geruh commented on code in PR #3119: URL: https://github.com/apache/iceberg-python/pull/3119#discussion_r2998079255
########## pyiceberg/io/fileformat.py: ########## @@ -0,0 +1,182 @@ +# 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. + +"""File Format API for writing Iceberg data files.""" + +from __future__ import annotations + +from abc import ABC, abstractmethod +from dataclasses import dataclass +from typing import TYPE_CHECKING, Any + +from pyiceberg.io import OutputFile +from pyiceberg.manifest import FileFormat +from pyiceberg.partitioning import PartitionField, PartitionSpec, partition_record_value +from pyiceberg.schema import Schema +from pyiceberg.typedef import Properties, Record + +if TYPE_CHECKING: + import pyarrow as pa + + from pyiceberg.io.pyarrow import StatsAggregator + + +@dataclass(frozen=True) +class DataFileStatistics: + record_count: int + column_sizes: dict[int, int] + value_counts: dict[int, int] + null_value_counts: dict[int, int] + nan_value_counts: dict[int, int] + column_aggregates: dict[int, StatsAggregator] + split_offsets: list[int] + + def _partition_value(self, partition_field: PartitionField, schema: Schema) -> Any: + if partition_field.source_id not in self.column_aggregates: + return None + + source_field = schema.find_field(partition_field.source_id) + iceberg_transform = partition_field.transform + + if not iceberg_transform.preserves_order: + raise ValueError( + f"Cannot infer partition value from parquet metadata for a non-linear Partition Field: " + f"{partition_field.name} with transform {partition_field.transform}" + ) + + transform_func = iceberg_transform.transform(source_field.field_type) + + lower_value = transform_func( + partition_record_value( + partition_field=partition_field, + value=self.column_aggregates[partition_field.source_id].current_min, + schema=schema, + ) + ) + upper_value = transform_func( + partition_record_value( + partition_field=partition_field, + value=self.column_aggregates[partition_field.source_id].current_max, + schema=schema, + ) + ) + if lower_value != upper_value: + raise ValueError( + f"Cannot infer partition value from parquet metadata as there are more than one partition values " + f"for Partition Field: {partition_field.name}. {lower_value=}, {upper_value=}" + ) + + return lower_value + + def partition(self, partition_spec: PartitionSpec, schema: Schema) -> Record: + return Record(*[self._partition_value(field, schema) for field in partition_spec.fields]) + + def to_serialized_dict(self) -> dict[str, Any]: + lower_bounds = {} + upper_bounds = {} + + for k, agg in self.column_aggregates.items(): + _min = agg.min_as_bytes() + if _min is not None: + lower_bounds[k] = _min + _max = agg.max_as_bytes() + if _max is not None: + upper_bounds[k] = _max + return { + "record_count": self.record_count, + "column_sizes": self.column_sizes, + "value_counts": self.value_counts, + "null_value_counts": self.null_value_counts, + "nan_value_counts": self.nan_value_counts, + "lower_bounds": lower_bounds, + "upper_bounds": upper_bounds, + "split_offsets": self.split_offsets, + } + + +class FileFormatWriter(ABC): + """Writes data to a single file in a specific format.""" + + _result: DataFileStatistics | None = None + + @abstractmethod + def write(self, table: pa.Table) -> None: + """Write a batch of data. May be called multiple times.""" + + @abstractmethod + def close(self) -> DataFileStatistics: + """Finalize the file and return statistics.""" + + def result(self) -> DataFileStatistics: + """Return statistics from a previous close() call.""" + if self._result is None: + raise RuntimeError("Writer has not been closed yet") + return self._result + + def __enter__(self) -> FileFormatWriter: + """Enter the context manager.""" + return self + + def __exit__(self, exc_type: Any, exc_val: Any, exc_tb: Any) -> None: + """Exit the context manager, closing the writer and caching statistics.""" + if exc_type is not None: Review Comment: Is this inverted? The exception branch calls close() and will swallow the error and return meaning _result gets populated with stats from a bad write. What if instead we do something like: ``` def __exit__(self, exc_type, exc_val, exc_tb): if exc_type is not None: try: self.close() except Exception: pass return self._result = self.close() ``` Similar to our logic in the manifest writer exit method. ########## tests/io/test_fileformat.py: ########## @@ -0,0 +1,45 @@ +# 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. + +from typing import Any + +import pytest + +from pyiceberg.io.fileformat import DataFileStatistics, FileFormatWriter + + +def test_backward_compat_import() -> None: Review Comment: Can we add some tests for the new abtractions, like getting an unknown format and duplicates like mentioned above. Also if there is an easy way to do the happy path roundtrip? ########## pyiceberg/io/fileformat.py: ########## @@ -0,0 +1,182 @@ +# 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. + +"""File Format API for writing Iceberg data files.""" + +from __future__ import annotations + +from abc import ABC, abstractmethod +from dataclasses import dataclass +from typing import TYPE_CHECKING, Any + +from pyiceberg.io import OutputFile +from pyiceberg.manifest import FileFormat +from pyiceberg.partitioning import PartitionField, PartitionSpec, partition_record_value +from pyiceberg.schema import Schema +from pyiceberg.typedef import Properties, Record + +if TYPE_CHECKING: + import pyarrow as pa + + from pyiceberg.io.pyarrow import StatsAggregator + + +@dataclass(frozen=True) +class DataFileStatistics: + record_count: int + column_sizes: dict[int, int] + value_counts: dict[int, int] + null_value_counts: dict[int, int] + nan_value_counts: dict[int, int] + column_aggregates: dict[int, StatsAggregator] Review Comment: I still don't know how I feel about this. I think for now it's okay since we are working with mostly parquet. but then in ORC it would use the stripe metadata. What we know is that the `_partition_value()` and `partition()` methods currently depend on `column_aggregates` to infer partition values from min/max. These could work from the serialized bounds instead but if refactoring is too much alternatively we could keep the `DataFileStatistics` in pyarrow class and introduce the shared type in your next phase as mentioned when parquet writer is actually extracted. ########## pyiceberg/io/fileformat.py: ########## @@ -0,0 +1,182 @@ +# 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. + +"""File Format API for writing Iceberg data files.""" + +from __future__ import annotations + +from abc import ABC, abstractmethod +from dataclasses import dataclass +from typing import TYPE_CHECKING, Any + +from pyiceberg.io import OutputFile +from pyiceberg.manifest import FileFormat +from pyiceberg.partitioning import PartitionField, PartitionSpec, partition_record_value +from pyiceberg.schema import Schema +from pyiceberg.typedef import Properties, Record + +if TYPE_CHECKING: + import pyarrow as pa + + from pyiceberg.io.pyarrow import StatsAggregator + + +@dataclass(frozen=True) +class DataFileStatistics: + record_count: int + column_sizes: dict[int, int] + value_counts: dict[int, int] + null_value_counts: dict[int, int] + nan_value_counts: dict[int, int] + column_aggregates: dict[int, StatsAggregator] + split_offsets: list[int] + + def _partition_value(self, partition_field: PartitionField, schema: Schema) -> Any: + if partition_field.source_id not in self.column_aggregates: + return None + + source_field = schema.find_field(partition_field.source_id) + iceberg_transform = partition_field.transform + + if not iceberg_transform.preserves_order: + raise ValueError( + f"Cannot infer partition value from parquet metadata for a non-linear Partition Field: " + f"{partition_field.name} with transform {partition_field.transform}" + ) + + transform_func = iceberg_transform.transform(source_field.field_type) + + lower_value = transform_func( + partition_record_value( + partition_field=partition_field, + value=self.column_aggregates[partition_field.source_id].current_min, + schema=schema, + ) + ) + upper_value = transform_func( + partition_record_value( + partition_field=partition_field, + value=self.column_aggregates[partition_field.source_id].current_max, + schema=schema, + ) + ) + if lower_value != upper_value: + raise ValueError( + f"Cannot infer partition value from parquet metadata as there are more than one partition values " + f"for Partition Field: {partition_field.name}. {lower_value=}, {upper_value=}" + ) + + return lower_value + + def partition(self, partition_spec: PartitionSpec, schema: Schema) -> Record: + return Record(*[self._partition_value(field, schema) for field in partition_spec.fields]) + + def to_serialized_dict(self) -> dict[str, Any]: + lower_bounds = {} + upper_bounds = {} + + for k, agg in self.column_aggregates.items(): + _min = agg.min_as_bytes() + if _min is not None: + lower_bounds[k] = _min + _max = agg.max_as_bytes() + if _max is not None: + upper_bounds[k] = _max + return { + "record_count": self.record_count, + "column_sizes": self.column_sizes, + "value_counts": self.value_counts, + "null_value_counts": self.null_value_counts, + "nan_value_counts": self.nan_value_counts, + "lower_bounds": lower_bounds, + "upper_bounds": upper_bounds, + "split_offsets": self.split_offsets, + } + + +class FileFormatWriter(ABC): + """Writes data to a single file in a specific format.""" + + _result: DataFileStatistics | None = None + + @abstractmethod + def write(self, table: pa.Table) -> None: + """Write a batch of data. May be called multiple times.""" + + @abstractmethod + def close(self) -> DataFileStatistics: + """Finalize the file and return statistics.""" + + def result(self) -> DataFileStatistics: + """Return statistics from a previous close() call.""" + if self._result is None: + raise RuntimeError("Writer has not been closed yet") + return self._result + + def __enter__(self) -> FileFormatWriter: + """Enter the context manager.""" + return self + + def __exit__(self, exc_type: Any, exc_val: Any, exc_tb: Any) -> None: + """Exit the context manager, closing the writer and caching statistics.""" + if exc_type is not None: + try: + self._result = self.close() + except Exception: + pass + return + self._result = self.close() + + +class FileFormatModel(ABC): + """Represents a file format's capabilities. Creates writers.""" + + @property + @abstractmethod + def format(self) -> FileFormat: ... + + @abstractmethod + def file_extension(self) -> str: + """Return file extension without dot, e.g. 'parquet', 'orc'.""" + + @abstractmethod + def create_writer( + self, + output_file: OutputFile, + file_schema: Schema, + properties: Properties, + ) -> FileFormatWriter: ... + + +class FileFormatFactory: + """Registry of FileFormatModel implementations.""" + + _registry: dict[FileFormat, FileFormatModel] = {} + + @classmethod + def register(cls, model: FileFormatModel) -> None: + cls._registry[model.format] = model Review Comment: Will this logic overwrite the models we support once we add that logic? It seems like the java implementation throws. https://github.com/apache/iceberg/blob/8f30d8350bdac64e67e3778cc9489f07a57bc2e7/core/src/main/java/org/apache/iceberg/formats/FormatModelRegistry.java#L88-L101 -- 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. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
