pkp / ots

PKP XML Parsing Service
GNU General Public License v3.0
32 stars 19 forks source link

CrossRef API changes causing reference parsing issues? #105

Closed axfelix closed 6 years ago

axfelix commented 6 years ago

I don't think these were introduced by a recent commit -- hoping it wasn't changes to CrossRef's API -- but we're getting "Couldn't find the stage document" failures in BibtexreferencesJob and, if that's bypassed, Empty string failures in HtmlConversionJob:

PHP Warning: DOMDocument::loadXML(): Emptystring supplied as input in /var/www/html/module/ReferencesConversion/src/ReferencesConversion/Model/Converter/References.php on line 136 PHP Warning: DOMDocument::loadXML(): Empty string supplied as input in /var/www/html/module/NERExtraction/src/NERExtraction/Model/Converter/NERExtractor.php on line 113 PHP Warning: DOMDocument::loadXML(): Empty string supplied as input in /var/www/html/module/HtmlConversion/src/HtmlConversion/Model/Converter/Html.php on line82

@kaschioudi, let me know if you have a chance to look into this, I might not get to it until Thurs.

axfelix commented 6 years ago

OK, the empty string failures were due to https://github.com/pkp/ots/commit/6620d19ab01d8f92c196dba9222e1ece003e978f, dumb bug on my part and now fixed. Still throwing Couldn't find the stage document in Bibtexreferences for any document with references detected though and I can't see any recent changes that could be related to...

axfelix commented 6 years ago

I wonder if Perl/Parscit just broke on here or something due to upstream Ubuntu packaging... now that I fixed the other silly issue, I'm just seeing <ref><mixed-citation>Citation</mixed-citation></ref> in the doc. It's that or CrossRef API stuff I suppose...

axfelix commented 6 years ago

Nope, ParsCit itself still executes fine...

axfelix commented 6 years ago

https://github.com/CrossRef/rest-api-doc#good-manners--more-reliable-service

Probably relevant, I know we haven't looked at this OTS codepath since this change