OpenKM 5.1.5 - WebDAV issues

[bgergens]

Hi, I'm very new to OpenKM and I'm trying to set it up for my small company.

I haven't done much beyond creating a folder and uploading a doc_u_ment to it. I haven't even created any users yet, as I'm still just trying to test the base functionality.

I enabled WebDAV and was able to successfully connect to it (as okmAdmin) on my Mac, using both Finder and Cyberduck. However, trying to upload a file of any kind results in the following error:

Code: Select all

14:01:18,164 WARN  [RegisteredExtractors] There was a problem extracting text from '/okm:root/IT Docs/time_entries_2011-06-06.xlsx'
14:01:18,388 ERROR [DavResourceImpl] Error while importing resource: java.io.IOException: okm:author

I thought the problem might be caused by trying to upload Office format files so I tried uploading a simple .txt file, and the error changed:

Code: Select all

14:02:04,555 WARN  [RegisteredExtractors] Mark not supported in java.io.InputStreamReader
14:02:04,640 ERROR [DavResourceImpl] Error while importing resource: java.io.IOException: okm:author

I'm very new to this so I'm not sure what to check.

[jllort]

have been logged at least one time with the user are you using to connect to OpenKM webdav ?
about which openkm version are we talking ?

[kernel]

Installed latest openkm (5.1.6),
enabled webdav,
system.ocr=""

After uploading .tif image, or an rtf/doc file the line below appears in log, and server response is 500 internal server error, which the client isn't so happy about.

[DavResourceImpl] Error while importing resource: java.io.IOException: okm:author

Also when uploading an unknown file type, this appears in the logs:

[RegisteredExtractors] There was a problem extracting text from '/okm:root/test.avi"

in my case text extraction isn't necessary, moreover, there will be huge video and image uploads (50Mb-1G in size), so processing of these files would eat the server's performace.
Is there a way completely disable everything ocr/text extraction related processing?

Any help would be appreciated

[jllort]

You can disable extractors from administration tab, configuration, and take a look at registered.text.extractors, you can remove from there what you don't like or all if you want all document will not be indexed.

[kernel]

Thanks for you help jllort, i really appreciate your quick reply

[kernel]

Hello again.

It seems that there will be documents(doc, pdf) uploaded to our system along with video and other files, which need to be extracted/indexed.
That means turning off text extraction for all file types isn't an option.

When uploading a .doc file through webdav, in the server log the aforementioned message appears:

[DavResourceImpl] Error while importing resource: java.io.IOException: okm:author

I've downloaded and compiled OpenKM from source(I'm using the latest binary openkm5.1, source is branches/5.1/openkm/ from svn), so i could try to debug the problem.
It's clear, that after upload, the error arises in the DefaultHandler.importContent method, when it deals with previous versions of this document.These are newly uploaded files, so there are no previous versions of the documents(i guess), but the code still does something i can't seem to figure out
(DefaultHandler.java:290):

String author = frozenNode.getProperty(Document.AUTHOR).getString();

The getProperty method throws a PathNotFoundException because there is no such property, then the whole process returns with http status 500, which i guess means the indexing part won't happen, and the webdav client stops uploading after the first file.

I'd appreciate if you could take a look, what might be wrong or what i'm doing wrong.
cheers,

[jllort]

I've forwarded this post to technical who's designed webdav to take a look about it.

[pavila]

Triaged and added this issue http://issues.openkm.com/view.php?id=1710.