public/tt-rss
1
0
Fork 0
mirror of https://git.tt-rss.org/git/tt-rss.git synced 2025-12-13 22:45:56 +00:00
Code Issues Releases Wiki Activity

af_readability: add a workaround for meta charset html pages

Browse Source
This commit is contained in:
Andrew Dolgov
2015-07-07 10:15:08 +03:00
parent 6322fc6870
commit b7d1306b19
1 changed files with 13 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

13
plugins/af_readability/init.php
View File

@@ -101,6 +101,19 @@ class Af_Readability extends Plugin {
$tmp = fetch_file_contents($article["link"]); $tmp = fetch_file_contents($article["link"]);
if ($tmp) { if ($tmp) {
$tmpdoc = new DOMDocument("1.0", "UTF-8");
$tmpdoc->loadHTML($tmp);
if ($tmpdoc->encoding != 'UTF-8') {
$tmpxpath = new DOMXPath($tmpdoc);
foreach ($tmpxpath->query("//meta") as $elem) {
$elem->parentNode->removeChild($elem);
}
$tmp = $tmpdoc->saveHTML();
}
$r = new Readability($tmp, $article["link"]); $r = new Readability($tmp, $article["link"]);
if ($r->init()) { if ($r->init()) {