public/tt-rss
1
0
Fork 0
mirror of https://git.tt-rss.org/git/tt-rss.git synced 2025-12-13 08:15:55 +00:00
Code Issues Releases Wiki Activity

domdocument: remove old meta charset unicode hacks, replace with shorter xml preamble utf8 hack (on loadhtml where it makes sense)

Browse Source 
af_readability: better (?) charset hack for non-unicode pages
This commit is contained in:
Andrew Dolgov
2019-03-21 21:08:02 +03:00
parent 3bd3324e5a
commit 671f4cee65
8 changed files with 15 additions and 46 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
plugins/af_fsckportal/init.php
View File

@@ -19,11 +19,7 @@ class Af_Fsckportal extends Plugin {
$doc = new DOMDocument();
$charset_hack = '<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8"/>
</head>';
@$doc->loadHTML($charset_hack . $article["content"]);
@$doc->loadHTML('<?xml encoding="UTF-8">' . $article["content"]);
if ($doc) {
$xpath = new DOMXPath($doc);

11
plugins/af_readability/init.php
View File

@@ -172,14 +172,10 @@ class Af_Readability extends Plugin {
if (!$tmpdoc->loadHTML($tmp))
return false;
// this is the worst hack yet :(
if (strtolower($tmpdoc->encoding) != 'utf-8') {
$tmpxpath = new DOMXPath($tmpdoc);
foreach ($tmpxpath->query("//meta") as $elem) {
$elem->parentNode->removeChild($elem);
}
$tmp = $tmpdoc->saveHTML();
$tmp = preg_replace("/<meta.*?charset.*?\/>/i", "", $tmp);
$tmp = mb_convert_encoding($tmp, 'utf-8', $tmpdoc->encoding);
}
try {
@@ -210,7 +206,6 @@ class Af_Readability extends Plugin {
} catch (Exception $e) {
return false;
}
}
return false;

8
plugins/af_tumblr_1280/init.php
View File

@@ -25,12 +25,8 @@ class Af_Tumblr_1280 extends Plugin {
if (!function_exists("curl_init") || ini_get("open_basedir"))
return $article;
$charset_hack = '<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8"/>
</head>';
$doc = new DOMDocument();
$doc->loadHTML($charset_hack . $article["content"]);
$doc->loadHTML('<?xml encoding="UTF-8">' . $article["content"]);
$found = false;
@@ -92,4 +88,4 @@ class Af_Tumblr_1280 extends Plugin {
return 2;
}
}
}

2
plugins/af_zz_imgproxy/init.php
View File

@@ -155,7 +155,7 @@ class Af_Zz_ImgProxy extends Plugin {
$proxy_all = $this->host->get($this, "proxy_all");
$doc = new DOMDocument();
if (@$doc->loadHTML($article["content"])) {
if (@$doc->loadHTML('<?xml encoding="UTF-8">' . $article["content"])) {
$xpath = new DOMXPath($doc);
$imgs = $xpath->query("//img[@src]");

6
plugins/cache_starred_images/init.php
View File

@@ -190,12 +190,8 @@ class Cache_Starred_Images extends Plugin implements IHandler {
return;
}
$charset_hack = '<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8"/>
</head>';
$doc = new DOMDocument();
$doc->loadHTML($charset_hack . $content);
$doc->loadHTML('<?xml encoding="UTF-8">' . $content);
$xpath = new DOMXPath($doc);
$entries = $xpath->query('(//img[@src])|(//video/source[@src])');