feat(api): add describe method to compute summary stats of table expressions

[jitingxu1]

Description of changes

Add summary stats for a table. Only calculate stats for numeric, string, and bool columns.

• name
• type
• count (counts and nulls)

For numeric columns:

• min
• mean
• std
• percentile
• max

For category columns:

• mode
• unique

For Boolean columns:

• mean

My questions:

• Not sure if it is a good practice to use execute() here. I have another way to implement this, but have to handle the null columns.
• Is unit test necessary for this, noticed we do not have tests for info()

Issues closed

Resolves #8459

[jitingxu1]

@cpcloud soft ping for the second review after rewriting the implementation. Thanks a lot for your feedback and time.

[cpcloud]

Can you add a test for this somewhere? Ideally wherever we test the info method (I think this is in test_generic.py).

[jitingxu1]

Can you add a test for this somewhere? Ideally wherever we test the info method (I think this is in test_generic.py).

Changes made:

• added unit test.
• Stats calculation now exclusively targets numeric, string, and boolean columns, disregarding others.

TODOs:
Certain backends encounter issues due to inadequate support for functionalities like quantile, mode, or standdev.

• The following backends need to implement mode to facilitate the full functionality of describe:
  clickhouse, pyspark, clickhouse, risingwave, impala, oracle
• Both mode and StandardDev are required for these backends: exasol, druid

Other backends require both mode and quantile: "datafusion", "impala", "trino", "mysql", "mssql", "trino", "flink"

[cpcloud]

LGTM. I'll rebase it and merge on green.