Re: proposal: minscale, rtrim, btrim functions for numeric - Mailing list pgsql-hackers

From Pavel Stehule
Subject Re: proposal: minscale, rtrim, btrim functions for numeric
Date
Msg-id CAFj8pRCsteYAbo_o3=xhhQ0RshH+8GHi-QLWqzJUwb=zBByOXQ@mail.gmail.com
Whole thread Raw
In response to Re: proposal: minscale, rtrim, btrim functions for numeric  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-hackers


po 6. 1. 2020 v 18:22 odesílatel Tom Lane <tgl@sss.pgh.pa.us> napsal:
Pavel Stehule <pavel.stehule@gmail.com> writes:
> út 10. 12. 2019 v 13:56 odesílatel Karl O. Pinc <kop@meme.com> napsal:
>> I'm marking it ready for a committer.

> Thank you for review

Pushed with minor adjustments.  Notably, I didn't like having
get_min_scale() depend on its callers having stripped trailing zeroes
to avoid getting into a tight infinite loop.  That's just trouble
waiting to happen, especially since non-stripped numerics are seldom
seen in practice (ones coming into the SQL-level functions should
never look like that, ie the strip_var calls you had are almost
certainly dead code).  If we did have a code path where the situation
could occur, and somebody forgot the strip_var call, the omission
could easily escape notice.  So I got rid of the strip_var calls and
made get_min_scale() defend itself against the case.  It's hardly
any more code, and it should be a shade faster than strip_var anyway.

Thank you very much

Maybe this issue was part of ToDo list

Pavel


                        regards, tom lane

pgsql-hackers by date:

Previous
From: Tom Lane
Date:
Subject: Re: [PATCH] fix a performance issue with multiple logical-decoding walsenders
Next
From: Tom Lane
Date:
Subject: Re: Removing pg_pltemplate and creating "trustable" extensions