friendica/dir
6
0
Fork 0
Code Issues Pull requests Releases Wiki Activity

parse_url: Removing of warnings with invalid charsets

Browse source
This commit is contained in:
Michael Vogel 2014-07-24 22:50:56 +02:00
parent c3c64d33bc
commit dbcfe58ea8
1 changed files with 7 additions and 2 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

9
mod/parse_url.php
View file

@ -100,7 +100,7 @@ function parseurl_getsiteinfo($url, $no_guessing = false, $do_oembed = true, $co
// Fetch the first mentioned charset. Can be in body or header // Fetch the first mentioned charset. Can be in body or header
$charset = ""; $charset = "";
if (preg_match('/charset=(.*?)['."'".'"\s\n]/', $header, $matches)) if (preg_match('/charset=(.*?)['."'".'"\s\n]/', $header, $matches))
$charset = trim(array_pop($matches)); $charset = trim(trim(trim(array_pop($matches)), ';,'));
if ($charset == "") if ($charset == "")
$charset = "utf-8"; $charset = "utf-8";
@ -112,7 +112,12 @@ function parseurl_getsiteinfo($url, $no_guessing = false, $do_oembed = true, $co
else else
$body = $header; $body = $header;
$body = mb_convert_encoding($body, "UTF-8", $charset); if (($charset != '') AND (strtoupper($charset) != "UTF-8")) {
logger("parseurl_getsiteinfo: detected charset ".$charset, LOGGER_DEBUG);
//$body = mb_convert_encoding($body, "UTF-8", $charset);
$body = iconv($charset, "UTF-8//TRANSLIT", $body);
}
$body = mb_convert_encoding($body, 'HTML-ENTITIES', "UTF-8"); $body = mb_convert_encoding($body, 'HTML-ENTITIES', "UTF-8");
$doc = new DOMDocument(); $doc = new DOMDocument();