Files
dl-organizer/tests/test_analysis.py
T

547 lines
22 KiB
Python
Raw Normal View History

2026-02-09 17:43:35 +08:00
"""Unit tests for the analysis engine.
Tests series completeness analysis, duplicate detection, and quality comparison.
"""
import pytest
from pathlib import Path
from datetime import datetime, timezone
2026-02-09 17:43:35 +08:00
from vlm.models import SeriesIdentity, SeasonCompleteness, MovieIdentity, VideoFile, DuplicateGroup
from vlm.analysis import analyze_series_completeness, detect_duplicates, compare_quality
class TestSeriesCompletenessAnalysis:
"""Test series completeness analysis functionality."""
def test_detect_single_gap(self):
"""Test detection of a single missing episode."""
episodes = [
SeriesIdentity("Show Name", 1, [1], 0.9, False, "Show.Name.S01E01.mkv"),
SeriesIdentity("Show Name", 1, [2], 0.9, False, "Show.Name.S01E02.mkv"),
SeriesIdentity("Show Name", 1, [4], 0.9, False, "Show.Name.S01E04.mkv"),
SeriesIdentity("Show Name", 1, [5], 0.9, False, "Show.Name.S01E05.mkv"),
]
result = analyze_series_completeness(episodes)
assert len(result) == 1
assert result[0].series_title == "Show Name"
assert result[0].season == 1
assert result[0].episodes_found == [1, 2, 4, 5]
assert result[0].episodes_missing == [3]
def test_detect_multiple_gaps(self):
"""Test detection of multiple missing episodes."""
episodes = [
SeriesIdentity("Show Name", 1, [1], 0.9, False, "Show.Name.S01E01.mkv"),
SeriesIdentity("Show Name", 1, [3], 0.9, False, "Show.Name.S01E03.mkv"),
SeriesIdentity("Show Name", 1, [5], 0.9, False, "Show.Name.S01E05.mkv"),
SeriesIdentity("Show Name", 1, [7], 0.9, False, "Show.Name.S01E07.mkv"),
]
result = analyze_series_completeness(episodes)
assert len(result) == 1
assert result[0].episodes_found == [1, 3, 5, 7]
assert result[0].episodes_missing == [2, 4, 6]
def test_no_gaps_returns_empty(self):
"""Test that complete seasons are not included in results."""
episodes = [
SeriesIdentity("Show Name", 1, [1], 0.9, False, "Show.Name.S01E01.mkv"),
SeriesIdentity("Show Name", 1, [2], 0.9, False, "Show.Name.S01E02.mkv"),
SeriesIdentity("Show Name", 1, [3], 0.9, False, "Show.Name.S01E03.mkv"),
]
result = analyze_series_completeness(episodes)
assert len(result) == 0
def test_multi_season_independence(self):
"""Test that gap detection for one season doesn't affect others."""
episodes = [
# Season 1 - has gap at episode 2
SeriesIdentity("Show Name", 1, [1], 0.9, False, "Show.Name.S01E01.mkv"),
SeriesIdentity("Show Name", 1, [3], 0.9, False, "Show.Name.S01E03.mkv"),
# Season 2 - complete
SeriesIdentity("Show Name", 2, [1], 0.9, False, "Show.Name.S02E01.mkv"),
SeriesIdentity("Show Name", 2, [2], 0.9, False, "Show.Name.S02E02.mkv"),
# Season 3 - has gap at episode 5
SeriesIdentity("Show Name", 3, [4], 0.9, False, "Show.Name.S03E04.mkv"),
SeriesIdentity("Show Name", 3, [6], 0.9, False, "Show.Name.S03E06.mkv"),
]
result = analyze_series_completeness(episodes)
# Should have 2 results (seasons 1 and 3 with gaps)
assert len(result) == 2
# Find season 1 result
season1 = next(r for r in result if r.season == 1)
assert season1.episodes_found == [1, 3]
assert season1.episodes_missing == [2]
# Find season 3 result
season3 = next(r for r in result if r.season == 3)
assert season3.episodes_found == [4, 6]
assert season3.episodes_missing == [5]
def test_multi_episode_files(self):
"""Test handling of multi-episode files."""
episodes = [
SeriesIdentity("Show Name", 1, [1, 2], 0.9, False, "Show.Name.S01E01-E02.mkv"),
SeriesIdentity("Show Name", 1, [4], 0.9, False, "Show.Name.S01E04.mkv"),
]
result = analyze_series_completeness(episodes)
assert len(result) == 1
assert result[0].episodes_found == [1, 2, 4]
assert result[0].episodes_missing == [3]
def test_different_series_separate_analysis(self):
"""Test that different series are analyzed separately."""
