[weld-issues] [JBoss JIRA] Commented: (WELD-834) URLScanner can not handle paths containing spaces

[Mickael Sauvee (JIRA)]

    [ https://issues.jboss.org/browse/WELD-834?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12594697#comment-12594697 ] 

Mickael Sauvee commented on WELD-834:
-------------------------------------

This issue does also occur for web apps deployed in tomcat 6 with # in name of the context (multi level context). The hashes are replaced by %23.
This is a regression from 1.0.1-final and is suggest a different change (similar to the code in 1.0.1-final) than the previous one:
In URLScanner of 1.1.1-final, add the following line #90 :
            urlPath = URLDecoder.decode(urlPath, "UTF-8");


> URLScanner can not handle paths containing spaces
> -------------------------------------------------
>
>                 Key: WELD-834
>                 URL: https://issues.jboss.org/browse/WELD-834
>             Project: Weld
>          Issue Type: Bug
>          Components: Bootstrap and Metamodel API
>    Affects Versions: 1.1.0.Final
>         Environment: Windows, Jetty 7.2.2
>            Reporter: Štěpán Mík
>            Assignee: Ales Justin
>
> When actual classpath  includes folder that contains spaces, URLScanner throws an exception (java.io.FileNotFoundException). Problem is at line 93 where string holding URL is "converted" to path just by stripping URL scheme prefix. Unfortunately, spaces in URL are expressed either by plus sing or %20 escape sequence, which remains in file path instead of proper URL decoding. For example "file:/C:/with%20space" is converted to "/C:/with%20space" instead of "C:/with space".

--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira