Closed tarrow closed 7 years ago
Don't suppose you'd be able to bump the version and push it to pypy if you have a a moment?
On Fri, Jun 16, 2017 at 10:29 PM, Gregory Stupp notifications@github.com wrote:
Closed #10 https://github.com/SuLab/WikidataIntegrator/issues/10 via ea40f8e https://github.com/SuLab/WikidataIntegrator/commit/ea40f8e97d0d849c454c0279f9891ae91d57001c .
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/SuLab/WikidataIntegrator/issues/10#event-1127540891, or mute the thread https://github.com/notifications/unsubscribe-auth/AHA028qpgm_aEL_dF6OXr81ii1OwmUQlks5sEvOhgaJpZM4N8K5O .
Done
It seems that at least for some items the query string used to get metadata from EuropePMC is malformed for content in PubMedCentral.
For example if you init with:
wdi_helpers.PubmedItem('PMC3139704', id_type='PMC')
Then the query string to the EBI appears to be:
http://www.ebi.ac.uk/europepmc/webservices/rest/search?query=EXT_ID:PMC3139704%20AND%20SRC:PMC&resulttype=core&format=json
This returns no results.
I think the correct string should be like:
http://www.ebi.ac.uk/europepmc/webservices/rest/search?query=PMCID:PMC3139704&resulttype=core&format=json
Something similar has already been done for DOI. I guess this may need to be done for the other identifiers in PubmedItem.id_types.
I think EXT_ID and SRC may only hold one value corresponding to whichever database EuropePMC got the metadata from first. This means that if you don't search by an item's 'main' id then you get no results.