Street names with directions (North, South, East, West) are often abbreviated with a single letter, or with a single letter and a dot. It seems to me that parse should return the same normalized address regardless of the input format, but this is not the case. For example:
> StreetAddress::US.parse("555 North Fake Street", :informal=>true)
=> 555 N Fake St
> StreetAddress::US.parse("555 N. Fake Street", :informal=>true)
=> 555 N. Fake St
> StreetAddress::US.parse("555 N Fake Street", :informal=>true)
=> 555 N Fake St
Note that if a dot is included, the dot is left. Whereas if the dot is not included, or the full directional name is provided, the result is an abbreviation without the dot. In my opinion, the dotted version should normalize to the non dotted version, thus:
> StreetAddress::US.parse("555 N. Fake Street", :informal=>true)
=> 555 N Fake St
Street names with directions (North, South, East, West) are often abbreviated with a single letter, or with a single letter and a dot. It seems to me that
parse
should return the same normalized address regardless of the input format, but this is not the case. For example:Note that if a dot is included, the dot is left. Whereas if the dot is not included, or the full directional name is provided, the result is an abbreviation without the dot. In my opinion, the dotted version should normalize to the non dotted version, thus: