Field Labeled CSV Reader

Parameters and methods of FieldLabeledCSVReader, which converts each CSV row into a field-labeled document.

FieldLabeledCSVReader is a reader class that converts CSV rows into field-labeled text documents.

Parameters

Parameter	Type	Default	Description
chunk_title	Optional[Union[str, List[str]]]	None	Title added at the top of each entry. A list rotates through entries
field_names	Optional[List[str]]	None	Custom labels for CSV fields. Column headers are used when not provided
format_headers	bool	True	Format column headers by replacing underscores with spaces and applying title case
skip_empty_fields	bool	True	Skip fields with empty values in the output

FieldLabeledCSVReader also accepts the base Reader constructor parameters, except chunk and chunking_strategy. Chunking is disabled; each CSV row becomes one document.

Methods

read()

Reads a CSV file and returns one field-labeled document per row.

Parameter	Type	Default	Description
file	Union[Path, IO[bytes]]	Required	Path to the CSV file or a binary file-like object
delimiter	str	","	Character used to separate fields in the CSV
quotechar	str	'"'	Character used to quote fields in the CSV
name	Optional[str]	None	Name override for the resulting documents

Open non-Path inputs in binary mode (rb). The reader decodes the returned bytes with its configured encoding.

async_read()

The async variant of read(). It accepts the same parameters, plus:

Parameter	Type	Default	Description
page_size	int	1000	Number of rows per page when paginating large files