|
|
@@ -14,7 +14,7 @@ namespace Readability;
|
|
|
|
* More information: http://fivefilters.org/content-only/
|
|
|
|
* More information: http://fivefilters.org/content-only/
|
|
|
|
* License: Apache License, Version 2.0
|
|
|
|
* License: Apache License, Version 2.0
|
|
|
|
* Requires: PHP version 5.2.0+
|
|
|
|
* Requires: PHP version 5.2.0+
|
|
|
|
* Date: 2013-08-02
|
|
|
|
* Date: 2013-08-02.
|
|
|
|
*
|
|
|
|
*
|
|
|
|
* Differences between the PHP port and the original
|
|
|
|
* Differences between the PHP port and the original
|
|
|
|
* ------------------------------------------------------
|
|
|
|
* ------------------------------------------------------
|
|
|
@@ -47,11 +47,11 @@ namespace Readability;
|
|
|
|
*/
|
|
|
|
*/
|
|
|
|
class Readability
|
|
|
|
class Readability
|
|
|
|
{
|
|
|
|
{
|
|
|
|
public $version = '1.7.2-without-multi-page';
|
|
|
|
|
|
|
|
public $convertLinksToFootnotes = false;
|
|
|
|
public $convertLinksToFootnotes = false;
|
|
|
|
public $revertForcedParagraphElements = true;
|
|
|
|
public $revertForcedParagraphElements = true;
|
|
|
|
public $articleTitle;
|
|
|
|
public $articleTitle;
|
|
|
|
public $articleContent;
|
|
|
|
public $articleContent;
|
|
|
|
|
|
|
|
public $original_html;
|
|
|
|
public $dom;
|
|
|
|
public $dom;
|
|
|
|
public $url = null; // optional - URL where HTML was retrieved
|
|
|
|
public $url = null; // optional - URL where HTML was retrieved
|
|
|
|
public $lightClean = true; // preserves more content (experimental)
|
|
|
|
public $lightClean = true; // preserves more content (experimental)
|
|
|
@@ -75,7 +75,7 @@ class Readability
|
|
|
|
'divToPElements' => '/<(?:blockquote|code|div|article|footer|aside|img|p|pre|dl|ol|ul)/mi',
|
|
|
|
'divToPElements' => '/<(?:blockquote|code|div|article|footer|aside|img|p|pre|dl|ol|ul)/mi',
|
|
|
|
'killBreaks' => '/(<br\s*\/?>([ \r\n\s]| ?)*)+/',
|
|
|
|
'killBreaks' => '/(<br\s*\/?>([ \r\n\s]| ?)*)+/',
|
|
|
|
'media' => '!//(?:[^\.\?/]+\.)?(?:youtu(?:be)?|soundcloud|dailymotion|vimeo|pornhub|xvideos|twitvid|rutube|viddler)\.(?:com|be|org|net)/!i',
|
|
|
|
'media' => '!//(?:[^\.\?/]+\.)?(?:youtu(?:be)?|soundcloud|dailymotion|vimeo|pornhub|xvideos|twitvid|rutube|viddler)\.(?:com|be|org|net)/!i',
|
|
|
|
'skipFootnoteLink' => '/^\s*(\[?[a-z0-9]{1,2}\]?|^|edit|citation needed)\s*$/i'
|
|
|
|
'skipFootnoteLink' => '/^\s*(\[?[a-z0-9]{1,2}\]?|^|edit|citation needed)\s*$/i',
|
|
|
|
);
|
|
|
|
);
|
|
|
|
public $tidy_config = array(
|
|
|
|
public $tidy_config = array(
|
|
|
|
'tidy-mark' => false,
|
|
|
|
'tidy-mark' => false,
|
|
|
@@ -100,7 +100,7 @@ class Readability
|
|
|
|
// 'merge-spans' => true,
|
|
|
|
// 'merge-spans' => true,
|
|
|
|
'input-encoding' => '????',
|
|
|
|
'input-encoding' => '????',
|
|
|
|
'output-encoding' => 'utf8',
|
|
|
|
'output-encoding' => 'utf8',
|
|
|
|
'hide-comments' => true
|
|
|
|
'hide-comments' => true,
|
|
|
|
);
|
|
|
|
);
|
|
|
|
// raw HTML filters
|
|
|
|
// raw HTML filters
|
|
|
|
protected $pre_filters = array(
|
|
|
|
protected $pre_filters = array(
|
|
|
@@ -110,7 +110,7 @@ class Readability
|
|
|
|
'!<font[^>]*>\s*\[AD\]\s*</font>!is' => '', // HACK: firewall-filtered content
|
|
|
|
'!<font[^>]*>\s*\[AD\]\s*</font>!is' => '', // HACK: firewall-filtered content
|
|
|
|
'!(<br[^>]*>[ \r\n\s]*){2,}!i' => '</p><p>', // HACK: replace linebreaks plus br's with p's
|
|
|
|
'!(<br[^>]*>[ \r\n\s]*){2,}!i' => '</p><p>', // HACK: replace linebreaks plus br's with p's
|
|
|
|
//'!</?noscript>!is' => '', // replace noscripts
|
|
|
|
//'!</?noscript>!is' => '', // replace noscripts
|
|
|
|
'!<(/?)font[^>]*>!is' => '<\\1span>' // replace fonts to spans
|
|
|
|
'!<(/?)font[^>]*>!is' => '<\\1span>', // replace fonts to spans
|
|
|
|
);
|
|
|
|
);
|
|
|
|
// output HTML filters
|
|
|
|
// output HTML filters
|
|
|
|
protected $post_filters = array(
|
|
|
|
protected $post_filters = array(
|
|
|
@@ -120,7 +120,7 @@ class Readability
|
|
|
|
"/\n+/" => "\n", //single newlines cleanup
|
|
|
|
"/\n+/" => "\n", //single newlines cleanup
|
|
|
|
'!<pre[^>]*>\s*<code!is' => '<pre', // modern web...
