Reference for CSVParser

The CSVParser is dedicated to the parsing of comma-separated Value file (.csv). By default it will attempt to infer the delimiter used (comma, semicolon, ...). Otherwise you may specify the delimiter it should use.

Bases: AbstractParser

Parser for Comma-Separated Values file (.csv)

__init__(csv_delimiter=None, output_format='json_lines')

Initializes a sheet parser

Parameters:

Name	Type	Description	Default
csv_delimiter	str | None	The delimiter to consider to parse the .csv files. If None, we will try to guess what the delimiter is. Defaults to None.	None
output_format	Literal["markdown_table", "json_lines"]	the output format of the parsed document. - markdown_table : uses tabula to build a markdown-formatted table. - json_lines : each row of the table will be output as a JSON line. NOTE : consumes way more tokens as column names are repeated at each row. But easier to read for LLMs. Defaults to "json_lines".	'json_lines'

convert_df_to_json_lines(df) staticmethod

Converts a DataFrame to json lines.

Parameters:

Name	Type	Description	Default
df	DataFrame	the dataframe to convert.	required

Returns:

Name	Type	Description
str	str	the json lines.

convert_df_to_markdown_table(df) staticmethod

Converts a DataFrame to markdown. Wraps tabula's method pd.DataFrame.to_markdown() between pre and post processing. Preprocess : - Remove in text columns PostProcess : - Replace multiple spaces with 2 spaces.

   Args:
       df (pd.DataFrame): the dataframe to convert.

   Returns:
       str: a markdown formatted table.

parse_file(filepath)

Parses a csv file to markdown.

Parameters:

Name	Type	Description	Default
filepath	str	the path to the csv file.	required

Returns:

Name	Type	Description
MarkdownDoc	MarkdownDoc	the markdown-formatted csv.

parse_string(string)

Parses a string representing a csv file to markdown.

Parameters:

Name	Type	Description	Default
string	str	the csv-formatted string.	required

Returns:

Name	Type	Description
MarkdownDoc	MarkdownDoc	the markdown-formatted csv.

read_file(filepath)

Read the provided filepath. For a list of handled filetypes, refer to https://pandas.pydata.org/docs/reference/api/pandas.read_excel.html.

Parameters:

Name	Type	Description	Default
filepath	str	path to the file.	required

Returns:

Name	Type	Description
str	str	the csv file content as a string.