feat: parse decimal in expr.

[youngsofun]

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

trying to split this pr, a drawback is that effect of some changes need this pr to test.

• fix(function): fix function register order #10343 vip
• fix: function point_in_polygon require float64. #10338
• fix(function): decimal divide handle 'divide by 0'. #10339
• fix(expression): fix MAX/MIN of Decimal. #10340
• ci: list_all_builtin_functions_ordered. #10341
• fix: decimal to decimal bugs #10345

Summary

• choose the minimal type for numeric literal, large number try to fit into decimal (may truncate trailing digits after the point), if overflow, use float64
• 1 + -1 result in plus(1, -1) instead of plus(1, minus(1))
• add auto cast rule Decimal->Float32/64
• to_variant(decimal) result in number.

notice:

• agg func like std(1.0), manually cast in test for now.

not in this pr

• decimal to int (not auto)
• string to decimal

Closes #10220
#10262

[vercel]

The latest updates on your projects. Learn more about Vercel for Git ↗︎

1 Ignored Deployment

Name	Status	Preview	Comments	Updated
databend	⬜️ Ignored (Inspect)	Visit Preview		Mar 5, 2023 at 2:23AM (UTC)

[youngsofun]

float64 to float32 is here https://github.com/datafuselabs/databend/blob/main/src/query/functions/src/scalars/arithmetic.rs#L301

it comes last in the number_to_number functions (not include decimal)

[youngsofun]

@b41sh @sundy-li it was my mistake(not formaily with jsonb and variant): to_variant is fixed here

1941669