|
|
|
|
'!<pre[^>]*>\s*<code!is' => '<pre', // modern web...
|
|
|
|
'!</code>\s*</pre>!is' => '</pre>',
|
|
|
|
'!</code>\s*</pre>!is' => '</pre>',
|
|
|
|
'!<[hb]r>!is' => '<\\1 />'
|
|
|
|
'!<[hb]r>!is' => '<\\1 />',
|
|
|
|
);
|
|
|
|
);
|
|
|
|
// flags
|
|
|
|
// flags
|
|
|
|
const FLAG_STRIP_UNLIKELYS = 1;
|
|
|
|
const FLAG_STRIP_UNLIKELYS = 1;
|
|
|
@@ -138,11 +138,12 @@ class Readability
|
|
|
|
const MIN_NODE_LENGTH = 80;
|
|
|
|
const MIN_NODE_LENGTH = 80;
|
|
|
|
const MAX_LINK_DENSITY = 0.25;
|
|
|
|
const MAX_LINK_DENSITY = 0.25;
|
|
|
|
/**
|
|
|
|
/**
|
|
|
|
* Create instance of Readability
|
|
|
|
* Create instance of Readability.
|
|
|
|
|
|
|
|
*
|
|
|
|
* @param string UTF-8 encoded string
|
|
|
|
* @param string UTF-8 encoded string
|
|
|
|
* @param string (optional) URL associated with HTML (for footnotes)
|
|
|
|
* @param string (optional) URL associated with HTML (for footnotes)
|
|
|
|
* @param string (optional) Which parser to use for turning raw HTML into a DOMDocument
|
|
|
|
* @param string (optional) Which parser to use for turning raw HTML into a DOMDocument
|
|
|
|
* @param boolean (optional) Use tidy
|
|
|
|
* @param bool (optional) Use tidy
|
|
|
|
*/
|
|
|
|
*/
|
|
|
|
public function __construct($html, $url = null, $parser = 'libxml', $use_tidy = true)
|
|
|
|
public function __construct($html, $url = null, $parser = 'libxml', $use_tidy = true)
|
|
|
|
{
|
|
|
|
{
|
|
|
@@ -153,9 +154,9 @@ class Readability
|
|
|
|
$this->domainRegExp = '/'.strtr(preg_replace('/www\d*\./', '', parse_url($url, PHP_URL_HOST)), array('.' => '\.')).'/';
|
|
|
|
$this->domainRegExp = '/'.strtr(preg_replace('/www\d*\./', '', parse_url($url, PHP_URL_HOST)), array('.' => '\.')).'/';
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
mb_internal_encoding("UTF-8");
|
|
|
|
mb_internal_encoding('UTF-8');
|
|
|
|
mb_http_output("UTF-8");
|
|
|
|
mb_http_output('UTF-8');
|
|
|
|
mb_regex_encoding("UTF-8");
|
|
|
|
mb_regex_encoding('UTF-8');
|
|
|
|
|
|
|
|
|
|
|
|
// HACK: dirty cleanup to replace some stuff; shouldn't use regexps with HTML but well...
|
|
|
|
// HACK: dirty cleanup to replace some stuff; shouldn't use regexps with HTML but well...
|
|
|
|
if (!$this->flagIsActive(self::FLAG_DISABLE_PREFILTER)) {
|
|
|
|
if (!$this->flagIsActive(self::FLAG_DISABLE_PREFILTER)) {
|
|
|
@@ -169,7 +170,7 @@ class Readability
|
|
|
|
$html = '<html></html>';
|
|
|
|
$html = '<html></html>';
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
/**
|
|
|
|
/*
|
|
|
|
* Use tidy (if it exists).
|
|
|
|
* Use tidy (if it exists).
|
|
|
|
* This fixes problems with some sites which would otherwise trouble DOMDocument's HTML parsing.
|
|
|
|
* This fixes problems with some sites which would otherwise trouble DOMDocument's HTML parsing.
|
|
|
|
* Although sometimes it makes matters worse, which is why there is an option to disable it.
|
|
|
|
* Although sometimes it makes matters worse, which is why there is an option to disable it.
