Do not remove html tag attributes

They might contains useful information (at least language)
Merge pull request #8 from j0k3r/nofollow
2026-09-27 06:26:17 +00:00 · 2015-09-23 21:09:38 +02:00 · 2015-09-23 17:58:32 +02:00 · 2015-09-22 19:25:57 +02:00 · 2015-09-20 21:05:55 +02:00 · 2015-09-18 19:19:26 +02:00
4 changed files with 341 additions and 64 deletions
@@ -3,7 +3,7 @@
 [![Build Status](https://travis-ci.org/j0k3r/php-readability.svg?branch=master)](https://travis-ci.org/j0k3r/php-readability)
 [![Code Coverage](https://scrutinizer-ci.com/g/j0k3r/php-readability/badges/coverage.png?b=master)](https://scrutinizer-ci.com/g/j0k3r/php-readability/?branch=master)
-This is an extract of the Readability class from the [full-text-rss](https://github.com/Dither/full-text-rss) fork. It kind be defined as a better version of the original [php-readability](http://code.fivefilters.org/php-readability).
+This is an extract of the Readability class from the [full-text-rss](https://github.com/Dither/full-text-rss) fork. It kind be defined as a better version of the original [php-readability](https://bitbucket.org/fivefilters/php-readability/overview).
 ## Differences
@@ -47,12 +47,14 @@ class JSLikeHTMLElement extends \DOMElement
    {
        if ($name == 'innerHTML') {
            // first, empty the element
-            for ($x = $this->childNodes->length - 1; $x >= 0; $x--) {
+            for ($x = $this->childNodes->length - 1; $x >= 0; --$x) {
                $this->removeChild($this->childNodes->item($x));
            }
            // $value holds our new inner HTML
            if ($value != '') {
                $f = $this->ownerDocument->createDocumentFragment();
                // appendXML() expects well-formed markup (XHTML)
                $result = @$f->appendXML($value); // @ to suppress PHP warnings
                if ($result) {
@@ -63,12 +65,14 @@ class JSLikeHTMLElement extends \DOMElement
                    // $value is probably ill-formed
                    $f = new \DOMDocument();
                    $value = mb_convert_encoding($value, 'HTML-ENTITIES', 'UTF-8');
                    // Using <htmlfragment> will generate a warning, but so will bad HTML
                    // (and by this point, bad HTML is what we've got).
                    // We use it (and suppress the warning) because an HTML fragment will
                    // be wrapped around <html><body> tags which we don't really want to keep.
                    // Note: despite the warning, if loadHTML succeeds it will return true.
                    $result = @$f->loadHTML('<htmlfragment>'.$value.'</htmlfragment>');
                    if ($result) {
                        $import = $f->getElementsByTagName('htmlfragment')->item(0);
                        foreach ($import->childNodes as $child) {
@@ -63,6 +63,7 @@ class Readability
    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 processing 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.
@@ -122,6 +123,7 @@ class Readability
        '!</code>\s*</pre>!is' => '</pre>',
        '!<[hb]r>!is' => '<\\1 />',
    );
    // flags
    const FLAG_STRIP_UNLIKELYS = 1;
    const FLAG_WEIGHT_ATTRIBUTES = 2;
@@ -137,6 +139,7 @@ class Readability
    const MIN_ARTICLE_LENGTH = 200;
    const MIN_NODE_LENGTH = 80;
    const MAX_LINK_DENSITY = 0.25;
    /**
     * Create instance of Readability.
     *
@@ -183,15 +186,16 @@ class Readability
                $this->original_html = $html;
                $this->tidied = true;
                $html = $tidy->value;
                $html = preg_replace('/<html[^>]+>/i', '<html>', $html);
                $html = preg_replace('/[\r\n]+/is', "\n", $html);
            }
            unset($tidy);
        }
        $html = mb_convert_encoding($html, 'HTML-ENTITIES', 'UTF-8');
        if (!($parser == 'html5lib' && ($this->dom = \HTML5_Parser::parse($html)))) {
            libxml_use_internal_errors(true);
            $this->dom = new \DOMDocument();
            $this->dom->preserveWhiteSpace = false;
@@ -206,6 +210,7 @@ class Readability
        $this->dom->registerNodeClass('DOMElement', 'Readability\JSLikeHTMLElement');
    }
    /**
     * Get article title element.
     *
@@ -215,6 +220,7 @@ class Readability
    {
        return $this->articleTitle;
    }
    /**
     * Get article content element.
     *
@@ -224,6 +230,7 @@ class Readability
    {
        return $this->articleContent;
    }
    /**
     * Add pre filter for raw input HTML processing.
     *
@@ -234,6 +241,7 @@ class Readability
    {
        $this->pre_filters[$filter] = $replacer;
    }
    /**
     * Add post filter for raw output HTML processing.
     *
@@ -244,6 +252,7 @@ class Readability
    {
        $this->post_filters[$filter] = $replacer;
    }
    /**
     * Runs readability.
     *
@@ -265,6 +274,7 @@ class Readability
        // Assume successful outcome
        $this->success = true;
        $bodyElems = $this->dom->getElementsByTagName('body');
        // WTF multiple body nodes?
        if ($this->bodyCache == null) {
            $this->bodyCache = '';
@@ -272,32 +282,40 @@ class Readability
                $this->bodyCache .= trim($bodyNode->innerHTML);
            }
        }
        if ($bodyElems->length > 0 && $this->body == null) {
            $this->body = $bodyElems->item(0);
        }
        $this->prepDocument();
        // Build readability's DOM tree.
        $overlay = $this->dom->createElement('div');
        $innerDiv = $this->dom->createElement('div');
-        $articleTitle   = $this->getArticleTitle();
+        $articleTitle = $this->getArticleTitle();
        $articleContent = $this->grabArticle();
        if (!$articleContent) {
            $this->success = false;
            $articleContent = $this->dom->createElement('div');
-            $articleContent->setAttribute('id', 'readability-content');
+            $articleContent->setAttribute('class', 'readability-content');
            $articleContent->innerHTML = '<p>Sorry, Readability was unable to parse this page for content.</p>';
        }
-        $overlay->setAttribute('id', 'readOverlay');
+
-        $innerDiv->setAttribute('id', 'readInner');
+        $overlay->setAttribute('class', 'readOverlay');
        $innerDiv->setAttribute('class', '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);
        $this->body->removeAttribute('style');
        $this->postProcessContent($articleContent);
        // Set title and content instance variables.
        $this->articleTitle = $articleTitle;
        $this->articleContent = $articleContent;
@@ -305,6 +323,7 @@ class Readability
        return $this->success;
    }
    /**
     * Debug.
     */
@@ -325,6 +344,7 @@ class Readability
            syslog(6, $this->debugText); // 1 - error 6 - info
        }
    }
    /**
     * Run any post-process modifications to article content as necessary.
     *
@@ -336,6 +356,7 @@ class Readability
            $this->addFootnotes($articleContent);
        }
    }
    /**
     * Get the article title as an H1.
