Files
moodle/lib/htmlpurifier/HTMLPurifier/URIScheme/data.php
Daniel Ziegenberg be7f6d4834 MDL-74823 lib: normalize line endings for HTMLPurifier
Prior to this change, all the line endings in the imported HTMLPurifier
library were using CRLF (\r\n aka Windows style), but the HTMLPurifier
source and also the downloadable artefacts use LF (\n aka Linux style)
as line endings. This has been the case since
510d190382 when with the commit
"MDL-38672 import HTML Purifier 4.5.0" all line endings were changed
from LF to CRLF. There was no comment in the commit on why this change
was done.

As the original source uses LF, this commit partly reverts
510d190382 and goes back to LF as line
endings.

Signed-off-by: Daniel Ziegenberg <daniel@ziegenberg.at>
2022-11-17 18:26:26 +01:00

137 lines
4.3 KiB
PHP

<?php
/**
* Implements data: URI for base64 encoded images supported by GD.
*/
class HTMLPurifier_URIScheme_data extends HTMLPurifier_URIScheme
{
/**
* @type bool
*/
public $browsable = true;
/**
* @type array
*/
public $allowed_types = array(
// you better write validation code for other types if you
// decide to allow them
'image/jpeg' => true,
'image/gif' => true,
'image/png' => true,
);
// this is actually irrelevant since we only write out the path
// component
/**
* @type bool
*/
public $may_omit_host = true;
/**
* @param HTMLPurifier_URI $uri
* @param HTMLPurifier_Config $config
* @param HTMLPurifier_Context $context
* @return bool
*/
public function doValidate(&$uri, $config, $context)
{
$result = explode(',', $uri->path, 2);
$is_base64 = false;
$charset = null;
$content_type = null;
if (count($result) == 2) {
list($metadata, $data) = $result;
// do some legwork on the metadata
$metas = explode(';', $metadata);
while (!empty($metas)) {
$cur = array_shift($metas);
if ($cur == 'base64') {
$is_base64 = true;
break;
}
if (substr($cur, 0, 8) == 'charset=') {
// doesn't match if there are arbitrary spaces, but
// whatever dude
if ($charset !== null) {
continue;
} // garbage
$charset = substr($cur, 8); // not used
} else {
if ($content_type !== null) {
continue;
} // garbage
$content_type = $cur;
}
}
} else {
$data = $result[0];
}
if ($content_type !== null && empty($this->allowed_types[$content_type])) {
return false;
}
if ($charset !== null) {
// error; we don't allow plaintext stuff
$charset = null;
}
$data = rawurldecode($data);
if ($is_base64) {
$raw_data = base64_decode($data);
} else {
$raw_data = $data;
}
if ( strlen($raw_data) < 12 ) {
// error; exif_imagetype throws exception with small files,
// and this likely indicates a corrupt URI/failed parse anyway
return false;
}
// XXX probably want to refactor this into a general mechanism
// for filtering arbitrary content types
if (function_exists('sys_get_temp_dir')) {
$file = tempnam(sys_get_temp_dir(), "");
} else {
$file = tempnam("/tmp", "");
}
file_put_contents($file, $raw_data);
if (function_exists('exif_imagetype')) {
$image_code = exif_imagetype($file);
unlink($file);
} elseif (function_exists('getimagesize')) {
set_error_handler(array($this, 'muteErrorHandler'));
$info = getimagesize($file);
restore_error_handler();
unlink($file);
if ($info == false) {
return false;
}
$image_code = $info[2];
} else {
trigger_error("could not find exif_imagetype or getimagesize functions", E_USER_ERROR);
}
$real_content_type = image_type_to_mime_type($image_code);
if ($real_content_type != $content_type) {
// we're nice guys; if the content type is something else we
// support, change it over
if (empty($this->allowed_types[$real_content_type])) {
return false;
}
$content_type = $real_content_type;
}
// ok, it's kosher, rewrite what we need
$uri->userinfo = null;
$uri->host = null;
$uri->port = null;
$uri->fragment = null;
$uri->query = null;
$uri->path = "$content_type;base64," . base64_encode($raw_data);
return true;
}
/**
* @param int $errno
* @param string $errstr
*/
public function muteErrorHandler($errno, $errstr)
{
}
}