refactor: compact duplicate detection output format and simplify CLI

[jmlrt]

Summary

Simplifies duplicate detection commands by removing separate CSV output functionality and optimizing terminal output format. Output is now comma-separated (CSV-compatible) with color-coded fields and truncated precision, improving both readability and line length in terminals.

Changes

Output Format:

• Change separator from - to , (saves ~14 chars/line)
• Truncate similarity scores to 2 decimals (e.g., 91.67 instead of 91.66666666666666)
• Remove ratio_type from find-duplicate-names output (unnecessary verbosity)
• Add ANSI color codes for terminal readability

CLI:

• Keep -o/--output on global parser for other commands (find-tracks, find-relinked-tracks)
• Remove output_file parameter from find_duplicate_ids() and find_duplicate_names()
• Deprecate CSV output for duplicate detection (pipe to files instead)

Code:

• Update docstrings to document CSV-compatible output format
• Remove unused write_duplicates_csv() and write_similarity_csv() functions
• Remove tests for CSV output (dead code)
• Fix pre-commit linting issues

Testing:

• All 38 tests passing
• Tested both commands work with new output format

Usage

# View colored output in terminal
spfm spotify find-duplicate-ids
spfm spotify find-duplicate-names

# Save to file (includes ANSI codes)
spfm spotify find-duplicate-ids > file.csv

# Strip ANSI codes for clean CSV
spfm spotify find-duplicate-ids | sed 's/\x1b\[[0-9;]*m//g' > file.csv

Resolved Issues

✅ Fixed Copilot review: restored -o option for other commands
✅ Fixed breaking changes: updated tests, restored compatibility
✅ Improved output readability: better use of screen real estate with colors

🤖 Generated with Claude Code

[Copilot]

Pull request overview

This PR removes the -o/--output option from the Spotify duplicate-detection CLI commands and changes duplicate output to be printed directly to the terminal (with ANSI coloring) rather than written to a file.

Changes:

• Removes -o/--output from the Spotify CLI parser and stops passing output_file into duplicate detection functions.
• Updates find_duplicate_ids() / find_duplicate_names() to print “CSV-like” colored rows to stdout and adjusts sorting/normalization.
• Modifies the duplicate CSV writer schema for ID duplicates.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 10 comments.

File	Description
spotfm/spotify/dupes.py	Removes output_file params for duplicate detection, adds ANSI-colored terminal output, and adjusts playlist normalization/sorting.
spotfm/cli.py	Removes -o/--output from Spotify CLI args and updates duplicate command dispatch accordingly.

Comments suppressed due to low confidence (2)

spotfm/cli.py:303

• Removing the global -o/--output argument from spotify_parser will break other Spotify commands that still reference args.output (e.g., find-relinked-tracks and find-tracks in spotify_cli), causing an AttributeError at runtime. Either keep -o/--output on the Spotify parser (and ignore it for the duplicate commands), or refactor to per-command subparsers so only duplicate commands lose the option.

    spotify_parser.add_argument(
        "-p", "--playlists", nargs="+", help="Playlist ID(s) or name pattern(s) (use %% as wildcard for LIKE syntax)"
    )
    spotify_parser.add_argument("-f", "--file")
    spotify_parser.add_argument("--start-date", help="Filter by album release start date (YYYY-MM-DD)")
    spotify_parser.add_argument("--end-date", help="Filter by album release end date (YYYY-MM-DD)")
    spotify_parser.add_argument("--genre", help="Filter by genre using a regex pattern")
    spotify_parser.add_argument(

spotfm/spotify/dupes.py:125

• This is a breaking API change: find_duplicate_ids no longer accepts output_file, but the current test suite (e.g., tests/test_dupes.py:228) and any external callers using this parameter will fail. If the CSV-output path is being removed, the tests and any callers should be updated accordingly, or consider keeping output_file temporarily with a deprecation path.

def find_duplicate_ids(excluded_playlist_ids=None):
    """Find tracks that appear multiple times (exact ID match).

    Args:
        excluded_playlist_ids: List of playlist IDs to exclude

    Returns:
        List of dicts with duplicate track information
    """
    if excluded_playlist_ids is None:
        excluded_playlist_ids = []

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

The CLI flag removal here changes user-facing behavior, but repo docs still reference -o for these commands (e.g., README.md:84 and SPEC.md:357 show spfm spotify find-duplicate-ids -o ...). Please update the documentation (and any changelog notes) to match the new CLI behavior.

[Copilot]

MAGENTA and BLUE are defined but unused in this module. Please remove unused color constants (or start using them) to avoid dead code and keep linting clean.

Suggested change

	MAGENTA = "\033[35m"
	BLUE = "\033[34m"

[Copilot]

The terminal output is not reliably CSV-parsable because fields are separated with commas while playlists is also a comma-joined list, and playlist/artist/track names can themselves contain commas/newlines. Consider emitting CSV via csv.writer (with quoting) to stdout, or using a delimiter that cannot appear in the data (consistent with the existing ;-delimited CSV writer).

[Copilot]

This output emits a different number of comma-separated fields depending on whether artists is empty (2 fields vs 3), which makes downstream parsing brittle. For CSV-compatible output, always emit the same columns (e.g., print an empty artists field when missing).

Suggested change

	if dup["artists"]:
	print(f"{playlists_str},{artists_str},{track_str}")
	else:
	print(f"{playlists_str},{track_str}")
	print(f"{playlists_str},{artists_str},{track_str}")

[Copilot]

ANSI color codes are always included in the printed output. When users pipe to a file, this produces non-CSV content by default and contradicts the “CSV-compatible” claim. Consider enabling colors only when sys.stdout.isatty() (and/or honoring NO_COLOR) so piped output is plain CSV.

[Copilot]

playlists1_str/playlists2_str are built from the raw playlist order coming from SQL GROUP_CONCAT (no ORDER BY), so the joined strings can be nondeterministic and the if playlists1_str > playlists2_str swap can flip between runs for the same pair. Sort playlist*_names before joining (and ideally compare normalized lists) to make output stable and normalization correct.

[Copilot]

Same determinism issue as in pass 1: playlists1_str/playlists2_str are joined from an undefined playlist ordering, so the normalization swap can vary between runs. Sort playlist names (or IDs) before joining to make the output stable and the alphabetical normalization meaningful.

[Copilot]

The printed output here uses commas as separators but does not quote/escape fields; artists* values are comma-separated (and track/playlist names can contain commas), so consumers cannot reliably parse this as CSV. If CSV compatibility is desired, write via csv.writer to stdout with proper quoting (and optionally disable ANSI when not a TTY).

[Copilot]

write_duplicates_csv now expects each duplicate dict to contain artists and track_name keys and no longer writes the previous Type/Track columns. This will break existing callers (and current tests in tests/test_dupes.py that pass the older shape). Either keep backward compatibility (accept both key shapes / preserve old columns) or update all callers/tests that rely on the old CSV schema.

[Copilot]

This signature change removes output_file, but the existing tests (e.g., tests/test_dupes.py:501) and any external callers passing output_file will break. If file output is being removed, the tests/callers should be updated; otherwise consider keeping output_file with a deprecation plan.