From: Claudio V. C. <cv...@us...> - 2001-07-21 08:04:05
|
> -----Original Message----- > From: fir...@li... > [mailto:fir...@li...]On Behalf Of Tjelvar > Eriksson > Sent: Viernes 20 de Julio de 2001 17:19 > > In the ib_udf.sql the substr is commented to handle > strings of 32768 chars, but in the definition of the > function five rows later down, the funcion is declared > with the usual 80 chars. Is the ib_udf ok and can I > only fix the declaration? This function is so ill-behaved that I added substrlen that works with more rationale. You can create a declaration with higher limit and it should work. However, there's no reason to continue using such udf call. Firebird has implemented substring, built-in, ANSI SQL native, able to work with any charset and even able to take blobs of type binary or containing single-byte character sets. C. |