use optional dependencies

[thorinaboenke]

Task

#176

Description

• Moves all LLM-parser dependencies (openai, tenacity, scipy, scikit-learn, tiktoken, pandas) out of core into a new llm optional extra
• Moves dataframe backend dependencies (pandas, polars) into the dataframes optional extra
• Update docs

One code fix needed
persistency_saver.py and persistency/init.py were eagerly importingEventDataFrame/ChunkedEventDataFrame at module level, pulling pandas into every import of the core library. Fixed with a lazy_get_backend_cls() helper

How Has This Been Tested?

Core install only (uv sync, no extras): import detectmatelibrary succeeds; import openai, import pandas, and import LogBatcherParser all raise ModuleNotFoundError

llm extra (uv sync --extra llm): from detectmatelibrary.parsers.logbatcher import LogBatcherParser succeeds.

dataframes extra (uv sync --extra dataframes): import EventDataFrame succeeds.

Checklist

• This Pull-Request goes to the development branch.
• I have successfully run prek locally.
• I have added tests to cover my changes.
• I have linked the issue-id to the task-description.
• I have performed a self-review of my own code.