emily.DocSearch
===============

.. py:module:: emily.DocSearch

.. autoapi-nested-parse::

   Created on Thu May  7 12:26:18 2026

   @author: Dr Peter J Bleackley



Attributes
----------

.. autoapisummary::

   emily.DocSearch.splitter


Classes
-------

.. autoapisummary::

   emily.DocSearch.DocSearch


Functions
---------

.. autoapisummary::

   emily.DocSearch.parse


Module Contents
---------------

.. py:data:: splitter

.. py:function:: parse(sentence: str) -> list[str]

   Splits sentence into lower-case words with punctuation removed

   :param sentence: Sentence to be parsed
   :type sentence: str

   :returns: Sentence as a list of lower-case words
   :rtype: list[str]


.. py:class:: DocSearch(embedding_url: str, reranking_url: str, vector_dir: str, collection_name: str, index_dir: Optional[str] = None)

   Overall search component


   .. py:attribute:: vector_index


   .. py:attribute:: reranker


   .. py:attribute:: sentence_tokenizer


   .. py:method:: sentences(text: str) -> list[list[str]]

      Splits a text into sentences, and then each sentence into a list of lower-case strings

      :param text: A document to be split into sentences.
      :type text: str

      :returns: Each sentence in the document is represented by a list of strings
      :rtype: list[list[str]]



   .. py:method:: add_documents(corpus: collections.abc.AsyncIterable[tuple[str, str]])
      :async:


      Adds documents to the database

      :param corpus: Documents as tuples of (filename,text)
      :type corpus: AsyncIterable[tuple[str,str]]

      :rtype: None.



   .. py:method:: __call__(query: str, top_k: int = 10) -> pandas.Series
      :async:


      Searches for documents relevant to query. Finds the 2*top_k best matches
      from the vector database, the automatically thresholded reposnses from
      each of the Okapi and Cooccurrence indices, retrieves the text and reranks them

      :param query: Text to query for.
      :type query: str
      :param top_k: Number of results to return. The default is 10.
      :type top_k: int, optional

      :returns: The reranker scores of the top_k best matching documents, indexed by their filenames.
      :rtype: pd.Series



   .. py:method:: save(path: str)

      Saves search indices

      :param path: Directory in which to save indices.
      :type path: str

      :rtype: None.



   .. py:method:: clear()
      :async:


      Clears indices

      :rtype: None.