mirror of
https://github.com/ezyang/htmlpurifier.git
synced 2025-08-01 11:50:28 +02:00
Release 2.0.0, merged in 1026 to HEAD.
git-svn-id: http://htmlpurifier.org/svnroot/htmlpurifier/branches/strict@1179 48356398-32a2-884e-a903-53898d9a118a
This commit is contained in:
@@ -5,6 +5,8 @@ require_once 'HTMLPurifier/HTMLDefinition.php';
|
||||
require_once 'HTMLPurifier/Generator.php';
|
||||
require_once 'HTMLPurifier/TagTransform.php';
|
||||
|
||||
require_once 'HTMLPurifier/AttrValidator.php';
|
||||
|
||||
HTMLPurifier_ConfigSchema::define(
|
||||
'Core', 'RemoveInvalidImg', true, 'bool',
|
||||
'This directive enables pre-emptive URI checking in <code>img</code> '.
|
||||
@@ -13,6 +15,14 @@ HTMLPurifier_ConfigSchema::define(
|
||||
'since 1.3.0, revert to pre-1.3.0 behavior by setting to false.'
|
||||
);
|
||||
|
||||
HTMLPurifier_ConfigSchema::define(
|
||||
'Core', 'RemoveScriptContents', true, 'bool', '
|
||||
This directive enables HTML Purifier to remove not only script tags
|
||||
but all of their contents. This directive has been available since 2.0.0,
|
||||
revert to pre-2.0.0 behavior by setting to false.
|
||||
'
|
||||
);
|
||||
|
||||
/**
|
||||
* Removes all unrecognized tags from the list of tokens.
|
||||
*
|
||||
@@ -28,33 +38,27 @@ class HTMLPurifier_Strategy_RemoveForeignElements extends HTMLPurifier_Strategy
|
||||
$definition = $config->getHTMLDefinition();
|
||||
$generator = new HTMLPurifier_Generator();
|
||||
$result = array();
|
||||
|
||||
$escape_invalid_tags = $config->get('Core', 'EscapeInvalidTags');
|
||||
$remove_invalid_img = $config->get('Core', 'RemoveInvalidImg');
|
||||
$remove_script_contents = $config->get('Core', 'RemoveScriptContents');
|
||||
|
||||
$attr_validator = new HTMLPurifier_AttrValidator();
|
||||
|
||||
// removes tokens until it reaches a closing tag with its value
|
||||
$remove_until = false;
|
||||
|
||||
foreach($tokens as $token) {
|
||||
if ($remove_until) {
|
||||
if (empty($token->is_tag) || $token->name !== $remove_until) {
|
||||
continue;
|
||||
}
|
||||
}
|
||||
if (!empty( $token->is_tag )) {
|
||||
// DEFINITION CALL
|
||||
if (isset($definition->info[$token->name])) {
|
||||
// leave untouched, except for a few special cases:
|
||||
|
||||
// hard-coded image special case, pre-emptively drop
|
||||
// if not available. Probably not abstract-able
|
||||
if ( $token->name == 'img' && $remove_invalid_img ) {
|
||||
if (!isset($token->attr['src'])) {
|
||||
continue;
|
||||
}
|
||||
if (!isset($definition->info['img']->attr['src'])) {
|
||||
continue;
|
||||
}
|
||||
$token->attr['src'] =
|
||||
$definition->
|
||||
info['img']->
|
||||
attr['src']->
|
||||
validate($token->attr['src'],
|
||||
$config, $context);
|
||||
if ($token->attr['src'] === false) continue;
|
||||
}
|
||||
|
||||
} elseif (
|
||||
|
||||
// before any processing, try to transform the element
|
||||
if (
|
||||
isset($definition->info_tag_transform[$token->name])
|
||||
) {
|
||||
// there is a transformation for this tag
|
||||
@@ -62,12 +66,45 @@ class HTMLPurifier_Strategy_RemoveForeignElements extends HTMLPurifier_Strategy
|
||||
$token = $definition->
|
||||
info_tag_transform[$token->name]->
|
||||
transform($token, $config, $context);
|
||||
}
|
||||
|
||||
if (isset($definition->info[$token->name])) {
|
||||
|
||||
// mostly everything's good, but
|
||||
// we need to make sure required attributes are in order
|
||||
if (
|
||||
$definition->info[$token->name]->required_attr &&
|
||||
($token->name != 'img' || $remove_invalid_img) // ensure config option still works
|
||||
) {
|
||||
$token = $attr_validator->validateToken($token, $config, $context);
|
||||
$ok = true;
|
||||
foreach ($definition->info[$token->name]->required_attr as $name) {
|
||||
if (!isset($token->attr[$name])) {
|
||||
$ok = false;
|
||||
break;
|
||||
}
|
||||
}
|
||||
if (!$ok) continue;
|
||||
$token->armor['ValidateAttributes'] = true;
|
||||
}
|
||||
|
||||
} elseif ($escape_invalid_tags) {
|
||||
// invalid tag, generate HTML and insert in
|
||||
$token = new HTMLPurifier_Token_Text(
|
||||
$generator->generateFromToken($token, $config, $context)
|
||||
);
|
||||
} else {
|
||||
// check if we need to destroy all of the tag's children
|
||||
// CAN BE GENERICIZED
|
||||
if ($token->name == 'script' && $remove_script_contents) {
|
||||
if ($token->type == 'start') {
|
||||
$remove_until = $token->name;
|
||||
} elseif ($token->type == 'empty') {
|
||||
// do nothing: we're still looking
|
||||
} else {
|
||||
$remove_until = false;
|
||||
}
|
||||
}
|
||||
continue;
|
||||
}
|
||||
} elseif ($token->type == 'comment') {
|
||||
|
Reference in New Issue
Block a user