episodes = [
# Series A - has gap
SeriesIdentity("Series A", 1, [1], 0.9, False, "Series.A.S01E01.mkv"),
SeriesIdentity("Series A", 1, [3], 0.9, False, "Series.A.S01E03.mkv"),
# Series B - complete
SeriesIdentity("Series B", 1, [1], 0.9, False, "Series.B.S01E01.mkv"),
SeriesIdentity("Series B", 1, [2], 0.9, False, "Series.B.S01E02.mkv"),
]
result = analyze_series_completeness(episodes)
# Only Series A should be in results
assert len(result) == 1
assert result[0].series_title == "Series A"
assert result[0].episodes_missing == [2]
def test_skip_episodes_without_season(self):
"""Test that episodes with season=None are excluded from analysis."""
episodes = [
SeriesIdentity("Show Name", 1, [1], 0.9, False, "Show.Name.S01E01.mkv"),
SeriesIdentity("Show Name", 1, [2], 0.9, False, "Show.Name.S01E02.mkv"),
SeriesIdentity("Show Name", None, [1], 0.3, True, "Show.Name.Episode.1.mkv"),
]
result = analyze_series_completeness(episodes)
# Should only analyze season 1, which is complete
assert len(result) == 0
def test_skip_episodes_with_empty_episode_list(self):
"""Test that episodes with empty episode list are excluded from analysis."""
episodes = [
SeriesIdentity("Show Name", 1, [1], 0.9, False, "Show.Name.S01E01.mkv"),
SeriesIdentity("Show Name", 1, [3], 0.9, False, "Show.Name.S01E03.mkv"),
SeriesIdentity("Show Name", 1, [], 0.3, True, "Show.Name.S01.mkv"),
]
result = analyze_series_completeness(episodes)
# Should detect gap at episode 2
assert len(result) == 1
assert result[0].episodes_missing == [2]
def test_non_sequential_start(self):
"""Test gap detection when episodes don't start at 1."""
episodes = [
SeriesIdentity("Show Name", 1, [5], 0.9, False, "Show.Name.S01E05.mkv"),
SeriesIdentity("Show Name", 1, [6], 0.9, False, "Show.Name.S01E06.mkv"),
SeriesIdentity("Show Name", 1, [8], 0.9, False, "Show.Name.S01E08.mkv"),
]
result = analyze_series_completeness(episodes)
# Should detect gap at episode 7 in range [5, 8]
assert len(result) == 1
assert result[0].episodes_found == [5, 6, 8]
assert result[0].episodes_missing == [7]
def test_empty_input(self):
"""Test handling of empty episode list."""
result = analyze_series_completeness([])
assert len(result) == 0
def test_single_episode_no_gap(self):
"""Test that a single episode has no gaps."""
episodes = [
SeriesIdentity("Show Name", 1, [1], 0.9, False, "Show.Name.S01E01.mkv"),
]
result = analyze_series_completeness(episodes)
# Single episode has no gaps
assert len(result) == 0
class TestDuplicateDetection:
"""Test duplicate detection functionality."""
def test_detect_movie_duplicates(self):
"""Test detection of duplicate movies with identical title and year."""
identities = [
MovieIdentity("The Matrix", 1999, 0.9, False, "The.Matrix.1999.1080p.mkv"),
MovieIdentity("The Matrix", 1999, 0.9, False, "The.Matrix.1999.720p.mkv"),
MovieIdentity("Inception", 2010, 0.9, False, "Inception.2010.mkv"),
]
files = [
VideoFile(
Path("/movies/The.Matrix.1999.1080p.mkv"),
"The.Matrix.1999.1080p.mkv",
2000000000,
datetime.now(timezone.utc),
2026-02-09 17:43:35 +08:00
"movie",
resolution="1920x1080",
codec="h264"
),
VideoFile(
Path("/movies/The.Matrix.1999.720p.mkv"),
"The.Matrix.1999.720p.mkv",
1000000000,
datetime.now(timezone.utc),
2026-02-09 17:43:35 +08:00
"movie",
resolution="1280x720",
codec="h264"
),
VideoFile(
Path("/movies/Inception.2010.mkv"),
"Inception.2010.mkv",
1500000000,
datetime.now(timezone.utc),
2026-02-09 17:43:35 +08:00
"movie"
),
]
result = detect_duplicates(list(zip(identities, files)))
2026-02-09 17:43:35 +08:00
# Should find one duplicate group (The Matrix)
assert len(result) == 1
assert isinstance(result[0].identity, MovieIdentity)
assert result[0].identity.title == "The Matrix"
assert result[0].identity.year == 1999
assert len(result[0].files) == 2
assert len(result[0].quality_comparison) == 2
def test_detect_series_duplicates(self):
"""Test detection of duplicate series episodes."""
