hypothesis-csv

hypothesis-csv is an extension to the hypothesis framework. The goal of this framework is to offer a flexible tool to perform generative-testing/property-based testing for software that accepts CSV files as an input.

Description

hypothesis-csv is designed with two main use cases in mind:

• test software that accepts a wide spectrum of CSV formats and doesn't make assumptions on the content (i.e. CSV parsers)
• test software that accepts very specific CSV formats and makes assumption on the content, types and header fields.

It provides two strategies, data_rows and csv, to pursue this goal, both contained in the strategies module. Please refer to their documentation and the examples for more details.

Examples

Generate arbitrary, non-empty CSV

from hypothesis_csv.strategies import csv
@given(csv=csv())
def test_my_csv_parse(csv):
    parsed_csv=my_csv_parser(csv)
    assert ...

Generate CSV of a given size (5 columns x 20 rows)

from hypothesis_csv.strategies import csv
@given(csv=csv(lines=20,header=5))
def test_my_csv_parse(csv):
    parsed_csv=my_csv_parser(csv)
    assert parsed_csv.num_columns == 5
    assert parsed_csv.num_rows == 20

Generate CSV with a header

from hypothesis_csv.strategies import csv
@given(csv=csv(header=["timestamp","val_1","val_2"]))
def test_my_csv_parse(csv):
    parsed_csv=my_csv_parser(csv)
    assert parsed_csv.num_columns == 3

Generate CSV with columns of a given type

from hypothesis_csv.strategies import csv
@given(csv=csv(columns=[text(),int(),float()]))
def test_my_csv_parse(csv):
    parsed_csv=my_csv_parser(csv)
    assert parsed_csv.num_columns == 3

Generate CSV with a given dialect

from hypothesis_csv.strategies import csv
@given(csv=csv(columns=[text(),int(),float()], dialect="excel-tab"))
def test_my_csv_parse(csv):
    ...

Generate CSV with drawns dialects

from hypothesis_csv.strategies import csv
@given(csv=csv(columns=[text(),int(),float()], dialect=None))
def test_my_csv_parse(csv):
    ...