Postgres Pro
Downloads
Home   >   mailing lists

Thread: format_datum debugging function

format_datum debugging function

From
Paul Jungwirth
Date:
Hi Hackers,

Often in a debugger I've wanted to way to print Datums, in particular non-trivial ones like range 
types. This came up a lot when I was working on multiranges, and I've wished for it lately while 
working on UPDATE/DELETE FOR PORTION OF. But all the obvious approaches are inlined functions or 
preprocessor macros, so they aren't available. Usually I wind up giving up on gdb and adding elog 
statements. Once or twice I've copy/pasted from the three or four levels of nested macros to make 
gdb do what I wanted, but it's a pain.

I assumed printing a Datum was easy, and I was the only one who didn't know how to do it. But 
perhaps not. The conversation about print.c [1] made me think I should propose a way to make it 
easier. This function takes a Datum and the appropriate out function, and returns a char *. So you 
can do this:

(gdb) call format_datum(range_out, $1)
$2 = 0x59162692d938 "[1,4)"

I assume a patch like this doesn't need documentation. Does it need a test? Anything else?

[1] https://www.postgresql.org/message-id/flat/7d023c20-6679-44bd-b5f7-44106659bd5a%40eisentraut.org

Yours,

-- 
Paul              ~{:-)
pj@illuminatedcomputing.com

Re: format_datum debugging function

From
Aleksander Alekseev
Date:
Hi Paul,

> [...] This function takes a Datum and the appropriate out function, and returns a char *. So you
> can do this:
>
> (gdb) call format_datum(range_out, $1)
> $2 = 0x59162692d938 "[1,4)"
>
> I assume a patch like this doesn't need documentation. Does it need a test? Anything else?

I think you forgot to attach the patch. Or is it just a proposal?

--
Best regards,
Aleksander Alekseev

Re: format_datum debugging function

From
Paul Jungwirth
Date:
On 8/12/24 04:32, Aleksander Alekseev wrote:
>> [...] This function takes a Datum and the appropriate out function, and returns a char *. So you
>> can do this:
>>
>> (gdb) call format_datum(range_out, $1)
>> $2 = 0x59162692d938 "[1,4)"
>>
>> I assume a patch like this doesn't need documentation. Does it need a test? Anything else?
> 
> I think you forgot to attach the patch. Or is it just a proposal?

Sorry, patch attached here.

Yours,

-- 
Paul              ~{:-)
pj@illuminatedcomputing.com
Attachment

Re: format_datum debugging function

From
Jelte Fennema-Nio
Date:
On Mon, 12 Aug 2024 at 23:15, Paul Jungwirth
<pj@illuminatedcomputing.com> wrote:
> On 8/12/24 04:32, Aleksander Alekseev wrote:
> >> (gdb) call format_datum(range_out, $1)
> >> $2 = 0x59162692d938 "[1,4)"
> >>
> >> I assume a patch like this doesn't need documentation. Does it need a test? Anything else?
> >
> > I think you forgot to attach the patch. Or is it just a proposal?
>
> Sorry, patch attached here.

+1 for the idea. And the code looks trivial enough. I think this
should also contain a print_datum function too though.

Re: format_datum debugging function

From
Peter Eisentraut
Date:
On 12.08.24 23:15, Paul Jungwirth wrote:
> On 8/12/24 04:32, Aleksander Alekseev wrote:
>>> [...] This function takes a Datum and the appropriate out function, 
>>> and returns a char *. So you
>>> can do this:
>>>
>>> (gdb) call format_datum(range_out, $1)
>>> $2 = 0x59162692d938 "[1,4)"
>>>
>>> I assume a patch like this doesn't need documentation. Does it need a 
>>> test? Anything else?
>>
>> I think you forgot to attach the patch. Or is it just a proposal?
> 
> Sorry, patch attached here.

I don't think it's safe to call output functions at arbitrary points 
from a debugger.  But if you're okay with that during development, say, 
then I think you could just call OidOutputFunctionCall(F_RANGE_OUT, $1)?