dl-organizer/tests/test_analysis_properties.py

"""Property-based tests for the analysis engine.

Tests universal correctness properties using Hypothesis with minimum 100 iterations.
Each test validates a specific property from the design document.
"""

from datetime import datetime, timezone
from pathlib import Path

from hypothesis import given, settings
from hypothesis import strategies as st

from vlm.analysis import analyze_series_completeness, detect_duplicates
from vlm.models import DuplicateGroup, MovieIdentity, SeasonCompleteness, SeriesIdentity, VideoFile
from vlm.reports import (
    generate_completeness_report,
    generate_duplicate_report,
    generate_summary_report,
)


def _movie(title="Movie", year=2020, **kw):
    return MovieIdentity(
        title=title, year=year, confidence=kw.pop("confidence", 0.9),
        needs_review=kw.pop("needs_review", False),
        original_filename=kw.pop("original_filename", f"{title}.{year}.mkv"),
        **kw,
    )


def _series(title="Show", season=1, episodes=None, **kw):
    if episodes is None:
        episodes = [1]
    return SeriesIdentity(
        title=title, season=season, episodes=episodes,
        confidence=kw.pop("confidence", 0.9),
        needs_review=kw.pop("needs_review", False),
        original_filename=kw.pop(
            "original_filename",
            f"{title.replace(' ', '.')}.S{season:02d}E01.mkv"
            if season is not None
            else f"{title.replace(' ', '.')}.E01.mkv",
        ),
        **kw,
    )


def _video(filename="file.mkv", size=1000, category="movie", **kw):
    return VideoFile(
        path=kw.pop("path", Path(f"/tmp/{filename}")),
        filename=filename, size_bytes=size,
        modified_timestamp=kw.pop("modified_timestamp", datetime(2023, 1, 1)),
        category=category, **kw,
    )


# Custom strategies for generating test data

@st.composite
def series_identity_strategy(draw, title=None, season=None):
    """Generate a SeriesIdentity with optional fixed title and season."""
    if title is None:
        title = draw(st.text(min_size=1, max_size=50, alphabet=st.characters(
            whitelist_categories=('Lu', 'Ll', 'Nd'), whitelist_characters=' '
        )))

    if season is None:
        season = draw(st.integers(min_value=1, max_value=20))

    # Generate 1-3 episode numbers
    episode_count = draw(st.integers(min_value=1, max_value=3))
    episodes = draw(st.lists(
        st.integers(min_value=1, max_value=50),
        min_size=episode_count,
        max_size=episode_count,
        unique=True
    ))

    confidence = draw(st.floats(min_value=0.5, max_value=1.0))
    original_filename = f"{title.replace(' ', '.')}.S{season:02d}E{episodes[0]:02d}.mkv"

    return _series(title, season, sorted(episodes),
                   confidence=confidence,
                   original_filename=original_filename)


@st.composite
def movie_identity_strategy(draw, title=None, year=None):
    """Generate a MovieIdentity with optional fixed title and year."""
    if title is None:
        title = draw(st.text(min_size=1, max_size=50, alphabet=st.characters(
            whitelist_categories=('Lu', 'Ll', 'Nd'), whitelist_characters=' '
        )))

    if year is None:
        year = draw(st.integers(min_value=1900, max_value=2030))

    confidence = draw(st.floats(min_value=0.5, max_value=1.0))
    original_filename = f"{title.replace(' ', '.')}.{year}.mkv"

    return _movie(title, year, confidence=confidence,
                  original_filename=original_filename)


@st.composite
def video_file_strategy(draw, filename=None, category="movie"):
    """Generate a VideoFile with optional fixed filename."""
    if filename is None:
        filename = draw(st.text(min_size=5, max_size=50, alphabet=st.characters(
            whitelist_categories=('Lu', 'Ll', 'Nd'), whitelist_characters='.-_'
        ))) + ".mkv"

    size_bytes = draw(st.integers(min_value=1000000, max_value=10000000000))
    now = datetime.now(timezone.utc)

    # Optional metadata
    has_metadata = draw(st.booleans())
    if has_metadata:
        resolution = draw(st.sampled_from(["1920x1080", "1280x720", "3840x2160", "720x480"]))
        codec = draw(st.sampled_from(["h264", "h265", "vp9", "av1"]))
        duration_seconds = draw(st.floats(min_value=300, max_value=10800))
        bitrate_kbps = draw(st.integers(min_value=500, max_value=20000))
        return _video(
            filename, size_bytes, category,
            modified_timestamp=now, path=Path(f"/{category}/{filename}"),
            resolution=resolution, codec=codec,
            duration_seconds=duration_seconds, bitrate_kbps=bitrate_kbps,
        )
    else:
        return _video(
            filename, size_bytes, category,
            modified_timestamp=now, path=Path(f"/{category}/{filename}"),
        )


