Description

• Add an optional filter_fn constructor argument to ExtracterNode

• If defined the filter_fn receives a DataFrame and returns a list[bool] mask to indicate which rows the LLM Engine should operate on. The mask will be saved into the LLM Context, the task handler then re-applies the mask to set the correct output rows

• This is different than our current FilterDetectionsStage impl in a few key ways:

1. The purpose of FilterDetectionsStage is to exclude certain rows from the output.
2. FilterDetectionsStage filteres according to a threshold
3. FilterDetectionsStage is specific to MultiMessage and it's slice operations.

Includes code changes from PRs #1538 & #1544

The changes specific to this PR are in :

• morpheus/_lib/include/morpheus/llm/llm_context.hpp
• morpheus/_lib/src/llm/llm_context.cpp
• morpheus/_lib/llm/module.cpp
• morpheus/llm/nodes/extracter_node.py
• morpheus/llm/task_handlers/simple_task_handler.py

By Submitting this PR I confirm:

• I am familiar with the Contributing Guidelines.
• When the PR is ready for review, new or existing tests cover these changes.
• When the PR is ready for review, the documentation is up to date with these changes.