Scottish name matching

[ajparsons]

From slack:

it looks like the current parser spots speakers with an ID and ones without and records them differently in parsed/ but then the convert script ignores the ID (apart from for a cache) and does a name lookup on everything, which then causes this issue here. Presumably if it doesn’t have a speaker_scot_id we shouldn’t even bother trying to match on it?

Also a miss on 2025-06-18 - fo r “5. Douglas Lumsden (North East Scotland) (Con)” which is valid and needs to be fixed

[dracos]

Currently the name matching is in convert.py, parse.py only takes whatever is there (and the Douglas Lumsden type liens do indeed have a name="0" in the source, it's not that we're losing it). So I guess most consistent in that would be changing convert.py to a) use ID if present, b) try name match (stripping prefix/suffix) if not, c) not caring if it still doesn't match. But I could also see adding name matching to parse.py to get the ID into the parsed/ file at that point.