\r
- }\r
- }\r
- }\r
- return false;\r
-}\r
-\r
-// based on content-type http header, decide what to do\r
-// param: HTTP headers string\r
-// return: array with keys: 'mime', 'type', 'subtype', 'action', 'name'\r
-// e.g. array('mime'=>'image/jpeg', 'type'=>'image', 'subtype'=>'jpeg', 'action'=>'link', 'name'=>'Image')\r
-function get_mime_action_info($headers) {\r
- global $options;\r
- // check if action defined for returned Content-Type\r
- $info = array();\r
- if (preg_match('!^Content-Type:\s*(([-\w]+)/([-\w\+]+))!im', $headers, $match)) {\r
- // look for full mime type (e.g. image/jpeg) or just type (e.g. image)\r
- // match[1] = full mime type, e.g. image/jpeg\r
- // match[2] = first part, e.g. image\r
- // match[3] = last part, e.g. jpeg\r
- $info['mime'] = strtolower(trim($match[1]));\r
- $info['type'] = strtolower(trim($match[2]));\r
- $info['subtype'] = strtolower(trim($match[3]));\r
- foreach (array($info['mime'], $info['type']) as $_mime) {\r
- if (isset($options->content_type_exc[$_mime])) {\r
- $info['action'] = $options->content_type_exc[$_mime]['action'];\r
- $info['name'] = $options->content_type_exc[$_mime]['name'];\r
- break;\r
- }\r
- }\r
- }\r
- return $info;\r
-}\r
-\r
-function remove_url_cruft($url) {\r
- // remove google analytics for the time being\r
- // regex adapted from http://navitronic.co.uk/2010/12/removing-google-analytics-cruft-from-urls/\r
- // https://gist.github.com/758177\r
- return preg_replace('/(\?|\&)utm_[a-z]+=[^\&]+/', '', $url);\r
-}\r
-\r
-function make_substitutions($string) {\r
- if ($string == '') return $string;\r
- global $item, $effective_url;\r
- $string = str_replace('{url}', htmlspecialchars($item->get_permalink()), $string);\r
- $string = str_replace('{effective-url}', htmlspecialchars($effective_url), $string);\r
- return $string;\r
-}\r
-\r
-function get_cache() {\r
- global $options, $valid_key;\r
- static $cache = null;\r
- if ($cache === null) {\r
- $frontendOptions = array(\r
- 'lifetime' => 10*60, // cache lifetime of 10 minutes\r
- 'automatic_serialization' => false,\r
- 'write_control' => false,\r
- 'automatic_cleaning_factor' => $options->cache_cleanup,\r
- 'ignore_user_abort' => false\r
- );\r
- $backendOptions = array(\r
- 'cache_dir' => ($valid_key) ? $options->cache_dir.'/rss-with-key/' : $options->cache_dir.'/rss/', // directory where to put the cache files\r
- 'file_locking' => false,\r
- 'read_control' => true,\r
- 'read_control_type' => 'strlen',\r
- 'hashed_directory_level' => $options->cache_directory_level,\r
- 'hashed_directory_perm' => 0777,\r
- 'cache_file_perm' => 0664,\r
- 'file_name_prefix' => 'ff'\r
- );\r
- // getting a Zend_Cache_Core object\r
- $cache = Zend_Cache::factory('Core', 'File', $frontendOptions, $backendOptions);\r
- }\r
- return $cache;\r
-}\r
-\r
-function debug($msg) {\r
- global $debug_mode;\r
- if ($debug_mode) {\r
- echo '* ',$msg,"\n";\r
- ob_flush();\r
- flush();\r
- }\r
-}
\ No newline at end of file
projects / github / wallabag / wallabag.git / blobdiff