added function to log file content

Author: sanderegg

services/dask-sidecar/src/simcore_service_dask_sidecar/computational_sidecar/core.py

@@ -7,7 +7,7 @@
 from pathlib import Path
 from pprint import pformat
 from types import TracebackType
-from typing import Final, cast
+from typing import cast
 from uuid import uuid4
 
 from aiodocker import Docker
@@ -32,6 +32,7 @@
 from ..utils.dask import TaskPublisher
 from ..utils.files import (
     check_need_unzipping,
+    log_partial_file_content,
     pull_file_from_remote,
     push_file_to_remote,
 )
@@ -49,7 +50,7 @@
 
 _logger = logging.getLogger(__name__)
 CONTAINER_WAIT_TIME_SECS = 2
-_TASK_PROCESSING_PROGRESS_WEIGHT: Final[float] = 0.99
+MAX_LOGGED_FILE_CHARS = 40
 
 
 @dataclass(kw_only=True, frozen=True, slots=True)
@@ -147,11 +148,14 @@ async def _retrieve_output_data(
             upload_tasks = []
             for output_params in output_data.values():
                 if isinstance(output_params, FileUrl):
-                    assert (  # nosec
+                    assert (
                         output_params.file_mapping
                     ), f"{output_params.model_dump_json(indent=1)} expected resolved in TaskOutputData.from_task_output"
 
                     src_path = task_volumes.outputs_folder / output_params.file_mapping
+                    await log_partial_file_content(
+                        src_path, _logger, MAX_LOGGED_FILE_CHARS
+                    )
                     upload_tasks.append(
                         push_file_to_remote(
                             src_path,

services/dask-sidecar/src/simcore_service_dask_sidecar/utils/files.py

@@ -330,3 +330,17 @@ async def push_file_to_remote(
         f"Upload of '{src_path.name}' to '{dst_url.path.strip('/')}' complete",
         logging.INFO,
     )
+
+
+async def log_partial_file_content(
+    src_path: Path, logger: logging.Logger, log_level: int, max_chars: int
+) -> None:
+    if not src_path.exists():
+        logger.log(log_level, "file does not exist: %s", src_path)
+        return
+    async with aiofiles.open(src_path, encoding="utf-8") as f:
+        content = await f.read(max_chars + 1)
+    if len(content) > max_chars:
+        logger.log(log_level, "file content (truncated): %s...", content[:max_chars])
+    else:
+        logger.log(log_level, "file content: %s", content)

services/dask-sidecar/tests/unit/test_utils_files.py

@@ -4,6 +4,7 @@
 
 import asyncio
 import hashlib
+import logging
 import mimetypes
 import zipfile
 from collections.abc import AsyncIterable
@@ -21,6 +22,7 @@
 from settings_library.s3 import S3Settings
 from simcore_service_dask_sidecar.utils.files import (
     _s3fs_settings_from_s3_settings,
+    log_partial_file_content,
     pull_file_from_remote,
     push_file_to_remote,
 )
@@ -511,3 +513,44 @@ async def test_push_file_to_remote_creates_reproducible_zip_archive(
     assert dst_path2.exists()
 
     assert _compute_hash(dst_path1) == _compute_hash(dst_path2)
+
+
+async def test_log_partial_file_content(
+    tmp_path: Path, caplog: pytest.LogCaptureFixture
+):
+    # Create a file with known content
+    file_content = "abcdefghijklmnopqrstuvwxyz0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ"
+    file_path = tmp_path / "testfile.txt"
+    file_path.write_text(file_content)
+    logger = logging.getLogger("pytest.utils.files")
+
+    # Case 1: file longer than max_chars
+    with caplog.at_level(logging.DEBUG, logger=logger.name):
+        await log_partial_file_content(file_path, logger, logging.DEBUG, max_chars=10)
+    assert any(
+        "file content (truncated): abcdefghij..." in record.getMessage()
+        for record in caplog.records
+    )
+
+    # Case 2: file shorter than max_chars
+    caplog.clear()
+    short_content = "short"
+    short_file = tmp_path / "short.txt"
+    short_file.write_text(short_content)
+    with caplog.at_level(logging.DEBUG, logger=logger.name):
+        await log_partial_file_content(short_file, logger, logging.DEBUG, max_chars=10)
+    assert any(
+        "file content: short" in record.getMessage() for record in caplog.records
+    )
+
+    # Case 3: file does not exist
+    caplog.clear()
+    non_existent = tmp_path / "doesnotexist.txt"
+    with caplog.at_level(logging.DEBUG, logger=logger.name):
+        await log_partial_file_content(
+            non_existent, logger, logging.DEBUG, max_chars=10
+        )
+    assert any(
+        f"file does not exist: {non_existent}" in record.getMessage()
+        for record in caplog.records
+    )