arw2019 commented on a change in pull request #8816: URL: https://github.com/apache/arrow/pull/8816#discussion_r541158203
########## File path: python/pyarrow/tests/parquet/common.py ########## @@ -0,0 +1,317 @@ +# Licensed to the Apache Software Foundation (ASF) under one +# or more contributor license agreements. See the NOTICE file +# distributed with this work for additional information +# regarding copyright ownership. The ASF licenses this file +# to you under the Apache License, Version 2.0 (the +# "License"); you may not use this file except in compliance +# with the License. You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, +# software distributed under the License is distributed on an +# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY +# KIND, either express or implied. See the License for the +# specific language governing permissions and limitations +# under the License. + +import io +import os + +import numpy as np +import pytest + +import pyarrow as pa +from pyarrow.filesystem import LocalFileSystem +from pyarrow.tests import util + +parametrize_legacy_dataset = pytest.mark.parametrize( + "use_legacy_dataset", + [True, pytest.param(False, marks=pytest.mark.dataset)]) +parametrize_legacy_dataset_not_supported = pytest.mark.parametrize( + "use_legacy_dataset", [True, pytest.param(False, marks=pytest.mark.skip)]) +parametrize_legacy_dataset_fixed = pytest.mark.parametrize( + "use_legacy_dataset", [pytest.param(True, marks=pytest.mark.xfail), + pytest.param(False, marks=pytest.mark.dataset)]) + +# Marks all of the tests in this module +# Ignore these with pytest ... -m 'not parquet' +pytestmark = pytest.mark.parquet + + +def _write_table(table, path, **kwargs): + # So we see the ImportError somewhere + import pyarrow.parquet as pq + from pyarrow.pandas_compat import _pandas_api + + if _pandas_api.is_data_frame(table): + table = pa.Table.from_pandas(table) + + pq.write_table(table, path, **kwargs) + return table + + +def _read_table(*args, **kwargs): + import pyarrow.parquet as pq + + table = pq.read_table(*args, **kwargs) + table.validate(full=True) + return table + + +def _roundtrip_table(table, read_table_kwargs=None, + write_table_kwargs=None, use_legacy_dataset=True): + read_table_kwargs = read_table_kwargs or {} + write_table_kwargs = write_table_kwargs or {} + + writer = pa.BufferOutputStream() + _write_table(table, writer, **write_table_kwargs) + reader = pa.BufferReader(writer.getvalue()) + return _read_table(reader, use_legacy_dataset=use_legacy_dataset, + **read_table_kwargs) + + +def _check_roundtrip(table, expected=None, read_table_kwargs=None, + use_legacy_dataset=True, **write_table_kwargs): + if expected is None: + expected = table + + read_table_kwargs = read_table_kwargs or {} + + # intentionally check twice + result = _roundtrip_table(table, read_table_kwargs=read_table_kwargs, + write_table_kwargs=write_table_kwargs, + use_legacy_dataset=use_legacy_dataset) + assert result.equals(expected) + result = _roundtrip_table(result, read_table_kwargs=read_table_kwargs, + write_table_kwargs=write_table_kwargs, + use_legacy_dataset=use_legacy_dataset) + assert result.equals(expected) + + +def _roundtrip_pandas_dataframe(df, write_kwargs, use_legacy_dataset=True): + table = pa.Table.from_pandas(df) + result = _roundtrip_table( + table, write_table_kwargs=write_kwargs, + use_legacy_dataset=use_legacy_dataset) + return result.to_pandas() + + +def _test_read_common_metadata_files(fs, base_path): Review comment: Fair - moved it to `test_dataset.py` ########## File path: python/pyarrow/tests/parquet/test_metadata.py ########## @@ -0,0 +1,475 @@ +# Licensed to the Apache Software Foundation (ASF) under one +# or more contributor license agreements. See the NOTICE file +# distributed with this work for additional information +# regarding copyright ownership. The ASF licenses this file +# to you under the Apache License, Version 2.0 (the +# "License"); you may not use this file except in compliance +# with the License. You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, +# software distributed under the License is distributed on an +# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY +# KIND, either express or implied. See the License for the +# specific language governing permissions and limitations +# under the License. + +import datetime +import io +from collections import OrderedDict + +import numpy as np +import pyarrow as pa +import pytest +from pyarrow.filesystem import LocalFileSystem +from pyarrow.tests.parquet.common import _check_roundtrip, make_sample_file + +try: + import pyarrow.parquet as pq + from pyarrow.tests.parquet.common import _write_table +except ImportError: + pq = None + + +try: + import pandas as pd + import pandas.testing as tm + from pyarrow.tests.parquet.common import alltypes_sample +except ImportError: + pd = tm = None + + [email protected] +def test_parquet_metadata_api(): + df = alltypes_sample(size=10000) + df = df.reindex(columns=sorted(df.columns)) + df.index = np.random.randint(0, 1000000, size=len(df)) + + fileh = make_sample_file(df) + ncols = len(df.columns) + + # Series of sniff tests + meta = fileh.metadata + repr(meta) + assert meta.num_rows == len(df) + assert meta.num_columns == ncols + 1 # +1 for index + assert meta.num_row_groups == 1 + assert meta.format_version == '2.0' + assert 'parquet-cpp' in meta.created_by + assert isinstance(meta.serialized_size, int) + assert isinstance(meta.metadata, dict) + + # Schema + schema = fileh.schema + assert meta.schema is schema + assert len(schema) == ncols + 1 # +1 for index + repr(schema) + + col = schema[0] + repr(col) + assert col.name == df.columns[0] + assert col.max_definition_level == 1 + assert col.max_repetition_level == 0 + assert col.max_repetition_level == 0 + + assert col.physical_type == 'BOOLEAN' + assert col.converted_type == 'NONE' + + with pytest.raises(IndexError): + schema[ncols + 1] # +1 for index + + with pytest.raises(IndexError): + schema[-1] + + # Row group + for rg in range(meta.num_row_groups): + rg_meta = meta.row_group(rg) + assert isinstance(rg_meta, pq.RowGroupMetaData) + repr(rg_meta) + + for col in range(rg_meta.num_columns): + col_meta = rg_meta.column(col) + assert isinstance(col_meta, pq.ColumnChunkMetaData) + repr(col_meta) + + with pytest.raises(IndexError): + meta.row_group(-1) + + with pytest.raises(IndexError): + meta.row_group(meta.num_row_groups + 1) + + rg_meta = meta.row_group(0) + assert rg_meta.num_rows == len(df) + assert rg_meta.num_columns == ncols + 1 # +1 for index + assert rg_meta.total_byte_size > 0 + + with pytest.raises(IndexError): + col_meta = rg_meta.column(-1) + + with pytest.raises(IndexError): + col_meta = rg_meta.column(ncols + 2) + + col_meta = rg_meta.column(0) + assert col_meta.file_offset > 0 + assert col_meta.file_path == '' # created from BytesIO + assert col_meta.physical_type == 'BOOLEAN' + assert col_meta.num_values == 10000 + assert col_meta.path_in_schema == 'bool' + assert col_meta.is_stats_set is True + assert isinstance(col_meta.statistics, pq.Statistics) + assert col_meta.compression == 'SNAPPY' + assert col_meta.encodings == ('PLAIN', 'RLE') + assert col_meta.has_dictionary_page is False + assert col_meta.dictionary_page_offset is None + assert col_meta.data_page_offset > 0 + assert col_meta.total_compressed_size > 0 + assert col_meta.total_uncompressed_size > 0 + with pytest.raises(NotImplementedError): + col_meta.has_index_page + with pytest.raises(NotImplementedError): + col_meta.index_page_offset + + +def test_parquet_metadata_lifetime(tempdir): + # ARROW-6642 - ensure that chained access keeps parent objects alive + table = pa.table({'a': [1, 2, 3]}) + pq.write_table(table, tempdir / 'test_metadata_segfault.parquet') + dataset = pq.ParquetDataset(tempdir / 'test_metadata_segfault.parquet') + dataset.pieces[0].get_metadata().row_group(0).column(0).statistics + + [email protected] [email protected]( + ( + 'data', + 'type', + 'physical_type', + 'min_value', + 'max_value', + 'null_count', + 'num_values', + 'distinct_count' + ), + [ + ([1, 2, 2, None, 4], pa.uint8(), 'INT32', 1, 4, 1, 4, 0), + ([1, 2, 2, None, 4], pa.uint16(), 'INT32', 1, 4, 1, 4, 0), + ([1, 2, 2, None, 4], pa.uint32(), 'INT32', 1, 4, 1, 4, 0), + ([1, 2, 2, None, 4], pa.uint64(), 'INT64', 1, 4, 1, 4, 0), + ([-1, 2, 2, None, 4], pa.int8(), 'INT32', -1, 4, 1, 4, 0), + ([-1, 2, 2, None, 4], pa.int16(), 'INT32', -1, 4, 1, 4, 0), + ([-1, 2, 2, None, 4], pa.int32(), 'INT32', -1, 4, 1, 4, 0), + ([-1, 2, 2, None, 4], pa.int64(), 'INT64', -1, 4, 1, 4, 0), + ( + [-1.1, 2.2, 2.3, None, 4.4], pa.float32(), + 'FLOAT', -1.1, 4.4, 1, 4, 0 + ), + ( + [-1.1, 2.2, 2.3, None, 4.4], pa.float64(), + 'DOUBLE', -1.1, 4.4, 1, 4, 0 + ), + ( + ['', 'b', chr(1000), None, 'aaa'], pa.binary(), + 'BYTE_ARRAY', b'', chr(1000).encode('utf-8'), 1, 4, 0 + ), + ( + [True, False, False, True, True], pa.bool_(), + 'BOOLEAN', False, True, 0, 5, 0 + ), + ( + [b'\x00', b'b', b'12', None, b'aaa'], pa.binary(), + 'BYTE_ARRAY', b'\x00', b'b', 1, 4, 0 + ), + ] +) +def test_parquet_column_statistics_api(data, type, physical_type, min_value, + max_value, null_count, num_values, + distinct_count): + df = pd.DataFrame({'data': data}) + schema = pa.schema([pa.field('data', type)]) + table = pa.Table.from_pandas(df, schema=schema, safe=False) + fileh = make_sample_file(table) + + meta = fileh.metadata + + rg_meta = meta.row_group(0) + col_meta = rg_meta.column(0) + + stat = col_meta.statistics + assert stat.has_min_max + assert _close(type, stat.min, min_value) + assert _close(type, stat.max, max_value) + assert stat.null_count == null_count + assert stat.num_values == num_values + # TODO(kszucs) until parquet-cpp API doesn't expose HasDistinctCount + # method, missing distinct_count is represented as zero instead of None + assert stat.distinct_count == distinct_count + assert stat.physical_type == physical_type + + +# ARROW-6339 [email protected] +def test_parquet_raise_on_unset_statistics(): + df = pd.DataFrame({"t": pd.Series([pd.NaT], dtype="datetime64[ns]")}) + meta = make_sample_file(pa.Table.from_pandas(df)).metadata + + assert not meta.row_group(0).column(0).statistics.has_min_max + assert meta.row_group(0).column(0).statistics.max is None + + +def _close(type, left, right): + if type == pa.float32(): + return abs(left - right) < 1E-7 + elif type == pa.float64(): + return abs(left - right) < 1E-13 + else: + return left == right + + +def test_statistics_convert_logical_types(tempdir): + # ARROW-5166, ARROW-4139 + + # (min, max, type) + cases = [(10, 11164359321221007157, pa.uint64()), + (10, 4294967295, pa.uint32()), + ("ähnlich", "öffentlich", pa.utf8()), + (datetime.time(10, 30, 0, 1000), datetime.time(15, 30, 0, 1000), + pa.time32('ms')), + (datetime.time(10, 30, 0, 1000), datetime.time(15, 30, 0, 1000), + pa.time64('us')), + (datetime.datetime(2019, 6, 24, 0, 0, 0, 1000), + datetime.datetime(2019, 6, 25, 0, 0, 0, 1000), + pa.timestamp('ms')), + (datetime.datetime(2019, 6, 24, 0, 0, 0, 1000), + datetime.datetime(2019, 6, 25, 0, 0, 0, 1000), + pa.timestamp('us'))] + + for i, (min_val, max_val, typ) in enumerate(cases): + t = pa.Table.from_arrays([pa.array([min_val, max_val], type=typ)], + ['col']) + path = str(tempdir / ('example{}.parquet'.format(i))) + pq.write_table(t, path, version='2.0') + pf = pq.ParquetFile(path) + stats = pf.metadata.row_group(0).column(0).statistics + assert stats.min == min_val + assert stats.max == max_val + + +def test_parquet_write_disable_statistics(tempdir): + table = pa.Table.from_pydict( + OrderedDict([ + ('a', pa.array([1, 2, 3])), + ('b', pa.array(['a', 'b', 'c'])) + ]) + ) + _write_table(table, tempdir / 'data.parquet') + meta = pq.read_metadata(tempdir / 'data.parquet') + for col in [0, 1]: + cc = meta.row_group(0).column(col) + assert cc.is_stats_set is True + assert cc.statistics is not None + + _write_table(table, tempdir / 'data2.parquet', write_statistics=False) + meta = pq.read_metadata(tempdir / 'data2.parquet') + for col in [0, 1]: + cc = meta.row_group(0).column(col) + assert cc.is_stats_set is False + assert cc.statistics is None + + _write_table(table, tempdir / 'data3.parquet', write_statistics=['a']) + meta = pq.read_metadata(tempdir / 'data3.parquet') + cc_a = meta.row_group(0).column(0) + cc_b = meta.row_group(0).column(1) + assert cc_a.is_stats_set is True + assert cc_b.is_stats_set is False + assert cc_a.statistics is not None + assert cc_b.statistics is None + + [email protected] +def test_pass_separate_metadata(): + # ARROW-471 + df = alltypes_sample(size=10000) + + a_table = pa.Table.from_pandas(df) + + buf = io.BytesIO() + _write_table(a_table, buf, compression='snappy', version='2.0') + + buf.seek(0) + metadata = pq.read_metadata(buf) + + buf.seek(0) + + fileh = pq.ParquetFile(buf, metadata=metadata) + + tm.assert_frame_equal(df, fileh.read().to_pandas()) + + +def test_field_id_metadata(): + # ARROW-7080 + table = pa.table([pa.array([1], type='int32'), + pa.array([[]], type=pa.list_(pa.int32())), + pa.array([b'boo'], type='binary')], + ['f0', 'f1', 'f2']) + + bio = pa.BufferOutputStream() + pq.write_table(table, bio) + contents = bio.getvalue() + + pf = pq.ParquetFile(pa.BufferReader(contents)) + schema = pf.schema_arrow + + # Expected Parquet schema for reference + # + # required group field_id=0 schema { + # optional int32 field_id=1 f0; + # optional group field_id=2 f1 (List) { + # repeated group field_id=3 list { + # optional int32 field_id=4 item; + # } + # } + # optional binary field_id=5 f2; + # } + + field_name = b'PARQUET:field_id' + assert schema[0].metadata[field_name] == b'1' + + list_field = schema[1] + assert list_field.metadata[field_name] == b'2' + + list_item_field = list_field.type.value_field + assert list_item_field.metadata[field_name] == b'4' + + assert schema[2].metadata[field_name] == b'5' + + [email protected] +def test_read_metadata_files(tempdir): Review comment: moved them both over to `test_dataset.py` ########## File path: python/pyarrow/tests/parquet/test_basic.py ########## @@ -0,0 +1,1063 @@ +# Licensed to the Apache Software Foundation (ASF) under one +# or more contributor license agreements. See the NOTICE file +# distributed with this work for additional information +# regarding copyright ownership. The ASF licenses this file +# to you under the Apache License, Version 2.0 (the +# "License"); you may not use this file except in compliance +# with the License. You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, +# software distributed under