|
|
|
@@ -179,7 +180,7 @@ class Readability
|
|
|
|
$this->debugText .= 'Tidying document'."\n";
|
|
|
|
$this->debugText .= 'Tidying document'."\n";
|
|
|
|
$tidy = tidy_parse_string($html, $this->tidy_config, 'UTF8');
|
|
|
|
$tidy = tidy_parse_string($html, $this->tidy_config, 'UTF8');
|
|
|
|
if (tidy_clean_repair($tidy)) {
|
|
|
|
if (tidy_clean_repair($tidy)) {
|
|
|
|
$original_html = $html;
|
|
|
|
$this->original_html = $html;
|
|
|
|
$this->tidied = true;
|
|
|
|
$this->tidied = true;
|
|
|
|
$html = $tidy->value;
|
|
|
|
$html = $tidy->value;
|
|
|
|
$html = preg_replace('/<html[^>]+>/i', '<html>', $html);
|
|
|
|
$html = preg_replace('/<html[^>]+>/i', '<html>', $html);
|
|
|
@@ -187,20 +188,27 @@ class Readability
|
|
|
|
}
|
|
|
|
}
|
|
|
|
unset($tidy);
|
|
|
|
unset($tidy);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
$html = mb_convert_encoding($html, 'HTML-ENTITIES', "UTF-8");
|
|
|
|
$html = mb_convert_encoding($html, 'HTML-ENTITIES', 'UTF-8');
|
|
|
|
|
|
|
|
|
|
|
|
if (!($parser == 'html5lib' && ($this->dom = \HTML5_Parser::parse($html)))) {
|
|
|
|
if (!($parser == 'html5lib' && ($this->dom = \HTML5_Parser::parse($html)))) {
|
|
|
|
libxml_use_internal_errors(true);
|
|
|
|
libxml_use_internal_errors(true);
|
|
|
|
$this->dom = new \DOMDocument();
|
|
|
|
$this->dom = new \DOMDocument();
|
|
|
|
$this->dom->preserveWhiteSpace = false;
|
|
|
|
$this->dom->preserveWhiteSpace = false;
|
|
|
|
@$this->dom->loadHTML($html, LIBXML_NOBLANKS | LIBXML_COMPACT | LIBXML_NOERROR);
|
|
|
|
|
|
|
|
|
|
|
|
if (PHP_VERSION_ID >= 50400) {
|
|
|
|
|
|
|
|
$this->dom->loadHTML($html, LIBXML_NOBLANKS | LIBXML_COMPACT | LIBXML_NOERROR);
|
|
|
|
|
|
|
|
} else {
|
|
|
|
|
|
|
|
$this->dom->loadHTML($html);
|
|
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
libxml_use_internal_errors(false);
|
|
|
|
libxml_use_internal_errors(false);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
$this->dom->registerNodeClass('DOMElement', 'Readability\JSLikeHTMLElement');
|
|
|
|
$this->dom->registerNodeClass('DOMElement', 'Readability\JSLikeHTMLElement');
|
|
|
|
}
|
|
|
|
}
|
|
|
|
/**
|
|
|
|
/**
|
|
|
|
* Get article title element
|
|
|
|
* Get article title element.
|
|
|
|
|
|
|
|
*
|
|
|
|
* @return DOMElement
|
|
|
|
* @return DOMElement
|
|
|
|
*/
|
|
|
|
*/
|
|
|
|
public function getTitle()
|
|
|
|
public function getTitle()
|
|
|
@@ -208,7 +216,8 @@ class Readability
|
|
|
|
return $this->articleTitle;
|
|
|
|
return $this->articleTitle;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
/**
|
|
|
|
/**
|
|
|
|
* Get article content element
|
|
|
|
* Get article content element.
|
|
|
|
|
|
|
|
*
|
|
|
|
* @return DOMElement
|
|
|
|
* @return DOMElement
|
|
|
|
*/
|
|
|
|
*/
|
|
|
|
public function getContent()
|
|
|
|
public function getContent()
|
|
|
@@ -216,7 +225,8 @@ class Readability
|
|
|
|
return $this->articleContent;
|
|
|
|
return $this->articleContent;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
/**
|
|
|
|
/**
|
|
|
|
* Add pre filter for raw input HTML processing
|
|
|
|
* Add pre filter for raw input HTML processing.
|
|
|
|
|
|
|
|
*
|
|
|
|
* @param string RegExp for replace
|
|
|
|
* @param string RegExp for replace
|
|
|
|
* @param string (optional) Replacer
|
|
|
|
* @param string (optional) Replacer
|
|
|
|
*/
|
|
|
|
*/
|
|
|
@@ -225,7 +235,8 @@ class Readability
|
|
|
|
$this->pre_filters[$filter] = $replacer;
|
|
|
|
$this->pre_filters[$filter] = $replacer;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
/**
|
|
|
|
/**
|
|
|
|
* Add post filter for raw output HTML processing
|
|
|
|
* Add post filter for raw output HTML processing.
|
|
|
|
|
|
|
|
*
|
|
|
|
* @param string RegExp for replace
|
|
|
|
* @param string RegExp for replace
|
|
|
|
* @param string (optional) Replacer
|
|
|
|
* @param string (optional) Replacer
|
|
|
|
*/
|
|
|
|
*/
|
|
|
@@ -243,7 +254,7 @@ class Readability
|
|
|
|
* 4. Replace the current DOM tree with the new one.
|
|
|
|
* 4. Replace the current DOM tree with the new one.
|
|
|
|
* 5. Read peacefully.
|
|
|
|
* 5. Read peacefully.
|
|
|
|
*
|
|
|
|
*
|
|
|
|
* @return boolean true if we found content, false otherwise
|
|
|
|
* @return bool true if we found content, false otherwise
|
|
|
|
*/
|
|
|
|
*/
|
|
|
|
public function init()
|
|
|
|
public function init()
|
|
|
|
{
|
|
|
|
{
|
|
|
@@ -258,7 +269,7 @@ class Readability
|
|
|
|
if ($this->bodyCache == null) {
|
|
|
|
if ($this->bodyCache == null) {
|
|
|
|
$this->bodyCache = '';
|
|
|
|
$this->bodyCache = '';
|
|
|
|
foreach ($bodyElems as $bodyNode) {
|
|
|
|
foreach ($bodyElems as $bodyNode) {
|
|
|
|
$this->bodyCache += $bodyNode->innerHTML;
|
|
|
|
$this->bodyCache .= trim($bodyNode->innerHTML);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
if ($bodyElems->length > 0 && $this->body == null) {
|
|
|
|
if ($bodyElems->length > 0 && $this->body == null) {
|
|
|
@@ -295,7 +306,7 @@ class Readability
|
|
|
|
return $this->success;
|
|
|
|
return $this->success;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
/**
|
|
|
|
/**
|
|
|
|
* Debug
|
|
|
|
* Debug.
|
|
|
|
*/
|
|
|
|
*/
|
|
|
|
protected function dbg($msg) //, $error=false)
|
|
|
|
protected function dbg($msg) //, $error=false)
|
|
|
|
{
|
|
|
|
{
|
|
|
@@ -305,12 +316,12 @@ class Readability
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
/**
|
|
|
|
/**
|
|
|
|
* Dump debug info
|
|
|
|
* Dump debug info.
