Fokko commented on code in PR #6775:
URL: https://github.com/apache/iceberg/pull/6775#discussion_r1192079490
##########
python/pyiceberg/io/pyarrow.py:
##########
@@ -498,6 +504,49 @@ def expression_to_pyarrow(expr: BooleanExpression) ->
pc.Expression:
return boolean_expression_visit(expr, _ConvertToArrowExpression())
+@lru_cache
+def _get_file_format(file_format: FileFormat, **kwargs: Dict[str, Any]) ->
ds.FileFormat:
+ if file_format == FileFormat.PARQUET:
+ return ds.ParquetFileFormat(**kwargs)
+ else:
+ raise ValueError(f"Unsupported file format: {file_format}")
+
+
+def _construct_fragment(fs: FileSystem, data_file: DataFile,
file_format_kwargs: Dict[str, Any] = EMPTY_DICT) -> ds.Fragment:
+ _, path = PyArrowFileIO.parse_location(data_file.file_path)
+ return _get_file_format(data_file.file_format,
**file_format_kwargs).make_fragment(path, fs)
+
+
+def _read_deletes(fs: FileSystem, data_file: DataFile) -> Dict[str,
pa.ChunkedArray]:
+ delete_fragment = _construct_fragment(
+ fs, data_file, file_format_kwargs={"dictionary_columns":
("file_path",), "pre_buffer": True, "buffer_size": ONE_MEGABYTE}
+ )
+ table = ds.Scanner.from_fragment(fragment=delete_fragment).to_table()
+ table.unify_dictionaries()
+ return {
+ file.as_py(): table.filter(pc.field("file_path") == file).column("pos")
+ for file in table.column("file_path").chunks[0].dictionary
+ }
+
+
+def _create_positional_deletes_indices(positional_deletes:
List[pa.ChunkedArray], fn_rows: Callable[[], int]) -> pa.Array:
+ sorted_deleted = merge(*positional_deletes)
+
+ def generator() -> Generator[int, None, None]:
+ deleted_pos = next(sorted_deleted).as_py() # type: ignore
+ for pos in range(fn_rows()):
+ if deleted_pos == pos:
+ try:
+ deleted_pos = next(sorted_deleted).as_py() # type: ignore
+ except StopIteration:
+ deleted_pos = -1
+ else:
+ yield pos
+
+ # Filter on the positions
+ return pa.array(generator(), type=pa.int64())
Review Comment:
Iceberg has two types of deletes:
- Positional deletes
- Equality deletes
This is the logic to handle positional deletes. There can be one or more
delete files that tell which positions in the data file are considered deleted.
Since there can be more files, we need to merge them. We read the delete file
using PyArrow, and then we don't want to create a Python array with all the
combined delete positions, therefore we have a generator that combines all the
delete files into a single PyArrow array.
--
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]