     *
@@ -345,10 +366,12 @@ class Readability
    {
        $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) {
@@ -365,15 +388,18 @@ class Readability
                $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.
@@ -388,17 +414,21 @@ class Readability
            $this->body = $this->dom->createElement('body');
            $this->dom->documentElement->appendChild($this->body);
        }
-        $this->body->setAttribute('id', 'readabilityBody');
+
        $this->body->setAttribute('class', 'readabilityBody');
        // Remove all style tags in head.
        $styleTags = $this->dom->getElementsByTagName('style');
-        for ($i = $styleTags->length - 1; $i >= 0; $i--) {
+        for ($i = $styleTags->length - 1; $i >= 0; --$i) {
            $styleTags->item($i)->parentNode->removeChild($styleTags->item($i));
        }
        $linkTags = $this->dom->getElementsByTagName('link');
-        for ($i = $linkTags->length - 1; $i >= 0; $i--) {
+        for ($i = $linkTags->length - 1; $i >= 0; --$i) {
            $linkTags->item($i)->parentNode->removeChild($linkTags->item($i));
        }
    }
    /**
     * For easier reading, convert this document to have footnotes at the bottom rather than inline links.
     *
@@ -407,52 +437,61 @@ class Readability
    public function addFootnotes($articleContent)
    {
        $footnotesWrapper = $this->dom->createElement('footer');
-        $footnotesWrapper->setAttribute('id', 'readability-footnotes');
+        $footnotesWrapper->setAttribute('class', 'readability-footnotes');
        $footnotesWrapper->innerHTML = '<h3>References</h3>';
        $articleFootnotes = $this->dom->createElement('ol');
-        $articleFootnotes->setAttribute('id', 'readability-footnotes-list');
+        $articleFootnotes->setAttribute('class', 'readability-footnotes-list');
        $footnotesWrapper->appendChild($articleFootnotes);
        $articleLinks = $articleContent->getElementsByTagName('a');
        $linkCount = 0;
-        for ($i = 0; $i < $articleLinks->length; $i++) {
+
-            $articleLink  = $articleLinks->item($i);
+        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);
+            $linkDomain = @parse_url($footnoteLink->getAttribute('href'), PHP_URL_HOST);
            if (!$linkDomain && isset($this->url)) {
                $linkDomain = @parse_url($this->url, PHP_URL_HOST);
            }
            $linkText = $this->getInnerText($articleLink);
            if ((strpos($articleLink->getAttribute('class'), 'readability-DoNotFootnote') !== false) || preg_match($this->regexps['skipFootnoteLink'], $linkText)) {
                continue;
            }
-            $linkCount++;
+
            ++$linkCount;
            // Add a superscript reference after the article link.
            $refLink->setAttribute('href', '#readabilityFootnoteLink-'.$linkCount);
            $refLink->innerHTML = '<small><sup>['.$linkCount.']</sup></small>';
            $refLink->setAttribute('class', 'readability-DoNotFootnote');
            $refLink->setAttribute('style', 'color: inherit;');
            if ($articleLink->parentNode->lastChild->isSameNode($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);
            $footnote->innerHTML = '<small><sup><a href="#readabilityLink-'.$linkCount.'" title="Jump to Link in Article">^</a></sup></small> ';
            $footnoteLink->innerHTML = ($footnoteLink->getAttribute('title') != '' ? $footnoteLink->getAttribute('title') : $linkText);
            $footnoteLink->setAttribute('name', 'readabilityFootnoteLink-'.$linkCount);
            $footnote->appendChild($footnoteLink);
            if ($linkDomain) {
                $footnote->innerHTML = $footnote->innerHTML.'<small> ('.$linkDomain.')</small>';
            }
            $articleFootnotes->appendChild($footnote);
        }
        if ($linkCount > 0) {
            $articleContent->appendChild($footnotesWrapper);
        }
    }
    /**
     * Prepare the article node for display. Clean out any inline styles,
     * iframes, forms, strip extraneous <p> tags, etc.
@@ -466,31 +505,29 @@ class Readability
        } else {
            $this->dbg('Standard clean enabled.');
        }
        $this->cleanStyles($articleContent);
        $this->killBreaks($articleContent);
        $xpath = new \DOMXPath($articleContent->ownerDocument);
        if ($this->revertForcedParagraphElements) {
            /*
             * Reverts P elements with class 'readability-styled' to text nodes:
             * which is what they were before.
             */
            $elems = $xpath->query('.//p[@data-readability-styled]', $articleContent);
-            for ($i = $elems->length - 1; $i >= 0; $i--) {
+            for ($i = $elems->length - 1; $i >= 0; --$i) {
                $e = $elems->item($i);
                $e->parentNode->replaceChild($articleContent->ownerDocument->createTextNode($e->textContent), $e);
            }
        }
        // Remove service data-candidate attribute.
        $elems = $xpath->query('.//*[@data-candidate]', $articleContent);
-        for ($i = $elems->length - 1; $i >= 0; $i--) {
+        for ($i = $elems->length - 1; $i >= 0; --$i) {
            $elems->item($i)->removeAttribute('data-candidate');
        }
-        // Remove unrelated links and other unneded stuff.
+
        // (not(*) and not(text()[normalize-space()])) or // What's wrong here?
        $elems = $xpath->query('.//a[@rel="nofollow"]', $articleContent);
        for ($i = $elems->length - 1; $i >= 0; $i--) {
            $elems->item($i)->parentNode->removeChild($elems->item($i));
        }
        // Clean out junk from the article content.
        $this->clean($articleContent, 'input');
        $this->clean($articleContent, 'button');
@@ -508,29 +545,35 @@ class Readability
        if ($h2s->length == 1 && mb_strlen($this->getInnerText($h2s->item(0), true, true)) < 100) {
            $this->clean($articleContent, 'h2');
        }
        $this->cleanHeaders($articleContent);
        // Do these last as the previous stuff may have removed junk that will affect these.
        $this->cleanConditionally($articleContent, 'form');
        $this->cleanConditionally($articleContent, 'table');
        $this->cleanConditionally($articleContent, 'ul');
        //if (!$this->lightClean)
        $this->cleanConditionally($articleContent, 'div');
        // Remove extra paragraphs.
