Run 2023-05-21 updates from wikidata_update script

[mattpaz]

This update contains:

• updates from the wikidata_update.py script.
• additional error handling for the wikidata_update.py script to better handle upstream errors in votesmart data entries sourced from wikidata.

---

Incidentally, I observed that there are 81 records in legislators-current.yaml that don't have an opensecrets_id. At first glance, they all seem to be new entrants from the 118th congress. If there isn't a special mechanism for updating those ids, I might look at sporadically populating them on wikidata and creating a subsequent PR after each incremental update.

Secondarily, I wonder if a github action that would run wikidata_update.py on cron and create an automated PR might be interesting?

[JoshData]

Thanks!

The two JSON files in the alternate_formats directory should be removed. They don't exist on the main branch. They're generated by CI on the gh-pages branch only.

81 records in legislators-current.yaml that don't have an opensecrets_id. At first glance, they all seem to be new entrants from the 118th congress. If there isn't a special mechanism for updating those ids, I might look at sporadically populating them on wikidata and creating a subsequent PR after each incremental update.

Yeah that would be great. Sometimes legislators don't have opensecrets IDs when they are sworn in, usually after special elections. And in other cases I may have just been too busy to look it up when adding the legislator.

I wonder if a github action that would run wikidata_update.py on cron and create an automated PR might be interesting?

Not only that but many of the scripts need to be run periodically. So it could be helpful!

[mattpaz]

The two JSON files in the alternate_formats directory should be removed. They don't exist on the main branch. They're generated by CI on the gh-pages branch only.

Yes, I should have caught that. Just updated to remove them accordingly.

Yeah that would be great. Sometimes legislators don't have opensecrets IDs when they are sworn in, usually after special elections. And in other cases I may have just been too busy to look it up when adding the legislator.

Makes sense. No worries. Once this gets committed, I may circle back to this myself. Which do you think would be more valuable -- work on #886 or work on gaps in opensecrets ids?

I wonder if a github action that would run wikidata_update.py on cron and create an automated PR might be interesting?

Not only that but many of the scripts need to be run periodically. So it could be helpful!

Glad to hear I was on the right track. If I get time, I might circle back to this and see if I can chip in on that.

[JoshData]

Thanks.

work on #886 or work on gaps in opensecrets ids?

I'm ambivalent. 🤷🏻‍♂️ It's unclear if either are impactful at the moment. But always nice to have complete data.