tree: fix decimal strings for 0 with exponents < -6

[otan]

Release note (bug fix): Fix a bug where 0 with exponents < -6 would display
as 0E(exponent) instead of printing all 0s, e.g. 0E-7 should be 0.0000000.

Informs #102217

[cockroach-teamcity]

This change is 

[rafiss]

@otan thanks for picking this up! are there any risks here, or should we just go for this approach?

[otan]

i'm fine with it if you are ;)

[otan]

(but we'll want to merge cockroachdb/apd#128 first)

[rafiss]

does this fix #102217 or are there more cases to deal with?

i'm surprised the ::decimal cases in pkg/cmd/generate-binary/main.go weren't affected. could we add a a test there that shows the fix in action?

Reviewable status: complete! 0 of 0 LGTMs obtained (waiting on @otan)

---

pkg/sql/logictest/testdata/logic_test/decimal line 480 at r1 (raw file):

0          -1         -0                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                      0     -0
0          4.2        0E+1                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                    0.0   0
0          Infinity   0.000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000   0     0

i tried this query on postgres and got this result for this row:

         0 |  Infinity |                       0 |    0 |         0

[otan]

does this fix #102217 or are there more cases to deal with?

this fixes the display issue, yeah

i'm surprised the ::decimal cases in pkg/cmd/generate-binary/main.go weren't affected. could we add a a test there that shows the fix in action?

yep

Reviewable status: complete! 0 of 0 LGTMs obtained (waiting on @rafiss)

---

pkg/sql/logictest/testdata/logic_test/decimal line 480 at r1 (raw file):

Previously, rafiss (Rafi Shamim) wrote…

i tried this query on postgres and got this result for this row:

         0 |  Infinity |                       0 |    0 |         0

this is a bug in our implementation more than anything, and is unrelated to this PR.

pg:

otan=# select 0::decimal / 'infinity'::decimal;
 ?column?
----------
        0
(1 row)

crdb, before this pr:

root@:26257/defaultdb> select 0::decimal / 'infinity'::decimal;
  ?column?
------------
  0E-2019
(1 row)

i'm guessing our use of HighPrecisionContext to divide Infinity is incorrect (it's also incorrect for all the other cases). Seems outside the scope of what I need out of this PR though.

[rafiss]

Reviewable status: complete! 0 of 0 LGTMs obtained (waiting on @otan)

---

pkg/sql/logictest/testdata/logic_test/decimal line 480 at r1 (raw file):

Previously, otan (Oliver Tan) wrote…

this is a bug in our implementation more than anything, and is unrelated to this PR.

pg:

otan=# select 0::decimal / 'infinity'::decimal;
 ?column?
----------
        0
(1 row)

crdb, before this pr:

root@:26257/defaultdb> select 0::decimal / 'infinity'::decimal;
  ?column?
------------
  0E-2019
(1 row)

i'm guessing our use of HighPrecisionContext to divide Infinity is incorrect (it's also incorrect for all the other cases). Seems outside the scope of what I need out of this PR though.

hm ok - but do you have a sense of what is needed in apd to make 0.00000...0 display as 0? would be nice to address later

[otan]

pkg/sql/logictest/testdata/logic_test/decimal line 480 at r1 (raw file):

Previously, rafiss (Rafi Shamim) wrote…

hm ok - but do you have a sense of what is needed in apd to make 0.00000...0 display as 0? would be nice to address later

it's not apd we need to change. it's all the execution stuff that makes 0 / infinity not use HighScalePrecision to divide.

[otan]

pkg/sql/logictest/testdata/logic_test/decimal line 480 at r1 (raw file):

Previously, otan (Oliver Tan) wrote…

it's not apd we need to change. it's all the execution stuff that makes 0 / infinity not use HighScalePrecision to divide.

(we need to change the builtin or whatever that does this)

[otan]

bors r=rafiss

[otan]

filed #102299

[otan]

pkg/sql/logictest/testdata/logic_test/decimal line 480 at r1 (raw file):

Previously, otan (Oliver Tan) wrote…

it's not apd we need to change. it's all the execution stuff that makes 0 / infinity not use HighScalePrecision to divide.

actually it may be Divide in apd, stuff like 0::decimal / 5 performs correctly.

[craig]

Build succeeded:

• Bazel Essential CI (Cockroach)

[blathers-crl]

Encountered an error creating backports. Some common things that can go wrong:

1. The backport branch might have already existed.
2. There was a merge conflict.
3. The backport branch contained merge commits.

You might need to create your backport manually using the backport tool.

---

error setting reviewers, but backport branch blathers/backport-release-23.1-102299 is ready: POST https://api.github.com/repos/cockroachdb/cockroach/pulls/103640/requested_reviewers: 422 Reviews may only be requested from collaborators. One or more of the teams you specified is not a collaborator of the cockroachdb/cockroach repository. []

Backport to branch 23.1.x failed. See errors above.

---

_{🦉 Hoot! I am a Blathers, a bot for CockroachDB. My owner is dev-inf.}