[WICKET-6012] Confused character encodings in JavaScriptPackageResource and CssPackageResource

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Duplicate
• Affects Version/s: 6.20.0
• Component/s: wicket
• Environment: Windows server; Tomcat 7; Wicket 6.20.0
• Assignee: Unassigned
• Reporter: Jon Mann
• Created: 28/Oct/15 14:58
• Updated: 29/Oct/15 07:43
• Resolved: 29/Oct/15 07:43

Description

We encountered a character encoding problem with a Javascript resource provided by Wicket 6.20.0 from our application hosted by Tomcat 7 on a Windows server (with a JVM default charset of "windows-1252").

The source Javascript resource is UTF-8 encoded with a leading BOM mark. The leading bytes of the resource received by the browser are broken.

We tested several different solutions which all worked (with various consequences):

1. Remove the BOM mark from the source Javascript file
2. Set the JVM "file.encoding" property to "UTF-8".
3. Disable DefaultJavaScriptCompressor in Wicket
4. Change the code in JavaScriptPackageResource to "return output.getBytes("UTF-8")"

We solved this problem using solution #1.

We think you should consider changing the code in JavaScriptPackageResource and CssPackageResource per solution #4.

Please see the attached images and an example quickstart test case (which demonstrates the problem on my Windows machine where the JVM default character encoding is "windows-1252").

Thanks to the Wicket team for a great framework.

Issue Links

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