the License is distributed on an +# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY +# KIND, either express or implied. See the License for the +# specific language governing permissions and limitations +# under the License. + +import decimal +import io +import os + +import numpy as np +import pyarrow as pa +import pytest +from pyarrow import fs +from pyarrow.filesystem import LocalFileSystem +from pyarrow.tests import util +from pyarrow.tests.parquet.common import (_check_roundtrip, _roundtrip_table, + make_sample_file, + parametrize_legacy_dataset) + +try: + import pyarrow.parquet as pq + from pyarrow.tests.parquet.common import (_read_table, _test_dataframe, + _write_table) +except ImportError: + pq = None + + +try: + import pandas as pd + import pandas.testing as tm + from pyarrow.tests.pandas_examples import (dataframe_with_arrays, + dataframe_with_lists) + from pyarrow.tests.parquet.common import alltypes_sample +except ImportError: + pd = tm = None + + +def test_large_binary(): + data = [b'foo', b'bar'] * 50 + for type in [pa.large_binary(), pa.large_string()]: + arr = pa.array(data, type=type) + table = pa.Table.from_arrays([arr], names=['strs']) + for use_dictionary in [False, True]: + _check_roundtrip(table, use_dictionary=use_dictionary) + + [email protected]_memory +def test_large_binary_huge(): + s = b'xy' * 997 + data = [s] * ((1 << 33) // len(s)) + for type in [pa.large_binary(), pa.large_string()]: + arr = pa.array(data, type=type) + table = pa.Table.from_arrays([arr], names=['strs']) + for use_dictionary in [False, True]: + _check_roundtrip(table, use_dictionary=use_dictionary) + del arr, table + + [email protected]_memory +def test_large_binary_overflow(): + s = b'x' * (1 << 31) + arr = pa.array([s], type=pa.large_binary()) + table = pa.Table.from_arrays([arr], names=['strs']) + for use_dictionary in [False, True]: + writer = pa.BufferOutputStream() + with pytest.raises( + pa.ArrowInvalid, + match="Parquet cannot store strings with size 2GB or more"): + _write_table(table, writer, use_dictionary=use_dictionary) + + +@parametrize_legacy_dataset [email protected]('dtype', [int, float]) +def test_single_pylist_column_roundtrip(tempdir, dtype, use_legacy_dataset): + filename = tempdir / 'single_{}_column.parquet'.format(dtype.__name__) + data = [pa.array(list(map(dtype, range(5))))] + table = pa.Table.from_arrays(data, names=['a']) + _write_table(table, filename) + table_read = _read_table(filename, use_legacy_dataset=use_legacy_dataset) + for i in range(table.num_columns): + col_written = table[i] + col_read = table_read[i] + assert table.field(i).name == table_read.field(i).name + assert col_read.num_chunks == 1 + data_written = col_written.chunk(0) + data_read = col_read.chunk(0) + assert data_written.equals(data_read) + + +def test_parquet_invalid_version(tempdir): + table = pa.table({'a': [1, 2, 3]}) + with pytest.raises(ValueError, match="Unsupported Parquet format version"): + _write_table(table, tempdir / 'test_version.parquet', version="2.2") + with pytest.raises(ValueError, match="Unsupported Parquet data page " + + "version"): + _write_table(table, tempdir / 'test_version.parquet', + data_page_version="2.2") + + +@parametrize_legacy_dataset +def test_set_data_page_size(use_legacy_dataset): + arr = pa.array([1, 2, 3] * 100000) + t = pa.Table.from_arrays([arr], names=['f0']) + + # 128K, 512K + page_sizes = [2 << 16, 2 << 18] + for target_page_size in page_sizes: + _check_roundtrip(t, data_page_size=target_page_size, + use_legacy_dataset=use_legacy_dataset) + + [email protected] +@parametrize_legacy_dataset +def test_chunked_table_write(use_legacy_dataset): + # ARROW-232 + tables = [] + batch = pa.RecordBatch.from_pandas(alltypes_sample(size=10)) + tables.append(pa.Table.from_batches([batch] * 3)) + df, _ = dataframe_with_lists() + batch = pa.RecordBatch.from_pandas(df) + tables.append(pa.Table.from_batches([batch] * 3)) + + for data_page_version in ['1.0', '2.0']: + for use_dictionary in [True, False]: + for table in tables: + _check_roundtrip( + table, version='2.0', + use_legacy_dataset=use_legacy_dataset, + data_page_version=data_page_version, + use_dictionary=use_dictionary) + + [email protected] +@parametrize_legacy_dataset +def test_memory_map(tempdir, use_legacy_dataset): + df = alltypes_sample(size=10) + + table = pa.Table.from_pandas(df) + _check_roundtrip(table, read_table_kwargs={'memory_map': True}, + version='2.0', use_legacy_dataset=use_legacy_dataset) + + filename = str(tempdir / 'tmp_file') + with open(filename, 'wb') as f: + _write_table(table, f, version='2.0') + table_read = pq.read_pandas(filename, memory_map=True, + use_legacy_dataset=use_legacy_dataset) + assert table_read.equals(table) + + [email protected] +@parametrize_legacy_dataset +def test_enable_buffered_stream(tempdir, use_legacy_dataset): + df = alltypes_sample(size=10) + + table = pa.Table.from_pandas(df) + _check_roundtrip(table, read_table_kwargs={'buffer_size': 1025}, + version='2.0', use_legacy_dataset=use_legacy_dataset) + + filename = str(tempdir / 'tmp_file') + with open(filename, 'wb') as f: + _write_table(table, f, version='2.0') + table_read = pq.read_pandas(filename, buffer_size=4096, + use_legacy_dataset=use_legacy_dataset) + assert table_read.equals(table) + + +@parametrize_legacy_dataset +def test_special_chars_filename(tempdir, use_legacy_dataset): + table = pa.Table.from_arrays([pa.array([42])], ["ints"]) + filename = "foo # bar" + path = tempdir / filename + assert not path.exists() + _write_table(table, str(path)) + assert path.exists() + table_read = _read_table(str(path), use_legacy_dataset=use_legacy_dataset) + assert table_read.equals(table) + + [email protected] +def test_file_with_over_int16_max_row_groups(): + # PARQUET-1857: Parquet encryption support introduced a INT16_MAX upper + # limit on the number of row groups, but this limit only impacts files with + # encrypted row group metadata because of the int16 row group ordinal used + # in the Parquet Thrift metadata. Unencrypted files are not impacted, so + # this test checks that it works (even if it isn't a good idea) + t = pa.table([list(range(40000))], names=['f0']) + _check_roundtrip(t, row_group_size=1) + + +@parametrize_legacy_dataset +def test_nested_list_nonnullable_roundtrip_bug(use_legacy_dataset): + # Reproduce failure in ARROW-5630 + typ = pa.list_(pa.field("item", pa.float32(), False)) + num_rows = 10000 + t = pa.table([ + pa.array(([[0] * ((i + 5) % 10) for i in range(0, 10)] * + (num_rows // 10)), type=typ) + ], ['a']) + _check_roundtrip( + t, data_page_size=4096, use_legacy_dataset=use_legacy_dataset) + + [email protected] +@parametrize_legacy_dataset +def test_multiple_path_types(tempdir, use_legacy_dataset): + # Test compatibility with PEP 519 path-like objects + path = tempdir / 'zzz.parquet' + df = pd.DataFrame({'x': np.arange(10, dtype=np.int64)}) + _write_table(df, path) + table_read = _read_table(path, use_legacy_dataset=use_legacy_dataset) + df_read = table_read.to_pandas() + tm.assert_frame_equal(df, df_read) + + # Test compatibility with plain string paths + path = str(tempdir) + 'zzz.parquet' + df = pd.DataFrame({'x': np.arange(10, dtype=np.int64)}) + _write_table(df, path) + table_read = _read_table(path, use_legacy_dataset=use_legacy_dataset) + df_read = table_read.to_pandas() + tm.assert_frame_equal(df, df_read) + + [email protected] +@parametrize_legacy_dataset [email protected]("filesystem", [ + None, fs.LocalFileSystem(), LocalFileSystem._get_instance() +]) +def test_relative_paths(tempdir, use_legacy_dataset, filesystem): + # reading and writing from relative paths + table = pa.table({"a": [1, 2, 3]}) + + # reading + pq.write_table(table, str(tempdir / "data.parquet")) + with util.change_cwd(tempdir): + result = pq.read_table("data.parquet", filesystem=filesystem, + use_legacy_dataset=use_legacy_dataset) + assert result.equals(table) + + # writing + with util.change_cwd(tempdir): + pq.write_table(table, "data2.parquet", filesystem=filesystem) + result = pq.read_table(tempdir / "data2.parquet") + assert result.equals(table) + + +@parametrize_legacy_dataset +def test_read_non_existing_file(use_legacy_dataset): + # ensure we have a proper error message + with pytest.raises(FileNotFoundError): + pq.read_table('i-am-not-existing.parquet') + + +@parametrize_legacy_dataset +def test_parquet_read_from_buffer(tempdir, use_legacy_dataset): + # reading from a buffer from python's open() + table = pa.table({"a": [1, 2, 3]}) + pq.write_table(table, str(tempdir / "data.parquet")) + + with open(str(tempdir / "data.parquet"), "rb") as f: + result = pq.read_table(f, use_legacy_dataset=use_legacy_dataset) + assert result.equals(table) + + with open(str(tempdir / "data.parquet"), "rb") as f: + result = pq.read_table(pa.PythonFile(f), + use_legacy_dataset=use_legacy_dataset) + assert result.equals(table) + + [email protected] +@parametrize_legacy_dataset +def test_parquet_incremental_file_build(tempdir, use_legacy_dataset): + df = _test_dataframe(100) + df['unique_id'] = 0 + + arrow_table = pa.Table.from_pandas(df, preserve_index=False) + out = pa.BufferOutputStream() + + writer = pq.ParquetWriter(out, arrow_table.schema, version='2.0') + + frames = [] + for i in range(10): + df['unique_id'] = i + arrow_table = pa.Table.from_pandas(df, preserve_index=False) + writer.write_table(arrow_table) + + frames.append(df.copy()) + + writer.close() + + buf = out.getvalue() + result = _read_table( + pa.BufferReader(buf), use_legacy_dataset=use_legacy_dataset) + + expected = pd.concat(frames, ignore_index=True) + tm.assert_frame_equal(result.to_pandas(), expected) + + +@parametrize_legacy_dataset +def test_byte_stream_split(use_legacy_dataset): + # This is only a smoke test. + arr_float = pa.array(list(map(float, range(100)))) + arr_int = pa.array(list(map(int, range(100)))) + data_float = [arr_float, arr_float] + table = pa.Table.from_arrays(data_float, names=['a', 'b']) + + # Check with byte_stream_split for both columns. + _check_roundtrip(table, expected=table, compression="gzip", + use_dictionary=False, use_byte_stream_split=True) + + # Check with byte_stream_split for column 'b' and dictionary + # for column 'a'. + _check_roundtrip(table, expected=table, compression="gzip", + use_dictionary=['a'], + use_byte_stream_split=['b']) + + # Check with a collision for both columns. + _check_roundtrip(table, expected=table, compression="gzip", + use_dictionary=['a', 'b'], + use_byte_stream_split=['a', 'b']) + + # Check with mixed column types. + mixed_table = pa.Table.from_arrays([arr_float, arr_int], + names=['a', 'b']) + _check_roundtrip(mixed_table, expected=mixed_table, + use_dictionary=['b'], + use_byte_stream_split=['a']) + + # Try to use the wrong data type with the byte_stream_split encoding. + # This should throw an exception. + table = pa.Table.from_arrays([arr_int], names=['tmp']) + with pytest.raises(IOError): + _check_roundtrip(table, expected=table, use_byte_stream_split=True, + use_dictionary=False, + use_legacy_dataset=use_legacy_dataset) + + +@parametrize_legacy_dataset +def test_compression_level(use_legacy_dataset): + arr = pa.array(list(map(int, range(1000)))) + data = [arr, arr] + table = pa.Table.from_arrays(data, names=['a', 'b']) + + # Check one compression level. + _check_roundtrip(table, expected=table, compression="gzip", + compression_level=1, + use_legacy_dataset=use_legacy_dataset) + + # Check another one to make sure that compression_level=1 does not + # coincide with the default one in Arrow. + _check_roundtrip(table, expected=table, compression="gzip", + compression_level=5, + use_legacy_dataset=use_legacy_dataset) + + # Check that the user can provide a compression per column + _check_roundtrip(table, expected=table, + compression={'a': "gzip", 'b': "snappy"}, + use_legacy_dataset=use_legacy_dataset) + + # Check that the user can provide a compression level per column + _check_roundtrip(table, expected=table, compression="gzip", + compression_level={'a': 2, 'b': 3}, + use_legacy_dataset=use_legacy_dataset) + + # Check that specifying a compression level for a codec which does allow + # specifying one, results into an error. + # Uncompressed, snappy, lz4 and lzo do not support specifying a compression + # level. + # GZIP (zlib) allows for specifying a compression level but as of up + # to version 1.2.11 the valid range is [-1, 9]. + invalid_combinations = [("snappy", 4), ("lz4", 5), ("gzip", -1337), + ("None", 444), ("lzo", 14)] + buf = io.BytesIO() + for (codec, level) in invalid_combinations: + with pytest.raises((ValueError, OSError)): + _write_table(table, buf, compression=codec, + compression_level=level) + + [email protected] +def test_compare_schemas(): + df = alltypes_sample(size=10000) + + fileh = make_sample_file(df) + fileh2 = make_sample_file(df) + fileh3 = make_sample_file(df[df.columns[::2]]) + + # ParquetSchema + assert isinstance(fileh.schema, pq.ParquetSchema) + assert fileh.schema.equals(fileh.schema) + assert fileh.schema == fileh.schema + assert fileh.schema.equals(fileh2.schema) + assert fileh.schema == fileh2.schema + assert fileh.schema != 'arbitrary object' + assert not fileh.schema.equals(fileh3.schema) + assert fileh.schema != fileh3.schema + + # ColumnSchema + assert isinstance(fileh.schema[0], pq.ColumnSchema) + assert fileh.schema[0].equals(fileh.schema[0]) + assert fileh.schema[0] == fileh.schema[0] + assert not fileh.schema[0].equals(fileh.schema[1]) + assert fileh.schema[0] != fileh.schema[1] + assert fileh.schema[0] != 'arbitrary object' + + +def test_validate_schema_write_table(tempdir): + # ARROW-2926 + simple_fields = [ + pa.field('POS', pa.uint32()), + pa.field('desc', pa.string()) + ] + + simple_schema = pa.schema(simple_fields) + + # simple_table schema does not match simple_schema + simple_from_array = [pa.array([1]), pa.array(['bla'])] + simple_table = pa.Table.from_arrays(simple_from_array, ['POS', 'desc']) + + path = tempdir / 'simple_validate_schema.parquet' + + with pq.ParquetWriter(path, simple_schema, + version='2.0', + compression='snappy', flavor='spark') as w: + with pytest.raises(ValueError): + w.write_table(simple_table) + + [email protected] +def test_column_of_arrays(tempdir): + df, schema = dataframe_with_arrays() + + filename = tempdir / 'pandas_roundtrip.parquet' + arrow_table = pa.Table.from_pandas(df, schema=schema) + _write_table(arrow_table, filename, version="2.0", coerce_timestamps='ms') + table_read = _read_table(filename) + df_read = table_read.to_pandas() + tm.assert_frame_equal(df, df_read) + + [email