Open mraross opened 3 years ago
After investigation, it was decided not to define these abbreviations as they they will cause incorrect matches or impact performance or both.
This seems to relate to #170 - if garbage collection doesn't solve these, then it likely isn't enough for the correctly spelled versions, either. Also, if we want to enable spell-correction on postal elements, moving the handling to the lexer and parser would allow for that. Really, the initial postal handling was very specific, and for something like a postal code, a regex makes the most sense. However, with other changes that have been made, most or all of the postal junk logic would likely work better in the standard lexer and parser.
Agreed.
The following non-street-type abbreviations were identified in the rejected address analysis: