feat: DH-19489: Support structured Filters in Core Py client

[jmao-denver]

No description provided.

[github-actions]

No docs changes detected for be53c2e

[Copilot]

Pull request overview

This PR adds support for structured Filter objects in the Deephaven Core Python client, allowing users to construct type-safe table filters programmatically instead of using only string-based filter expressions.

Key changes:

• New filters.py module with Filter abstract base class and concrete implementations for comparison, logical, and pattern-matching operations
• Support for datetime types (Python datetime, numpy datetime64, pandas Timestamp) in filter values
• Case-sensitivity control for string comparison operations
• Integration of Filter objects into Table.where() and Query.where() methods

Reviewed changes

Copilot reviewed 8 out of 8 changed files in this pull request and generated no comments.

Show a summary per file

File	Description
py/client/pydeephaven/filters.py	New module implementing Filter classes and factory functions for creating structured filters
py/client/tests/test_filters.py	Comprehensive unit and integration tests for the filters module
py/client/pydeephaven/_table_ops.py	Added FilterOp class to handle structured Filter objects in table operations
py/client/pydeephaven/_table_interface.py	Updated where() method to accept both string and Filter objects
py/client/pydeephaven/table.py	Updated Table.where() signature to include Filter types
py/client/pydeephaven/query.py	Updated Query.where() signature to include Filter types
py/client/setup.py	Added pandas as a required dependency
py/client/requirements.txt	Added pandas to requirements

Comments suppressed due to low confidence (1)

py/client/pydeephaven/filters.py:1

• The FilterOp class is defined in filters.py but should be in _table_ops.py based on the diff context. This appears to be a misplaced code block - the FilterOp class definition shown in the diff for _table_ops.py (lines 356-383) is the correct location.

#

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[cpwright]

Should we make np and pandas a soft dependency? I think this is just for the time types in a filter? We are very loose with the version, so I don't think it actually is that awful; but we might consider it.

[jmao-denver]

In theory we can make them soft deps, but the code will be quite a bit messier and probably not worth the benefits.