        $articleParagraphs = $articleContent->getElementsByTagName('p');
-        for ($i = $articleParagraphs->length - 1; $i >= 0; $i--) {
+
        for ($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;
            $videoCount = $articleParagraphs->item($i)->getElementsByTagName('video')->length;
            $audioCount = $articleParagraphs->item($i)->getElementsByTagName('audio')->length;
            $iframeCount = $articleParagraphs->item($i)->getElementsByTagName('iframe')->length;
            if ($iframeCount === 0 && $imgCount === 0 && $embedCount === 0 && $objectCount === 0 && $videoCount === 0 && $audioCount === 0 && mb_strlen(preg_replace('/\s+/is', '', $this->getInnerText($articleParagraphs->item($i), false, false))) === 0) {
                $articleParagraphs->item($i)->parentNode->removeChild($articleParagraphs->item($i));
            }
            // add extra text to iframe tag to avoid an auto-closing iframe and then break the html code
            if ($iframeCount) {
                $iframe = $articleParagraphs->item($i)->getElementsByTagName('iframe');
-                $iframe->item(0)->nodeValue = '&nbsp;';
+                $iframe->item(0)->nodeValue = ' ';
                $articleParagraphs->item($i)->parentNode->replaceChild($iframe->item(0), $articleParagraphs->item($i));
            }
@@ -547,6 +590,7 @@ class Readability
            }
        }
    }
    /**
     * Initialize a node with the readability object. Also checks the
     * className/id for special names to add to its score.
@@ -612,8 +656,10 @@ class Readability
                $readability->value -= 5;
                break;
        }
        $readability->value += $this->getWeight($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.
@@ -625,43 +671,51 @@ class Readability
        if (!$page) {
            $page = $this->dom;
        }
        $xpath = null;
        $nodesToScore = array();
        if ($page instanceof \DOMDocument && isset($page->documentElement)) {
            $xpath = new \DOMXPath($page);
        }
        $allElements = $page->getElementsByTagName('*');
-        for ($nodeIndex = 0; ($node = $allElements->item($nodeIndex)); $nodeIndex++) {
+
        for ($nodeIndex = 0; ($node = $allElements->item($nodeIndex)); ++$nodeIndex) {
            $tagName = $node->tagName;
            // Some well known site uses sections as paragraphs.
            if (strcasecmp($tagName, 'p') === 0 || strcasecmp($tagName, 'td') === 0 || strcasecmp($tagName, 'section') === 0) {
                $nodesToScore[] = $node;
            }
            // Turn divs into P tags where they have been used inappropriately
            //  (as in, where they contain no other block level elements).
            if (strcasecmp($tagName, 'div') === 0 || strcasecmp($tagName, 'article') === 0 || strcasecmp($tagName, 'section') === 0) {
                if (!preg_match($this->regexps['divToPElements'], $node->innerHTML)) {
                    //$this->dbg('Altering '.$node->getNodePath().' to p');
                    $newNode = $this->dom->createElement('p');
                    try {
                        $newNode->innerHTML = $node->innerHTML;
                        // It's easier to debug using original attributes.
                        //$newNode->setAttribute('class', $node->getAttribute('class'));
                        //$newNode->setAttribute('id', $node->getAttribute('id'));
                        $node = $node->parentNode->replaceChild($newNode, $node);
-                        $nodeIndex--;
+                        --$nodeIndex;
                        $nodesToScore[] = $newNode;
                    } catch (Exception $e) {
                        $this->dbg('Could not alter div/article to p, reverting back to div: '.$e->getMessage());
                    }
                } else {
                    // Will change these P elements back to text nodes after processing.
-                    for ($i = 0, $il = $node->childNodes->length; $i < $il; $i++) {
+                    for ($i = 0, $il = $node->childNodes->length; $i < $il; ++$i) {
                        $childNode = $node->childNodes->item($i);
                        if (is_object($childNode) && get_class($childNode) === 'DOMProcessingInstruction') { //executable tags (<?php or <?xml) warning
                            $childNode->parentNode->removeChild($childNode);
                            continue;
                        }
                        if ($childNode->nodeType == 3) { // XML_TEXT_NODE
                            //$this->dbg('replacing text node with a P tag with the same content.');
                            $p = $this->dom->createElement('p');
@@ -673,6 +727,7 @@ class Readability
                }
            }
        }
        /*
         * Loop through all paragraphs, and assign a score to them based on how content-y they look.
         * Then add their score to their parent node.
@@ -680,23 +735,27 @@ class Readability
         * A score is determined by things like number of commas, class names, etc.
         * Maybe eventually link density.
         */
-        for ($pt = 0, $scored = count($nodesToScore); $pt < $scored; $pt++) {
+        for ($pt = 0, $scored = count($nodesToScore); $pt < $scored; ++$pt) {
            $parentNode = $nodesToScore[$pt]->parentNode;
            // No parent node? Move on...
            if (!$parentNode) {
                continue;
            }
            $grandParentNode = ($parentNode->parentNode instanceof DOMElement) ? $parentNode->parentNode : null;
            $innerText = $this->getInnerText($nodesToScore[$pt]);
            // If this paragraph is less than MIN_PARAGRAPH_LENGTH (default:20) characters, don't even count it.
            if (mb_strlen($innerText) < self::MIN_PARAGRAPH_LENGTH) {
                continue;
            }
            // Initialize readability data for the parent.
            if (!$parentNode->hasAttribute('readability')) {
                $this->initializeNode($parentNode);
                $parentNode->setAttribute('data-candidate', 'true');
            }
            // Initialize readability data for the grandparent.
            if ($grandParentNode && !$grandParentNode->hasAttribute('readability') && isset($grandParentNode->tagName)) {
                $this->initializeNode($grandParentNode);
@@ -723,6 +782,7 @@ class Readability
            }
            $score = floor($score);
            $contentScore += max(min($score, 3), -3);/**/
            // Add the score to the parent. The grandparent gets half.
            $parentNode->getAttributeNode('readability')->value += $contentScore;
            if ($grandParentNode) {
@@ -735,7 +795,8 @@ class Readability
         */
        if ($this->flagIsActive(self::FLAG_STRIP_UNLIKELYS) && $xpath) {
            $candidates = $xpath->query('.//*[(self::footer and count(//footer)<2) or (self::aside and count(//aside)<2)]', $page->documentElement);
-            for ($node = null, $c = $candidates->length - 1; $c >= 0; $c--) {
+
            for ($node = null, $c = $candidates->length - 1; $c >= 0; --$c) {
                $node = $candidates->item($c);
                // node should be readable but not inside of an article otherwise it's probably non-readable block
                if ($node->hasAttribute('readability') && (int) $node->getAttributeNode('readability')->value < 40 && ($node->parentNode ? strcasecmp($node->parentNode->tagName, 'article') !== 0 : true)) {
@@ -743,8 +804,10 @@ class Readability
                    $node->parentNode->removeChild($node);
                }
            }
            $candidates = $xpath->query('.//*[not(self::body) and (@class or @id or @style) and ((number(@readability) < 40) or not(@readability))]', $page->documentElement);
-            for ($node = null, $c = $candidates->length - 1; $c >= 0; $c--) {
+
            for ($node = null, $c = $candidates->length - 1; $c >= 0; --$c) {
                $node = $candidates->item($c);
                $tagName = $node->tagName;
                /* Remove unlikely candidates */
@@ -756,11 +819,12 @@ class Readability
                ) {
                    $this->dbg('Removing unlikely candidate '.$node->getNodePath().' by "'.$unlikelyMatchString.'" with readability '.($node->hasAttribute('readability') ? (int) $node->getAttributeNode('readability')->value : 0));
                    $node->parentNode->removeChild($node);
-                    $nodeIndex--;
+                    --$nodeIndex;
                }
            }
            unset($candidates);
        }
        /*
         * After we've calculated scores, loop through all of the possible candidate nodes we found
         * and find the one with the highest score.
