Re: ICU for global collation - Mailing list pgsql-hackers

From Finnerty, Jim
Subject Re: ICU for global collation
Date
Msg-id 3D155AD2-1500-40CC-9003-465FE6280B6A@amazon.com
Whole thread Raw
In response to Re: ICU for global collation  (Robert Haas <robertmhaas@gmail.com>)
Responses Re: ICU for global collation  ("Daniel Verite" <daniel@manitou-mail.org>)
List pgsql-hackers
Can we get some more consistent terminology around the term "locale"?

In ICU, the "locale" is just the first part of what we can pass to the "locale" parameter in CREATE COLLATION - the
partbefore the optional '@' delimiter.  The ICU locale does not include the secondary or tertiary properties, so it is
usuallyjust the country and the language, e.g. en_US (or en-US), but it can also be something like es_TRADITIONAL for
traditionalSpanish.  
 

I think it would be an improvement in clarity if we consistently use the term 'locale' to mean the same thing that ICU
meansby that term, and not to have the thing that we call the "locale" also include collation modifiers, or to imply
thata locale is the same thing as a collation.
 

    /Jim





pgsql-hackers by date:

Previous
From: Greg Stark
Date:
Subject: Re: Tablesync early exit
Next
From: "Daniel Verite"
Date:
Subject: Re: ICU for global collation