Re: LIKE optimization and locale

Bruce Momjian writes:

> Why don't we just spin through all 255 locale values, and find the
> lowest value that is greater than comparison target.

The issue is not that the 255 extended ASCII characters have a different
ordering in various locales (although that's part of it).  The real
problem lies with multi-character collating elements, context dependent
collation order, multi-pass sorting algorithms, etc.  I'm almost convinced
that it is not possible to do any such optimization as we had for the most
general case.

-- 
Peter Eisentraut      peter_e@gmx.net       http://yi.org/peter-e/