Wikidata importer from ID

[LPorcaro]

It seems that there is an inconstitencies among methods for importing data from Wikidata.
Indeed, in
https://github.com/trompamusic/ce-data-import/blob/master/ceimport/loader.py#L294
it uses the wikidata URL to get the person info

while in
https://github.com/trompamusic/ce-data-import/blob/master/ceimport/loader.py#L460
it uses the wikidata ID

I assume that every method the wikidata URL should be passed as input

[alastair]

It looks like I may have made some mistake when merging the two repositories. I'll have a look at it again.

As an overview, there are two ways of loading data, but that's because we get data in different ways from different sites.

• the main source of data should be wikidata. We get a description from the first paragraph of a related wikipedia article
• imslp and cpdl have links to wikipedia
• musicbrainz has a link to wikidata
• when we have a wikipedia link, we convert it to a wikidata link
• then we load the data from wikidata
• then we load data from wikipedia, using the wikidata id and language code to find the article, not the original wikipedia url

what these two loaders should do is:

1. given a wikipedia url, look up its wikidata id
2. import the person from wikidata
3. import the person from wikipedia (using the wikidata id + language code to find the article)

but it seems that they're not doing the right thing, because they're just calling the same method. I'll update it.