[EP-tech] Re: Eprints 3.1 to 3.3 upgrade not working

Tim Brody tdb2 at ecs.soton.ac.uk
Tue Jan 22 16:47:46 GMT 2013


On Tue, 2013-01-22 at 10:19 +0000, David Whitehead wrote:
> Hi All,
> 
>  
> 
> I have been assigned the task of trying to convert the current EPrints
> repository we have into 3.3. So far what I have managed to do is
> install 3.2 and then upgrade the current EPrints database and data
> into 3.2.4. I have installed 3.3 and I am attempting to do the same
> thing but It’s not working and I get quite a lot of various different
> errors.
> 
>  
> 
> Below is a general list of errors I seem to be getting, I was
> wondering if you could tell me either what might be causing them or
> any kind of known solution for them:-
> 
>  
> 
>  
> 
> ·         Error: Expected the default config, but wasn’t able to find
> it, or it isn’t a dictionary.
> 
> ·         Error: no current point in closepath.
> 
> ·         Error: Failed parsing page 85 using hint tables.
> 
> ·         Error: Failed to parse xref entry.
> 
> ·         Error:Top-level pages object is wrong type (null).
> 
> ·         Error: Object (253, 0) is not a pageDict.
> 
> ·         Error: Invalid number of first page shared object groups.
> 
> ·         Error Opening /tmp/ep-convertK7cXi/Drawing Perspective
> Remembering the basics Lecture.txt: No such file or directory
> at /usr/share/eprints3/perl_lib/Eprints/Plugin/Convert/IndexCodes.pm
> line 57 (This stops the upgrade process all together)
> 
>  
> 
> Note: I have been on the EPrints wiki and I have tried the
> subject_reindex method on both 3.2 and 3.3 but I still cannot get past
> 3.3 without the errors.
> 
> http://wiki.eprints.org/w/Upgrading_from_3.1

Hi,

It looks like the text isn't being extracted from a PDF and, rather than
the PDF convert plugin gracefully returning, it trips a file missing
check in Convert::IndexCodes.

Can you compare your set of Convert plugins with:
http://trac.eprints.org/eprints/browser/tags/3.3.7/system/perl_lib/EPrints/Plugin/Convert

It doesn't look like you should get to the problem you have, so I'm
wondering if you have an old PDF text extraction plugin that's getting
in the way (instead of Convert::DocPDF and the 'antiwordpdf' utility).

-- 
All the best,
Tim
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 490 bytes
Desc: This is a digitally signed message part
Url : http://mailman.ecs.soton.ac.uk/pipermail/eprints-tech/attachments/20130122/c2fe7f5c/attachment.bin 


More information about the Eprints-tech mailing list