feat(expr): add support for make_date/time/timestamp

[KeXiangWang]

I hereby agree to the terms of the RisingWave Labs, Inc. Contributor License Agreement.

What's changed and what's your intention?

As title
https://www.postgresql.org/docs/current/functions-datetime.html

Checklist

• I have written necessary rustdoc comments
• I have added necessary unit tests and integration tests
• I have added test labels as necessary. See details.
• I have added fuzzing tests or opened an issue to track them. (Optional, recommended for new SQL features Sqlsmith: Sql feature generation #7934).
• My PR contains breaking changes. (If it deprecates some features, please create a tracking issue to remove them in the future).
• All checks passed in ./risedev check (or alias, ./risedev c)
• My PR changes performance-critical code. (Please run macro/micro-benchmarks and show the results.)

• My PR contains critical fixes that are necessary to be merged into the latest release. (Please check out the details)

Documentation

• My PR needs documentation updates. (Please use the Release note section below to summarize the impact on users)

Release note

If this PR includes changes that directly affect users or other significant modifications relevant to the community, kindly draft a release note to provide a concise summary of these changes. Please prioritize highlighting the impact these changes will have on users.

make_date ( year int, month int, day int ) → date
Create date from year, month and day fields (negative years signify BC)
make_date(2024, 1, 31) → 2024-01-31

make_time ( hour int, min int, sec double precision ) → time
Create time from hour, minute and seconds fields
make_time(1, 45, 30.2) → 01:45:30.2

make_timestamp ( year int, month int, day int, hour int, min int, sec double precision ) → timestamp
Create timestamp from year, month, day, hour, minute and seconds fields (negative years signify BC)
make_timestamp(2024, 1, 31, 1, 45, 30.2) → 2024-01-31 01:45:30.2

[wangrunji0408]

Generally LGTM.

@xiangjinwu please double check. 🥹

[xiangjinwu]

Rest LGTM!

[xiangjinwu]

Did not quite get this part. Date / timestamp / timestamptz shall all follow the same rule:

• In PG SQL, the year before 1 is -1. There is no year 0. It is printed as 1 BC.
• In ISO 8601 (chrono crate), the year before 1 is 0. It can be printed as 1 BC (year_ce instead of year)

Practically it is okay to be incompatible with PG for BC years. But the logic here is both wrong and more complicated

In short, year + 1 as long as year < 0 when passing a PG SQL input to the chrono crate. This yields the correct value in memory during input. Whether showing X+1 BC or -X shall be adjusted on the output side. (same issue discovered multiple times recently #14832 (comment))

[KeXiangWang]

After discussion with xiangjin, we conduct +1 for both Date and Timestamp/Timestamptz. However, as our ToText for Date and Timestamp/Timestamptz are different, the output of make_date and make_timestamp/timestamptz are different.

SELECT make_timestamp(-5, 02, 29, 08, 15, 55.33);
----
-0004-02-29 08:15:55.330

SELECT make_date(-5, 2, 29);
----
0005-02-29 BC

We'll fix it later in another PR by modify the impl of ToText forTimestamp/Timestamptz.

[xiangjinwu]

LGTM

[BugenZhao]

In postgres we have date field value out of range: -4-02-29. Note the -4 here. Are we behaving correctly?

[KeXiangWang]

> SELECT make_date(-4, 02, 29);
ERROR:  date field value out of range: -3-02-29
> SELECT make_timestamp(-4, 02, 29, 08, 15, 55.33);
ERROR:  date field value out of range: -4-02-29

Pg behaves differently on make_date and make_timestamp. Seems to be a misaligned design. So I think it's OK if we don't follow it strictly.

[BugenZhao]

Rest LGTM