58 lines
2.6 KiB
Gherkin
58 lines
2.6 KiB
Gherkin
Feature: Dataset availability validation
|
|
Scenario: Dataset exists on HuggingFace
|
|
Given a dataset id "example-dataset"
|
|
And the HuggingFace API reports the dataset exists
|
|
When I validate dataset existence
|
|
Then the dataset should be reported as existing
|
|
|
|
Scenario: Dataset can be downloaded from HuggingFace
|
|
Given a dataset id "example-dataset"
|
|
And the HuggingFace download succeeds
|
|
When I validate dataset download
|
|
Then the dataset should be reported as downloadable
|
|
|
|
Scenario: Dataset file is a parquet file
|
|
When I validate the dataset file format
|
|
Then the dataset file should be reported as parquet
|
|
|
|
Scenario: Dataset file has the required columns
|
|
When I validate the dataset columns
|
|
Then the dataset should have the required columns
|
|
|
|
Scenario: Collect source strings from download files
|
|
Given a dataset id "example-dataset"
|
|
And a dataset download directory with CSV and N-Triples data
|
|
When I collect source strings
|
|
Then the source strings should include "hello"
|
|
And the source strings should include "http://example.org/s"
|
|
And the source strings should not include " "
|
|
|
|
Scenario: Collect parquet strings from output file
|
|
Given a parquet file with sample data
|
|
When I collect parquet strings
|
|
Then the parquet strings should include "http://example.org/s"
|
|
And the parquet strings should include "http://example.org/p"
|
|
And the parquet strings should include "object"
|
|
And the parquet strings should not include " "
|
|
|
|
Scenario: Bloom filter subset check with identical elements
|
|
Given two bloom filters with the same element
|
|
When I check if the first bloom filter is a subset of the second
|
|
And I check if the second bloom filter is a subset of the first
|
|
Then the first bloom filter should be a subset of the second
|
|
And the second bloom filter should be a subset of the first
|
|
|
|
Scenario: Bloom filter subset check with partial overlap
|
|
Given two bloom filters with one shared element and one unique to the second
|
|
When I check if the first bloom filter is a subset of the second
|
|
And I check if the second bloom filter is a subset of the first
|
|
Then the first bloom filter should be a subset of the second
|
|
And the second bloom filter should not be a subset of the first
|
|
|
|
Scenario: Bloom filter subset check with different elements
|
|
Given two bloom filters with different elements
|
|
When I check if the first bloom filter is a subset of the second
|
|
And I check if the second bloom filter is a subset of the first
|
|
Then the first bloom filter should not be a subset of the second
|
|
And the second bloom filter should not be a subset of the first
|