rag.core.query
==============
.. py:module:: rag.core.query
.. autoapi-nested-parse::
Query / chat engine over the papers Chroma index.
Attributes
----------
.. autoapisummary::
rag.core.query.RETRIEVAL_OVERFETCH
rag.core.query.PAPERS_TEXT_QA_TEMPLATE
rag.core.query.PAPERS_REFINE_TEMPLATE
rag.core.query.CORPUS_SYNTHESIS_PROMPT
rag.core.query.AUTHOR_SYNTHESIS_PROMPT
Classes
-------
.. autoapisummary::
rag.core.query.ProseNodePostprocessor
rag.core.query.QueryResult
Functions
---------
.. autoapisummary::
rag.core.query.require_openai_api_key
rag.core.query.build_llm
rag.core.query.build_query_engine
rag.core.query.ask_llamaindex
rag.core.query.ask
Module Contents
---------------
.. py:data:: RETRIEVAL_OVERFETCH
:type: Final
:value: 3
.. py:data:: PAPERS_TEXT_QA_TEMPLATE
.. py:data:: PAPERS_REFINE_TEMPLATE
.. py:class:: ProseNodePostprocessor(/, **data: Any)
Bases: :py:obj:`llama_index.core.postprocessor.types.BaseNodePostprocessor`
Keep retrieved chunks that look like readable paper prose.
Author questions drop acknowledgement / thanks chunks and prefer
earlier pages (title-page author lists over end-matter).
:ivar keep: Maximum number of prose nodes to retain after filtering.
:vartype keep: int
.. py:attribute:: keep
:type: int
:value: None
.. py:class:: QueryResult
Answer plus grounding citations.
:ivar answer: Model response text.
:vartype answer: str
:ivar citations: Retrieved sources.
:vartype citations: list of Citation
:ivar citations_markdown: Pre-formatted markdown for UIs.
:vartype citations_markdown: str
.. py:attribute:: answer
:type: str
.. py:attribute:: citations
:type: list[rag.core.citations.Citation]
.. py:attribute:: citations_markdown
:type: str
.. py:function:: require_openai_api_key() -> str
Load ``.env`` if present and return ``OPENAI_API_KEY``.
:returns: API key value.
:rtype: str
:raises EnvironmentError: If the key is missing or empty.
.. py:function:: build_llm(*, config: rag.core.config.RagConfig, api_key: str | None = None) -> llama_index.llms.openai.OpenAI
Construct the OpenAI chat LLM.
:param config: Provides ``llm_model_name``.
:type config: RagConfig
:param api_key: Explicit key; otherwise read from the environment.
:type api_key: str or None
:returns: LlamaIndex OpenAI LLM wrapper.
:rtype: OpenAI
.. py:function:: build_query_engine(*, config: rag.core.config.RagConfig, embed_model: llama_index.core.embeddings.BaseEmbedding | None = None, llm: Any | None = None, index: llama_index.core.VectorStoreIndex | None = None) -> llama_index.core.query_engine.BaseQueryEngine
Build a retrieval-augmented query engine.
:param config: Retrieval and model settings.
:type config: RagConfig
:param embed_model: Defaults to the HuggingFace model from config.
:type embed_model: BaseEmbedding or None
:param llm: Defaults to OpenAI from config + env key.
:type llm: Any or None
:param index: Defaults to loading the persistent Chroma index.
:type index: VectorStoreIndex or None
:returns: Configured LlamaIndex query engine.
:rtype: BaseQueryEngine
.. py:data:: CORPUS_SYNTHESIS_PROMPT
:type: Final
:value: Multiline-String
.. raw:: html
Show Value
.. code-block:: python
"""You are summarizing a researcher's full PDF library.
The catalog below is the COMPLETE set of papers ({paper_count} files), not a retrieval sample.
---------------------
{catalog_markdown}
---------------------
If the files span multiple application domains (medical imaging, networks, environment, finance, news, engineering, …) but share supervised learning, deep learning, graph ML, or similar methods, the unifying theme is applied machine learning. State that umbrella first, then name several distinct domains from the filenames/titles. Do not treat one paper as the whole corpus.
Question: {query_str}
Answer: """
.. raw:: html
Show Value
.. code-block:: python
"""You are identifying authors from a researcher's PDF library.
The catalog below is the COMPLETE set of papers ({paper_count} files).
Each line is a filename plus opening-page text (title and usually authors).
---------------------
{catalog_markdown}
---------------------
SPIE footers like 'edited by' name volume editors, not the paper's authors. People thanked for revising a manuscript are not authors. If one person is first author or appears on most papers, they are the main author of this library — say so, then name frequent co-authors. If the question names a specific paper or topic, answer for that paper only. Do not invent names that are not in the catalog.
Question: {query_str}
Answer: """
.. raw:: html