[FLINK-13284] [table-planner-blink] Correct some builtin functions' r…

[lincoln-lil]

What is the purpose of the change

Correct several builtin temporal functions' return type inference in Blink planner which should cascade a SqlTypeTransforms.FORCE_NULLABLE, otherwise it may derive wrong nullable info for constant input arg(s) that will cause a incorrect expression reduction.

Brief change log

• Update the type inference for FlinkSqlOperatorTable
• Add more test cases
• fix a NPE in SqlDateTimeUtils

Verifying this change

This change added tests and can be verified as follows:

• Added cases for TemporalTypesTest

Does this pull request potentially affect one of the following parts:

• Dependencies (does it add or upgrade a dependency): ( no)
• The public API, i.e., is any changed class annotated with @Public(Evolving): ( no)
• The serializers: ( no )
• The runtime per-record code paths (performance sensitive): ( no )
• Anything that affects deployment or recovery: JobManager (and its components), Checkpointing, Yarn/Mesos, ZooKeeper: ( no)
• The S3 file system connector: ( no)

Documentation

• Does this pull request introduce a new feature? ( no)
• If yes, how is the feature documented? (not applicable )

[flinkbot]

Thanks a lot for your contribution to the Apache Flink project. I'm the @flinkbot. I help the community
to review your pull request. We will use this comment to track the progress of the review.

Review Progress

• ❓ 1. The [description] looks good.
• ❓ 2. There is [consensus] that the contribution should go into to Flink.
• ❓ 3. Needs [attention] from.
• ❓ 4. The change fits into the overall [architecture].
• ❓ 5. Overall code [quality] is good.

Please see the Pull Request Review Guide for a full explanation of the review process.

The Bot is tracking the review progress through labels. Labels are applied according to the order of the review items. For consensus, approval by a Flink committer of PMC member is required

Bot commands

The @flinkbot bot supports the following commands:

• @flinkbot approve description to approve one or more aspects (aspects: description, consensus, architecture and quality)
• @flinkbot approve all to approve all aspects
• @flinkbot approve-until architecture to approve everything until architecture
• @flinkbot attention @username1 [@username2 ..] to require somebody's attention
• @flinkbot disapprove architecture to remove an approval you gave earlier

[flinkbot]

CI report:

• c41363e : FAILURE Build
• e4725a5 : SUCCESS Build

[godfreyhe]

Thanks for this PR @lincoln-lil . for stream job, I think "return null" is the general case (maybe for some special cases, users also want "throw exception" when processing illegal data). while for batch job, "throw exception" is more general ?! for the long term, maybe we should add a config to let users choose the mode. for this PR, it looks good to me.

[JingsongLi]

Thanks for this PR @lincoln-lil . for stream job, I think "return null" is the general case (maybe for some special cases, users also want "throw exception" when processing illegal data). while for batch job, "throw exception" is more general ?! for the long term, maybe we should add a config to let users choose the mode. for this PR, it looks good to me.

For batch job, "return null" is more popular too in hadoop sql engines.

[lincoln-lil]

@godfreyhe @JingsongLi Agree with you that we should offer a deterministic semantic for those 'dirty data', I think we can achieve this for two steps:

1. unify all the builtin functions' exception handling behavior for blink planner(since it differs with flink planner), I found two exception functions and will create another issue to fix it.
2. add a global configuration to support something like MySQL's strict/non-strict sql mode for exception handling includes numeric out-of-range and overflow and illegal inputs for sources. We can start a new thread to discuss it, what do you think?

[godfreyhe]

@godfreyhe @JingsongLi Agree with you that we should offer a deterministic semantic for those 'dirty data', I think we can achieve this for two steps:

1. unify all the builtin functions' exception handling behavior for blink planner(since it differs with flink planner), I found two exception functions and will create another issue to fix it.
2. add a global configuration to support something like MySQL's strict/non-strict sql mode for exception handling includes numeric out-of-range and overflow and illegal inputs for sources. We can start a new thread to discuss it, what do you think?

sounds good, look forward the discussion

[wuchong]

LGTM.

I also agree with starting a new thread to discuss invalid input data.