protected] +def test_column_of_lists(tempdir): + df, schema = dataframe_with_lists(parquet_compatible=True) + + filename = tempdir / 'pandas_roundtrip.parquet' + arrow_table = pa.Table.from_pandas(df, schema=schema) + _write_table(arrow_table, filename, version='2.0') + table_read = _read_table(filename) + df_read = table_read.to_pandas() + + tm.assert_frame_equal(df, df_read) + + +def test_large_list_records(): + # This was fixed in PARQUET-1100 + + list_lengths = np.random.randint(0, 500, size=50) + list_lengths[::10] = 0 + + list_values = [list(map(int, np.random.randint(0, 100, size=x))) + if i % 8 else None + for i, x in enumerate(list_lengths)] + + a1 = pa.array(list_values) + + table = pa.Table.from_arrays([a1], ['int_lists']) + _check_roundtrip(table) + + +def test_sanitized_spark_field_names(): + a0 = pa.array([0, 1, 2, 3, 4]) + name = 'prohib; ,\t{}' + table = pa.Table.from_arrays([a0], [name]) + + result = _roundtrip_table(table, write_table_kwargs={'flavor': 'spark'}) + + expected_name = 'prohib______' + assert result.schema[0].name == expected_name + + +def test_fixed_size_binary(): + t0 = pa.binary(10) + data = [b'fooooooooo', None, b'barooooooo', b'quxooooooo'] + a0 = pa.array(data, type=t0) + + table = pa.Table.from_arrays([a0], + ['binary[10]']) + _check_roundtrip(table) + + [email protected] +@parametrize_legacy_dataset +def test_multithreaded_read(use_legacy_dataset): + df = alltypes_sample(size=10000) + + table = pa.Table.from_pandas(df) + + buf = io.BytesIO() + _write_table(table, buf, compression='SNAPPY', version='2.0') + + buf.seek(0) + table1 = _read_table( + buf, use_threads=True, use_legacy_dataset=use_legacy_dataset) + + buf.seek(0) + table2 = _read_table( + buf, use_threads=False, use_legacy_dataset=use_legacy_dataset) + + assert table1.equals(table2) + + [email protected] +@parametrize_legacy_dataset +def test_min_chunksize(use_legacy_dataset): + data = pd.DataFrame([np.arange(4)], columns=['A', 'B', 'C', 'D']) + table = pa.Table.from_pandas(data.reset_index()) + + buf = io.BytesIO() + _write_table(table, buf, chunk_size=-1) + + buf.seek(0) + result = _read_table(buf, use_legacy_dataset=use_legacy_dataset) + + assert result.equals(table) + + with pytest.raises(ValueError): + _write_table(table, buf, chunk_size=0) + + [email protected] +def test_read_single_row_group(): + # ARROW-471 + N, K = 10000, 4 + df = alltypes_sample(size=N) + + a_table = pa.Table.from_pandas(df) + + buf = io.BytesIO() + _write_table(a_table, buf, row_group_size=N / K, + compression='snappy', version='2.0') + + buf.seek(0) + + pf = pq.ParquetFile(buf) + + assert pf.num_row_groups == K + + row_groups = [pf.read_row_group(i) for i in range(K)] + result = pa.concat_tables(row_groups) + tm.assert_frame_equal(df, result.to_pandas()) + + [email protected] +def test_read_single_row_group_with_column_subset(): + N, K = 10000, 4 + df = alltypes_sample(size=N) + a_table = pa.Table.from_pandas(df) + + buf = io.BytesIO() + _write_table(a_table, buf, row_group_size=N / K, + compression='snappy', version='2.0') + + buf.seek(0) + pf = pq.ParquetFile(buf) + + cols = list(df.columns[:2]) + row_groups = [pf.read_row_group(i, columns=cols) for i in range(K)] + result = pa.concat_tables(row_groups) + tm.assert_frame_equal(df[cols], result.to_pandas()) + + # ARROW-4267: Selection of duplicate columns still leads to these columns + # being read uniquely. + row_groups = [pf.read_row_group(i, columns=cols + cols) for i in range(K)] + result = pa.concat_tables(row_groups) + tm.assert_frame_equal(df[cols], result.to_pandas()) + + [email protected] +def test_read_multiple_row_groups(): + N, K = 10000, 4 + df = alltypes_sample(size=N) + + a_table = pa.Table.from_pandas(df) + + buf = io.BytesIO() + _write_table(a_table, buf, row_group_size=N / K, + compression='snappy', version='2.0') + + buf.seek(0) + + pf = pq.ParquetFile(buf) + + assert pf.num_row_groups == K + + result = pf.read_row_groups(range(K)) + tm.assert_frame_equal(df, result.to_pandas()) + + [email protected] +def test_read_multiple_row_groups_with_column_subset(): + N, K = 10000, 4 + df = alltypes_sample(size=N) + a_table = pa.Table.from_pandas(df) + + buf = io.BytesIO() + _write_table(a_table, buf, row_group_size=N / K, + compression='snappy', version='2.0') + + buf.seek(0) + pf = pq.ParquetFile(buf) + + cols = list(df.columns[:2]) + result = pf.read_row_groups(range(K), columns=cols) + tm.assert_frame_equal(df[cols], result.to_pandas()) + + # ARROW-4267: Selection of duplicate columns still leads to these columns + # being read uniquely. + result = pf.read_row_groups(range(K), columns=cols + cols) + tm.assert_frame_equal(df[cols], result.to_pandas()) + + [email protected] +def test_scan_contents(): + N, K = 10000, 4 + df = alltypes_sample(size=N) + a_table = pa.Table.from_pandas(df) + + buf = io.BytesIO() + _write_table(a_table, buf, row_group_size=N / K, + compression='snappy', version='2.0') + + buf.seek(0) + pf = pq.ParquetFile(buf) + + assert pf.scan_contents() == 10000 + assert pf.scan_contents(df.columns[:4]) == 10000 + + [email protected] +def test_write_error_deletes_incomplete_file(tempdir): + # ARROW-1285 + df = pd.DataFrame({'a': list('abc'), + 'b': list(range(1, 4)), + 'c': np.arange(3, 6).astype('u1'), + 'd': np.arange(4.0, 7.0, dtype='float64'), + 'e': [True, False, True], + 'f': pd.Categorical(list('abc')), + 'g': pd.date_range('20130101', periods=3), + 'h': pd.date_range('20130101', periods=3, + tz='US/Eastern'), + 'i': pd.date_range('20130101', periods=3, freq='ns')}) + + pdf = pa.Table.from_pandas(df) + + filename = tempdir / 'tmp_file' + try: + _write_table(pdf, filename) + except pa.ArrowException: + pass + + assert not filename.exists() + + +@parametrize_legacy_dataset +def test_read_non_existent_file(tempdir, use_legacy_dataset): + path = 'non-existent-file.parquet' + try: + pq.read_table(path, use_legacy_dataset=use_legacy_dataset) + except Exception as e: + assert path in e.args[0] + + +@parametrize_legacy_dataset +def test_read_table_doesnt_warn(datadir, use_legacy_dataset): + with pytest.warns(None) as record: + pq.read_table(datadir / 'v0.7.1.parquet', + use_legacy_dataset=use_legacy_dataset) + + assert len(record) == 0 + + [email protected]_memory +def test_large_table_int32_overflow(): + size = np.iinfo('int32').max + 1 + + arr = np.ones(size, dtype='uint8') + + parr = pa.array(arr, type=pa.uint8()) + + table = pa.Table.from_arrays([parr], names=['one']) + f = io.BytesIO() + _write_table(table, f) + + +def _simple_table_roundtrip(table, use_legacy_dataset=False, **write_kwargs): + stream = pa.BufferOutputStream() + _write_table(table, stream, **write_kwargs) + buf = stream.getvalue() + return _read_table(buf, use_legacy_dataset=use_legacy_dataset) + + [email protected]_memory +@parametrize_legacy_dataset +def test_byte_array_exactly_2gb(use_legacy_dataset): + # Test edge case reported in ARROW-3762 + val = b'x' * (1 << 10) + + base = pa.array([val] * ((1 << 21) - 1)) + cases = [ + [b'x' * 1023], # 2^31 - 1 + [b'x' * 1024], # 2^31 + [b'x' * 1025] # 2^31 + 1 + ] + for case in cases: + values = pa.chunked_array([base, pa.array(case)]) + t = pa.table([values], names=['f0']) + result = _simple_table_roundtrip( + t, use_legacy_dataset=use_legacy_dataset, use_dictionary=False) + assert