# Property 10: Gap detection
# Feature: video-library-manager, Property 10: Gap detection
@settings(max_examples=100)
@given(
    title=st.text(min_size=1, max_size=30, alphabet=st.characters(
        whitelist_categories=('Lu', 'Ll'), whitelist_characters=' '
    )),
    season=st.integers(min_value=1, max_value=10),
    # Generate a list of episode numbers with guaranteed gaps
    episodes_data=st.lists(
        st.integers(min_value=1, max_value=30),
        min_size=3,
        max_size=15,
        unique=True
    )
)
def test_property_10_gap_detection(title, season, episodes_data):
    """Property 10: For any set of episodes within season, analysis SHALL detect
    missing episode numbers in range [min, max].

    Validates: Requirements 4.1, 4.2
    """
    # Sort episodes and ensure there's at least one gap
    sorted_episodes = sorted(episodes_data)

    # Create episodes, intentionally removing one to create a gap
    if len(sorted_episodes) >= 3:
        # Remove a middle episode to guarantee a gap
        gap_index = len(sorted_episodes) // 2
        removed_episode = sorted_episodes[gap_index]
        episodes_with_gap = sorted_episodes[:gap_index] + sorted_episodes[gap_index + 1:]

        # Create SeriesIdentity objects
        episode_identities = [
            _series(title, season, [ep],
                    original_filename=f"{title}.S{season:02d}E{ep:02d}.mkv")
            for ep in episodes_with_gap
        ]

        # Analyze completeness
        result = analyze_series_completeness(episode_identities)

        # Should detect the gap
        if len(result) > 0:
            assert result[0].series_title == title
            assert result[0].season == season

            # The missing episode should be in the detected gaps
            min_ep = min(episodes_with_gap)
            max_ep = max(episodes_with_gap)
            expected_missing = set(range(min_ep, max_ep + 1)) - set(episodes_with_gap)

            assert set(result[0].episodes_missing) == expected_missing
            assert removed_episode in result[0].episodes_missing


# Property 11: Multi-season independence
# Feature: video-library-manager, Property 11: Multi-season independence
@settings(max_examples=100)
@given(
    title=st.text(min_size=1, max_size=30, alphabet=st.characters(
        whitelist_categories=('Lu', 'Ll'), whitelist_characters=' '
    )),
    season1_episodes=st.lists(st.integers(min_value=1, max_value=20), min_size=2, max_size=10, unique=True),
    season2_episodes=st.lists(st.integers(min_value=1, max_value=20), min_size=2, max_size=10, unique=True),
)
def test_property_11_multi_season_independence(title, season1_episodes, season2_episodes):
    """Property 11: For any series with multiple seasons, gap detection of one
    season SHALL not affect others.

    Validates: Requirements 4.4
    """
    # Create episodes for season 1 with a gap
    s1_sorted = sorted(season1_episodes)
    if len(s1_sorted) >= 3:
        gap_index = len(s1_sorted) // 2
        s1_with_gap = s1_sorted[:gap_index] + s1_sorted[gap_index + 1:]
        s1_missing = s1_sorted[gap_index]
    else:
        s1_with_gap = s1_sorted
        s1_missing = None

    # Create episodes for season 2 (complete, no gaps)
    s2_sorted = sorted(season2_episodes)
    s2_complete = list(range(min(s2_sorted), max(s2_sorted) + 1))

    # Create SeriesIdentity objects
    episode_identities = []
    for ep in s1_with_gap:
        episode_identities.append(
            _series(title, 1, [ep],
                    original_filename=f"{title}.S01E{ep:02d}.mkv")
        )
    for ep in s2_complete:
        episode_identities.append(
            _series(title, 2, [ep],
                    original_filename=f"{title}.S02E{ep:02d}.mkv")
        )

    # Analyze completeness
    result = analyze_series_completeness(episode_identities)

    # Season 2 should not appear in results (it's complete)
    season2_results = [r for r in result if r.season == 2]
    assert len(season2_results) == 0

    # Season 1 should appear if there's a gap
    if s1_missing is not None:
        season1_results = [r for r in result if r.season == 1]
        if len(season1_results) > 0:
            assert s1_missing in season1_results[0].episodes_missing


# Property 12: Duplicate detection for movies
# Feature: video-library-manager, Property 12: Duplicate detection for movies
@settings(max_examples=100)
@given(
    title=st.text(min_size=1, max_size=30, alphabet=st.characters(
        whitelist_categories=('Lu', 'Ll'), whitelist_characters=' '
    )),
    year=st.integers(min_value=1900, max_value=2030),
    duplicate_count=st.integers(min_value=2, max_value=5)
)
def test_property_12_duplicate_detection_movies(title, year, duplicate_count):
    """Property 12: For any set of movies with identical normalized titles and years,
    all SHALL be grouped as duplicates.

