Github user wuchong commented on the issue:
https://github.com/apache/flink/pull/2653
Hi @fhueske @twalthr , I have addressed all the comments and made the
following changes:
1. Forbid TableFunction implemented by Scala object, since the `collect` is
called on a singleton. It will be error-prone in some concurrent cases.
2. Check correct errors if a SQL query refers to a function in FROM that
has not been registered or which is a ScalarFunction.
3. Make `TableFunction` and `ScalarFunction` clean, **remove
`UserDefinedFunction`**, and move eval relative functions and
`createSqlFunction` to utils.
4. Rename `ScalarFunctions` to `SqlFunctions` because contains also
TableFunction logic.
5. Restructure tests. Test Java Table API by comparing the RelNode of two
tables. And check SQL API's DataSetRel and DataStreamRel via `TableTestBase`
utils. The tests reduced into `stream/UserDefinedTableFunctionITCase`,
`batch/UserDefinedTableFunctionITCase`, and
`stream/UserDefinedTableFunctionTest` , `batch/UserDefinedTableFunctionTest`.
6. Scala Table API implicitly convert `TableFunction` into
`TableFunctionCall`. `TableFunctionCall` is not an `Expression` or
`LogicalNode`, but is like `GroupWindow`, can be visible to the users (contains
API such `as(...)`).
7. Fix the hierarchy type extraction problem.
8. Rebase the code and fix conflicts.
9. ...
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