t.equals(result) + + [email protected] [email protected]_memory +@parametrize_legacy_dataset +def test_binary_array_overflow_to_chunked(use_legacy_dataset): + # ARROW-3762 + + # 2^31 + 1 bytes + values = [b'x'] + [ + b'x' * (1 << 20) + ] * 2 * (1 << 10) + df = pd.DataFrame({'byte_col': values}) + + tbl = pa.Table.from_pandas(df, preserve_index=False) + read_tbl = _simple_table_roundtrip( + tbl, use_legacy_dataset=use_legacy_dataset) + + col0_data = read_tbl[0] + assert isinstance(col0_data, pa.ChunkedArray) + + # Split up into 2GB chunks + assert col0_data.num_chunks == 2 + + assert tbl.equals(read_tbl) + + [email protected] [email protected]_memory +@parametrize_legacy_dataset +def test_list_of_binary_large_cell(use_legacy_dataset): + # ARROW-4688 + data = [] + + # TODO(wesm): handle chunked children + # 2^31 - 1 bytes in a single cell + # data.append([b'x' * (1 << 20)] * 2047 + [b'x' * ((1 << 20) - 1)]) + + # A little under 2GB in cell each containing approximately 10MB each + data.extend([[b'x' * 1000000] * 10] * 214) + + arr = pa.array(data) + table = pa.Table.from_arrays([arr], ['chunky_cells']) + read_table = _simple_table_roundtrip( + table, use_legacy_dataset=use_legacy_dataset) + assert table.equals(read_table) + + [email protected] +@parametrize_legacy_dataset +def test_parquet_nested_convenience(tempdir, use_legacy_dataset): + # ARROW-1684 + df = pd.DataFrame({ + 'a': [[1, 2, 3], None, [4, 5], []], + 'b': [[1.], None, None, [6., 7.]], + }) + + path = str(tempdir / 'nested_convenience.parquet') + + table = pa.Table.from_pandas(df, preserve_index=False) + _write_table(table, path) + + read = pq.read_table( + path, columns=['a'], use_legacy_dataset=use_legacy_dataset) + tm.assert_frame_equal(read.to_pandas(), df[['a']]) + + read = pq.read_table( + path, columns=['a', 'b'], use_legacy_dataset=use_legacy_dataset) + tm.assert_frame_equal(read.to_pandas(), df) + + [email protected] +@parametrize_legacy_dataset +def test_decimal_roundtrip(tempdir, use_legacy_dataset): + num_values = 10 + + columns = {} + for precision in range(1, 39): + for scale in range(0, precision + 1): + with util.random_seed(0): + random_decimal_values = [ + util.randdecimal(precision, scale) + for _ in range(num_values) + ] + column_name = ('dec_precision_{:d}_scale_{:d}' + .format(precision, scale)) + columns[column_name] = random_decimal_values + + expected = pd.DataFrame(columns) + filename = tempdir / 'decimals.parquet' + string_filename = str(filename) + table = pa.Table.from_pandas(expected) + _write_table(table, string_filename) + result_table = _read_table( + string_filename, use_legacy_dataset=use_legacy_dataset) + result = result_table.to_pandas() + tm.assert_frame_equal(result, expected) + + [email protected] [email protected]( + raises=pa.ArrowException, reason='Parquet does not support negative scale' +) +def test_decimal_roundtrip_negative_scale(tempdir): + expected = pd.DataFrame({'decimal_num': [decimal.Decimal('1.23E4')]}) + filename = tempdir / 'decimals.parquet' + string_filename = str(filename) + t = pa.Table.from_pandas(expected) + _write_table(t, string_filename) + result_table = _read_table(string_filename) + result = result_table.to_pandas() + tm.assert_frame_equal(result, expected) + + [email protected] +@parametrize_legacy_dataset +def test_parquet_writer_context_obj(tempdir, use_legacy_dataset): + df = _test_dataframe(100) + df['unique_id'] = 0 + + arrow_table = pa.Table.from_pandas(df, preserve_index=False) + out = pa.BufferOutputStream() + + with pq.ParquetWriter(out, arrow_table.schema, version='2.0') as writer: + + frames = [] + for i in range(10): + df['unique_id'] = i + arrow_table = pa.Table.from_pandas(df, preserve_index=False) + writer.write_table(arrow_table) + + frames.append(df.copy()) + + buf = out.getvalue() + result = _read_table( + pa.BufferReader(buf), use_legacy_dataset=use_legacy_dataset) + + expected = pd.concat(frames, ignore_index=True) + tm.assert_frame_equal(result.to_pandas(), expected) + + [email protected] +@parametrize_legacy_dataset +def test_parquet_writer_context_obj_with_exception( + tempdir, use_legacy_dataset +): + df = _test_dataframe(100) + df['unique_id'] = 0 + + arrow_table = pa.Table.from_pandas(df, preserve_index=False) + out = pa.BufferOutputStream() + error_text = 'Artificial Error' + + try: + with pq.ParquetWriter(out, + arrow_table.schema, + version='2.0') as writer: + + frames = [] + for i in range(10): + df['unique_id'] = i + arrow_table = pa.Table.from_pandas(df, preserve_index=False) + writer.write_table(arrow_table) + frames.append(df.copy()) + if i == 5: + raise ValueError(error_text) + except Exception as e: + assert str(e) == error_text + + buf = out.getvalue() + result = _read_table( + pa.BufferReader(buf), use_legacy_dataset=use_legacy_dataset) + + expected = pd.concat(frames, ignore_index=True) + tm.assert_frame_equal(result.to_pandas(), expected) + + [email protected] +@parametrize_legacy_dataset +def test_zlib_compression_bug(use_legacy_dataset): + # ARROW-3514: "zlib deflate failed, output buffer too small" + table = pa.Table.from_arrays([pa.array(['abc', 'def'])], ['some_col']) + f = io.BytesIO() + pq.write_table(table, f, compression='gzip') + + f.seek(0) + roundtrip = pq.read_table(f, use_legacy_dataset=use_legacy_dataset) + tm.assert_frame_equal(roundtrip.to_pandas(), table.to_pandas()) + + +def test_parquet_file_pass_directory_instead_of_file(tempdir): + # ARROW-7208 + path = tempdir / 'directory' + os.mkdir(str(path)) + + with pytest.raises(IOError, match="Expected file path"): + pq.ParquetFile(path) + + +def test_read_column_invalid_index(): + table = pa.table([pa.array([4, 5]), pa.array(["foo", "bar"])], + names=['ints', 'strs']) + bio = pa.BufferOutputStream() + pq.write_table(table, bio) + f = pq.ParquetFile(bio.getvalue()) + assert f.reader.read_column(0).to_pylist() == [4, 5] + assert f.reader.read_column(1).to_pylist() == ["foo", "bar"] + for index in (-1, 2): + with pytest.raises((ValueError, IndexError)): + f.reader.read_column(index) + + +@parametrize_legacy_dataset +def test_parquet_file_too_small(tempdir, use_legacy_dataset): + path = str(tempdir / "test.parquet") + # TODO(dataset) with datasets API it raises OSError instead + with pytest.raises((pa.ArrowInvalid, OSError), + match='size is 0 bytes'): + with open(path, 'wb') as f: + pass + pq.read_table(path, use_legacy_dataset=use_legacy_dataset) + + with pytest.raises((pa.ArrowInvalid, OSError), + match='size is 4 bytes'): + with open(path, 'wb') as f: + f.write(b'ffff') + pq.read_table(path, use_legacy_dataset=use_legacy_dataset) + + +@parametrize_legacy_dataset [email protected] +def test_filter_before_validate_schema(tempdir, use_legacy_dataset): Review comment: Moved it to `test_dataset` ########## File path: python/pyarrow/tests/parquet/common.py ########## @@ -0,0 +1,313 @@ +# Licensed to the Apache Software Foundation (ASF) under one +# or more contributor license agreements. See the NOTICE file +# distributed with this work for additional information +# regarding copyright ownership. The ASF licenses this file +# to you under the Apache License, Version 2.0 (the +# "License"); you may not use this file except in compliance +# with the