@@ -769,25 +833,30 @@ class Readability
        if ($xpath) {
            // Using array of DOMElements after deletion is a path to DOOMElement.
            $candidates = $xpath->query('.//*[@data-candidate]', $page->documentElement);
-            for ($c = $candidates->length - 1; $c >= 0; $c--) {
+
            for ($c = $candidates->length - 1; $c >= 0; --$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.
                // If not for this we would have used XPath to find maximum @readability.
                $readability = $candidates->item($c)->getAttributeNode('readability');
                $readability->value = round($readability->value * (1 - $this->getLinkDensity($candidates->item($c))), 0, PHP_ROUND_HALF_UP);
                if (!$topCandidate || $readability->value > (int) $topCandidate->getAttribute('readability')) {
                    $this->dbg('Candidate: '.$candidates->item($c)->getNodePath().' ('.$candidates->item($c)->getAttribute('class').':'.$candidates->item($c)->getAttribute('id').') with score '.$readability->value);
                    $topCandidate = $candidates->item($c);
                }
            }
            unset($candidates);
        }
        /*
         * If we still have no top candidate, just use the body as a last resort.
         * We also have to copy the body node so it is something we can modify.
         */
        if ($topCandidate === null || strcasecmp($topCandidate->tagName, 'body') === 0) {
            $topCandidate = $this->dom->createElement('div');
            if ($page instanceof \DOMDocument) {
                if (!isset($page->documentElement)) {
                    // we don't have a body either? what a mess! :)
@@ -796,6 +865,7 @@ class Readability
                    $this->dbg('Setting body to a raw HTML of original page!');
                    $topCandidate->innerHTML = $page->documentElement->innerHTML;
                    $page->documentElement->innerHTML = '';
                    $this->reinitBody();
                    $page->documentElement->appendChild($topCandidate);
                }
            } else {
@@ -803,81 +873,102 @@ class Readability
                $page->innerHTML = '';
                $page->appendChild($topCandidate);
            }
            $this->initializeNode($topCandidate);
        }
        // Set table as the main node if resulted data is table element.
        $tagName = $topCandidate->tagName;
        if (strcasecmp($tagName, 'td') === 0 || strcasecmp($tagName, 'tr') === 0) {
            $up = $topCandidate;
            if ($up->parentNode instanceof DOMElement) {
                $up = $up->parentNode;
                if (strcasecmp($up->tagName, 'table') === 0) {
                    $topCandidate = $up;
                }
            }
        }
        $this->dbg('Top candidate: '.$topCandidate->getNodePath());
        /*
         * 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');
+        $articleContent->setAttribute('class', '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; $s++) {
+
        for ($s = 0, $sl = $siblingNodes->length; $s < $sl; ++$s) {
            $siblingNode = $siblingNodes->item($s);
            $siblingNodeName = $siblingNode->nodeName;
            $append = false;
            $this->dbg('Looking at sibling node: '.$siblingNode->getNodePath().(($siblingNode->nodeType === XML_ELEMENT_NODE && $siblingNode->hasAttribute('readability')) ? (' with score '.$siblingNode->getAttribute('readability')) : ''));
            //$this->dbg('Sibling has score ' . ($siblingNode->readability ? siblingNode.readability.contentScore : 'Unknown'));
            if ($siblingNode->isSameNode($topCandidate)) {
                $append = true;
            }
            $contentBonus = 0;
            // Give a bonus if sibling nodes and top candidates have the same classname.
            if ($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) {
                $append = true;
            }
            if (strcasecmp($siblingNodeName, 'p') === 0) {
                $linkDensity = $this->getLinkDensity($siblingNode);
                $nodeContent = $this->getInnerText($siblingNode, true, true);
-                $nodeLength  = mb_strlen($nodeContent);
+                $nodeLength = mb_strlen($nodeContent);
                if ($nodeLength > self::MIN_NODE_LENGTH && $linkDensity < self::MAX_LINK_DENSITY) {
                    $append = true;
                } elseif ($nodeLength < self::MIN_NODE_LENGTH && $linkDensity === 0 && preg_match('/\.( |$)/', $nodeContent)) {
                    $append = true;
                }
            }
            if ($append) {
                $this->dbg('Appending node: '.$siblingNode->getNodePath());
                $nodeToAppend = null;
                if (strcasecmp($siblingNodeName, 'div') !== 0 && strcasecmp($siblingNodeName, 'p') !== 0) {
                    /* 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. */
                    $this->dbg('Altering siblingNode '.$siblingNodeName.' to div.');
                    $nodeToAppend = $this->dom->createElement('div');
                    try {
-                        $nodeToAppend->setAttribute('id', $siblingNode->getAttribute('id'));
+                        if ($siblingNode->getAttribute('id')) {
                            $nodeToAppend->setAttribute('id', $siblingNode->getAttribute('id'));
                        }
                        $nodeToAppend->setAttribute('alt', $siblingNodeName);
                        $nodeToAppend->innerHTML = $siblingNode->innerHTML;
                    } catch (Exception $e) {
                        $this->dbg('Could not alter siblingNode '.$siblingNodeName.' to div, reverting to original.');
                        $nodeToAppend = $siblingNode;
-                        $s--;
+                        --$s;
-                        $sl--;
+                        --$sl;
                    }
                } else {
                    $nodeToAppend = $siblingNode;
-                    $s--;
+                    --$s;
-                    $sl--;
+                    --$sl;
                }
                // To ensure a node does not interfere with readability styles, remove its classnames & ids.
                // Now done via RegExp post_filter.
                //$nodeToAppend->removeAttribute('class');
@@ -886,9 +977,12 @@ class Readability
                $articleContent->appendChild($nodeToAppend);
            }
        }
        unset($xpath);
        // 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
@@ -896,10 +990,8 @@ class Readability
         * finding the -right- content.
         */
        if (mb_strlen($this->getInnerText($articleContent, false)) < self::MIN_ARTICLE_LENGTH) {
-            if (!$this->body->hasChildNodes()) {
+            $this->reinitBody();
-                $this->body = $this->dom->createElement('body');
+
            }
            $this->body->innerHTML = $this->bodyCache;
            if ($this->flagIsActive(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");
@@ -915,13 +1007,14 @@ class Readability
                $this->dbg('...content is shorter than '.self::MIN_ARTICLE_LENGTH." letters, trying not to clean at all.\n");
                return $this->grabArticle($this->body);
            } else {
                return false;
            }
            return false;
        }
        return $articleContent;
    }
    /**
     * Get the inner text of a node.
     * This also strips out any excess whitespace to be found.
