Class ElCarroDocumentSaver (0.2.0) 

ElCarroDocumentSaver(
    elcarro_engine: langchain_google_el_carro.engine.ElCarroEngine,
    table_name: str,
    content_column: typing.Optional[str] = None,
    metadata_json_column: typing.Optional[str] = None,
)

A class for saving langchain documents into an ElCarro-Oracle database table.

Methods

ElCarroDocumentSaver

ElCarroDocumentSaver(
    elcarro_engine: langchain_google_el_carro.engine.ElCarroEngine,
    table_name: str,
    content_column: typing.Optional[str] = None,
    metadata_json_column: typing.Optional[str] = None,
)

ElCarroDocumentSaver allows for saving of langchain documents in an ElCarro-Oracle database. If the table doesn't exist, a table with default schema will be created. The default schema:

  • page_content (type: text)
  • langchain_metadata (type: JSON)
Parameters
Name Description
elcarro_engine ElCarroEngine

ElCarroEngine object to connect to the ElCarro-Oracle database.
table_name str

The name of table for saving documents.
content_column str

The column to store document content. Default: page_content. Optional.
metadata_json_column str

The name of the JSON column to use as the metadata’s base dictionary. Default: langchain_metadata. Optional.

_parse_row_from_doc

_parse_row_from_doc(
    column_names: typing.Iterable[str],
    doc: langchain_core.documents.base.Document,
    content_column: str = "page_content",
    metadata_json_column: str = "langchain_metadata",
    json_as_string=True,
) -> typing.Dict

Convert a Document into a dictionary according to the table schema: Store Doc.page_content in result[content_column] If a Doc metadata fields exist in the Table Schema, add them to the result

If there's a JSON metadata column in the table, unpack remaining Doc metadata
    to result[metadata_json_column]

add_documents

add_documents(docs: typing.List[langchain_core.documents.base.Document]) -> None

Save documents in the DocumentSaver table. Document’s metadata is added to columns if found or stored in langchain_metadata JSON column.

Args: docs (List[langchain_core.documents.Document]): a list of documents to be saved.

delete

delete(docs: typing.List[langchain_core.documents.base.Document]) -> None

Delete all instances of a document from the DocumentSaver table by matching the entire Document object.

Parameter
Name Description
docs List[langchain_core.documents.Document]

a list of documents