License. You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, +# software distributed under the License is distributed on an +# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY +# KIND, either express or implied. See the License for the +# specific language governing permissions and limitations +# under the License. + +import io +import os + +import numpy as np +import pyarrow as pa +import pytest +from pyarrow.filesystem import LocalFileSystem +from pyarrow.tests import util + +parametrize_legacy_dataset = pytest.mark.parametrize( + "use_legacy_dataset", + [True, pytest.param(False, marks=pytest.mark.dataset)]) +parametrize_legacy_dataset_not_supported = pytest.mark.parametrize( + "use_legacy_dataset", [True, pytest.param(False, marks=pytest.mark.skip)]) +parametrize_legacy_dataset_fixed = pytest.mark.parametrize( + "use_legacy_dataset", [pytest.param(True, marks=pytest.mark.xfail), + pytest.param(False, marks=pytest.mark.dataset)]) + +# Marks all of the tests in this module +# Ignore these with pytest ... -m 'not parquet' +pytestmark = pytest.mark.parquet Review comment: It works for me locally but I have added the mark in each file just in case ########## File path: python/pyarrow/tests/parquet/test_basic.py ########## @@ -0,0 +1,1063 @@ +# Licensed to the Apache Software Foundation (ASF) under one +# or more contributor license agreements. See the NOTICE file +# distributed with this work for additional information +# regarding copyright ownership. The ASF licenses this file +# to you under the Apache License, Version 2.0 (the +# "License"); you may not use this file except in compliance +# with the License. You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, +# software distributed under the License is distributed on an +# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY +# KIND, either express or implied. See the License for the +# specific language governing permissions and limitations +# under the License. + +import decimal +import io +import os + +import numpy as np +import pyarrow as pa +import pytest +from pyarrow import fs +from pyarrow.filesystem import LocalFileSystem +from pyarrow.tests import util +from pyarrow.tests.parquet.common import (_check_roundtrip, _roundtrip_table, + make_sample_file, + parametrize_legacy_dataset) + +try: + import pyarrow.parquet as pq + from pyarrow.tests.parquet.common import (_read_table, _test_dataframe, + _write_table) +except ImportError: + pq = None + + +try: + import pandas as pd + import pandas.testing as tm + from pyarrow.tests.pandas_examples import (dataframe_with_arrays, + dataframe_with_lists) + from pyarrow.tests.parquet.common import alltypes_sample +except ImportError: + pd = tm = None + + +def test_large_binary(): + data = [b'foo', b'bar'] * 50 + for type in [pa.large_binary(), pa.large_string()]: + arr = pa.array(data, type=type) + table = pa.Table.from_arrays([arr], names=['strs']) + for use_dictionary in [False, True]: + _check_roundtrip(table, use_dictionary=use_dictionary) + + [email protected]_memory +def test_large_binary_huge(): + s = b'xy' * 997 + data = [s] * ((1 << 33) // len(s)) + for type in [pa.large_binary(), pa.large_string()]: + arr = pa.array(data, type=type) + table = pa.Table.from_arrays([arr], names=['strs']) + for use_dictionary in [False, True]: + _check_roundtrip(table, use_dictionary=use_dictionary) + del arr, table + + [email protected]_memory +def test_large_binary_overflow(): + s = b'x' * (1 << 31) + arr = pa.array([s], type=pa.large_binary()) + table = pa.Table.from_arrays([arr], names=['strs']) + for use_dictionary in [False, True]: + writer = pa.BufferOutputStream() + with pytest.raises( + pa.ArrowInvalid, + match="Parquet cannot store strings with size 2GB or more"): + _write_table(table, writer, use_dictionary=use_dictionary) + + +@parametrize_legacy_dataset [email protected]('dtype', [int, float]) +def test_single_pylist_column_roundtrip(tempdir, dtype, use_legacy_dataset): + filename = tempdir / 'single_{}_column.parquet'.format(dtype.__name__) + data = [pa.array(list(map(dtype, range(5))))] + table = pa.Table.from_arrays(data, names=['a']) + _write_table(table, filename) + table_read = _read_table(filename, use_legacy_dataset=use_legacy_dataset) + for i in range(table.num_columns): + col_written = table[i] + col_read = table_read[i] + assert table.field(i).name == table_read.field(i).name + assert col_read.num_chunks == 1 + data_written = col_written.chunk(0) + data_read = col_read.chunk(0) + assert data_written.equals(data_read) + + +def test_parquet_invalid_version(tempdir): + table = pa.table({'a': [1, 2, 3]}) + with pytest.raises(ValueError, match="Unsupported Parquet format version"): + _write_table(table, tempdir / 'test_version.parquet', version="2.2") + with pytest.raises(ValueError, match="Unsupported Parquet data page " + + "version"): + _write_table(table, tempdir / 'test_version.parquet', + data_page_version="2.2") + + +@parametrize_legacy_dataset +def test_set_data_page_size(use_legacy_dataset): + arr = pa.array([1, 2, 3] * 100000) + t = pa.Table.from_arrays([arr], names=['f0']) + + # 128K, 512K + page_sizes = [2 << 16, 2 << 18] + for target_page_size in page_sizes: + _check_roundtrip(t, data_page_size=target_page_size, + use_legacy_dataset=use_legacy_dataset) + + [email protected] +@parametrize_legacy_dataset +def test_chunked_table_write(use_legacy_dataset): + # ARROW-232 + tables = [] + batch = pa.RecordBatch.from_pandas(alltypes_sample(size=10)) + tables.append(pa.Table.from_batches([batch] * 3)) + df, _ = dataframe_with_lists() + batch = pa.RecordBatch.from_pandas(df) + tables.append(pa.Table.from_batches([batch] * 3)) + + for data_page_version in ['1.0', '2.0']: + for use_dictionary in [True, False]: + for table in tables: + _check_roundtrip( + table, version='2.0', + use_legacy_dataset=use_legacy_dataset, + data_page_version=data_page_version, + use_dictionary=use_dictionary) + + [email protected] +@parametrize_legacy_dataset +def test_memory_map(tempdir, use_legacy_dataset): + df = alltypes_sample(size=10) + + table = pa.Table.from_pandas(df) + _check_roundtrip(table, read_table_kwargs={'memory_map': True}, + version='2.0', use_legacy_dataset=use_legacy_dataset) + + filename = str(tempdir / 'tmp_file') + with open(filename, 'wb') as f: + _write_table(table, f, version='2.0') + table_read = pq.read_pandas(filename, memory_map=True, + use_legacy_dataset=use_legacy_dataset) + assert table_read.equals(table) + + [email protected] +@parametrize_legacy_dataset +def test_enable_buffered_stream(tempdir, use_legacy_dataset): + df = alltypes_sample(size=10) + + table = pa.Table.from_pandas(df) + _check_roundtrip(table, read_table_kwargs={'buffer_size': 1025}, + version='2.0', use_legacy_dataset=use_legacy_dataset) + + filename = str(tempdir / 'tmp_file') + with open(filename, 'wb') as f: + _write_table(table, f, version='2.0') + table_read = pq.read_pandas(filename, buffer_size=4096, + use_legacy_dataset=use_legacy_dataset) + assert table_read.equals(table) + + +@parametrize_legacy_dataset +def test_special_chars_filename(tempdir, use_legacy_dataset): + table = pa.Table.from_arrays([pa.array([42])], ["ints"]) + filename = "foo # bar" + path = tempdir / filename + assert not path.exists() + _write_table(table, str(path)) + assert