Ambiguous timezone abbreviations - parsing fails.

[guyboertje]

Some timezone abbreviations are ambiguous and Joda does not know which offset to use.

See https://www.timeanddate.com/time/zones/

One solution is to use mutate/gsub the abbreviation for the long form timezone, e.g. IST -> Asia/Kolkata then use ZZZ in the match pattern.

I propose to add a setting tz_abbreviation_gsub so a user can supply a map so replacements to be done before the matching is executed.