Apparently since "{,#}" is not a valid regexp expression the engine simply
interprets it as a literal and says 'AAA' != 'A{,4}'

http://www.postgresql.org/docs/9.3/interactive/functions-matching.html#FUNCTIONS-POSIX-REGEXP

Table 9-13. Regular Expression Quantifiers

Note the all of the { } expressions have a lower bound (whether explicit or
implied).

David J.




--
View this message in context: http://postgresql.1045698.n5.nabble.com/Regular-expression-question-with-Postgres-tp5812777p5812778.html
Sent from the PostgreSQL - general mailing list archive at Nabble.com.

Our regex documentation lists the following variants of bounds syntax:
{m}
{m,}
{m,n}
Nothing about {,n}. I rather imagine that the engine is deciding that
that's just literal text and not a bounds constraint ...

regression=# SELECT 'A{,4}' ~ '^A{,4}$';
?column?
----------
t
(1 row)

... yup, apparently so.

A look at the POSIX standard says that it has the same idea of what
is a valid bounds constraint:

When an ERE matching a single character or an ERE enclosed in
parentheses is followed by an interval expression of the format
"{m}", "{m,}", or "{m,n}", together with that interval expression
it shall match what repeated consecutive occurrences of the ERE
would match. The values of m and n are decimal integers in the
range 0 <= m<= n<= {RE_DUP_MAX}, where m specifies the exact or
minimum number of occurrences and n specifies the maximum number
of occurrences. The expression "{m}" matches exactly m occurrences
of the preceding ERE, "{m,}" matches at least m occurrences, and
"{m,n}" matches any number of occurrences between m and n,
inclusive.

regards, tom lane