|
[Rivet] rivet-findid KeyErrorDavid Grellscheid david.grellscheid at durham.ac.ukTue Jun 23 09:00:10 BST 2015
Hi Dmitri, I agree that the test was very useful, but a bug like this specific one shouldn't be blocking the rest of the Genser workflow. Maybe you can distinguish serious errors from advisory ones? Thanks, David On 23/06/15 06:19, Dmitri Konstantinov wrote: > Hello Andy, hello Benedikt, > > I would say that such kind of tests cost nothing for us. > As it takes a second to execute them. Then why not? :) > > Cheers, > Dima > > Sent from my iPad > >> On 23 Jun 2015, at 00:55, Andy Buckley <andy.buckley at cern.ch> wrote: >> >> Thanks for the report, Grigory: I've fixed it on the Rivet development >> branch and it'll be in the next release. >> >> I agree with Benedikt that this isn't obviously functionality that you >> need to be testing, but am still grateful that it flagged up a problem >> (due to a change in the Inspire content, apparently)! >> >> Cheers, >> Andy >> >> >>> On 22/06/15 16:31, Grigory Latyshev wrote: >>> Dear rivet developers, >>> >>> In GENSER project we provide a number of LCG installations of rivet for >>> experiments. For each version we have a set (or just one) of tests which >>> notify us in case of errors. >>> >>> Starting from 18-06-2015 we have found that rivet-findid test fails >>> (more detail on >>> http://cdash.cern.ch/testDetails.php?test=8919703&build=135824)! >>> The test is very simple: >>> rivet-findid 1003.0694 >>> >>> It should print details about Rivet user manual: >>> >>> $ rivet-findid 1003.0694 >>> title Rivet user manual >>> arxiv 1003.0694 >>> arxiv_url http://arxiv.org/abs/1003.0694 >>> spires 8576807 >>> inspire 847552 >>> inspire_url http://inspirehep.net/record/847552 >>> bibtex Buckley:2010ar >>> >>> >>> But now it fails with the following error: >>> >>> Traceback (most recent call last): >>> File "/build/jenkins/workspace/lcg_experimental/BUILDTYPE/Release/COMPILER/gcc48/LABEL/lcgapp-slc6-physical1/x86_64-slc6-gcc48-opt-install/MCGenerators/rivet/2.2.1/x86_64-slc6-gcc48-opt/bin/rivet-findid", line 187, in <module> >>> main() >>> File "/build/jenkins/workspace/lcg_experimental/BUILDTYPE/Release/COMPILER/gcc48/LABEL/lcgapp-slc6-physical1/x86_64-slc6-gcc48-opt-install/MCGenerators/rivet/2.2.1/x86_64-slc6-gcc48-opt/bin/rivet-findid", line 44, in main >>> RESULT = try_arxiv(id) >>> File "/build/jenkins/workspace/lcg_experimental/BUILDTYPE/Release/COMPILER/gcc48/LABEL/lcgapp-slc6-physical1/x86_64-slc6-gcc48-opt-install/MCGenerators/rivet/2.2.1/x86_64-slc6-gcc48-opt/bin/rivet-findid", line 109, in try_arxiv >>> ret = _search_inspire(url) >>> File "/build/jenkins/workspace/lcg_experimental/BUILDTYPE/Release/COMPILER/gcc48/LABEL/lcgapp-slc6-physical1/x86_64-slc6-gcc48-opt-install/MCGenerators/rivet/2.2.1/x86_64-slc6-gcc48-opt/bin/rivet-findid", line 163, in _search_inspire >>> result['bibtex'] = entries['z'] >>> KeyError: 'z' >>> >>> I have made a little investigation and found that rivet-findid looks on >>> page http://inspirehep.net/search?p=eprint+1003.0694&of=xm and tries to >>> find key 'z' in <subfield> in the following block ... >>> >>> <datafield tag="035" ind1=" " ind2=" "> >>> <subfield code="9">SPIRESTeX</subfield> >>> <subfield code="a">Buckley:2010ar</subfield> >>> </datafield> >>> >>> ... but now the code is "a". >>> >>> Manual replacement >>> >>> result['bibtex'] = entries['z'] -> result['bibtex'] = entries['a'] >>> >>> resolves the problem locally. >>> >>> But it practically impossible to reinstall all rivet versions in LCG stack. >>> >>> Could you please take a look? >>> It is not critical issue for us. >>> >>> Thank you! >>> >>> -- >>> Best regards, >>> Grigory Latyshev behalf on GENSER team >>> >> >> >> -- >> Dr Andy Buckley, Lecturer / Royal Society University Research Fellow >> Particle Physics Expt Group, University of Glasgow > _______________________________________________ > Rivet mailing list > Rivet at projects.hepforge.org > https://www.hepforge.org/lists/listinfo/rivet >
More information about the Rivet mailing list |