|
|
|
|
*/
|
|
|
|
*/
|
|
|
|
protected function dump_dbg()
|
|
|
|
protected function dump_dbg()
|
|
|
|
{
|
|
|
|
{
|
|
|
|
if ($this->debug) {
|
|
|
|
if ($this->debug) {
|
|
|
|
openlog("Readability PHP ", LOG_PID | LOG_PERROR, 0);
|
|
|
|
openlog('Readability PHP ', LOG_PID | LOG_PERROR, 0);
|
|
|
|
syslog(6, $this->debugText); // 1 - error 6 - info
|
|
|
|
syslog(6, $this->debugText); // 1 - error 6 - info
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
@@ -318,7 +329,6 @@ class Readability
|
|
|
|
* Run any post-process modifications to article content as necessary.
|
|
|
|
* Run any post-process modifications to article content as necessary.
|
|
|
|
*
|
|
|
|
*
|
|
|
|
* @param DOMElement
|
|
|
|
* @param DOMElement
|
|
|
|
* @return void
|
|
|
|
|
|
|
|
*/
|
|
|
|
*/
|
|
|
|
public function postProcessContent($articleContent)
|
|
|
|
public function postProcessContent($articleContent)
|
|
|
|
{
|
|
|
|
{
|
|
|
@@ -337,7 +347,8 @@ class Readability
|
|
|
|
$origTitle = '';
|
|
|
|
$origTitle = '';
|
|
|
|
try {
|
|
|
|
try {
|
|
|
|
$curTitle = $origTitle = $this->getInnerText($this->dom->getElementsByTagName('title')->item(0));
|
|
|
|
$curTitle = $origTitle = $this->getInnerText($this->dom->getElementsByTagName('title')->item(0));
|
|
|
|
} catch (Exception $e) {}
|
|
|
|
} catch (Exception $e) {
|
|
|
|
|
|
|
|
}
|
|
|
|
if (preg_match('/ [\|\-] /', $curTitle)) {
|
|
|
|
if (preg_match('/ [\|\-] /', $curTitle)) {
|
|
|
|
$curTitle = preg_replace('/(.*)[\|\-] .*/i', '$1', $origTitle);
|
|
|
|
$curTitle = preg_replace('/(.*)[\|\-] .*/i', '$1', $origTitle);
|
|
|
|
if (count(explode(' ', $curTitle)) < 3) {
|
|
|
|
if (count(explode(' ', $curTitle)) < 3) {
|
|
|
@@ -366,12 +377,10 @@ class Readability
|
|
|
|
/**
|
|
|
|
/**
|
|
|
|
* Prepare the HTML document for readability to scrape it.
|
|
|
|
* Prepare the HTML document for readability to scrape it.
|
|
|
|
* This includes things like stripping javascript, CSS, and handling terrible markup.
|
|
|
|
* This includes things like stripping javascript, CSS, and handling terrible markup.
|
|
|
|
*
|
|
|
|
|
|
|
|
* @return void
|
|
|
|
|
|
|
|
*/
|
|
|
|
*/
|
|
|
|
protected function prepDocument()
|
|
|
|
protected function prepDocument()
|
|
|
|
{
|
|
|
|
{
|
|
|
|
/**
|
|
|
|
/*
|
|
|
|
* In some cases a body element can't be found (if the HTML is totally hosed for example)
|
|
|
|
* In some cases a body element can't be found (if the HTML is totally hosed for example)
|
|
|
|
* so we create a new body node and append it to the document.
|
|
|
|
* so we create a new body node and append it to the document.
|
|
|
|
*/
|
|
|
|
*/
|
|
|
@@ -392,9 +401,8 @@ class Readability
|
|
|
|
}
|
|
|
|
}
|
|
|
|
/**
|
|
|
|
/**
|
|
|
|
* For easier reading, convert this document to have footnotes at the bottom rather than inline links.
|
|
|
|
* For easier reading, convert this document to have footnotes at the bottom rather than inline links.
|
|
|
|
* @see http://www.roughtype.com/archives/2010/05/experiments_in.php
|
|
|
|
|
|
|
|
*
|
|
|
|
*
|
|
|
|
* @return void
|
|
|
|
* @see http://www.roughtype.com/archives/2010/05/experiments_in.php
|
|
|
|
*/
|
|
|
|
*/
|
|
|
|
public function addFootnotes($articleContent)
|
|
|
|
public function addFootnotes($articleContent)
|
|
|
|
{
|
|
|
|
{
|
|
|
@@ -450,7 +458,6 @@ class Readability
|
|
|
|
* iframes, forms, strip extraneous <p> tags, etc.
|
|
|
|
* iframes, forms, strip extraneous <p> tags, etc.
|
|
|
|
*
|
|
|
|
*
|
|
|
|
* @param DOMElement
|
|
|
|
* @param DOMElement
|
|
|
|
* @return void
|
|
|
|
|
|
|
|
*/
|
|
|
|
*/
|
|
|
|
public function prepArticle($articleContent)
|
|
|
|
public function prepArticle($articleContent)
|
|
|
|
{
|
|
|
|
{
|
|
|
@@ -463,7 +470,7 @@ class Readability
|
|
|
|
$this->killBreaks($articleContent);
|
|
|
|
$this->killBreaks($articleContent);
|
|
|
|
$xpath = new \DOMXPath($articleContent->ownerDocument);
|
|
|
|
$xpath = new \DOMXPath($articleContent->ownerDocument);
|
|
|
|
if ($this->revertForcedParagraphElements) {
|
|
|
|
if ($this->revertForcedParagraphElements) {
|
|
|
|
/**
|
|
|
|
/*
|
|
|
|
* Reverts P elements with class 'readability-styled' to text nodes:
|
|
|
|
* Reverts P elements with class 'readability-styled' to text nodes:
|
|
|
|
* which is what they were before.
