[enhancement](function) floor/ceil/round/round_bankers can use column as scale argument #34391
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
Thank you for your contribution to Apache Doris. Since 2024-03-18, the Document has been moved to doris-website. |
superdiaodiao
force-pushed
the
enhance_round
branch
from
a804e99 to
908a495
Compare
![github-actions[bot]](https://avatars.githubusercontent.com/in/15368?s=60&v=4){kind=small}
superdiaodiao
force-pushed
the
enhance_round
branch
from
df5f558 to
6313577
Compare
superdiaodiao
force-pushed
the
enhance_round
branch
from
6313577 to
9ed83f2
Compare
Contributor
|
clang-tidy review says "All clean, LGTM! 👍" |
superdiaodiao
force-pushed
the
enhance_round
branch
from
9ed83f2 to
472f4db
Compare
Contributor
Author
|
run buildall |
Contributor
|
clang-tidy review says "All clean, LGTM! 👍" |
|
TeamCity be ut coverage result: |
TPC-H: Total hot run time: 41175 ms |
TPC-DS: Total hot run time: 186675 ms |
superdiaodiao
force-pushed
the
enhance_round
branch
from
472f4db to
ef697bb
Compare
Contributor
Author
|
run buildall |
|
TeamCity be ut coverage result: |
TPC-H: Total hot run time: 41854 ms |
superdiaodiao
force-pushed
the
enhance_round
branch
from
ef697bb to
3fc43b4
Compare
TPC-DS: Total hot run time: 185793 ms |
Contributor
|
clang-tidy review says "All clean, LGTM! 👍" |
superdiaodiao
force-pushed
the
enhance_round
branch
from
3fc43b4 to
adb88e9
Compare
Contributor
|
clang-tidy review says "All clean, LGTM! 👍" |
Contributor
Author
|
run buildall |
|
TeamCity be ut coverage result: |
TPC-DS: Total hot run time: 186675 ms |
Contributor
Author
Oh, you get the point. No wonder I just wanted to print the log but got nothing. Thanks~~~ |
Contributor
|
clang-tidy review says "All clean, LGTM! 👍" |
Contributor
Author
|
run buildall |
|
TeamCity be ut coverage result: |
TPC-DS: Total hot run time: 187296 ms |
Contributor
|
https://dev.mysql.com/doc/refman/8.0/en/mathematical-functions.html#function_ceil ceil function in mysql and pg do not support using second argument. So, @superdiaodiao could you please do a small modification so that we can detect the above situation for function ceil? Just need to modify the logic in planner and do not register two argument for ceil on BE. |
Contributor
Author
Doris has been support two arguments for a long time, I think we shouldn't delete this behavior in this enhancement PR. What's more, Clickhouse also supports two argument on ceil(). |
Contributor
ok, we can keep the code in this pr, and we can change behavior in another pr if necessary. |
Contributor
Author
Yes, I agree with this. |
zhiqiang-hhhh
approved these changes
May 6, 2024
Contributor
zhiqiang-hhhh
left a comment
zhiqiang-hhhh
left a comment
There was a problem hiding this comment.
LGTM, excellent job
Contributor
|
PR approved by anyone and no changes requested. |
Contributor
Author
|
run feut |
github-actions
bot
added
the
approved
Contributor
|
PR approved by at least one committer and no changes requested. |
ByteYue
pushed a commit
to ByteYue/doris
that referenced
this pull request
May 15, 2024
… as scale argument (apache#34391)
16 tasks
924060929
added a commit
that referenced
this pull request
Aug 5, 2025
1. fix compute wrong scale of round_bankers/truncate/round/dround/floor/dfloor/ceil/dceil, introduced by #32746 and #34391 this sql will throw exception: ```sql create table test_round_bankers_scale( id bigint, a decimal(32,6) ) properties ("replication_num" = "1"); select round_bankers(a, 1-2) from test_round_bankers_scale order by id (1105, 'errCode = 2, detailMessage = (172.20.48.119)[INTERNAL_ERROR]output type not match expr type , col name , expected type Nullable(Decimal(32, 6)) , real type Nullable(Decimal(32, 0))') ``` The scale of return type of the round_bankers function depends on the type of its arguments. If the second argument is an integer constant, that constant is used as the scale. If it is not a constant, the scale of the first argument is used instead. This leads to a problem: for expressions like `1 - 2`, the scale used in the computation differs before and after constant folding, which can cause the Backend to throw an exception: type mismatch So we need to fold the arguments into constants before binding the function, to ensure that the scale will not change 2. keep idempotent of function signature. when we select a signature for a function, we should not change signature after do some rewrite
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
7 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Proposed changes
Issue Number: #32755
Like PR: #32746, we enhanced floor/ceil/round/round_bankers can use column as scale argument.
Meanwhile, we moved these round based function register into new file named round.cpp to accelerate compile process and enable code more readable.
What's more, we made truncate() can run with single const argument avoiding BE crash, like branch2.0's behavior or Clickhouse's.
Docs refer PR: apache/doris-website#620
Further comments
If this is a relatively large or complex change, kick off the discussion at dev@doris.apache.org by explaining why you chose the solution you did and what alternatives you considered, etc...