import api: use edition ID from IA metadata if it exists

[scottbarnes]

Related: #8290 #8289

Fix

Technical

Currently ia_import() in the import API does not use Open Library edition IDs from Internet Archive metadata because it is using the wrong field -- metadata['openlibrary'] instead of metadata[openlibrary_edition].

This PR fixes that, and ensures we won't make a new edition when trying to import an IA record. Additionally, it avoids at least two further database calls that would have been made in an attempt to find a match.

Testing

Try to import an IA record and see that 'step 2' now works as intended and gets the existing Open Library edition ID from the Internet Archive metadata, rather than relying on finding a match by searching the Open Library database.

❯ http POST http://localhost:8080/api/import/ia identifier==007exoticlocatio0000arno Cookie:$OL_COOKIE
HTTP/1.1 200 OK
[...]

{
    "edition": {
        "key": "/books/OL8582004M",
        "status": "modified"
    },
    "success": true,
    "work": {
        "key": "/works/OL8839644W",
        "status": "matched"
    }
}

@cdrini

[mekarpeles]

➕ 1 -- openlibrary is a deprecated key name on archive.org, openlibrary_[edition|work] are preferred

[cdrini]

Lgtm!