Embed

New imagesBlacklist option to define the images you want to avoid (Thank, @Soullivaneuh). You can use * and ? special characters in this way:

$config = [
    'adapter' => [
        'config' => [
            'minImageWidth' => 16,
            'minImageHeight' => 16,
            'imagesBlacklist' => [
                'http://example.com/full/path/to/image.jpg', //images with this exact url
                'https?://ugglyimages.com/*', //images begining by "http://ugglyimages.com/" (http or https)
                '*/bad_image.gif', //image ending in "/bad_image.gif"
                'https?://test.*/*.png/', //more complex pattern
            ]
        ]
    ]
];

Fixed SoundCloud clips with no artwork causes error #52

Fixed Facebook adapter (missing namespace) #51

• Improved the capture of images in the html. The images with rel="nofollow" will be ignored
• More info provided by each image: width, height, mime, size, url and providers in which appears
• The $info->images array is numeric.
• Improved the way of choose the main image. Now the bigger image found in the first provided will be choosen by default.

New methods getImagesUrl() and getProviderIconsUrls() to retrieve only the image urls, without check the mimetype and dimmensions. This fixes #50

• Removed support for php 5.3. If you need support for this, use the 1.x version.

• Removed FastImage to get the images dimmension. Now we have Embed\ImageInfo\Curl that can execute parallel requests and uses getimagesizefromstring to get the image dimmensions.

• Changed the configuration array. Now it has subarrays to provide options for adapters, providers, resolvers and images. See readme for more info.

• The images and providerIcons properties returns not only the url but also the mimetype and dimmensions:

  //v.1x
  var_dump($info->images);
  /*
  array(
    'http://example.com/images/image1.jpg',
    'http://example.com/images/image2.jpg'
  );
  */
  
  //v2.x
  var_dump($info->images);
  /*
  array(
    'http://example.com/images/image1.jpg' => array(
      0 => 200,
      1 => 500,
      2 => 'image/jpeg',
    ),
    'http://example.com/images/image2.jpg' => array(
      0 => 320,
      1 => 545,
      2 => 'image/jpeg',
    )
  );
  */
  

• fixed Resolver that cannot be configured #46

• New adapter for slides.com #36
• New adapter for line.do #41
• Fixed soundcloud adapter with mobile urls #43 #44

Fixed composer configuration #39

• Changed to PSR-4 autoloader
• PSR-1/2 code style fixes
• Removed some deprecated adapters: Cadenaser, Marca, Parleys, Rtve, Yfrog
• Changes in the Embed\Request class and Embed\RequestResolvers\RequestResolverInterface
• A lot of fixes and new tests
• Changed the way to configure the resolver. Now it's configured with the rest of the options under the "resolver" subarray:

//Old style:
Embed\Request::setResolverConfig(array(
    'userAgent' => 'My spider',
    'connectionTimeout' => 10
));

$info = Embed\Embed::create($url);

//New style
$info = Embed\Embed::create($url, array(
    'resolver' => array(
        "options" => array(
            CURLOPT_USERAGENT => 'My spider',
            CURLOPT_CONNECTTIMEOUT => 10
        )
    )
));

• Changed the way to detect the bigger image when 'getBiggerImage' or 'getBiggerIcon' is true. Until now the php function arsort was used but it doesn't work fine with images with the same dimmensions. From now, it returns the first bigger image found (the first images are suppossed to be the more relevants).
• The options for the Curl request resolver have been changed to camelCase for more consistency with the other options
• Improvements in the demo page.

• By default, the facebook provider is disabled, to reduce the number of requests and avoid weird values. You can enable it again setting the config option facebookProvider to true
• If the url is to a binary file (image, pdf, etc) the Curl request resolver does not download the content, only the headers. This avoid download large files unnecesary.

• Added an adapter for giphy.com (the default adapter returns a not so good embed code)
• Added ignoreFacebookProvider option to not use the facebook scrapping (sometimes returns weird values)

Fixed a bug that return null in imageWith/imageHeight in some circunstances

Ignored the demos in the releases

Moved the demos to a subdirectory

• Updated facebook embed code #32
• Fixes in url redirections

• Fixed Rdio adapter (now it works with https) #28
• Fixed Facebook adapter (The embed code was mising in some urls) #29
• Now gists urls are type "rich" instead "url" #30
• Added support for urls with hash bangs #31

• Fixed some types of domains #25
• Added oembed provider of TED videos #26

Added automatic redirections for google url shortener and google translate