[WICKET-4460] Wrong string encoding in JavaScriptPackageResource

• Type: Bug
• Status: Resolved
• Priority: Minor
• Resolution: Fixed
• Affects Version/s: 1.5.5
• Fix Version/s: 1.5.6, 6.0.0-beta1
• Component/s: wicket
• Environment: Wicket 1.5.5, Jetty 7.6.2, Windows
• Assignee: Martin Tzvetanov Grigorov
• Reporter: Jonas
• Created: 19/Mar/12 12:00
• Updated: 08/Jul/15 13:41
• Resolved: 19/Mar/12 14:52

Description

org.apache.wicket.request.resource.JavaScriptPackageResource.processResponse(Attributes, byte[])
transforms javascript using an IJavaScriptCompressor if configured - which is the case by default in deployment mode, but not in development mode.
The problem is that the code uses UTF-8 encoding when creating the String from bytes, but when converting the result String of IJavaScriptCompressor back to bytes, the platform's default encoding is used (e.g. Cp1252 on Windows).
I propose to at least use the same character encoding while converting from/to bytes:

something like (JavaScriptPackageResource line 59 and following):

try

{ String nonCompressed = new String(processedResponse, "UTF-8"); return compressor.compress(nonCompressed).getBytes("UTF-8"); }

catch (Exception e)

{ log.error("Error while filtering content", e); return processedResponse; }

Of course it would be even better not to make any assumtions about the encoding of the .js file, but I'm not sure how one could determine that.

Issue Links

• is duplicated by: WICKET-5946 JavaScript/Css PackageResource should use the same charset for compressing