TheDJ has uploaded a new change for review.
https://gerrit.wikimedia.org/r/288173
Change subject: Strip the BOM character when we concatenate files
......................................................................
Strip the BOM character when we concatenate files
We are reading files and concatenating their contents, but some files
start with a BOM character and these BOM characters are only allowed at the
beginning of files, not half way.
Stripping them should be safe, since we already assume that everything
is UTF-8.
Change-Id: I14ad698a684e78976e873e9ae2c367475550a063
---
M includes/resourceloader/ResourceLoaderFileModule.php
1 file changed, 19 insertions(+), 4 deletions(-)
git pull ssh://gerrit.wikimedia.org:29418/mediawiki/core
refs/changes/73/288173/1
diff --git a/includes/resourceloader/ResourceLoaderFileModule.php
b/includes/resourceloader/ResourceLoaderFileModule.php
index 1e7329a..66c0f73 100644
--- a/includes/resourceloader/ResourceLoaderFileModule.php
+++ b/includes/resourceloader/ResourceLoaderFileModule.php
@@ -468,7 +468,7 @@
if ( !file_exists( $localPath ) ) {
throw new MWException( __METHOD__ . ": skip function
file not found: \"$localPath\"" );
}
- $contents = file_get_contents( $localPath );
+ $contents = $this->stripBom( file_get_contents( $localPath ) );
if ( $this->getConfig()->get( 'ResourceLoaderValidateStaticJS'
) ) {
$contents = $this->validateScriptFile( $localPath,
$contents );
}
@@ -810,7 +810,7 @@
if ( !file_exists( $localPath ) ) {
throw new MWException( __METHOD__ . ": script
file not found: \"$localPath\"" );
}
- $contents = file_get_contents( $localPath );
+ $contents = $this->stripBom( file_get_contents(
$localPath ) );
if ( $this->getConfig()->get(
'ResourceLoaderValidateStaticJS' ) ) {
// Static files don't really need to be checked
as often; unlike
// on-wiki module they shouldn't change
unexpectedly without
@@ -882,7 +882,7 @@
$style = $this->compileLessFile( $localPath, $context );
$this->hasGeneratedStyles = true;
} else {
- $style = file_get_contents( $localPath );
+ $style = $this->stripBom( file_get_contents( $localPath
) );
}
if ( $flip ) {
@@ -990,7 +990,7 @@
$localPath = $this->getLocalPath( $templatePath );
if ( file_exists( $localPath ) ) {
$content = file_get_contents( $localPath );
- $templates[$alias] = $content;
+ $templates[$alias] = $this->stripBom( $content
);
} else {
$msg = __METHOD__ . ": template file not found:
\"$localPath\"";
wfDebugLog( 'resourceloader', $msg );
@@ -999,4 +999,19 @@
}
return $templates;
}
+
+ /**
+ * Takes an input string and removes the UTF-8 BOM character if present
+ *
+ * We need to remove these after reading a file, because we concatenate
our files and
+ * the BOM character is not valid in the middle of a string.
+ * We already assume UTF-8 everywhere, so this should be safe.
+ *
+ * @return string input minus the intial BOM char
+ */
+ protected function stripBom( $input ) {
+ $bom = pack( 'H*', 'EFBBBF' );
+ $output = preg_replace( "/^$bom/", '', $input );
+ return $output;
+ }
}
--
To view, visit https://gerrit.wikimedia.org/r/288173
To unsubscribe, visit https://gerrit.wikimedia.org/r/settings
Gerrit-MessageType: newchange
Gerrit-Change-Id: I14ad698a684e78976e873e9ae2c367475550a063
Gerrit-PatchSet: 1
Gerrit-Project: mediawiki/core
Gerrit-Branch: master
Gerrit-Owner: TheDJ <[email protected]>
_______________________________________________
MediaWiki-commits mailing list
[email protected]
https://lists.wikimedia.org/mailman/listinfo/mediawiki-commits
