[jira] Commented: (FELIX-770) Bundle Fragment Support check too restrictive

["Richard S. Hall (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/FELIX-770?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12641145#action_12641145 ] 

Richard S. Hall commented on FELIX-770:
---------------------------------------

Karl, I looked into applying this patch. Is the proposed check sufficient? The ExtensionManager.isExtensionBundle() method actually tests the values returned by ManifestParser.parseExtensionBundleHeader(), not just for null. Are these testing the same thing or something different? If it is the same, then perhaps we could just come up with one way of testing for it.

> Bundle Fragment Support check too restrictive
> ---------------------------------------------
>
>                 Key: FELIX-770
>                 URL: https://issues.apache.org/jira/browse/FELIX-770
>             Project: Felix
>          Issue Type: Sub-task
>          Components: Framework
>    Affects Versions: felix-1.2.1

>            Reporter: Felix Meschberger
>             Fix For: felix-1.4.0
>
>         Attachments: FELIX-770.patch
>
>
> As indicated in [1], the validation of the Fragment-Host header in the ManifestParser.validateFragment is too restrictive in that it also throws or logs for Framework Extension bundles (fragments attached to the system bundle with an extension directive), because such Framework Extensions have long been supported.
> I think, just excluding such fragments from further inspection in the validateFragment method should do the trick.
> [1] http://markmail.org/message/55crz64j7qxi7c5u

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.