|
|
|
|
* which is what they were before.
|
|
|
|
*/
|
|
|
|
*/
|
|
|
@@ -493,7 +500,7 @@ class Readability
|
|
|
|
$this->clean($articleContent, 'canvas');
|
|
|
|
$this->clean($articleContent, 'canvas');
|
|
|
|
$this->clean($articleContent, 'h1');
|
|
|
|
$this->clean($articleContent, 'h1');
|
|
|
|
|
|
|
|
|
|
|
|
/**
|
|
|
|
/*
|
|
|
|
* If there is only one h2, they are probably using it as a main header, so remove it since we
|
|
|
|
* If there is only one h2, they are probably using it as a main header, so remove it since we
|
|
|
|
* already have a header.
|
|
|
|
* already have a header.
|
|
|
|
*/
|
|
|
|
*/
|
|
|
@@ -536,7 +543,7 @@ class Readability
|
|
|
|
}
|
|
|
|
}
|
|
|
|
unset($search, $replace);
|
|
|
|
unset($search, $replace);
|
|
|
|
} catch (Exception $e) {
|
|
|
|
} catch (Exception $e) {
|
|
|
|
$this->dbg("Cleaning output HTML failed. Ignoring: " . $e->getMessage());
|
|
|
|
$this->dbg('Cleaning output HTML failed. Ignoring: '.$e->getMessage());
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
@@ -545,7 +552,6 @@ class Readability
|
|
|
|
* className/id for special names to add to its score.
|
|
|
|
* className/id for special names to add to its score.
|
|
|
|
*
|
|
|
|
*
|
|
|
|
* @param Element
|
|
|
|
* @param Element
|
|
|
|
* @return void
|
|
|
|
|
|
|
|
*/
|
|
|
|
*/
|
|
|
|
protected function initializeNode($node)
|
|
|
|
protected function initializeNode($node)
|
|
|
|
{
|
|
|
|
{
|
|
|
@@ -667,7 +673,7 @@ class Readability
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
/**
|
|
|
|
/*
|
|
|
|
* Loop through all paragraphs, and assign a score to them based on how content-y they look.
|
|
|
|
* 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.
|
|
|
|
* Then add their score to their parent node.
|
|
|
|
*
|
|
|
|
*
|
|
|
@@ -723,7 +729,7 @@ class Readability
|
|
|
|
$grandParentNode->getAttributeNode('readability')->value += $contentScore / self::GRANDPARENT_SCORE_DIVISOR;
|
|
|
|
$grandParentNode->getAttributeNode('readability')->value += $contentScore / self::GRANDPARENT_SCORE_DIVISOR;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
/**
|
|
|
|
/*
|
|
|
|
* Node prepping: trash nodes that look cruddy (like ones with the class name "comment", etc).
|
|
|
|
* Node prepping: trash nodes that look cruddy (like ones with the class name "comment", etc).
|
|
|
|
* This is faster to do before scoring but safer after.
|
|
|
|
* This is faster to do before scoring but safer after.
|
|
|
|
*/
|
|
|
|
*/
|
|
|
@@ -742,7 +748,7 @@ class Readability
|
|
|
|
$node = $candidates->item($c);
|
|
|
|
$node = $candidates->item($c);
|
|
|
|
$tagName = $node->tagName;
|
|
|
|
$tagName = $node->tagName;
|
|
|
|
/* Remove unlikely candidates */
|
|
|
|
/* Remove unlikely candidates */
|
|
|
|
$unlikelyMatchString = $node->getAttribute('class')." ".$node->getAttribute('id')." ".$node->getAttribute('style');
|
|
|
|
$unlikelyMatchString = $node->getAttribute('class').' '.$node->getAttribute('id').' '.$node->getAttribute('style');
|
|
|
|
//$this->dbg('Processing '.$node->getNodePath().' by "'. $unlikelyMatchString.'" with readability '.($node->hasAttribute('readability') ? (int)$node->getAttributeNode('readability')->value : 0));
|
|
|
|
//$this->dbg('Processing '.$node->getNodePath().' by "'. $unlikelyMatchString.'" with readability '.($node->hasAttribute('readability') ? (int)$node->getAttributeNode('readability')->value : 0));
|
|
|
|
if (mb_strlen($unlikelyMatchString) > 3 && // don't process "empty" strings
|
|
|
|
if (mb_strlen($unlikelyMatchString) > 3 && // don't process "empty" strings
|
|
|
|
preg_match($this->regexps['unlikelyCandidates'], $unlikelyMatchString) &&
|
|
|
|
preg_match($this->regexps['unlikelyCandidates'], $unlikelyMatchString) &&
|
|
|
@@ -755,7 +761,7 @@ class Readability
|
|
|
|
}
|
|
|
|
}
|
|
|
|
unset($candidates);
|
|
|
|
unset($candidates);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
/**
|
|
|
|
/*
|
|
|
|
* After we've calculated scores, loop through all of the possible candidate nodes we found
|
|
|
|
* After we've calculated scores, loop through all of the possible candidate nodes we found
|
|
|
|
* and find the one with the highest score.
|
|
|
|
* and find the one with the highest score.
|
|
|
|
*/
|
|
|
|
*/
|
|
|
@@ -776,7 +782,7 @@ class Readability
|
|
|
|
}
|
|
|
|
}
|
|
|
|
unset($candidates);
|
|
|
|
unset($candidates);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
/**
|
|
|
|
/*
|
|
|
|
* If we still have no top candidate, just use the body as a last resort.
|
|
|
|
* 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.
|
|
|
|
* We also have to copy the body node so it is something we can modify.
|
|
|
|
*/
|
|
|
|
*/
|
|
|
@@ -811,7 +817,7 @@ class Readability
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
$this->dbg('Top candidate: '.$topCandidate->getNodePath());
|
|
|
|
$this->dbg('Top candidate: '.$topCandidate->getNodePath());
|
|
|
|
/**
|
|
|
|
/*
|
|
|
|
* Now that we have the top candidate, look through its siblings for content that might also be related.
|
|
|
|
* 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.
|
|
|
|
* Things like preambles, content split by ads that we removed, etc.
|
|
|
|
*/
|
|
|
|
*/
|
|
|
@@ -883,7 +889,7 @@ class Readability
|
|
|
|
unset($xpath);
|
|
|
|
unset($xpath);
|
|
|
|
// So we have all of the content that we need. Now we clean it up for presentation.
|
|
|
|
// So we have all of the content that we need. Now we clean it up for presentation.
|
|
|
|
$this->prepArticle($articleContent);
|
|
|
|
$this->prepArticle($articleContent);
|
|
|
|
/**
|
|
|
|
/*
|
|
|
|
* Now that we've gone through the full algorithm, check to see if we got any meaningful content.
|
|
|
|
* 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
|
|
|
|
* 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
|
|
|
|
* likelihood of finding the content, and the sieve approach gives us a higher likelihood of
|
|
|
@@ -896,17 +902,17 @@ class Readability
|
|
|
|
$this->body->innerHTML = $this->bodyCache;
|
|
|
|
$this->body->innerHTML = $this->bodyCache;
|
|
|
|
if ($this->flagIsActive(self::FLAG_STRIP_UNLIKELYS)) {
|
|
|
|
if ($this->flagIsActive(self::FLAG_STRIP_UNLIKELYS)) {
|
|
|
|
$this->removeFlag(self::FLAG_STRIP_UNLIKELYS);
|
|
|
|
$this->removeFlag(self::FLAG_STRIP_UNLIKELYS);
|
|
|
|
$this->dbg("...content is shorter than ".self::MIN_ARTICLE_LENGTH." letters, trying not to strip unlikely content.\n");
|
|
|
|
$this->dbg('...content is shorter than '.self::MIN_ARTICLE_LENGTH." letters, trying not to strip unlikely content.\n");
|
|
|
|
|
|
|
|
|
|
|
|
return $this->grabArticle($this->body);
|
|
|
|
return $this->grabArticle($this->body);
|
|
|
|
} elseif ($this->flagIsActive(self::FLAG_WEIGHT_ATTRIBUTES)) {
|
|
|
|
} elseif ($this->flagIsActive(self::FLAG_WEIGHT_ATTRIBUTES)) {
|
|
|
|
$this->removeFlag(self::FLAG_WEIGHT_ATTRIBUTES);
|
|
|
|
$this->removeFlag(self::FLAG_WEIGHT_ATTRIBUTES);
|
|
|
|
$this->dbg("...content is shorter than ".self::MIN_ARTICLE_LENGTH." letters, trying not to weight attributes.\n");
|
|
|
|
$this->dbg('...content is shorter than '.self::MIN_ARTICLE_LENGTH." letters, trying not to weight attributes.\n");
|
|
|
|
|
|
|
|
|
|
|
|
return $this->grabArticle($this->body);
|
|
|
|
return $this->grabArticle($this->body);
|
|
|
|
} elseif ($this->flagIsActive(self::FLAG_CLEAN_CONDITIONALLY)) {
|
|
|
|
} elseif ($this->flagIsActive(self::FLAG_CLEAN_CONDITIONALLY)) {
|
|
|
|
$this->removeFlag(self::FLAG_CLEAN_CONDITIONALLY);
|
|
|
|
$this->removeFlag(self::FLAG_CLEAN_CONDITIONALLY);
|
|
|
|
$this->dbg("...content is shorter than ".self::MIN_ARTICLE_LENGTH." letters, trying not to clean at all.\n");
|
|
|
|
$this->dbg('...content is shorter than '.self::MIN_ARTICLE_LENGTH." letters, trying not to clean at all.\n");
|
|
|
|
|
|
|
|
|
|
|
|
return $this->grabArticle($this->body);
|
|
|
|
return $this->grabArticle($this->body);
|
|
|
|
} else {
|
|
|
|
} else {
|
|
|
@@ -921,8 +927,9 @@ class Readability
|
|
|
|
* This also strips out any excess whitespace to be found.
|
|
|
|
* This also strips out any excess whitespace to be found.
|
|
|
|
*
|
|
|
|
*
|
|
|
|
* @param DOMElement $e
|
|
|
|
* @param DOMElement $e
|
|
|
|
* @param boolean $normalizeSpaces (default: true)
|
|
|
|
* @param bool $normalizeSpaces (default: true)
|
|
|
|
* @param boolean $flattenLines (default: false)
|
|
|
|
* @param bool $flattenLines (default: false)
|
|
|
|
|
|
|
|
*
|
|
|
|
* @return string
|
|
|
|
* @return string
|
|
|
|
*/
|
|
|
|
*/
|
|
|
|
public function getInnerText($e, $normalizeSpaces = true, $flattenLines = false)
|
|
|
|
public function getInnerText($e, $normalizeSpaces = true, $flattenLines = false)
|
|
|
@@ -943,7 +950,6 @@ class Readability
|
|
|
|
* Remove the style attribute on every $e and under.
|
|
|
|
* Remove the style attribute on every $e and under.
|
|
|
|
*
|
|
|
|
*
|
|
|
|
* @param DOMElement $e
|
|
|
|
* @param DOMElement $e
|
|
|
|
* @return void
|
|
|
|
|
|
|
|
*/
|
|
|
|
*/
|
|
|
|
public function cleanStyles($e)
|
|
|
|
public function cleanStyles($e)
|
|
|
|
{
|
|
|
|
{
|
|
|
@@ -959,6 +965,7 @@ class Readability
|
|
|
|
* Get comma number for a given text.
