Updated meta_from_soup_row to handle parsing names with two commas

[snake-plissken]

Fix for issue BurntSushi/nfldb#214. Added a max-split parameter to the string's split function so that player names are split once.

This solves the issue of unpacking the parts of a name, but now technically the player names in the database will be kind of weird, since the player's suffix will be a part of the first name. See the name which causes the problem: Leno, Charles, Jr

Edit: Thought about the further fix on the train this morning and looked into what would happen when there is only one comma, like most player names contain. A name with one comma would cause a ValueError. For some reason I thought suffix would just come back as None if there wasn't a second comma.

Potentially a further fix could be:

name = tds[1].a.get_text().strip()
    if ',' not in name:
        last_name, first_name = name, ''
    else:
        last_name, first_name, suffix = map(lambda s: s.strip(), name.split(','))
        last_name = last_name + ', ' + suffix