    Validates: Requirements 5.1
    """
    # Create multiple movie identities with same title and year
    identities = []
    files = []
    now = datetime.now(timezone.utc)

    for i in range(duplicate_count):
        filename = f"{title.replace(' ', '.')}.{year}.{i}.mkv"
        identities.append(_movie(title, year, original_filename=filename))
        files.append(_video(
            filename, 1_000_000_000 + i * 100_000_000,
            modified_timestamp=now, path=Path(f"/movies/{filename}"),
        ))

    # Detect duplicates
    result = detect_duplicates(list(zip(identities, files)))

    # Should find exactly one duplicate group
    assert len(result) == 1

    # The group should contain all files
    assert len(result[0].files) == duplicate_count

    # Identity should match
    assert result[0].identity.title == title
    assert result[0].identity.year == year


# Property 13: Duplicate detection for series
# Feature: video-library-manager, Property 13: Duplicate detection for series
@settings(max_examples=100)
@given(
    title=st.text(min_size=1, max_size=30, alphabet=st.characters(
        whitelist_categories=('Lu', 'Ll'), whitelist_characters=' '
    )),
    season=st.integers(min_value=1, max_value=10),
    episode=st.integers(min_value=1, max_value=30),
    duplicate_count=st.integers(min_value=2, max_value=5)
)
def test_property_13_duplicate_detection_series(title, season, episode, duplicate_count):
    """Property 13: For any set of series files with identical normalized titles,
    seasons, and episodes, all SHALL be grouped as duplicates.

    Validates: Requirements 5.2
    """
    # Create multiple series identities with same title, season, and episode
    identities = []
    files = []
    now = datetime.now(timezone.utc)

    for i in range(duplicate_count):
        filename = f"{title.replace(' ', '.')}.S{season:02d}E{episode:02d}.{i}.mkv"
        identities.append(
            _series(title, season, [episode], original_filename=filename)
        )
        files.append(_video(
            filename, 1_000_000_000 + i * 100_000_000, "series",
            modified_timestamp=now, path=Path(f"/series/{filename}"),
        ))

    # Detect duplicates
    result = detect_duplicates(list(zip(identities, files)))

    # Should find exactly one duplicate group
    assert len(result) == 1

    # The group should contain all files
    assert len(result[0].files) == duplicate_count

    # Identity should match
    assert result[0].identity.title == title
    assert result[0].identity.season == season
    assert episode in result[0].identity.episodes


# Property 14: Duplicate quality comparison
# Feature: video-library-manager, Property 14: Duplicate quality comparison
@settings(max_examples=100)
@given(
    title=st.text(min_size=1, max_size=30, alphabet=st.characters(
        whitelist_categories=('Lu', 'Ll'), whitelist_characters=' '
    )),
    year=st.integers(min_value=1900, max_value=2030),
    file_count=st.integers(min_value=2, max_value=4)
)
def test_property_14_duplicate_quality_comparison(title, year, file_count):
    """Property 14: For any duplicate group, comparison data SHALL include
    available metadata for each file.

    Validates: Requirements 5.3
    """
    # Create movie identities and files with varying metadata
    identities = []
    files = []
    now = datetime.now(timezone.utc)

    for i in range(file_count):
        filename = f"{title.replace(' ', '.')}.{year}.{i}.mkv"
        identities.append(_movie(title, year, original_filename=filename))

        # Some files have full metadata, some don't
        if i % 2 == 0:
            files.append(_video(
                filename, 1_000_000_000 + i * 100_000_000,
                modified_timestamp=now, path=Path(f"/movies/{filename}"),
                resolution="1920x1080", codec="h264",
                duration_seconds=7200.0, bitrate_kbps=5000,
            ))
        else:
            files.append(_video(
                filename, 1_000_000_000 + i * 100_000_000,
                modified_timestamp=now, path=Path(f"/movies/{filename}"),
            ))

    # Detect duplicates
    result = detect_duplicates(list(zip(identities, files)))

    # Should have quality comparison data
    assert len(result) == 1
    assert len(result[0].quality_comparison) == file_count

    # Each comparison entry should have at least filename and size
    for comparison in result[0].quality_comparison:
        assert 'filename' in comparison
        assert 'size_bytes' in comparison
        assert 'path' in comparison