identities = [
SeriesIdentity("Breaking Bad", 1, [1], 0.9, False, "Breaking.Bad.S01E01.1080p.mkv"),
SeriesIdentity("Breaking Bad", 1, [1], 0.9, False, "Breaking.Bad.S01E01.720p.mkv"),
SeriesIdentity("Breaking Bad", 1, [2], 0.9, False, "Breaking.Bad.S01E02.mkv"),
]
files = [
VideoFile(
Path("/series/Breaking.Bad.S01E01.1080p.mkv"),
"Breaking.Bad.S01E01.1080p.mkv",
1500000000,
datetime.now(timezone.utc),
2026-02-09 17:43:35 +08:00
"series",
resolution="1920x1080"
),
VideoFile(
Path("/series/Breaking.Bad.S01E01.720p.mkv"),
"Breaking.Bad.S01E01.720p.mkv",
800000000,
datetime.now(timezone.utc),
2026-02-09 17:43:35 +08:00
"series",
resolution="1280x720"
),
VideoFile(
Path("/series/Breaking.Bad.S01E02.mkv"),
"Breaking.Bad.S01E02.mkv",
1200000000,
datetime.now(timezone.utc),
2026-02-09 17:43:35 +08:00
"series"
),
]
result = detect_duplicates(list(zip(identities, files)))
2026-02-09 17:43:35 +08:00
# Should find one duplicate group (S01E01)
assert len(result) == 1
assert isinstance(result[0].identity, SeriesIdentity)
assert result[0].identity.title == "Breaking Bad"
assert result[0].identity.season == 1
assert 1 in result[0].identity.episodes
assert len(result[0].files) == 2
def test_no_duplicates(self):
"""Test that unique files are not flagged as duplicates."""
identities = [
MovieIdentity("Movie A", 2020, 0.9, False, "Movie.A.2020.mkv"),
MovieIdentity("Movie B", 2021, 0.9, False, "Movie.B.2021.mkv"),
]
files = [
VideoFile(Path("/movies/Movie.A.2020.mkv"), "Movie.A.2020.mkv", 1000000000, datetime.now(timezone.utc), "movie"),
VideoFile(Path("/movies/Movie.B.2021.mkv"), "Movie.B.2021.mkv", 1000000000, datetime.now(timezone.utc), "movie"),
2026-02-09 17:43:35 +08:00
]
result = detect_duplicates(list(zip(identities, files)))
2026-02-09 17:43:35 +08:00
assert len(result) == 0
def test_skip_movies_without_year(self):
"""Test that movies without year are excluded from duplicate detection."""
identities = [
MovieIdentity("Unknown Movie", None, 0.3, True, "Unknown.Movie.mkv"),
MovieIdentity("Unknown Movie", None, 0.3, True, "Unknown.Movie.2.mkv"),
]
files = [
VideoFile(Path("/movies/Unknown.Movie.mkv"), "Unknown.Movie.mkv", 1000000000, datetime.now(timezone.utc), "movie"),
VideoFile(Path("/movies/Unknown.Movie.2.mkv"), "Unknown.Movie.2.mkv", 1000000000, datetime.now(timezone.utc), "movie"),
2026-02-09 17:43:35 +08:00
]
result = detect_duplicates(list(zip(identities, files)))
2026-02-09 17:43:35 +08:00
# Should not detect duplicates for files needing review
assert len(result) == 0
def test_skip_series_without_season(self):
"""Test that series without season are excluded from duplicate detection."""
identities = [
SeriesIdentity("Unknown Show", None, [1], 0.3, True, "Unknown.Show.E01.mkv"),
SeriesIdentity("Unknown Show", None, [1], 0.3, True, "Unknown.Show.Episode.1.mkv"),
]
files = [
VideoFile(Path("/series/Unknown.Show.E01.mkv"), "Unknown.Show.E01.mkv", 1000000000, datetime.now(timezone.utc), "series"),
VideoFile(Path("/series/Unknown.Show.Episode.1.mkv"), "Unknown.Show.Episode.1.mkv", 1000000000, datetime.now(timezone.utc), "series"),
2026-02-09 17:43:35 +08:00
]
result = detect_duplicates(list(zip(identities, files)))
2026-02-09 17:43:35 +08:00
assert len(result) == 0
def test_skip_series_with_empty_episodes(self):
"""Test that series with empty episode list are excluded."""