path.exists() + table_read = _read_table(str(path), use_legacy_dataset=use_legacy_dataset) + assert table_read.equals(table) + + [email protected] +def test_file_with_over_int16_max_row_groups(): + # PARQUET-1857: Parquet encryption support introduced a INT16_MAX upper + # limit on the number of row groups, but this limit only impacts files with + # encrypted row group metadata because of the int16 row group ordinal used + # in the Parquet Thrift metadata. Unencrypted files are not impacted, so + # this test checks that it works (even if it isn't a good idea) + t = pa.table([list(range(40000))], names=['f0']) + _check_roundtrip(t, row_group_size=1) + + +@parametrize_legacy_dataset +def test_nested_list_nonnullable_roundtrip_bug(use_legacy_dataset): + # Reproduce failure in ARROW-5630 + typ = pa.list_(pa.field("item", pa.float32(), False)) + num_rows = 10000 + t = pa.table([ + pa.array(([[0] * ((i + 5) % 10) for i in range(0, 10)] * + (num_rows // 10)), type=typ) + ], ['a']) + _check_roundtrip( + t, data_page_size=4096, use_legacy_dataset=use_legacy_dataset) + + [email protected] +@parametrize_legacy_dataset +def test_multiple_path_types(tempdir, use_legacy_dataset): + # Test compatibility with PEP 519 path-like objects + path = tempdir / 'zzz.parquet' + df = pd.DataFrame({'x': np.arange(10, dtype=np.int64)}) + _write_table(df, path) + table_read = _read_table(path, use_legacy_dataset=use_legacy_dataset) + df_read = table_read.to_pandas() + tm.assert_frame_equal(df, df_read) + + # Test compatibility with plain string paths + path = str(tempdir) + 'zzz.parquet' + df = pd.DataFrame({'x': np.arange(10, dtype=np.int64)}) + _write_table(df, path) + table_read = _read_table(path, use_legacy_dataset=use_legacy_dataset) + df_read = table_read.to_pandas() + tm.assert_frame_equal(df, df_read) + + [email protected] +@parametrize_legacy_dataset [email protected]("filesystem", [ + None, fs.LocalFileSystem(), LocalFileSystem._get_instance() +]) +def test_relative_paths(tempdir, use_legacy_dataset, filesystem): + # reading and writing from relative paths + table = pa.table({"a": [1, 2, 3]}) + + # reading + pq.write_table(table, str(tempdir / "data.parquet")) + with util.change_cwd(tempdir): + result = pq.read_table("data.parquet", filesystem=filesystem, + use_legacy_dataset=use_legacy_dataset) + assert result.equals(table) + + # writing + with util.change_cwd(tempdir): + pq.write_table(table, "data2.parquet", filesystem=filesystem) + result = pq.read_table(tempdir / "data2.parquet") + assert result.equals(table) + + +@parametrize_legacy_dataset +def test_read_non_existing_file(use_legacy_dataset): + # ensure we have a proper error message + with pytest.raises(FileNotFoundError): + pq.read_table('i-am-not-existing.parquet') + + +@parametrize_legacy_dataset +def test_parquet_read_from_buffer(tempdir, use_legacy_dataset): + # reading from a buffer from python's open() + table = pa.table({"a": [1, 2, 3]}) + pq.write_table(table, str(tempdir / "data.parquet")) + + with open(str(tempdir / "data.parquet"), "rb") as f: + result = pq.read_table(f, use_legacy_dataset=use_legacy_dataset) + assert result.equals(table) + + with open(str(tempdir / "data.parquet"), "rb") as f: + result = pq.read_table(pa.PythonFile(f), + use_legacy_dataset=use_legacy_dataset) + assert result.equals(table) + + [email protected] +@parametrize_legacy_dataset +def test_parquet_incremental_file_build(tempdir, use_legacy_dataset): + df = _test_dataframe(100) + df['unique_id'] = 0 + + arrow_table = pa.Table.from_pandas(df, preserve_index=False) + out = pa.BufferOutputStream() + + writer = pq.ParquetWriter(out, arrow_table.schema, version='2.0') + + frames = [] + for i in range(10): + df['unique_id'] = i + arrow_table = pa.Table.from_pandas(df, preserve_index=False) + writer.write_table(arrow_table) + + frames.append(df.copy()) + + writer.close() + + buf = out.getvalue() + result = _read_table( + pa.BufferReader(buf), use_legacy_dataset=use_legacy_dataset) + + expected = pd.concat(frames, ignore_index=True) + tm.assert_frame_equal(result.to_pandas(), expected) + + +@parametrize_legacy_dataset +def test_byte_stream_split(use_legacy_dataset): + # This is only a smoke test. + arr_float = pa.array(list(map(float, range(100)))) + arr_int = pa.array(list(map(int, range(100)))) + data_float = [arr_float, arr_float] + table = pa.Table.from_arrays(data_float, names=['a', 'b']) + + # Check with byte_stream_split for both columns. + _check_roundtrip(table, expected=table, compression="gzip", + use_dictionary=False, use_byte_stream_split=True) + + # Check with byte_stream_split for column 'b' and dictionary + # for column 'a'. + _check_roundtrip(table, expected=table, compression="gzip", + use_dictionary=['a'], + use_byte_stream_split=['b']) + + # Check with a collision for both columns. + _check_roundtrip(table, expected=table, compression="gzip", + use_dictionary=['a', 'b'], + use_byte_stream_split=['a', 'b']) + + # Check with mixed column types. + mixed_table = pa.Table.from_arrays([arr_float, arr_int], + names=['a', 'b']) + _check_roundtrip(mixed_table, expected=mixed_table, + use_dictionary=['b'], + use_byte_stream_split=['a']) + + # Try to use the wrong data type with the byte_stream_split encoding. + # This should throw an exception. + table = pa.Table.from_arrays([arr_int], names=['tmp']) + with pytest.raises(IOError): + _check_roundtrip(table, expected=table, use_byte_stream_split=True, + use_dictionary=False, + use_legacy_dataset=use_legacy_dataset) + + +@parametrize_legacy_dataset +def test_compression_level(use_legacy_dataset): + arr = pa.array(list(map(int, range(1000)))) + data = [arr, arr] + table = pa.Table.from_arrays(data, names=['a', 'b']) + + # Check one compression level. + _check_roundtrip(table, expected=table, compression="gzip", + compression_level=1, + use_legacy_dataset=use_legacy_dataset) + + # Check another one to make sure that compression_level=1 does not + # coincide with the default one in Arrow. + _check_roundtrip(table, expected=table, compression="gzip", + compression_level=5, + use_legacy_dataset=use_legacy_dataset) + + # Check that the user can provide a compression per column + _check_roundtrip(table, expected=table, + compression={'a': "gzip", 'b': "snappy"}, + use_legacy_dataset=use_legacy_dataset) + + # Check that the user can provide a compression level per column + _check_roundtrip(table, expected=table, compression="gzip", + compression_level={'a': 2, 'b': 3}, + use_legacy_dataset=use_legacy_dataset) + + # Check that specifying a compression level for a codec which does allow + # specifying one, results into an error. + # Uncompressed, snappy, lz4 and lzo do not support specifying a compression + # level. + # GZIP (zlib) allows for specifying a compression level but as of up + # to version 1.2.11 the valid range is [-1, 9]. + invalid_combinations = [("snappy", 4), ("lz4", 5), ("gzip", -1337), + ("None", 444), ("lzo", 14)] + buf = io.BytesIO() + for (codec, level) in invalid_combinations: + with pytest.raises((ValueError, OSError)): + _write_table(table, buf, compression=codec, + compression_level=level) + + [email protected] +def test_compare_schemas(): Review comment: Moved it to `test_metadata.py` ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected]
