Run all generic tests on the provided transformed data. This should be called from test-extract.R in a montagu-imports import. It tests that the csv metadata is syntactically valid and as expected, and that querying related database tables has also worked. After calling stoner::stone_text_extract(extracted_data), you should write tests that are specific to the metadata for the particular import you are writing.

stone_test_transform(transformed_data)

Arguments

transformed_data

A list of data.frames produced by the transform function. They will have names matching database tables in montagu, and columns matching the columns of those tables.

Details

See the vignette for information on the stoner-specific tests.