[FLINK-38674][table] Support schema definition for Materialized tables while CREATE

[snuyanzin]

What is the purpose of the change

This PR is starting PR for FLIP-550. Adds support for schema definition while CREATE MATERIALIZED TABLE.
Also revisits a bit hierarchy of classes for materialized table converters in a way similar to tables.

After that it will allow to have syntax in way

CREATE MATERIALIZED TABLE [catalog_name.][db_name.]table_name
[( { <schema_definition> | <column_list> } )]
...
AS <select_statement>
<schema_definition>:
  [ <column_definition>[, ...n] ],
  [ <watermark_definition> ],
  [ <table_constraint>[, ...n] ]
  
<column_list>:
  col_name1 [, column_name2, ...]

Brief change log

parser and materialized table converters

Verifying this change

A number of tests added for parser, converters

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, Kubernetes/Yarn, ZooKeeper: (no)
• The S3 file system connector: (no)

Documentation

• Does this pull request introduce a new feature? (yes)
• If yes, how is the feature documented? (docs / JavaDocs)

[flinkbot]

CI report:

• 6cd8dcc Azure: SUCCESS

Bot commands

The @flinkbot bot supports the following commands:

• @flinkbot run azure re-run the last Azure build

[davidradl]

As this as called from validate, would it make sense to validate that all the columns are identifiers rather than assume?

[snuyanzin]

there is no need for that since it is not a validation.
The validation, that if first is identifier then others are as well, happens on parse level
I submitted a test for that
https://github.com/snuyanzin/flink/blob/9963205624c657c6943fc5e726d4b1c151dfd941/flink-table/flink-sql-parser/src/test/java/org/apache/flink/sql/parser/MaterializedTableStatementParserTest.java#L51-L57

[davidradl]

I am curious - can we have a watermark when there are no columns? I am not sure what a no column materialized table means - is this allowed?

[snuyanzin]

we still can have something like

CREATE MATERIALIZED TABLE users_shops (watermark for ts as ts - interval '2' SECOND)
 FRESHNESS = INTERVAL '30' SECOND
 AS SELECT current_timestamp() as ts, 1 as shop_id, 2 as user_id

[twalthr]

include a data type here? I think specifying both the name and data type is a very common task.

[snuyanzin]

added example below