@@ -937,7 +1030,9 @@ class Readability
        if (!isset($e->textContent) || $e->textContent === '') {
            return '';
        }
        $textContent = trim($e->textContent);
        if ($flattenLines) {
            $textContent = mb_ereg_replace('(?:[\r\n](?:\s|&nbsp;)*)+', '', $textContent);
        } elseif ($normalizeSpaces) {
@@ -946,6 +1041,7 @@ class Readability
        return $textContent;
    }
    /**
     * Remove the style attribute on every $e and under.
     *
@@ -956,11 +1052,14 @@ class Readability
        if (!is_object($e)) {
            return;
        }
        $elems = $e->getElementsByTagName('*');
        foreach ($elems as $elem) {
            $elem->removeAttribute('style');
        }
    }
    /**
     * Get comma number for a given text.
     *
@@ -972,6 +1071,7 @@ class Readability
    {
        return substr_count($text, ',');
    }
    /**
     * Get words number for a given text if words separated by a space.
     * Input string should be normalized.
@@ -984,6 +1084,7 @@ class Readability
    {
        return substr_count($text, ' ');
    }
    /**
     * 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.
@@ -999,18 +1100,21 @@ class Readability
        $links = $e->getElementsByTagName('a');
        $textLength = mb_strlen($this->getInnerText($e, true, true));
        $linkLength = 0;
-        for ($dRe = $this->domainRegExp, $i = 0, $il = $links->length; $i < $il; $i++) {
+
        for ($dRe = $this->domainRegExp, $i = 0, $il = $links->length; $i < $il; ++$i) {
            if ($excludeExternal && $dRe && !preg_match($dRe, $links->item($i)->getAttribute('href'))) {
                continue;
            }
            $linkLength += mb_strlen($this->getInnerText($links->item($i)));
        }
        if ($textLength > 0 && $linkLength > 0) {
            return $linkLength / $textLength;
        } else {
            return 0;
        }
        return 0;
    }
    /**
     * Get an element weight by attribute.
     * Uses regular expressions to tell if this element looks good or bad.
@@ -1026,6 +1130,7 @@ class Readability
            return 0;
        }
        $weight = 0;
        //$attribute_val = trim($element->getAttribute('class')." ".$element->getAttribute('id'));
        $attribute_val = trim($element->getAttribute($attribute));
        if ($attribute_val != '') {
@@ -1045,6 +1150,7 @@ class Readability
        return $weight;
    }
    /**
     * Get an element relative weight.
     *
@@ -1057,6 +1163,7 @@ class Readability
        if (!$this->flagIsActive(self::FLAG_WEIGHT_ATTRIBUTES)) {
            return 0;
        }
        $weight = 0;
        /* Look for a special classname */
        $weight += $this->weightAttribute($e, 'class');
@@ -1065,6 +1172,7 @@ class Readability
        return $weight;
    }
    /**
     * Remove extraneous break tags from a node.
     *
@@ -1076,6 +1184,7 @@ class Readability
        $html = preg_replace($this->regexps['killBreaks'], '<br />', $html);
        $node->innerHTML = $html;
    }
    /**
     * Clean a node of all elements of type "tag".
     * (Unless it's a youtube/vimeo video. People love movies.).
@@ -1089,23 +1198,29 @@ class Readability
    {
        $targetList = $e->getElementsByTagName($tag);
        $isEmbed = ($tag === 'audio' || $tag === 'video' || $tag === 'iframe' || $tag === 'object' || $tag === 'embed');
-        for ($cur_item = null, $y = $targetList->length - 1; $y >= 0; $y--) {
+
        for ($cur_item = null, $y = $targetList->length - 1; $y >= 0; --$y) {
            /* Allow youtube and vimeo videos through as people usually want to see those. */
            $cur_item = $targetList->item($y);
            if ($isEmbed) {
                $attributeValues = $cur_item->getAttribute('src').' '.$cur_item->getAttribute('href');
                /* First, check the elements attributes to see if any of them contain known media hosts */
                if (preg_match($this->regexps['media'], $attributeValues)) {
                    continue;
                }
                /* Then check the elements inside this element for the same. */
                if (preg_match($this->regexps['media'], $targetList->item($y)->innerHTML)) {
                    continue;
                }
            }
            $cur_item->parentNode->removeChild($cur_item);
        }
    }
    /**
     * Clean an element of all tags of type "tag" if they look fishy.
     * "Fishy" is an algorithm based on content length, classnames,
@@ -1119,20 +1234,23 @@ class Readability
        if (!$this->flagIsActive(self::FLAG_CLEAN_CONDITIONALLY)) {
            return;
        }
        $tagsList = $e->getElementsByTagName($tag);
        $curTagsLength = $tagsList->length;
        /*
         * Gather counts for other typical elements embedded within.
         * Traverse backwards so we can remove nodes at the same time without effecting the traversal.
         *
         * TODO: Consider taking into account original contentScore here.
         */
-        for ($node = null, $i = $curTagsLength - 1; $i >= 0; $i--) {
+        for ($node = null, $i = $curTagsLength - 1; $i >= 0; --$i) {
            $node = $tagsList->item($i);
            //$class = $node->getAttribute('class').' '.$node->getAttribute('id'); //debug
            $weight = $this->getWeight($node);
            $contentScore = ($node->hasAttribute('readability')) ? (int) $node->getAttribute('readability') : 0;
            $this->dbg('Start conditional cleaning of '.$node->getNodePath().' (class='.$node->getAttribute('class').'; id='.$node->getAttribute('id').')'.(($node->hasAttribute('readability')) ? (' with score '.$node->getAttribute('readability')) : ''));
            if ($weight + $contentScore < 0) {
                $this->dbg('Removing...');
                $node->parentNode->removeChild($node);
@@ -1148,20 +1266,24 @@ class Readability
                $a = $node->getElementsByTagName('a')->length;
                $embedCount = 0;
                $embeds = $node->getElementsByTagName('embed');
-                for ($ei = 0, $il = $embeds->length; $ei < $il; $ei++) {
+
                for ($ei = 0, $il = $embeds->length; $ei < $il; ++$ei) {
                    if (preg_match($this->regexps['media'], $embeds->item($ei)->getAttribute('src'))) {
-                        $embedCount++;
+                        ++$embedCount;
                    }
                }
                $embeds = $node->getElementsByTagName('iframe');
-                for ($ei = 0, $il = $embeds->length; $ei < $il; $ei++) {
+                for ($ei = 0, $il = $embeds->length; $ei < $il; ++$ei) {
                    if (preg_match($this->regexps['media'], $embeds->item($ei)->getAttribute('src'))) {
-                        $embedCount++;
+                        ++$embedCount;
                    }
                }
-                $linkDensity   = $this->getLinkDensity($node, true);
+
                $linkDensity = $this->getLinkDensity($node, true);
                $contentLength = mb_strlen($this->getInnerText($node));
                $toRemove = false;
                if ($this->lightClean) {
                    if ($li > $p && $tag != 'ul' && $tag != 'ol') {
                        $this->dbg(' too many <li> elements, and parent is not <ul> or <ol>');
@@ -1192,8 +1314,8 @@ class Readability
                    } elseif ($input > floor($p / 3)) {
                        $this->dbg('  too many <input> elements');
                        $toRemove = true;
-                    } elseif ($contentLength < 25 && ($img === 0 || $img > 2)) {
+                    } elseif ($contentLength < 10 && ($img === 0 || $img > 2)) {
-                        $this->dbg('  content length less than 25 chars and 0 images, or more than 2 images');
+                        $this->dbg('  content length less than 10 chars and 0 images, or more than 2 images');
                        $toRemove = true;
                    } elseif ($weight < 25 && $linkDensity > 0.2) {
                        $this->dbg('  weight is '.$weight.' lower than 0 and link density is '.sprintf('%.2f', $linkDensity).' > 0.2');
@@ -1206,6 +1328,7 @@ class Readability
                        $toRemove = true;
                    }
                }
                if ($toRemove) {
                    //$this->dbg('Removing: '.$node->innerHTML);
                    $this->dbg('Removing...');
@@ -1214,6 +1337,7 @@ class Readability
            }
        }
    }
    /**
     * Clean out spurious headers from an Element. Checks things like classnames and link density.
