Postgres normalize human name string

Question

Is there some function to normalize name string:

 select normalize('R.C. Bray') 

 'R. C. Bray'

Or the opposite direction?

The simple replace('. ', '.') or replace('.', '. ') is wrong and may be effect another data.

Answer 1

regexp_replace should be enough:

regexp_replace('R.C. Bray.', '([A-Z])(\.)(\w)', '\1\2 \3', 'g') 

This replaces a single upper case letter followed by a . directly followed by a "word character" by inserting a space between the dot and the character. It leaves dots that have a whitespace after them untouched.

Answer 2

select  regexp_replace(array_to_string(string_to_array('R.C. Bray', '.'), '. '), '[ ]+', ' ', 'g');

Can be work. But may be not a good method.