Merge pull request #10 from j0k3r/remove-tidy-constraint

Remove tidy from requirement

.gitattributes

@@ -0,0 +1,8 @@
+/.editorconfig export-ignore
+/.gitattributes export-ignore
+/.gitignore export-ignore
+/.scrutinizer.yml export-ignore
+/.travis.yml export-ignore
+/README.md export-ignore
+/phpunit.xml.dist export-ignore
+/tests export-ignore

README.md

@@ -3,11 +3,11 @@
 [![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](https://bitbucket.org/fivefilters/php-readability/overview).
+This is an extract of the Readability class from this [full-text-rss](https://github.com/Dither/full-text-rss) fork. It can be defined as a better version of the original [php-readability](https://bitbucket.org/fivefilters/php-readability/overview).
 
 ## Differences
 
-The default php-readability lib is really old and needs to be improved. I found a great fork of [full-text-rss](http://fivefilters.org/content-only/) from @Dither which improve the Readability class.
+The default php-readability lib is really old and needs to be improved. I found a great fork of full-text-rss from [@Dither](https://github.com/Dither/full-text-rss) which improve the Readability class.
 
  - I've extracted the class from its fork to be able to use it out of the box
  - I've added some simple tests
@@ -15,6 +15,12 @@ The default php-readability lib is really old and needs to be improved. I found
 
 **But** the code is still really hard to understand / read ...
 
+## Requirements
+
+By default, this lib will use the [Tidy extension](https://github.com/htacg/tidy-html5) if it's available. Tidy is only used to cleanup the given HTML and avoid problems with bad HTML structure, etc ..
+
+Since Composer doesn't support suggestion on PHP extension, I write this suggestion here.
+
 ## Usage
 
 ```php
@@ -26,6 +32,8 @@ $url = 'http://www.medialens.org/index.php/alerts/alert-archive/alerts-2013/729-
 $html = file_get_contents($url);
 
 $readability = new Readability($html, $url);
+// or without Tidy
+// $readability = new Readability($html, $url, 'libxml', false);
 $result = $readability->init();
 
 if ($result) {

composer.json

@@ -24,8 +24,7 @@
         "role": "Developer (original JS version)"
     }],
     "require": {
-        "php": ">=5.3.3",
+        "php": ">=5.3.3"
-        "ext-tidy": ">=1.2"
     },
     "autoload": {
         "psr-4": { "Readability\\": "src/" }

src/Readability.php

@@ -69,10 +69,10 @@ class Readability
      * Defined up here so we don't instantiate them repeatedly in loops.
      */
     public $regexps = array(
-        'unlikelyCandidates' => '/display\s*:\s*none|ignore|\binfo|annoy|clock|date|time|author|intro|links|hidd?e|about|archive|\bprint|bookmark|tags|share|search|social|robot|published|combx|comment|mast(?:head)|subscri|community|category|disqus|extra|head(?:er|note)|floor|foot(?:er|note)|menu|tool|function|nav|remark|rss|shoutbox|tool|widget|meta|banner|sponsor|adsense|inner-?ad|ad-|sponsor|\badv\b|\bads\b|agr?egate?|pager|sidebar|popup|tweet|twitter/i',
+        'unlikelyCandidates' => '/display\s*:\s*none|ignore|\binfo|annoy|clock|date|time|author|intro|links|hidd?e|about|archive|\bprint|bookmark|tags|share|search|social|robot|published|combx|comment|mast(?:head)|subscri|community|category|disqus|extra|head|head(?:er|note)|floor|foot(?:er|note)|menu|tool|function|nav|remark|rss|shoutbox|tool|widget|meta|banner|sponsor|adsense|inner-?ad|ad-|sponsor|\badv\b|\bads\b|agr?egate?|pager|sidebar|popup|tweet|twitter/i',
         'okMaybeItsACandidate' => '/article\b|contain|\bcontent|column|general|detail|shadow|lightbox|blog|body|entry|main|page/i',
         'positive' => '/read|full|article|body|\bcontent|contain|entry|main|markdown|page|attach|pagination|post|text|blog|story/i',
-        'negative' => '/bottom|stat|info|discuss|e[\-]?mail|comment|reply|log.{2}(n|ed)|sign|single|combx|com-|contact|_nav|link|media|\bout|promo|\bad-|related|scroll|shoutbox|sidebar|sponsor|shopping|teaser/i',
+        'negative' => '/bottom|stat|info|discuss|e[\-]?mail|comment|reply|log.{2}(n|ed)|sign|single|combx|com-|contact|_nav|link|media|\bout|promo|\bad-|related|scroll|shoutbox|sidebar|sponsor|shopping|teaser|recommend/i',
         'divToPElements' => '/<(?:blockquote|code|div|article|footer|aside|img|p|pre|dl|ol|ul)/mi',
         'killBreaks' => '/(<br\s*\/?>([ \r\n\s]|&nbsp;?)*)+/',
         'media' => '!//(?:[^\.\?/]+\.)?(?:youtu(?:be)?|soundcloud|dailymotion|vimeo|pornhub|xvideos|twitvid|rutube|viddler)\.(?:com|be|org|net)/!i',

tests/ReadabilityTest.php

@@ -19,6 +19,9 @@ class ReadabilityTested extends Readability
 
 class ReadabilityTest extends \PHPUnit_Framework_TestCase
 {
+    /**
+     * @requires extension tidy
+     */
     public function testConstructDefault()
     {
         $readability = new ReadabilityTested('');
@@ -30,6 +33,9 @@ class ReadabilityTest extends \PHPUnit_Framework_TestCase
         $this->assertInstanceOf('DomDocument', $readability->dom);
     }
 
+    /**
+     * @requires extension tidy
+     */
     public function testConstructSimple()
     {
         $readability = new ReadabilityTested('<html/>', 'http://0.0.0.0');
@@ -41,6 +47,28 @@ class ReadabilityTest extends \PHPUnit_Framework_TestCase
         $this->assertInstanceOf('DomDocument', $readability->dom);
     }
 
+    public function testConstructDefaultWithoutTidy()
+    {
+        $readability = new ReadabilityTested('', null, 'libxml', false);
+
+        $this->assertNull($readability->url);
+        $this->assertContains('Parsing URL', $readability->getDebugText());
+        $this->assertNotContains('Tidying document', $readability->getDebugText());
+        $this->assertNull($readability->getDomainRegexp());
+        $this->assertInstanceOf('DomDocument', $readability->dom);
+    }
+
+    public function testConstructSimpleWithoutTidy()
+    {
+        $readability = new ReadabilityTested('<html/>', 'http://0.0.0.0', 'libxml', false);
+
+        $this->assertEquals('http://0.0.0.0', $readability->url);
+        $this->assertContains('Parsing URL', $readability->getDebugText());
+        $this->assertNotContains('Tidying document', $readability->getDebugText());
+        $this->assertEquals('/0\.0\.0\.0/', $readability->getDomainRegexp());
+        $this->assertInstanceOf('DomDocument', $readability->dom);
+    }
+
     public function testInitNoContent()
     {
         $readability = new ReadabilityTested('<html/>', 'http://0.0.0.0');