Joe Conway wrote:
> Tom Lane wrote:
>> A couple of minor thoughts:
>> * You could reduce the ugliness of many of the tests by introducing a
>> variant strcmp function that does the "right" things with NULL inputs.
>> It might also be worth adding a variant pstrdup that takes a NULL.
> I had thoughts along those lines -- it would certainly make the code
> more readable. I'll go ahead and do that but it won't be in time for a
> 26 October beta2.
I'm not quite ready to commit this, mostly because I'd like to give the
rest of tablefunc.c the once-over for similar issues related to not
checking for NULL return values from SPI_getvalue(). But it is close
enough if needed for a beta2 tomorrow -- let me know if we plan to
bundle up beta2 and I'll get it in.
In response to
pgsql-patches by date
|Next:||From: Zdenek Kotala||Date: 2007-10-26 09:17:04|
|Subject: Fix ecpg dependency on postgres.h|
|Previous:||From: Joe Conway||Date: 2007-10-26 04:40:04|
|Subject: Re: [GENERAL] Crosstab Problems|
pgsql-general by date
|Next:||From: Ow Mun Heng||Date: 2007-10-26 06:24:46|
|Subject: Query_time SQL as a function w/o creating a new type|
|Previous:||From: Patrick TJ McPhee||Date: 2007-10-26 04:40:35|
|Subject: Re: Selecting K random rows - efficiently!|