        # Files with metadata should have those fields
        if comparison['filename'].endswith('.0.mkv') or comparison['filename'].endswith('.2.mkv'):
            assert 'resolution' in comparison
            assert 'codec' in comparison
            assert 'duration_seconds' in comparison
            assert 'bitrate_kbps' in comparison


# Property 42: Completeness report
# Feature: video-library-manager, Property 42: Completeness report
@settings(max_examples=100)
@given(
    series_count=st.integers(min_value=1, max_value=5),
    format=st.sampled_from(["text", "json"])
)
def test_property_42_completeness_report(series_count, format):
    """Property 42: For any set of analyzed series, completeness report SHALL
    include all series with detected gaps.

    Validates: Requirements 11.2
    """
    # Create series with gaps
    analysis_results = []

    for i in range(series_count):
        title = f"Series {i}"

        analysis_results.append(SeasonCompleteness(
            series_title=title,
            season=1,
            episodes_found=[1, 2, 4, 5],  # Gap at episode 3
            episodes_missing=[3]
        ))

    # Generate report
    library_root = Path("/test/library")
    report = generate_completeness_report(analysis_results, format, library_root)

    # Report should include all series
    for i in range(series_count):
        assert f"Series {i}" in report

    # Report should include metadata
    assert str(library_root) in report


# Property 43: Duplicate report grouping
# Feature: video-library-manager, Property 43: Duplicate report grouping
@settings(max_examples=100)
@given(
    duplicate_count=st.integers(min_value=1, max_value=5),
    format=st.sampled_from(["text", "json"])
)
def test_property_43_duplicate_report_grouping(duplicate_count, format):
    """Property 43: For any set of detected duplicates, duplicate report SHALL
    group files by identity with comparison data.

    Validates: Requirements 11.3
    """
    # Create duplicate groups
    duplicate_groups = []
    now = datetime.now(timezone.utc)

    for i in range(duplicate_count):
        title = f"Movie {i}"
        year = 2020 + i

        # Create 2 files for each duplicate group
        files = []
        quality_comparison = []

        for j in range(2):
            filename = f"{title.replace(' ', '.')}.{year}.{j}.mkv"
            file = _video(
                filename, 1_000_000_000 + j * 500_000_000,
                modified_timestamp=now, path=Path(f"/movies/{filename}"),
                resolution="1920x1080" if j == 0 else "1280x720",
                codec="h264",
            )
            files.append(file)
            quality_comparison.append({
                'filename': filename,
                'path': str(file.path),
                'size_bytes': file.size_bytes,
                'resolution': file.resolution,
                'codec': file.codec
            })

        identity = _movie(title, year, original_filename=files[0].filename)
        duplicate_groups.append(DuplicateGroup(
            identity=identity, files=files,
            quality_comparison=quality_comparison,
        ))

    # Generate report
    library_root = Path("/test/library")
    report = generate_duplicate_report(duplicate_groups, format, library_root)

    # Report should include all duplicate groups
    for i in range(duplicate_count):
        assert f"Movie {i}" in report

    # Report should include comparison data (file sizes, resolutions)
    assert "1920x1080" in report or "resolution" in report.lower()

    # Report should include metadata
    assert str(library_root) in report


# Property 44: Summary report accuracy
# Feature: video-library-manager, Property 44: Summary report accuracy
@settings(max_examples=100)
@given(
    file_count=st.integers(min_value=1, max_value=20),
    categories=st.lists(
        st.sampled_from(["movie", "series", "anime", "other"]),
        min_size=1,
        max_size=4
    )
)
def test_property_44_summary_report_accuracy(file_count, categories):
    """Property 44: For any scanned library, summary report SHALL contain
    accurate counts and sizes.

    Validates: Requirements 11.4
    """
    # Create video files
    files = []
    total_size = 0
    category_counts = {}
    now = datetime.now(timezone.utc)

    for i in range(file_count):
        category = categories[i % len(categories)]
        size = 1_000_000_000 + i * 100_000_000
        filename = f"file_{i}.mkv"

        files.append(_video(
            filename, size, category,
            modified_timestamp=now, path=Path(f"/{category}/{filename}"),
        ))

        total_size += size
        category_counts[category] = category_counts.get(category, 0) + 1

    # Generate summary report
    library_root = Path("/test/library")
    report = generate_summary_report(files, library_root)

    # Report should include total file count
    assert f"Total Files: {file_count}" in report

    # Report should include category breakdown
    for category, count in category_counts.items():
        assert category.capitalize() in report
        assert f"Files: {count}" in report

    # Report should include metadata
    assert str(library_root) in report