Zip files support
Brought to you by:
jhereth,
peterbecker
Menu
▾
▴
#7 Zip files support
Hello,
I presented Docco to my brother who wants to use it for
text-mining purposes in his Job... He has noted that it
would be great if Docco would support Zip format.
Once I will have more time... I might try to do this.
But I thought perhaps it would be still good to send
you this feature request :-)...
J
Logged In: YES
user_id=41603
Tickets are always welcome since they show interest and make
sure nothing gets forgotten -- unless intended of course ;-)
Recursing into ZIP files would be expensive, but should be
feasible. It might be worthwhile looking into this:
https://truezip.dev.java.net/
since it claims that it handles ZIP transparently, while the
standard JDK view would require handling the ZipInputStream
explicitely. Not that this is really hard, the OOo handler
already has some sample code for it:
http://svn.sourceforge.net/viewvc/tockit/trunk/java/applications/docco/source/org/tockit/docco/documenthandler/OpenOfficeDocumentHandler.java?revision=976&view=markup
The difference would be that handling generic zip files
would mean recursing back into the normal document
processing loop, which is based around Files, not InputStreams:
http://svn.sourceforge.net/viewvc/tockit/trunk/java/applications/docco/source/org/tockit/docco/indexer/Indexer.java?revision=967&view=markup
I don't think you could easily move to standard
InputStreams, since the file names are used in the index.
TrueZip claims to solve that issue.
Additionally a proper integration would have to be able to
handle the shell execute, too -- at least in a way that the
user gets notified about the difference and/or has the
option to open the containing folder. But I guess that would
be step 2.
Tika has solved this problem, so if we should move indexing to Tika (see https://sourceforge.net/tracker2/?func=detail&aid=2346619&group_id=21448&atid=371556 ), then this feature will be solved automatically.