identities = [
SeriesIdentity("Show Name", 1, [], 0.3, True, "Show.Name.S01.mkv"),
SeriesIdentity("Show Name", 1, [], 0.3, True, "Show.Name.Season.1.mkv"),
]
files = [
VideoFile(Path("/series/Show.Name.S01.mkv"), "Show.Name.S01.mkv", 1000000000, datetime.now(timezone.utc), "series"),
VideoFile(Path("/series/Show.Name.Season.1.mkv"), "Show.Name.Season.1.mkv", 1000000000, datetime.now(timezone.utc), "series"),
2026-02-09 17:43:35 +08:00
]
result = detect_duplicates(list(zip(identities, files)))
2026-02-09 17:43:35 +08:00
assert len(result) == 0
def test_quality_comparison_includes_all_metadata(self):
"""Test that quality comparison includes all available metadata."""
identities = [
MovieIdentity("Test Movie", 2020, 0.9, False, "Test.Movie.2020.1080p.mkv"),
MovieIdentity("Test Movie", 2020, 0.9, False, "Test.Movie.2020.720p.mkv"),
]
files = [
VideoFile(
Path("/movies/Test.Movie.2020.1080p.mkv"),
"Test.Movie.2020.1080p.mkv",
2000000000,
datetime.now(timezone.utc),
2026-02-09 17:43:35 +08:00
"movie",
resolution="1920x1080",
codec="h264",
duration_seconds=7200.0,
bitrate_kbps=5000
),
VideoFile(
Path("/movies/Test.Movie.2020.720p.mkv"),
"Test.Movie.2020.720p.mkv",
1000000000,
datetime.now(timezone.utc),
2026-02-09 17:43:35 +08:00
"movie",
resolution="1280x720",
codec="h264",
duration_seconds=7200.0,
bitrate_kbps=2500
),
]
result = detect_duplicates(list(zip(identities, files)))
2026-02-09 17:43:35 +08:00
assert len(result) == 1
comparison = result[0].quality_comparison
assert len(comparison) == 2
# Check first file comparison data
assert comparison[0]['filename'] == "Test.Movie.2020.1080p.mkv"
assert comparison[0]['size_bytes'] == 2000000000
assert comparison[0]['resolution'] == "1920x1080"
assert comparison[0]['codec'] == "h264"
assert comparison[0]['duration_seconds'] == 7200.0
assert comparison[0]['bitrate_kbps'] == 5000
# Check second file comparison data
assert comparison[1]['filename'] == "Test.Movie.2020.720p.mkv"
assert comparison[1]['size_bytes'] == 1000000000
assert comparison[1]['resolution'] == "1280x720"
def test_multi_episode_file_duplicates(self):
"""Test duplicate detection for multi-episode files."""
identities = [
SeriesIdentity("Show", 1, [1, 2], 0.9, False, "Show.S01E01-E02.mkv"),
SeriesIdentity("Show", 1, [1], 0.9, False, "Show.S01E01.mkv"),
SeriesIdentity("Show", 1, [2], 0.9, False, "Show.S01E02.mkv"),
]
files = [
VideoFile(Path("/series/Show.S01E01-E02.mkv"), "Show.S01E01-E02.mkv", 2000000000, datetime.now(timezone.utc), "series"),
VideoFile(Path("/series/Show.S01E01.mkv"), "Show.S01E01.mkv", 1000000000, datetime.now(timezone.utc), "series"),
VideoFile(Path("/series/Show.S01E02.mkv"), "Show.S01E02.mkv", 1000000000, datetime.now(timezone.utc), "series"),
2026-02-09 17:43:35 +08:00
]
result = detect_duplicates(list(zip(identities, files)))
2026-02-09 17:43:35 +08:00
# Should find duplicates for both E01 and E02
assert len(result) == 2
def test_different_years_not_duplicates(self):
"""Test that same title with different years are not duplicates."""
identities = [
MovieIdentity("The Thing", 1982, 0.9, False, "The.Thing.1982.mkv"),
MovieIdentity("The Thing", 2011, 0.9, False, "The.Thing.2011.mkv"),
]
files = [
VideoFile(Path("/movies/The.Thing.1982.mkv"), "The.Thing.1982.mkv", 1000000000, datetime.now(timezone.utc), "movie"),
VideoFile(Path("/movies/The.Thing.2011.mkv"), "The.Thing.2011.mkv", 1000000000, datetime.now(timezone.utc), "movie"),
2026-02-09 17:43:35 +08:00
]
result = detect_duplicates(list(zip(identities, files)))
2026-02-09 17:43:35 +08:00
assert len(result) == 0
def test_different_seasons_not_duplicates(self):
"""Test that same series/episode in different seasons are not duplicates."""
identities = [
SeriesIdentity("Show", 1, [1], 0.9, False, "Show.S01E01.mkv"),
SeriesIdentity("Show", 2, [1], 0.9, False, "Show.S02E01.mkv"),
]
files = [
VideoFile(Path("/series/Show.S01E01.mkv"), "Show.S01E01.mkv", 1000000000, datetime.now(timezone.utc), "series"),
VideoFile(Path("/series/Show.S02E01.mkv"), "Show.S02E01.mkv", 1000000000, datetime.now(timezone.utc), "series"),
2026-02-09 17:43:35 +08:00
]
result = detect_duplicates(list(zip(identities, files)))
2026-02-09 17:43:35 +08:00
assert len(result) == 0
class TestQualityComparison:
"""Test quality comparison functionality."""
def test_compare_quality_with_all_metadata(self):
"""Test quality comparison with all metadata available."""
files = [
VideoFile(
Path("/test/file1.mkv"),
"file1.mkv",
2000000000,
datetime.now(timezone.utc),
2026-02-09 17:43:35 +08:00
"movie",
resolution="1920x1080",
codec="h264",
duration_seconds=7200.0,
bitrate_kbps=5000
),
VideoFile(
Path("/test/file2.mkv"),
"file2.mkv",
1000000000,
datetime.now(timezone.utc),
2026-02-09 17:43:35 +08:00
"movie",
resolution="1280x720",
codec="h265",
duration_seconds=7200.0,
bitrate_kbps=2500
),
]
result = compare_quality(files)
assert len(result) == 2
assert result[0]['filename'] == "file1.mkv"
assert result[0]['size_bytes'] == 2000000000
assert result[0]['resolution'] == "1920x1080"
assert result[0]['codec'] == "h264"
assert result[0]['duration_seconds'] == 7200.0
assert result[0]['bitrate_kbps'] == 5000
assert result[1]['filename'] == "file2.mkv"
assert result[1]['size_bytes'] == 1000000000
assert result[1]['resolution'] == "1280x720"
assert result[1]['codec'] == "h265"
def test_compare_quality_with_partial_metadata(self):
"""Test quality comparison when some metadata is missing."""
files = [
VideoFile(
Path("/test/file1.mkv"),
"file1.mkv",
2000000000,
datetime.now(timezone.utc),
2026-02-09 17:43:35 +08:00
"movie",
resolution="1920x1080"
# codec, duration, bitrate not available
),
VideoFile(
Path("/test/file2.mkv"),
"file2.mkv",
1000000000,
datetime.now(timezone.utc),
2026-02-09 17:43:35 +08:00
"movie"
# No optional metadata
),
]
result = compare_quality(files)
assert len(result) == 2
assert result[0]['filename'] == "file1.mkv"
assert result[0]['size_bytes'] == 2000000000
assert result[0]['resolution'] == "1920x1080"
assert 'codec' not in result[0]
assert 'duration_seconds' not in result[0]
assert 'bitrate_kbps' not in result[0]
assert result[1]['filename'] == "file2.mkv"
assert result[1]['size_bytes'] == 1000000000
assert 'resolution' not in result[1]
assert 'codec' not in result[1]
def test_compare_quality_empty_list(self):
"""Test quality comparison with empty file list."""
result = compare_quality([])
assert len(result) == 0
def test_compare_quality_single_file(self):
"""Test quality comparison with single file."""
files = [
VideoFile(
Path("/test/file.mkv"),
"file.mkv",
1500000000,
datetime.now(timezone.utc),
2026-02-09 17:43:35 +08:00
"movie",
resolution="1920x1080",
codec="h264"
),
]
result = compare_quality(files)
assert len(result) == 1
assert result[0]['filename'] == "file.mkv"
assert result[0]['size_bytes'] == 1500000000