You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@commons.apache.org by "Stefan Bodewig (JIRA)" <ji...@apache.org> on 2013/05/20 16:09:15 UTC

[jira] [Created] (COMPRESS-227) duplicate entries may let ZipFile#getInputStream return null

Stefan Bodewig created COMPRESS-227:
---------------------------------------

             Summary: duplicate entries may let ZipFile#getInputStream return null
                 Key: COMPRESS-227
                 URL: https://issues.apache.org/jira/browse/COMPRESS-227
             Project: Commons Compress
          Issue Type: Bug
          Components: Archivers
    Affects Versions: 1.5
            Reporter: Stefan Bodewig
            Assignee: Stefan Bodewig
             Fix For: 1.6


Found while investigating https://issues.apache.org/bugzilla/show_bug.cgi?id=54967

If an archive contains two entries for the same file readCentralDirectory in ZipFile will only add the first entry to the entries map and only the last to the nameMap map - this is because entries is a LinkedHashMap and nameMap is a "plain" HashMap.

Normally this wouldn't matter since both ZipArchiveEntry instances are equal and thus it is irrelevant which of the two is used as a key when obtaining the InputStream.

Things get different, though, if the entry has data inside the local file header as only the first entry is dealt with in resolveLocalFileHeaderData - after this the two instances are no longer equal and nameMap.get(NAME) will return an instance that can no longer be found.

I intend to modify readCentralDirectory to only add the first entry to nameMap as well and document the behavior.  I'll also start a discussion thread on the dev list on whether we need to provide an additional API with multiple entries per name.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira