Open ksachs opened 6 years ago
I recognize the first one in the list, as we investigated it a bit with @salmanmaq and @michamos. It's due to the fact the matcher is not ignoring deleted records, so it detects an ambiguous match between https://labs.inspirehep.net/literature/9113 and https://labs.inspirehep.net/literature/431037, so it decides to assign the citation to none of them.
That cant be the only problem. E.g.
In [3]: search_pattern(p='773__p:"Eur.Phys.J." 773__v:"C75" 773__c:"1"')
intbitset([1300380])
In [7]: search_pattern(p="037:'1307.1347'")
intbitset([1241571])
In [8]: search_pattern(p='037:"CERN-2013-004"')
intbitset([1241571])
have only one record (as far as I see).
Mh. My best guess is that the cited record was not migrating successfully at the time of the experiment, so the matcher was not able to find it.
can we make sure these citations are resolvable if everything goes OK? Just to make sure there are no hidden bugs. And please fix the problem with the deleted records.
To be honest, I can't really tell the reason for not matching here. The cited records are present in the localhost on which I ran the experiment. :neutral_face:
I'll look in more detail though but nothing concrete so far.
Part of this issue (specifically the problem I mentioned in https://github.com/inspirehep/inspire-next/issues/3448#issuecomment-394729685) is addressed by https://github.com/inspirehep/inspire-next/pull/3462.
I don't see why labs doesn't find these citations