bug in categories

[vuillaut]

Given the following code and the master version of the library (commit 2a0e00f)

import arxivscraper.arxivscraper as ax
scraper = ax.Scraper(category='stat',
                     date_from='2010-01-01',
                     date_until='2012-02-01',
                     t=10, 
                     filters={'categories':['physics:hep-ex', 'physics:astro-ph'],
                              'abstract':['deep learning']})
output = scraper.scrape()

print(output[0]['categories'])

Returns

cs.lg cs.ai cs.cv cs.ir stat.ml

It seems the categories filter did not work.

[Mahdisadjadi]

Thank you for raising the issue. The current implementation of the filter is a logical OR. So if deep learning is in the abstract, it still saves the record even when categories are not among the specified categories. I'll add a note to README.

• I think this is the manuscript https://arxiv.org/abs/1108.3298 for future reference.