fix: Handling of column types for Presto, Trino, et al.

[john-bodley]

SUMMARY

This is a rehash of @brouberol's #26782 which was later reverted in #28613 as it was problematic with Trino, given both the PrestoEngineSpec and TrinoEngineSpec derive from the PrestoBaseEngineSpec.

The TL;DR is typically most SQLAlchemy dialects (including Trino) use native SQLAlchemy types for encoding column types, whereas PyHive is unconventional and uses str. I asked a similar question years ago and you can see here how these types are materialized.

I hypothesize that part of the confusion when @brouberol's was authoring their PR was that the ResultSetColumnType TypedDict was incorrect, alluding to the fact that the type field was Optional[str] as opposed to Optional[Union[str, TypeEngine, type[TypeEngine]]. This PR simply ensures that in the PrestoBaseEngineSpec class—which handles Hive, Presto, and Trino—that we only try to coerce str types to a SQLAlchemy type (if available).

BEFORE/AFTER SCREENSHOTS OR ANIMATED GIF

TESTING INSTRUCTIONS

Added unit tests.

ADDITIONAL INFORMATION

• Has associated issue:
  • Fixes [sqllab] Tableschemaview fails to load with presto using parquet format files #25636
  • Fixes Failure on creation of a new Presto dataset due to failure to load columns #25962
  • Fixes [sqllqb] Superset always fires "show create view" although it is a table which results in "error while fetching table metadata" #26740
• Required feature flags:
• Changes UI
• Includes DB Migration (follow approval process in SIP-59)
  • Migration is atomic, supports rollback & is backwards-compatible
  • Confirm DB migration upgrade and downgrade tested
  • Runtime estimates and downtime expectations provided
• Introduces new feature or API
• Removes existing feature or API

[codecov]

Codecov Report

Attention: Patch coverage is 87.50000% with 1 lines in your changes are missing coverage. Please review.

Project coverage is 83.50%. Comparing base (76d897e) to head (6a35459).
Report is 158 commits behind head on master.

Files	Patch %	Lines
superset/db_engine_specs/presto.py	80.00%	1 Missing ⚠️

Additional details and impacted files

@@             Coverage Diff             @@
##           master   #28653       +/-   ##
===========================================
+ Coverage   60.48%   83.50%   +23.01%     
===========================================
  Files        1931      522     -1409     
  Lines       76236    37497    -38739     
  Branches     8568        0     -8568     
===========================================
- Hits        46114    31313    -14801     
+ Misses      28017     6184    -21833     
+ Partials     2105        0     -2105     

Flag	Coverage Δ
hive	49.12% <37.50%> (-0.05%)	⬇️
javascript	?
mysql	77.16% <62.50%> (?)
postgres	77.30% <62.50%> (?)
presto	53.67% <37.50%> (-0.13%)	⬇️
python	83.50% <87.50%> (+20.02%)	⬆️
sqlite	76.75% <62.50%> (?)
unit	58.97% <75.00%> (+1.35%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[john-bodley]

Same logic as before just condensed.

[admsev]

@gooroodev review

[gooroodev]

1. Summary of Changes

The pull request addresses the handling of column types for Presto and Trino engines in the Superset project. The key changes include:

• Presto Engine:
  • Handling of column types in the get_children and where_latest_partition methods.
  • Added logic to convert string column types to SQLAlchemy types dynamically.
• Trino Engine:
  • Added a similar handling of column types in the where_latest_partition method.
  • Added unit tests for the where_latest_partition method.
• Typing Changes:
  • Introduced a new type alias SQLType in superset_typing.py to represent SQLAlchemy column types.
  • Updated ResultSetColumnType to use SQLType.

2. Issues, Bugs, or Typos

Issue 1:

• File: superset/db_engine_specs/presto.py
• Code:

  match = pattern.match(cast(str, column["type"]))

  Problem: The use of cast(str, column["type"]) is unnecessary if column["type"] is already a string.
  Improvement:

  match = pattern.match(column["type"])

Issue 2:

• File: superset/db_engine_specs/presto.py
• Code:

  if isinstance(col_type, str):
      if col_type_class := getattr(types, col_type, None):
          col_type = col_type_class()
      else:
          col_type = None

  Problem: The assignment of None to col_type if col_type_class is not found can be simplified.
  Improvement:

  if isinstance(col_type, str):
      col_type_class = getattr(types, col_type, None)
      col_type = col_type_class() if col_type_class else None

3. General Review of Code Quality and Style

• Code Quality: The code changes are well-structured and address the problem of handling different column types effectively. The use of type hinting and type checking improves code readability and maintainability.
• Code Style: The code adheres to PEP 8 standards, making it clean and easy to read. The use of modern Python features like type hinting and := operator (walrus operator) is commendable.
• Testing: The addition of unit tests for the where_latest_partition method in both Presto and Trino engines ensures that the changes are tested and validated.

Additional Suggestions

• Documentation: Consider adding docstrings to the modified methods to explain the purpose and functionality of the changes.
• Error Handling: The use of broad exception handling (raise Exception) could be improved by raising more specific exceptions to provide better context in case of errors.

Overall, the pull request is well-implemented and addresses the problem effectively. The suggested improvements are minor and aim to further enhance the code quality.

Yours, Gooroo.dev. To receive reviews automatically, install Github App