/**
* Extract the css files from a html code
*
* @param string $html
*
* @return string
*/
public function extractStyle($html)
{
// the CSS content
$style = ' ';
// extract the link tags, and remove them in the html code
preg_match_all('/<link([^>]*)>/isU', $html, $match);
$html = preg_replace('/<link[^>]*>/isU', '', $html);
$html = preg_replace('/<\\/link[^>]*>/isU', '', $html);
// analyse each link tag
foreach ($match[1] as $code) {
$tmp = $this->tagParser->extractTagAttributes($code);
// if type text/css => we keep it
if (isset($tmp['type']) && strtolower($tmp['type']) == 'text/css' && isset($tmp['href'])) {
// get the href
$url = $tmp['href'];
// get the content of the css file
$content = @file_get_contents($url);
// if "http://" in the url
if (strpos($url, 'http://') !== false) {
// get the domain "http://xxx/"
$url = str_replace('http://', '', $url);
$url = explode('/', $url);
$urlMain = 'http://' . $url[0] . '/';
// get the absolute url of the path
$urlSelf = $url;
unset($urlSelf[count($urlSelf) - 1]);
$urlSelf = 'http://' . implode('/', $urlSelf) . '/';
// adapt the url in the css content
$content = preg_replace('/url\\(([^\\\\][^)]*)\\)/isU', 'url(' . $urlSelf . '$1)', $content);
$content = preg_replace('/url\\((\\\\[^)]*)\\)/isU', 'url(' . $urlMain . '$1)', $content);
} else {
// @TODO correction on url in absolute on a local css content
// $content = preg_replace('/url\(([^)]*)\)/isU', 'url('.dirname($url).'/$1)', $content);
}
// add to the CSS content
$style .= $content . "\n";
}
}
// extract the style tags des tags style, and remove them in the html code
preg_match_all('/<style[^>]*>(.*)<\\/style[^>]*>/isU', $html, $match);
$html = preg_replace('/<style[^>]*>(.*)<\\/style[^>]*>/isU', '', $html);
// analyse each style tags
foreach ($match[1] as $code) {
// add to the CSS content
$code = str_replace('<!--', '', $code);
$code = str_replace('-->', '', $code);
$style .= $code . "\n";
}
//analyse the css content
$this->analyseStyle($style);
return $html;
}
/**
* TODO remove the reference on the $parents variable
*
* @param Token $token
* @param array $parents
*
* @return array
* @throws HtmlParsingException
*/
protected function getTagAction(Token $token, &$parents)
{
// tag that can be not closed
$tagsNotClosed = array('br', 'hr', 'img', 'col', 'input', 'link', 'option', 'circle', 'ellipse', 'path', 'rect', 'line', 'polygon', 'polyline');
// analyze the HTML code
$node = $this->tagParser->analyzeTag($token->getData());
// save the current position in the HTML code
$node->setLine($token->getLine());
$actions = array();
// if the tag must be closed
if (!in_array($node->getName(), $tagsNotClosed)) {
// if it is a closure tag
if ($node->isClose()) {
// HTML validation
if (count($parents) < 1) {
$e = new HtmlParsingException('Too many tag closures found for [' . $node->getName() . ']');
$e->setInvalidTag($node->getName());
$e->setHtmlLine($token->getLine());
throw $e;
} elseif (end($parents) != $node->getName()) {
$e = new HtmlParsingException('Tags are closed in a wrong order for [' . $node->getName() . ']');
$e->setInvalidTag($node->getName());
$e->setHtmlLine($token->getLine());
throw $e;
} else {
array_pop($parents);
}
} else {
// if it is an auto-closed tag
if ($node->isAutoClose()) {
// save the opened tag
$actions[] = $node;
// prepare the closed tag
$node = clone $node;
$node->setParams(array());
$node->setClose(true);
} else {
// else: add a child for validation
array_push($parents, $node->getName());
}
}
// if it is a <pre> tag (or <code> tag) not auto-closed => update the flag
if (($node->getName() == 'pre' || $node->getName() == 'code') && !$node->isAutoClose()) {
$this->tagPreIn = !$node->isClose();
}
}
// save the actions to convert
$actions[] = $node;
return $actions;
}
/**
* @param string $code
* @param array $expected
*
* @dataProvider tagAttributesProvider
*/
public function testExtractTagAttributes($code, $expected)
{
$result = $this->parser->extractTagAttributes($code);
$this->assertEquals($expected, $result);
}