|
|
|
|
* Get comma number for a given text.
|
|
|
|
*
|
|
|
|
*
|
|
|
|
* @param string $text
|
|
|
|
* @param string $text
|
|
|
|
|
|
|
|
*
|
|
|
|
* @return number (integer)
|
|
|
|
* @return number (integer)
|
|
|
|
*/
|
|
|
|
*/
|
|
|
|
public function getCommaCount($text)
|
|
|
|
public function getCommaCount($text)
|
|
|
@@ -970,6 +977,7 @@ class Readability
|
|
|
|
* Input string should be normalized.
|
|
|
|
* Input string should be normalized.
|
|
|
|
*
|
|
|
|
*
|
|
|
|
* @param string $text
|
|
|
|
* @param string $text
|
|
|
|
|
|
|
|
*
|
|
|
|
* @return number (integer)
|
|
|
|
* @return number (integer)
|
|
|
|
*/
|
|
|
|
*/
|
|
|
|
public function getWordCount($text)
|
|
|
|
public function getWordCount($text)
|
|
|
@@ -983,6 +991,7 @@ class Readability
|
|
|
|
*
|
|
|
|
*
|
|
|
|
* @param DOMElement $e
|
|
|
|
* @param DOMElement $e
|
|
|
|
* @param string $excludeExternal
|
|
|
|
* @param string $excludeExternal
|
|
|
|
|
|
|
|
*
|
|
|
|
* @return number (float)
|
|
|
|
* @return number (float)
|
|
|
|
*/
|
|
|
|
*/
|
|
|
|
public function getLinkDensity($e, $excludeExternal = false)
|
|
|
|
public function getLinkDensity($e, $excludeExternal = false)
|
|
|
@@ -1008,6 +1017,7 @@ class Readability
|
|
|
|
*
|
|
|
|
*
|
|
|
|
* @param DOMElement $element
|
|
|
|
* @param DOMElement $element
|
|
|
|
* @param string $attribute
|
|
|
|
* @param string $attribute
|
|
|
|
|
|
|
|
*
|
|
|
|
* @return number (Integer)
|
|
|
|
* @return number (Integer)
|
|
|
|
*/
|
|
|
|
*/
|
|
|
|
protected function weightAttribute($element, $attribute)
|
|
|
|
protected function weightAttribute($element, $attribute)
|
|
|
@@ -1039,6 +1049,7 @@ class Readability
|
|
|
|
* Get an element relative weight.
|
|
|
|
* Get an element relative weight.
|
|
|
|
*
|
|
|
|
*
|
|
|
|
* @param DOMElement $e
|
|
|
|
* @param DOMElement $e
|
|
|
|
|
|
|
|
*
|
|
|
|
* @return number (Integer)
|
|
|
|
* @return number (Integer)
|
|
|
|
*/
|
|
|
|
*/
|
|
|
|
public function getWeight($e)
|
|
|
|
public function getWeight($e)
|
|
|
@@ -1058,7 +1069,6 @@ class Readability
|
|
|
|
* Remove extraneous break tags from a node.
|
|
|
|
* Remove extraneous break tags from a node.
|
|
|
|
*
|
|
|
|
*
|
|
|
|
* @param DOMElement $node
|
|
|
|
* @param DOMElement $node
|
|
|
|
* @return void
|
|
|
|
|
|
|
|
*/
|
|
|
|
*/
|
|
|
|
public function killBreaks($node)
|
|
|
|
public function killBreaks($node)
|
|
|
|
{
|
|
|
|
{
|
|
|
@@ -1068,13 +1078,12 @@ class Readability
|
|
|
|
}
|
|
|
|
}
|
|
|
|
/**
|
|
|
|
/**
|
|
|
|
* Clean a node of all elements of type "tag".
|
|
|
|
* Clean a node of all elements of type "tag".
|
|
|
|
* (Unless it's a youtube/vimeo video. People love movies.)
|
|
|
|
* (Unless it's a youtube/vimeo video. People love movies.).
|
|
|
|
*
|
|
|
|
*
|
|
|
|
* Updated 2012-09-18 to preserve youtube/vimeo iframes
|
|
|
|
* Updated 2012-09-18 to preserve youtube/vimeo iframes
|
|
|
|
*
|
|
|
|
*
|
|
|
|
* @param DOMElement $e
|
|
|
|
* @param DOMElement $e
|
|
|
|
* @param string $tag
|
|
|
|
* @param string $tag
|
|
|
|
* @return void
|
|
|
|
|
|
|
|
*/
|
|
|
|
*/
|
|
|
|
public function clean($e, $tag)
|
|
|
|
public function clean($e, $tag)
|
|
|
|
{
|
|
|
|
{
|
|
|
@@ -1104,7 +1113,6 @@ class Readability
|
|
|
|
*
|
|
|
|
*
|
|
|
|
* @param DOMElement $e
|
|
|
|
* @param DOMElement $e
|
|
|
|
* @param string $tag
|
|
|
|
* @param string $tag
|
|
|
|
* @return void
|
|
|
|
|
|
|
|
*/
|
|
|
|
*/
|
|
|
|
public function cleanConditionally($e, $tag)
|
|
|
|
public function cleanConditionally($e, $tag)
|
|
|
|
{
|
|
|
|
{
|
|
|
@@ -1113,7 +1121,7 @@ class Readability
|
|
|
|
}
|
|
|
|
}
|
|
|
|
$tagsList = $e->getElementsByTagName($tag);
|
|
|
|
$tagsList = $e->getElementsByTagName($tag);
|
|
|
|
$curTagsLength = $tagsList->length;
|
|
|
|
$curTagsLength = $tagsList->length;
|
|
|
|
/**
|
|
|
|
/*
|
|
|
|
* Gather counts for other typical elements embedded within.
