From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
---|---|
To: | Stephan Szabo <sszabo(at)megazone(dot)bigpanda(dot)com> |
Cc: | pgsql-general(at)postgresql(dot)org |
Subject: | Re: SELECT CAST(123 AS char) -> 1 |
Date: | 2008-02-16 15:47:24 |
Message-ID: | 19594.1203176844@sss.pgh.pa.us |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
Stephan Szabo <sszabo(at)megazone(dot)bigpanda(dot)com> writes:
> On Tue, 12 Feb 2008, Tom Lane wrote:
>> Also, section 6.10 <cast specification> defines an explicit cast to
>> a fixed-length string type as truncating or padding to the target
>> length (LTD):
> Are you sure that's the correct section to be using? Isn't that 6.10
> General Rules 5c which is if the source type is a fixed or variable
> length character string? Wouldn't the correct place for an int->char
> conversion be 5a or am I misreading it?
Hm, good point, so really we ought to have a separate casting path for
numeric types to char(n). However, this section still doesn't offer
any support for the OP's desire to auto-size the result; it says
that you get an error if the result doesn't fit in the declared
length:
> iv) Otherwise, an exception condition is raised: data exception-
> string data, right truncation.
regards, tom lane
From | Date | Subject | |
---|---|---|---|
Next Message | Robbie Thurman | 2008-02-16 15:53:20 | Re: |
Previous Message | Tom Lane | 2008-02-16 15:39:39 | Re: Timestamp indexes (why ">" or "between" does not use index?) |