Re: [HACKERS] char_length()?

> Tatsuo Ishii <t-ishii(at)sra(dot)co(dot)jp> writes:
> > char_length()/octet_length() for char() datatype returns a character
> > length *except* the trailing blanks. Is this what the standard
> > expects? Oracle's length() returns 3 in the case below.
>
> AFAICT Oracle is right --- the spec just says
>
> i) If the data type of S is a character data type, then the
> result is the number of characters in the value of S.
>
> and I can't see anything there about stripping pad characters. You
> could ask for length(trim(S)) if you don't want to count blanks.

Ok, I have committed fixes for this. Maybe we should add this
incompatible changes to the release note:

char_length()/octet_length() for the char() datatype now returns a
character length (or byte length) including the trailing blanks.
--
Tatsuo Ishii