Missed TidRangePath in the print_path switch

Started by Andrei Lepikhovover 2 years ago2 messagesbugs
Jump to latest
#1Andrei Lepikhov
lepihov@gmail.com

Hi,

The TidRangePath node was introduced three years ago. However, path
printing machinery still does not support it. It isn't enjoyable to add
this string of code each time. I suggest adding it now.
Also, I propose not printing "???path" in the default case but throwing
a warning - see the patch in the attachment. It can help to avoid such
slip-ups next time.
It generally makes sense to generate such long switches automatically,
likewise nodeToString.

--
regards,
Andrey Lepikhov
Postgres Professional

Attachments:

tidrangepath.patchtext/plain; charset=UTF-8; name=tidrangepath.patchDownload+5-1
#2David Rowley
dgrowleyml@gmail.com
In reply to: Andrei Lepikhov (#1)
Re: Missed TidRangePath in the print_path switch

On Thu, 28 Sept 2023 at 18:07, Andrey Lepikhov
<a.lepikhov@postgrespro.ru> wrote:

The TidRangePath node was introduced three years ago. However, path
printing machinery still does not support it. It isn't enjoyable to add
this string of code each time. I suggest adding it now.
Also, I propose not printing "???path" in the default case but throwing
a warning - see the patch in the attachment. It can help to avoid such
slip-ups next time.
It generally makes sense to generate such long switches automatically,
likewise nodeToString.

Thanks for the report and patch.

I'll push the fix shortly. I'll leave off the WARNING thing for now.
Perhaps an Assert would draw more attention to the correct people. We
can debate that later as a master only change.

David