示例#1
0
def test_convert_parquet(parquet_schema):
    strings = [dt.string, dt.string, dt.string]

    # uint32, int8, int16 stored as upcasted types
    types = [dt.uint8, dt.uint16, dt.int64, dt.uint64,
             dt.int16, dt.int16, dt.int32,
             dt.int64, dt.float32, dt.float64, dt.boolean,
             dt.timestamp] + strings + [dt.binary, dt.int64]
    names = ['uint8', 'uint16', 'uint32', 'uint64',
             'int8', 'int16', 'int32',
             'int64', 'float32', 'float64', 'bool',
             'datetime', 'str', 'str_with_nulls', 'empty_str',
             'bytes']
    expected = ibis.schema(zip(names, types))

    result = ibis.infer_schema(parquet_schema)
    assert result == expected
示例#2
0
def test_convert_parquet(parquet_schema):
    # TODO(jreback)
    # not entirely sure this is correct
    # should these be strings in py2?
    if PY2:
        strings = [dt.binary, dt.binary, dt.binary]
    else:
        strings = [dt.string, dt.string, dt.string]

    # uint32, int8, int16 stored as upcasted types
    types = [
        dt.uint8, dt.uint16, dt.int64, dt.uint64, dt.int16, dt.int16, dt.int32,
        dt.int64, dt.float32, dt.float64, dt.boolean, dt.timestamp
    ] + strings + [dt.binary, dt.int64]
    names = [
        'uint8', 'uint16', 'uint32', 'uint64', 'int8', 'int16', 'int32',
        'int64', 'float32', 'float64', 'bool', 'datetime', 'str',
        'str_with_nulls', 'empty_str', 'bytes', '__index_level_0__'
    ]
    expected = ibis.schema(zip(names, types))

    result = ibis.infer_schema(parquet_schema)
    assert result == expected