Lee-W commented on code in PR #49470: URL: https://github.com/apache/airflow/pull/49470#discussion_r2151545734
########## airflow-core/src/airflow/utils/log/log_stream_accumulator.py: ########## @@ -0,0 +1,154 @@ +# +# 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 __future__ import annotations + +import os +import tempfile +from itertools import islice +from typing import IO, TYPE_CHECKING + +if TYPE_CHECKING: + from airflow.typing_compat import Self + from airflow.utils.log.file_task_handler import ( + LogHandlerOutputStream, + StructuredLogMessage, + StructuredLogStream, + ) + + +class LogStreamAccumulator: + """ + Memory-efficient log stream accumulator that tracks the total number of lines while preserving the original stream. + + This class captures logs from a stream and stores them in a buffer, flushing them to disk when the buffer + exceeds a specified threshold. This approach optimizes memory usage while handling large log streams. + + Usage: + + .. code-block:: python + + with LogStreamAccumulator(stream, threshold) as log_accumulator: + # Get total number of lines captured + total_lines = log_accumulator.get_total_lines() + + # Retrieve the original stream of logs + for log in log_accumulator.get_stream(): + print(log) + """ + + def __init__( + self, + stream: LogHandlerOutputStream, + threshold: int, + ) -> None: + """ + Initialize the LogStreamAccumulator. + + Args: + stream: The input log stream to capture and count. + threshold: Maximum number of lines to keep in memory before flushing to disk. + """ + self._stream = stream + self._threshold = threshold + self._buffer: list[StructuredLogMessage] = [] + self._disk_lines: int = 0 + self._tmpfile: IO[str] | None = None + + def _flush_buffer_to_disk(self) -> None: + """Flush the buffer contents to a temporary file on disk.""" + if self._tmpfile is None: + self._tmpfile = tempfile.NamedTemporaryFile(delete=False, mode="w+", encoding="utf-8") + + self._disk_lines += len(self._buffer) + self._tmpfile.writelines(f"{log.model_dump_json()}\n" for log in self._buffer) + self._tmpfile.flush() + self._buffer.clear() + + def _capture(self) -> None: + """Capture logs from the stream into the buffer, flushing to disk when threshold is reached.""" + while True: + # `islice` will try to get up to `self._threshold` lines from the stream. + self._buffer.extend(islice(self._stream, self._threshold)) + # If the buffer has reached the threshold, flush it to disk. + if len(self._buffer) >= self._threshold: + self._flush_buffer_to_disk() + else: # If there are no more lines to capture, exit the loop. + break Review Comment: hmmm... I'm a bit confused here. I thought what I suggested is basically the same thing, but in a different order. is there anything I missed 🤔 -- 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]
