Re: BUG #6185: Segmentation fault with NULL string columns

On Tue, Aug 30, 2011 at 8:21 PM, Heikki Linnakangas
<heikki(dot)linnakangas(at)enterprisedb(dot)com> wrote:
>
> to_ascii_encname is marked as STRICT, which means that it returns NULL
> on NULL input, without ever calling the underlying C function. The
> "to_ascii" function that you created is not marked as STRICT, so the
> to_ascii_encname C function is called on NULL input. It's not prepared
> for that, and crashes, because it's not supposed to be called on NULL
> input.

I figured that 5 minutes after reporting. I tried to sent a reply email
but it seems it didn't arrive. Thanks anyway.

> Why do you create your own 'internal' language function like that in
> the first place? Just use the existing to_ascii function.

The default to_ascii expects a string as a first argument. As our DB
encoding is UTF8 we need to filter everything through
convert_to(..., 'LATIN1'). Because convert_to function returns a
"bytea" type, we cannot use to_ascii in vanilla flavour.

We are not too fond with the PostgreSQL specifics (although the explicit
::text cast didn't work), so we found that function definition on a
wiki and it worked.

But now, unless you know a more correct way to do it, it works like a
charm by appending "STRICT" to the function definition.

Best regards and sorry for the noise.

--
Isaac Jurado
Internet Busines Solutions eConcept