|
|
|
|
* Gather counts for other typical elements embedded within.
|
|
|
|
* Traverse backwards so we can remove nodes at the same time without effecting the traversal.
|
|
|
|
* Traverse backwards so we can remove nodes at the same time without effecting the traversal.
|
|
|
|
*
|
|
|
|
*
|
|
|
@@ -1129,7 +1137,7 @@ class Readability
|
|
|
|
$this->dbg('Removing...');
|
|
|
|
$this->dbg('Removing...');
|
|
|
|
$node->parentNode->removeChild($node);
|
|
|
|
$node->parentNode->removeChild($node);
|
|
|
|
} elseif ($this->getCommaCount($this->getInnerText($node)) < self::MIN_COMMAS_IN_PARAGRAPH) {
|
|
|
|
} elseif ($this->getCommaCount($this->getInnerText($node)) < self::MIN_COMMAS_IN_PARAGRAPH) {
|
|
|
|
/**
|
|
|
|
/*
|
|
|
|
* If there are not very many commas, and the number of
|
|
|
|
* 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.
|
|
|
|
* non-paragraph elements is more than paragraphs or other ominous signs, remove the element.
|
|
|
|
*/
|
|
|
|
*/
|
|
|
@@ -1165,10 +1173,10 @@ class Readability
|
|
|
|
$this->dbg(' content length less than 6 chars, 0 embeds and either 0 images or more than 2 images');
|
|
|
|
$this->dbg(' content length less than 6 chars, 0 embeds and either 0 images or more than 2 images');
|
|
|
|
$toRemove = true;
|
|
|
|
$toRemove = true;
|
|
|
|
} elseif ($weight < 25 && $linkDensity > 0.25) {
|
|
|
|
} elseif ($weight < 25 && $linkDensity > 0.25) {
|
|
|
|
$this->dbg(' weight is '.$weight.' < 25 and link density is '.sprintf("%.2f", $linkDensity).' > 0.25');
|
|
|
|
$this->dbg(' weight is '.$weight.' < 25 and link density is '.sprintf('%.2f', $linkDensity).' > 0.25');
|
|
|
|
$toRemove = true;
|
|
|
|
$toRemove = true;
|
|
|
|
} elseif ($a > 2 && ($weight >= 25 && $linkDensity > 0.5)) {
|
|
|
|
} elseif ($a > 2 && ($weight >= 25 && $linkDensity > 0.5)) {
|
|
|
|
$this->dbg(' more than 2 links and weight is '.$weight.' > 25 but link density is '.sprintf("%.2f", $linkDensity).' > 0.5');
|
|
|
|
$this->dbg(' more than 2 links and weight is '.$weight.' > 25 but link density is '.sprintf('%.2f', $linkDensity).' > 0.5');
|
|
|
|
$toRemove = true;
|
|
|
|
$toRemove = true;
|
|
|
|
} elseif ($embedCount > 3) {
|
|
|
|
} elseif ($embedCount > 3) {
|
|
|
|
$this->dbg(' more than 3 embeds');
|
|
|
|
$this->dbg(' more than 3 embeds');
|
|
|
@@ -1188,10 +1196,10 @@ class Readability
|
|
|
|
$this->dbg(' content length less than 25 chars and 0 images, or more than 2 images');
|
|
|
|
$this->dbg(' content length less than 25 chars and 0 images, or more than 2 images');
|
|
|
|
$toRemove = true;
|
|
|
|
$toRemove = true;
|
|
|
|
} elseif ($weight < 25 && $linkDensity > 0.2) {
|
|
|
|
} elseif ($weight < 25 && $linkDensity > 0.2) {
|
|
|
|
$this->dbg(' weight is '.$weight.' lower than 0 and link density is '.sprintf("%.2f", $linkDensity).' > 0.2');
|
|
|
|
$this->dbg(' weight is '.$weight.' lower than 0 and link density is '.sprintf('%.2f', $linkDensity).' > 0.2');
|
|
|
|
$toRemove = true;
|
|
|
|
$toRemove = true;
|
|
|
|
} elseif ($weight >= 25 && $linkDensity > 0.5) {
|
|
|
|
} elseif ($weight >= 25 && $linkDensity > 0.5) {
|
|
|
|
$this->dbg(' weight above 25 but link density is '.sprintf("%.2f", $linkDensity).' > 0.5');
|
|
|
|
$this->dbg(' weight above 25 but link density is '.sprintf('%.2f', $linkDensity).' > 0.5');
|
|
|
|
$toRemove = true;
|
|
|
|
$toRemove = true;
|
|
|
|
} elseif (($embedCount == 1 && $contentLength < 75) || $embedCount > 1) {
|
|
|
|
} elseif (($embedCount == 1 && $contentLength < 75) || $embedCount > 1) {
|
|
|
|
$this->dbg(' 1 embed and content length smaller than 75 chars, or more than one embed');
|
|
|
|
$this->dbg(' 1 embed and content length smaller than 75 chars, or more than one embed');
|
|
|
@@ -1210,7 +1218,6 @@ class Readability
|
|
|
|
* Clean out spurious headers from an Element. Checks things like classnames and link density.
|
|
|
|
* Clean out spurious headers from an Element. Checks things like classnames and link density.
|
|
|
|
*
|
|
|
|
*
|
|
|
|
* @param DOMElement $e
|
|
|
|
* @param DOMElement $e
|
|
|
|
* @return void
|
|
|
|
|
|
|
|
*/
|
|
|
|
*/
|
|
|
|
public function cleanHeaders($e)
|
|
|
|
public function cleanHeaders($e)
|
|
|
|
{
|
|
|
|
{
|
|
|
|
Jeremy Benoist