Parsers (pre-processing) do not run if validation is disabled

[libklein]

I've noticed that parsers are not run if validation is disabled. (See code snippet)

I wonder if this is intended, or an oversight? The primary reason for disabling validation is, as outlined in the docs, avoiding expensive checks in e.g. production. Wouldn't we still want to run parsers if validation is disabled?

Happy to hear your opinions.

def test_parser_runs() -> None:
    import pandera.pandas as pa
    import pandas as pd
    from pandas.testing import assert_frame_equal
    from pandera.config import config_context

    with config_context(validation_enabled=False):
        schema = pa.DataFrameSchema(
            parsers=pa.Parser(lambda df: df.transform(lambda s: s*2)),
            columns={
                "a": pa.Column(float, parsers=pa.Parser(lambda s: s + 1)),
            }
        )

        data = pd.DataFrame({
            "a": [1.0]
        })
        expected_data = pd.DataFrame({
            "a": [5.0]
        })

        assert_frame_equal(expected_data, schema.validate(data)) # Fails - dataframe is not transformed.