bibstem parsing in match_w_metadata is flawed

[seasidesparrow]

The utility match_w_metadata.py has an embedded query of the journals api to check whether a bibstem is present in the database and has a valid pubtype. However, prior to the query, the bibstem is determined by taking the bibcode and extracting only bibcode[4:9]:

ADSDocMatchPipeline/adsdocmatch/match_w_metadata.py

Line 76 in 9d21568

bibstem = metadata.get('bibcode', '')[4:9].strip('.')

This will match journal-type bibstems, but will not match any conf, rept, book or other bibcodes other than the standard journal bibstem. If we are expecting docmatch pipeline to things other than journal articles, this bug will prevent that.