boyska/rss-bridge
1
0
Fork 0
forked from blallo/rss-bridge
Code Releases Activity

Corrected CourrierInternationalBridge

Browse source
This commit is contained in:
teromene@teromene.fr 2016-03-02 11:49:27 +00:00
parent 9553505bdd
commit 05c2e2aa20
1 changed files with 25 additions and 28 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

33
bridges/CourrierInternationalBridge.php
View file

@ -13,25 +13,13 @@ class CourrierInternationalBridge extends BridgeAbstract{
public function collectData(array $param){ public function collectData(array $param){
function fetchArticle($link) {
$page = file_get_html($link);
$contenu = $page->find(".article-text")[0];
return strip_tags($contenu);
}
$html = ''; $html = '';
$html = file_get_html('http://www.courrierinternational.com/article') or $this->returnError('Error.', 500); $html = file_get_html('http://www.courrierinternational.com/') or $this->returnError('Error.', 500);
$element = $html->find(".type-normal"); $element = $html->find("article");
$article_count = 1; $article_count = 1;
@ -39,11 +27,20 @@ class CourrierInternationalBridge extends BridgeAbstract{
$item = new \Item(); $item = new \Item();
$item->uri = "http://www.courrierinternational.com".$article->find("a")[0]->getAttribute("href"); $item->uri = $article->parent->getAttribute("href");
$item->content = fetchArticle("http://www.courrierinternational.com".$article->find("a")[0]->getAttribute("href"));
$item->title = strip_tags($article->find("h2")[0]);
$dateTime = date_parse($article->find("time")[0]); if(strpos($item->uri, "http") === FALSE) {
$item->uri = "http://courrierinternational.fr/".$item->uri;
}
$page = file_get_html($item->uri);
$cleaner = new HTMLSanitizer();
$item->content = $cleaner->sanitize($page->find("div.article-text")[0]);
$item->title = strip_tags($article->find(".title")[0]);
$dateTime = date_parse($page->find("time")[0]);
$item->timestamp = mktime( $item->timestamp = mktime(
$dateTime['hour'], $dateTime['hour'],