     *
@@ -1221,25 +1345,39 @@ class Readability
     */
    public function cleanHeaders($e)
    {
-        for ($headerIndex = 1; $headerIndex < 3; $headerIndex++) {
+        for ($headerIndex = 1; $headerIndex < 3; ++$headerIndex) {
            $headers = $e->getElementsByTagName('h'.$headerIndex);
-            for ($i = $headers->length - 1; $i >= 0; $i--) {
+            for ($i = $headers->length - 1; $i >= 0; --$i) {
                if ($this->getWeight($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;
    }
    /**
     * Will recreate previously deleted body property.
     */
    protected function reinitBody()
    {
        if (!isset($this->body->childNodes)) {
            $this->body = $this->dom->createElement('body');
            $this->body->innerHTML = $this->bodyCache;
        }
    }
 }
@@ -48,6 +48,8 @@ class ReadabilityTest extends \PHPUnit_Framework_TestCase
        $this->assertFalse($res);
        $this->assertInstanceOf('Readability\JSLikeHTMLElement', $readability->getContent());
        $this->assertInstanceOf('Readability\JSLikeHTMLElement', $readability->getTitle());
        $this->assertEmpty($readability->getTitle()->innerHTML);
        $this->assertContains('Sorry, Readability was unable to parse this page for content.', $readability->getContent()->innerHTML);
    }
@@ -58,7 +60,9 @@ class ReadabilityTest extends \PHPUnit_Framework_TestCase
        $this->assertTrue($res);
        $this->assertInstanceOf('Readability\JSLikeHTMLElement', $readability->getContent());
        $this->assertInstanceOf('Readability\JSLikeHTMLElement', $readability->getTitle());
        $this->assertContains('<div readability=', $readability->getContent()->innerHTML);
        $this->assertEmpty($readability->getTitle()->innerHTML);
        $this->assertContains('This is the awesome content :)', $readability->getContent()->innerHTML);
    }
@@ -69,7 +73,9 @@ class ReadabilityTest extends \PHPUnit_Framework_TestCase
        $this->assertTrue($res);
        $this->assertInstanceOf('Readability\JSLikeHTMLElement', $readability->getContent());
        $this->assertInstanceOf('Readability\JSLikeHTMLElement', $readability->getTitle());
        $this->assertContains('<div readability=', $readability->getContent()->innerHTML);
        $this->assertEmpty($readability->getTitle()->innerHTML);
        $this->assertContains('This is the awesome content :)', $readability->getContent()->innerHTML);
    }
@@ -81,7 +87,9 @@ class ReadabilityTest extends \PHPUnit_Framework_TestCase
        $this->assertTrue($res);
        $this->assertInstanceOf('Readability\JSLikeHTMLElement', $readability->getContent());
        $this->assertInstanceOf('Readability\JSLikeHTMLElement', $readability->getTitle());
        $this->assertContains('<div readability=', $readability->getContent()->innerHTML);
        $this->assertEmpty($readability->getTitle()->innerHTML);
        $this->assertContains('This is the awesome content :)', $readability->getContent()->innerHTML);
    }
@@ -94,7 +102,9 @@ class ReadabilityTest extends \PHPUnit_Framework_TestCase
        $this->assertTrue($res);
        $this->assertInstanceOf('Readability\JSLikeHTMLElement', $readability->getContent());
        $this->assertInstanceOf('Readability\JSLikeHTMLElement', $readability->getTitle());
        $this->assertContains('<div readability=', $readability->getContent()->innerHTML);
        $this->assertEmpty($readability->getTitle()->innerHTML);
        $this->assertContains('This is an awesome text with some links, here there are', $readability->getContent()->innerHTML);
        $this->assertContains('readabilityFootnoteLink', $readability->getContent()->innerHTML);
        $this->assertContains('readabilityLink-3', $readability->getContent()->innerHTML);
@@ -102,16 +112,18 @@ class ReadabilityTest extends \PHPUnit_Framework_TestCase
    public function testStandardClean()
    {
-        $readability = new ReadabilityTested('<div><h2>Title</h2>'.str_repeat('<p>This is an awesome text with some links, here there are: <a href="http://0.0.0.0/test.html">the awesome</a></p>', 7).'<a href="#nofollow" rel="nofollow">will be removed</a></div>', 'http://0.0.0.0');
+        $readability = new ReadabilityTested('<div><h2>Title</h2>'.str_repeat('<p>This is an awesome text with some links, here there are: <a href="http://0.0.0.0/test.html">the awesome</a></p>', 7).'<a href="#nofollow" rel="nofollow">will NOT be removed</a></div>', 'http://0.0.0.0');
        $readability->debug = true;
        $readability->lightClean = false;
        $res = $readability->init();
        $this->assertTrue($res);
        $this->assertInstanceOf('Readability\JSLikeHTMLElement', $readability->getContent());
        $this->assertInstanceOf('Readability\JSLikeHTMLElement', $readability->getTitle());
        $this->assertContains('<div readability=', $readability->getContent()->innerHTML);
        $this->assertEmpty($readability->getTitle()->innerHTML);
        $this->assertContains('This is an awesome text with some links, here there are', $readability->getContent()->innerHTML);
-        $this->assertNotContains('will be removed', $readability->getContent()->innerHTML);
+        $this->assertContains('will NOT be removed', $readability->getContent()->innerHTML);
        $this->assertNotContains('<h2>', $readability->getContent()->innerHTML);
    }
@@ -123,7 +135,9 @@ class ReadabilityTest extends \PHPUnit_Framework_TestCase
        $this->assertTrue($res);
        $this->assertInstanceOf('Readability\JSLikeHTMLElement', $readability->getContent());
        $this->assertInstanceOf('Readability\JSLikeHTMLElement', $readability->getTitle());
        $this->assertContains('<div readability=', $readability->getContent()->innerHTML);
        $this->assertEmpty($readability->getTitle()->innerHTML);
        $this->assertContains('This is an awesome text with some links, here there are', $readability->getContent()->innerHTML);
        $this->assertContains('nofollow', $readability->getContent()->innerHTML);
    }
@@ -136,7 +150,9 @@ class ReadabilityTest extends \PHPUnit_Framework_TestCase
        $this->assertTrue($res);
        $this->assertInstanceOf('Readability\JSLikeHTMLElement', $readability->getContent());
        $this->assertInstanceOf('Readability\JSLikeHTMLElement', $readability->getTitle());
        $this->assertContains('alt="article"', $readability->getContent()->innerHTML);
        $this->assertEmpty($readability->getTitle()->innerHTML);
