I don't know what SQL2008 says about SUBSTRING, but SQL2003 says in
ISO/IEC 9075-2:2003 (E), 6.29 , General Rules,
5), page 261:
d) If R [the regular expression] does not contain exactly two
occurrences of the two-character sequence consisting of E [the escape
character], each immediately fol
"Roman Kononov" writes:
> test=# select substring('34' from '(2|3)#"4#"' for '#');
> substring
> ---
> 3
> (1 row)
Hmm. I guess we need to translate ( and ) to non-capturing parens.
> test=# select substring('^' from '#"^#"' for '#');
> substring
> ---
>
> (1 row)
> test=#
The following bug has been logged online:
Bug reference: 5257
Logged by: Roman Kononov
Email address: kono...@ftml.net
PostgreSQL version: 8.4.2
Operating system: GNU/Linux x86_64
Description:wrong results of SUBSTRING with SQL regular expressions
Details:
test=# se
