Add expression nullable support

[junli1026]

I hereby agree to the terms of the CLA available at: https://databend.rs/policies/cla/

Summary

Add expression nullable support.

Changelog

• Improvement

Related Issues

Fixes #issue

Test Plan

Stateless Tests

[databend-bot]

Thanks for the contribution!
I have applied any labels matching special text in your PR Changelog.

Please review the labels and make any necessary changes.

[databend-bot]

Thanks for the contribution!
I have applied any labels matching special text in your PR Changelog.

Please review the labels and make any necessary changes.

[vercel]

This pull request is being automatically deployed with Vercel (learn more).
To see the status of your deployment, click below or on the icon next to each commit.

🔍 Inspect: https://vercel.com/databend/databend/Enz2YQm8UTFZteu3EBr3Q24aPmNB
✅ Preview: Canceled

[Deployment for 35957b0 canceled]

[codecov-commenter]

Codecov Report

Merging #3452 (35957b0) into main (ee13a4e) will increase coverage by 0%.
The diff coverage is 60%.

@@          Coverage Diff          @@
##            main   #3452   +/-   ##
=====================================
  Coverage     61%     61%           
=====================================
  Files        610     611    +1     
  Lines      34185   34220   +35     
=====================================
+ Hits       20996   21033   +37     
+ Misses     13189   13187    -2     

Impacted Files	Coverage Δ
query/src/sql/statements/statement_create_table.rs	66% <ø> (ø)
common/planners/src/plan_expression.rs	80% <60%> (-3%)	⬇️
query/src/clusters/cluster.rs	40% <0%> (-1%)	⬇️
cli/src/error.rs	32% <0%> (-1%)	⬇️
common/functions/src/scalars/strings/length.rs	0% <0%> (ø)
common/functions/src/scalars/strings/string.rs	97% <0%> (+<1%)	⬆️
...mmon/functions/src/aggregates/aggregate_min_max.rs	56% <0%> (+1%)	⬆️
metasrv/src/store/meta_raft_store.rs	80% <0%> (+2%)	⬆️
common/functions/src/aggregates/aggregate_sum.rs	89% <0%> (+3%)	⬆️
common/functions/src/aggregates/aggregate_avg.rs	94% <0%> (+3%)	⬆️
... and 2 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update ee13a4e...35957b0. Read the comment docs.

[sundy-li]

new line needed.

[sundy-li]

Nullable is annoyment.

ClickHouse takes Nullable as a separated datatype.
Yet MySQL takes Nullable as a property for datatype.

So in the function, there are two kinds of ways to handle nullable:

1. if args are not nullable, the result should not be nullable except some special functions (isNull.. )
2. if args are not nullable, the result can be nullable.

I prefer the first one, thus we can control the return type.

[sundy-li]

query group by nullable(col) may take much more memory/costs than query group by col.

[junli1026]

Nullable is annoyment.

ClickHouse takes Nullable as a separated datatype. Yet MySQL takes Nullable as a property for datatype.

So in the function, there are two kinds of ways to handle nullable:

1. if args are not nullable, the result should not be nullable except some special functions (isNull.. )

Sure, will address. Do we have an example of the 'isNull' ? -- the inputs are not nullable but output is null ?

2. if args are not nullable, the result can be nullable.

I prefer the first one, thus we can control the return type.

[sundy-li]

MySQL can produce nullable result by none nullable args.

[junli1026]

MySQL can produce nullable result by none nullable args.

Right, that is why I just check the function's nullable() return. My initial thoughts are just like, checking the input, if any of them is nullable, the output should be nullable. BTW, I think we return +inf for '3/0', which makes more sense than mysql.

Also aggregate function is a little special too, that, input can be nullable, but output is most not nullable.

[databend-bot]

Wait for another reviewer approval

[BohuTANG]

Thank you @junli1026 !

[databend-bot]

Wait for another reviewer approval