        $this->assertContains('This is an awesome text with some links, here there are', $readability->getContent()->innerHTML);
        $this->assertContains('nofollow', $readability->getContent()->innerHTML);
    }
@@ -149,7 +165,9 @@ class ReadabilityTest extends \PHPUnit_Framework_TestCase
        $this->assertTrue($res);
        $this->assertInstanceOf('Readability\JSLikeHTMLElement', $readability->getContent());
        $this->assertInstanceOf('Readability\JSLikeHTMLElement', $readability->getTitle());
        $this->assertContains('alt="article"', $readability->getContent()->innerHTML);
        $this->assertEmpty($readability->getTitle()->innerHTML);
        $this->assertContains('This is an awesome text with some links, here there are', $readability->getContent()->innerHTML);
        $this->assertNotContains('<aside>', $readability->getContent()->innerHTML);
        $this->assertContains('<footer/>', $readability->getContent()->innerHTML);
@@ -163,7 +181,9 @@ class ReadabilityTest extends \PHPUnit_Framework_TestCase
        $this->assertTrue($res);
        $this->assertInstanceOf('Readability\JSLikeHTMLElement', $readability->getContent());
        $this->assertInstanceOf('Readability\JSLikeHTMLElement', $readability->getTitle());
        $this->assertContains('alt="article"', $readability->getContent()->innerHTML);
        $this->assertEmpty($readability->getTitle()->innerHTML);
        $this->assertContains('This is an awesome text with some links, here there are', $readability->getContent()->innerHTML);
        $this->assertNotContains('This text should be removed', $readability->getContent()->innerHTML);
    }
@@ -176,7 +196,9 @@ class ReadabilityTest extends \PHPUnit_Framework_TestCase
        $this->assertTrue($res);
        $this->assertInstanceOf('Readability\JSLikeHTMLElement', $readability->getContent());
        $this->assertInstanceOf('Readability\JSLikeHTMLElement', $readability->getTitle());
        $this->assertContains('alt="tr"', $readability->getContent()->innerHTML);
        $this->assertEmpty($readability->getTitle()->innerHTML);
        $this->assertContains('This is an awesome text with some links, here there are', $readability->getContent()->innerHTML);
    }
@@ -188,7 +210,9 @@ class ReadabilityTest extends \PHPUnit_Framework_TestCase
        $this->assertTrue($res);
        $this->assertInstanceOf('Readability\JSLikeHTMLElement', $readability->getContent());
        $this->assertInstanceOf('Readability\JSLikeHTMLElement', $readability->getTitle());
        $this->assertContains('alt="article"', $readability->getContent()->innerHTML);
        $this->assertEmpty($readability->getTitle()->innerHTML);
        $this->assertContains('This is an awesome text with some links, here there are', $readability->getContent()->innerHTML);
        $this->assertContains('This text is also an awesome text and you should know that', $readability->getContent()->innerHTML);
    }
@@ -201,7 +225,69 @@ class ReadabilityTest extends \PHPUnit_Framework_TestCase
        $this->assertTrue($res);
        $this->assertInstanceOf('Readability\JSLikeHTMLElement', $readability->getContent());
        $this->assertInstanceOf('Readability\JSLikeHTMLElement', $readability->getTitle());
        $this->assertContains('alt="article"', $readability->getContent()->innerHTML);
        $this->assertEmpty($readability->getTitle()->innerHTML);
        $this->assertContains('This is an awesome text with some links, here there are', $readability->getContent()->innerHTML);
        $this->assertNotContains('This text is also an awesome text and you should know that', $readability->getContent()->innerHTML);
    }
    public function testTitle()
    {
        $readability = new ReadabilityTested('<title>this is my title</title><article class="awesomecontent">'.str_repeat('<p>This is an awesome text with some links, here there are the awesome</p>', 7).'<p></p></article>', 'http://0.0.0.0');
        $readability->debug = true;
        $res = $readability->init();
        $this->assertTrue($res);
        $this->assertInstanceOf('Readability\JSLikeHTMLElement', $readability->getContent());
        $this->assertInstanceOf('Readability\JSLikeHTMLElement', $readability->getTitle());
        $this->assertContains('alt="article"', $readability->getContent()->innerHTML);
        $this->assertEquals('this is my title', $readability->getTitle()->innerHTML);
        $this->assertContains('This is an awesome text with some links, here there are', $readability->getContent()->innerHTML);
        $this->assertNotContains('This text is also an awesome text and you should know that', $readability->getContent()->innerHTML);
    }
    public function testTitleWithDash()
    {
        $readability = new ReadabilityTested('<title> title2 - title3 </title><article class="awesomecontent">'.str_repeat('<p>This is an awesome text with some links, here there are the awesome</p>', 7).'<p></p></article>', 'http://0.0.0.0');
        $readability->debug = true;
        $res = $readability->init();
        $this->assertTrue($res);
        $this->assertInstanceOf('Readability\JSLikeHTMLElement', $readability->getContent());
        $this->assertInstanceOf('Readability\JSLikeHTMLElement', $readability->getTitle());
        $this->assertContains('alt="article"', $readability->getContent()->innerHTML);
        $this->assertEquals('title2 - title3', $readability->getTitle()->innerHTML);
        $this->assertContains('This is an awesome text with some links, here there are', $readability->getContent()->innerHTML);
        $this->assertNotContains('This text is also an awesome text and you should know that', $readability->getContent()->innerHTML);
    }
    public function testTitleWithDoubleDot()
    {
        $readability = new ReadabilityTested('<title> title2 : title3 </title><article class="awesomecontent">'.str_repeat('<p>This is an awesome text with some links, here there are the awesome</p>', 7).'<p></p></article>', 'http://0.0.0.0');
        $readability->debug = true;
        $res = $readability->init();
        $this->assertTrue($res);
        $this->assertInstanceOf('Readability\JSLikeHTMLElement', $readability->getContent());
        $this->assertInstanceOf('Readability\JSLikeHTMLElement', $readability->getTitle());
        $this->assertContains('alt="article"', $readability->getContent()->innerHTML);
        $this->assertEquals('title2 : title3', $readability->getTitle()->innerHTML);
        $this->assertContains('This is an awesome text with some links, here there are', $readability->getContent()->innerHTML);
