search

ffdev-info / wikidp-issues

An issues repository for resolving issues in Wikidata around the records relating to Digital Preservation
GNU General Public License v3.0
1 stars 0 forks source link

Have we described Wikidata parsing heuristics well-enough? (linting) #21

Open ross-spencer opened 3 years ago

ross-spencer commented 3 years ago

Description of problem

We need to describe how Siegfried's/Roy's Wikidata processing heuristic works.

Largely, any result we get from ./roy build -wikidatadebug in the debug output that returns a bad heuristic needs a rationale and the output doesn't show how we arrive at a decision yet. Other debug output should be self-describing but can also be audited to garner an opinion.

ross-spencer commented 3 years ago

Changing the title of this issue and I realized I have described the messages in the Wiki. the next question for me is whether the messages are described well-enough? What else do we need. (NB. I'd like to propose auditing the linting module and consider refactoring this as soon as the next Siegfried release is out. It feels like there could be some better de-coupling of the work at least).