apostrophes in names are ignored by tools/names.js parser

[kgilbert-cmu]

A commit from my complexNames branch (6bdbdb4) tried to manually edit these names, but it's been reverted so that the data/names.js file will always be in agreement with the output of tools/names.js

The parser has a bug and I have a burning passion to find it. I don't know if it's a data source issue or a parsing issue yet.

[kgilbert-cmu]

I'd like to see if there's a programmatic fix somewhere in the parser code, but worst-case scenario I think we just add some more apostrophe exceptions to the nameFixes.

[dumbmatter]

It's not a bug in the parser, the issue is I'm using data from DraftExpress which has names like http://www.draftexpress.com/profile/ATorri-Shine-78199/

If RealGM data is more comprehensive and cleaner, maybe it'd be worth switching..

[kgilbert-cmu]

Thanks for confirming it's a data source issue.

May I start writing some exceptions into nameFixes?

ATorri and NDoye are the two giant sore thumbs to me. At least I can almost believe DAndre and DMarius.

[kgilbert-cmu]

I added the apostrophe'd first names to the nameFixes because it looks like there's only one of each. I could move them down to the unused fnFixes if you think they're common enough but my search of DraftExpress didn't bring up more than one.

Last names happened a lot so I wrote a completely new lnFixes section, based on your fnFixes.

[dumbmatter]

Thanks! That part of your PR looks good. Not sure about the rest :)

Is this all the names you think need to be altered?

[kgilbert-cmu]

As of right now, yes. They're exceptions, and it seems all of the other Irish names came correctly. When new names get added to DraftExpress (and if you ever decide to re-run your crawler, say, every 18 months) then I would want to scan the output to see if there are new "error types".

Proof:

~/basketball/src/js/data $ egrep [A-Z]{2} names.js
"USA": [
        ["ATorri", 65],
        ["DAndre", 5058],
        ["DMarius", 5059],
        ["DMarr", 5060],
        ["MBaye", 168],
        ["NDoye", 201],
        ["NDiaye", 76],
        ["NDoye", 77],
    "USA": [
        ["OBrien", 15659],
            first = {USA: g.names.first};
            last = {USA: g.names.last};