        $this->assertNotContains('This text is also an awesome text and you should know that', $readability->getContent()->innerHTML);
    }
    public function testTitleTooShortUseH1()
    {
        $readability = new ReadabilityTested('<title>too short</title><h1>this is my h1 title !</h1><article class="awesomecontent">'.str_repeat('<p>This is an awesome text with some links, here there are the awesome</p>', 7).'<p></p></article>', 'http://0.0.0.0');
        $readability->debug = true;
        $res = $readability->init();
        $this->assertTrue($res);
        $this->assertInstanceOf('Readability\JSLikeHTMLElement', $readability->getContent());
        $this->assertInstanceOf('Readability\JSLikeHTMLElement', $readability->getTitle());
        $this->assertContains('alt="article"', $readability->getContent()->innerHTML);
        $this->assertEquals('this is my h1 title !', $readability->getTitle()->innerHTML);
        $this->assertContains('This is an awesome text with some links, here there are', $readability->getContent()->innerHTML);
        $this->assertNotContains('This text is also an awesome text and you should know that', $readability->getContent()->innerHTML);
    }
@@ -216,4 +302,53 @@ class ReadabilityTest extends \PHPUnit_Framework_TestCase
    //     $this->assertEquals('/0\.0\.0\.0/', $readability->getDomainRegexp());
    //     $this->assertInstanceOf('DomDocument', $readability->dom);
    // }
    // dummy function to be used to the next test
    public function error2Exception($code, $string, $file, $line, $context)
    {
        throw new \Exception($string, $code);
    }
    public function testAutoClosingIframeNotThrowingException()
    {
        error_reporting(E_ALL | E_STRICT);
        ini_set('display_errors', true);
        set_error_handler(array($this, 'error2Exception'), E_ALL | E_STRICT);
        $data = '<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
            <html xmlns="http://www.w3.org/1999/xhtml" lang="ru-RU" prefix="og: http://ogp.me/ns#">
            <head profile="http://gmpg.org/xfn/11">
            <meta http-equiv="Content-Type" content="text/html; charset=UTF-8" />
            <meta name="viewport" content="width=device-width, initial-scale=1.0" />
            </head>
            <body class="single single-post postid-22030 single-format-standard">
              <div id="wrapper">
                <div id="content">
                  <div class="post-22030 post type-post status-publish format-standard has-post-thumbnail hentry category-video category-reviews tag-193" id="post-22030">
                    <h1>3D Touch &#8212; будущее мобильных игр</h1>
                    <div class="postdate">Автор:  <strong>Сергей Пак</strong> | Просмотров: 1363 | Опубликовано:  14 сентября 2015 </div>
                    <div class="entry">
                      <p>Компания Apple представила новую технологию 3D Touch, которая является прямым потомком более ранней версии Force Touch &#8212; последняя, напомним, используется сейчас в трекпадах Macbook Pro и Macbook 2015. Теперь управлять устройством стало в разы проще, и Force Touch открывает перед пользователями новые возможности, но при этом 3D Touch &#8212; это про другое. Дело в том, что теперь и на мобильных устройствах интерфейс будет постепенно меняться, кардинальные перемены ждут мобильный гейминг, потому что здесь разработчики действительно могут разгуляться.<span id="more-22030"></span></p>
                      <p><iframe src="https://www.youtube.com/embed/PUep6xNeKjA" width="560" height="315" frameborder="0" allowfullscreen="allowfullscreen"></iframe></p>
                      <p>Итак, просто представьте себе, что iPhone 6S &#8212; это, по большому счету, отличная игровая приставка, которую вы носите с собой, а еще она может выдавать невероятной красоты картинку. Но проблема заключается, пожалуй, в том, что управлять персонажем в играх довольно трудно &#8212; он неповоротлив, обладает заторможенной реакцией, а игровой клиент зачастую требует перегруза интерфейса для того, чтобы обеспечить максимально большое количество возможностей. Благодаря трехуровневому нажатию можно избавиться от лишних кнопок и обеспечить более качественный обзор местности, и при этом пользователь будет закрывать пальцами минимальное пространство.</p>
                    </div>
                  </div>
                </div>
              </div>
            </body>
            </html>';
        $readability = new ReadabilityTested($data, 'http://iosgames.ru/?p=22030');
        $readability->debug = true;
        $res = $readability->init();
        $this->assertTrue($res);
        $this->assertInstanceOf('Readability\JSLikeHTMLElement', $readability->getContent());
        $this->assertInstanceOf('Readability\JSLikeHTMLElement', $readability->getTitle());
        $this->assertContains('<iframe src="https://www.youtube.com/embed/PUep6xNeKjA" width="560" height="315" frameborder="0" allowfullscreen="allowfullscreen"> </iframe>', $readability->getContent()->innerHTML);
        $this->assertContains('3D Touch', $readability->getTitle()->innerHTML);
    }
 }
Author	SHA1	Message	Date
Jeremy Benoist	f71c3a4196	Do not remove html tag attributes They might contains useful information (at least language)	2015-09-23 21:09:38 +02:00
Jeremy Benoist	255a2fc7bc	Merge pull request #8 from j0k3r/nofollow Do not remove nofollow links	2015-09-23 17:58:32 +02:00
Jeremy Benoist	b77876b30a	Do not remove nofollow links Most the time, they can be usefull. At least, it'll be a link to something unrelated. But we won't lose a link inside the content. Also, adding some extra space.	2015-09-22 19:25:57 +02:00
Jeremy Benoist	1830dc45d4	Merge pull request #7 from j0k3r/fix-nbsp Avoid error with	2015-09-20 21:05:55 +02:00
Jeremy Benoist	6be1f9b984	Fix link to fivefilters fork	2015-09-18 19:19:26 +02:00
Jeremy Benoist	175196d6c2	Avoid error with   Fix #5	2015-09-18 19:10:48 +02:00
Jeremy Benoist	2b5af601d5	Do not format output to avoid breaking apps It'll require to jump to 2.0.0 and I think it's to soon	2015-09-15 22:25:08 +02:00
Jeremy Benoist	d01eb2ac1e	Use class instead of id to avoid error It generates error like `ID XXX already defined`	2015-09-14 21:49:40 +02:00
Jeremy Benoist	c5a4a490e1	CS	2015-08-24 11:10:54 +02:00
Jeremy Benoist	908a49824f	Add test on title	2015-08-24 11:09:47 +02:00
Jeremy Benoist	c67189248e	Backport changes from wallabag https://github.com/wallabag/php-readability/commit/e9e4ff87f8fc56d406ccdd5a9a7f1d3d6af07e79	2015-08-24 11:09:38 +02:00