blallo/tt-rss
1
0
Fork 0
Code Issues Pull requests Releases Wiki Activity

Feedparser: Change handling of libxml error 9 (cycle all errors)

Browse source
This commit is contained in:
wltb 2013-09-09 00:44:55 +02:00
parent c6e089b969
commit 5fa36a6401
1 changed files with 20 additions and 17 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

37
classes/feedparser.php
View file

@ -23,18 +23,12 @@ class FeedParser {
$error = libxml_get_last_error(); $error = libxml_get_last_error();
// libxml compiled without iconv? // libxml compiled without iconv?
if ($error && ($error->code == 32 || $error->code == 9)) { if ($error && $error->code == 32) {
if (preg_match('/^(<\?xml[\t\n\r ].*?encoding=["\'])(.+?)(["\'].*?\?>)/s', $data, $matches) === 1) { if (preg_match('/^(<\?xml[\t\n\r ].*?encoding[\t\n\r ]*=[\t\n\r ]*["\'])(.+?)(["\'].*?\?>)/s', $data, $matches) === 1) {
$enc = $matches[2]; $data = mb_convert_encoding($data, 'UTF-8', $matches[2]);
$data = mb_convert_encoding($data, 'UTF-8', $enc);
$data = preg_replace('/^<\?xml[\t\n\r ].*?\?>/s', $matches[1] . "UTF-8" . $matches[3] , $data); $data = preg_replace('/^<\?xml[\t\n\r ].*?\?>/s', $matches[1] . "UTF-8" . $matches[3] , $data);
// apparently not all UTF-8 characters are valid for XML
$data = preg_replace('/[^\x{0009}\x{000a}\x{000d}\x{0020}-\x{D7FF}\x{E000}-\x{FFFD}]+/u', ' ', $data);
if ($data) { if ($data) {
libxml_clear_errors(); libxml_clear_errors();
@ -43,20 +37,29 @@ class FeedParser {
$error = libxml_get_last_error(); $error = libxml_get_last_error();
} }
} }
} }
// some terrible invalid unicode entity? // some terrible invalid unicode entity?
if ($error && $error->code == 9) { if ($error) {
$data = mb_convert_encoding($data, 'UTF-8', 'UTF-8'); foreach(libxml_get_errors() as $err) {
if ($err->code == 9) {
// remove dangling bytes
$data = mb_convert_encoding($data, 'UTF-8', 'UTF-8');
if ($data) { // apparently not all UTF-8 characters are valid for XML
libxml_clear_errors(); $data = preg_replace('/[^\x{0009}\x{000a}\x{000d}\x{0020}-\x{D7FF}\x{E000}-\x{FFFD}]+/u', ' ', $data);
$this->doc = new DOMDocument(); if ($data) {
$this->doc->loadXML($data); libxml_clear_errors();
$error = libxml_get_last_error(); $this->doc = new DOMDocument();
$this->doc->loadXML($data);
$error = libxml_get_last_error();
}
break;
}
} }
} }