Re: PoC/WIP: Extended statistics on expressions

On 6/11/21 6:55 AM, Noah Misch wrote:
> On Sun, Jun 06, 2021 at 09:13:17PM +0200, Tomas Vondra wrote:
>>
>> On 6/6/21 7:37 AM, Noah Misch wrote:
>>> On Sat, Mar 27, 2021 at 01:17:14AM +0100, Tomas Vondra wrote:
>>>> OK, pushed after a bit more polishing and testing.
>>>
>>> This added a "transformed" field to CreateStatsStmt, but it didn't mention
>>> that field in src/backend/nodes.  Should those functions handle the field?
>>>
>>
>> Yup, that's a mistake - it should do whatever CREATE INDEX is doing. Not
>> sure if it can result in error/failure or just inefficiency (due to
>> transforming the expressions repeatedly), but it should do whatever
>> CREATE INDEX is doing.
>>
>> Thanks for noticing! Fixed by d57ecebd12.
> 
> Great.  For future reference, this didn't need a catversion bump.  readfuncs.c
> changes need a catversion bump, since the catalogs might contain input for
> each read function.  Other src/backend/nodes functions don't face that.  Also,
> src/backend/nodes generally process fields in the order that they appear in
> the struct.  The order you used in d57ecebd12 is nicer, being more like
> IndexStmt, so I'm pushing an order change to the struct.
> 

OK, makes sense. Thanks!

-- 
Tomas Vondra
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company