problems with PICA2 translator in GBV catalogs
Hi
for a few weeks now there seems to be a problem with PICA translators in catalogs maintained by the GBV. I am not able to download metadata from the university libraries in Göttingen, Clausthal-Zellerfeld, TIB Hannover etc. using the PICA2 translator (try for example https://opac.sub.uni-goettingen.de/DB=1/XMLPRS=N/PPN?PPN=57161647X, http://opac.tib.uni-hannover.de/DB=1/XMLPRS=N/PPN?PPN=620088028 , http://lhclz.gbv.de/DB=1/XMLPRS=N/PPN?PPN=08727342X). It works with CoinS though.
To my surprise, there is no problem with the catalog of the Stabi in Berlin (see for example http://stabikat.de/DB=1/XMLPRS=N/PPN?PPN=57161647X ).
In the main GBV catalog there is no problem either, but here Zotero uses the "PICA" translator (and not "PICA2") (see for example http://gso.gbv.de/DB=2.1/PPNSET?PPN=57161647X ).
I had colleagues trying it out, and used Firefox 19 and 20 and Zotero 3.x and 4.0.3, but I don´t think the versions of the software have anything to do with it. Here is a report ID too, but I don´t think that it adds relevant information either:
Report ID 1135268122
Best
r2wagner
for a few weeks now there seems to be a problem with PICA translators in catalogs maintained by the GBV. I am not able to download metadata from the university libraries in Göttingen, Clausthal-Zellerfeld, TIB Hannover etc. using the PICA2 translator (try for example https://opac.sub.uni-goettingen.de/DB=1/XMLPRS=N/PPN?PPN=57161647X, http://opac.tib.uni-hannover.de/DB=1/XMLPRS=N/PPN?PPN=620088028 , http://lhclz.gbv.de/DB=1/XMLPRS=N/PPN?PPN=08727342X). It works with CoinS though.
To my surprise, there is no problem with the catalog of the Stabi in Berlin (see for example http://stabikat.de/DB=1/XMLPRS=N/PPN?PPN=57161647X ).
In the main GBV catalog there is no problem either, but here Zotero uses the "PICA" translator (and not "PICA2") (see for example http://gso.gbv.de/DB=2.1/PPNSET?PPN=57161647X ).
I had colleagues trying it out, and used Firefox 19 and 20 and Zotero 3.x and 4.0.3, but I don´t think the versions of the software have anything to do with it. Here is a report ID too, but I don´t think that it adds relevant information either:
Report ID 1135268122
Best
r2wagner
(if you're curious about the details - Zotero tries to access bibtex for the records in the PICA2 translator. In my tests all "Pica2" catalogs, i.e. the ones with trapezoid tabs in the middle rather than rectangular ones to the left, had bibtex export, but that's clearly not the case.)
Your version of Zotero will automatically update within 24hs, or you can update manually using the "Update Now" button in the "General" tab of the Zotero preferences. They'll all show up as PICA now.
Any further problems let us know & thanks for reporting
http://katalog.ub.uni-rostock.de/DB=1/PPNSET?PPN=254743781
<technical details, feel free to skip> so the reason this breaks for Rostock is that they have %D1 in the COinS, which breaks Firefox's decodeURIComponent, because it's not a valid utf-8 encoding (I suspect it's ISO 8859). If I understand the COinS specs correctly that isn't recommended, but it isn't strictly illegal (non utf-8 values _are_ illegal for the decodeURIComponent function, so FF is doing the right thing here. So two things:
1. We shouldn't have this break if the COinS throws an error
2. Is this something we can deal with or is non-utf encoding definitely wrong in COinS? My reading of http://www.niso.org/apps/group_public/download.php/6640/The OpenURL Framework for Context-Sensitive Services.pdf page 12 is that if they want to use non-utf they would need to specify that. <end technical details>
I'd still be curious what Aurimas or Simon have to say about the error itself.