Langroid

* Fix windows logger issue 341
* OpenAIGPT local model detection, use "local/" not "local", and "litellm/" not "litellm" (comments in 346)

DocChatAgentConfig.filter_fields param: specify which fields are "filterable"

[DocChatAgentConfig](https://github.com/langroid/langroid/blob/main/langroid/agent/special/doc_chat_agent.py): new param
`add_fields_to_content`.

This lets you add additional fields into the "main" content field of a document. E.g. if a movie-related Document has fields `year`,
`genre`, `rating`, these can be added to the main `content` field,
as key=value pairs, e.g. year=1989, genre=Crime. (The original fields are left intact, so can still be used for filtering).
Adding fields to the content can help with retrieval since these would be taken into account when creating embeddings,
as well as during keyword/lexical search.
See test `test_doc_chat_add_content_fields` here: https://github.com/langroid/langroid/blob/main/tests/main/test_doc_chat_agent.py

* New Task control signals: PASS_TO, SEND_TO (in addition to the existing PASS, DONE)
- PASS_TO:<recipient> => send current pending message to <recipient>
- SEND_TO:<recipient> <content> => send <content> to <recipient>
(Tests coming soon)
* Agent: tool_messages are now inserted into ChatDocument as soon as they are identified
* [VectorStore](https://github.com/langroid/langroid/blob/main/langroid/vector_store/base.py): `compute_from_docs` method to compute a
pandas-expression on documents
* [LanceRagTaskCreator](https://github.com/langroid/langroid/blob/main/langroid/agent/special/lance_rag/lance_rag_task.py): creates a
3-agent RAG system consisting of: QueryPlanner, QueryPlanCritic, LanceRAGAgent.
This illustrates a fairly complex 3-agent workflow and is a good reference.

* DocChatAgentConfig: remove max_context_tokens param
* misc: remove "nofunc" GPT models

LanceDB Related:
* LanceDB filtered search examples:
https://github.com/langroid/langroid/blob/main/examples/docqa/lance-filtered-movies.py
https://github.com/langroid/langroid/blob/main/examples/docqa/lance-filtered-gh-issues.py

These showcase:
- LanceDB ingest of pandas dataframe directly
- Langroid 2-agent setup where LanceFilterAgent decides on a filter (and retries if no answers) plus possibly rephrased query, and LanceDocChatAgent does the actual answering using RAG on the LanceDB, combining Lancedb-native filtering (via SQL conditions) vector-search, plus lexical (kw/bm25 search).

Minor:
* Tweaks in LanceDocChatAgent prompt
* RecipientTool: return err msg instead of raising