Once we parse a VIAF file into an intermediate format, we need some way of boiling down all the names and opinions into a ContributorData object that can be written to the database.

Here's what we want to get:

• We don't need a sort name because we got one from OCLC, but let's pick one just in case.
• We would really like a display name, but VIAF doesn't generally distinguish between sort and display names – it's just telling you the strings that certain libraries assigned to people. This is the tough part, but display name is the real prize.
• If we can find it, it would be good to get a person's wiki name, birth and death date.

Here are the rules/caveats I can think of:

• If more libraries use a name it's more likely to be a good name.
• Libraries from English-speaking countries are more likely to have a name we can use.
• If a name contains a comma, it's more likely to be a sort name than a display name (but some display names do contain commas)
• Superficially appealing display names sometimes have extra disambigiation stuff at the end, e.g. "Laura Wright romancière" or "Black & Decker (Firm)".
• UNIMARC records have split up the name into parts that can be assembled into a display name, which is incredibly useful., However my records indicate that UNIMARC records are flaky somehow – I don't know how.