bellingcat/auto-archiver
1
0
Fork 0
mirror of https://github.com/bellingcat/auto-archiver.git synced 2026-06-08 03:18:28 +03:00
Code Issues Packages Projects Releases Wiki Activity

thumbnail enricher always run probe by default

Browse Source
This commit is contained in:
msramalho
2025-06-17 20:28:20 +01:00
parent c5a2fd45f9
commit 2aec240128
2 changed files with 21 additions and 16 deletions
Download Patch File Download Diff File Expand all files Collapse all files

22
src/auto_archiver/modules/thumbnail_enricher/thumbnail_enricher.py
View File

@@ -35,16 +35,18 @@ class ThumbnailEnricher(Enricher):
logger.debug(f"generating thumbnails for {m.filename}")
duration = m.get("duration")
if duration is None:
try:
probe = ffmpeg.probe(m.filename)
duration = float(
next(stream for stream in probe["streams"] if stream["codec_type"] == "video")["duration"]
)
to_enrich.media[m_id].set("duration", duration)
except Exception as e:
logger.error(f"error getting duration of video {m.filename}: {e}")
return
try:
probe = ffmpeg.probe(m.filename)
duration = float(
next(stream for stream in probe["streams"] if stream["codec_type"] == "video")["duration"]
)
to_enrich.media[m_id].set("duration", duration)
except Exception as e:
logger.warning(f"failed to get duration with FFMPEG from {m.filename}: {e}")
if not duration or type(duration) not in [float, int] or duration <= 0:
logger.warning(f"cannot generate thumbnails for {m.filename} without valid duration")
continue
num_thumbs = int(min(max(1, (duration / 60) * self.thumbnails_per_minute), self.max_thumbnails))
timestamps = [duration / (num_thumbs + 1) * i for i in range(1, num_thumbs + 1)]

15
tests/enrichers/test_thumbnail_enricher.py
View File

@@ -74,12 +74,12 @@ def test_enrich_thumbnail_limits(
def test_enrich_handles_probe_failure(thumbnail_enricher, metadata_with_video, mocker):
mocker.patch("ffmpeg.probe", side_effect=Exception("Probe error"))
mocker.patch("os.makedirs")
mock_logger = mocker.patch("loguru.logger.error")
mock_logger = mocker.patch("loguru.logger.warning")
mocker.patch.object(Media, "is_video", return_value=True)
thumbnail_enricher.enrich(metadata_with_video)
# Ensure error was logged
mock_logger.assert_called_with("error getting duration of video video.mp4: Probe error")
mock_logger.assert_called_with("cannot generate thumbnails for video.mp4 without valid duration")
# Ensure no thumbnails were created
thumbnails = metadata_with_video.media[0].get("thumbnails")
assert thumbnails is None
@@ -126,11 +126,14 @@ def test_enrich_handles_short_video(
assert len(thumbnails) == expected_count
def test_uses_existing_duration(thumbnail_enricher, metadata_with_video, mock_ffmpeg_environment):
metadata_with_video.media[0].set("duration", 60)
def test_uses_existing_duration_on_exception(thumbnail_enricher, metadata_with_video, mock_ffmpeg_environment, mocker):
mock_logger = mocker.patch("loguru.logger.warning")
mock_probe = mocker.patch("ffmpeg.probe", side_effect=Exception("New probe error"))
metadata_with_video.media[0].set("duration", 3)
thumbnail_enricher.enrich(metadata_with_video)
mock_ffmpeg_environment["mock_probe"].assert_not_called()
assert mock_ffmpeg_environment["mock_output"].run.call_count == 4
mock_probe.assert_called_once()
mock_logger.assert_called_with("failed to get duration with FFMPEG from video.mp4: New probe error")
assert mock_ffmpeg_environment["mock_output"].run.call_count == 3
def test_enrich_metadata_structure(thumbnail_enricher, metadata_with_video, mock_ffmpeg_environment, mocker):