reaper-control/scripts/analyze_drumloop.py

"""Analyze drumloops from the library and output structured forensic data.

Usage:
    python scripts/analyze_drumloop.py [--count N] [--output PATH] [--json PATH]
"""

from __future__ import annotations

import argparse
import json
import sys
from pathlib import Path

PROJECT = Path(__file__).resolve().parent.parent
sys.path.insert(0, str(PROJECT))

from src.composer.drum_analyzer import DrumLoopAnalyzer


def load_drumloop_paths(index_path: Path, count: int = 5) -> list[dict]:
    with open(index_path, "r", encoding="utf-8") as f:
        data = json.load(f)

    drumloops = [s for s in data["samples"] if s.get("role") == "drumloop"]

    scored = []
    for d in drumloops:
        path = Path(d["original_path"])
        if not path.exists():
            continue
        dur = d.get("signal", {}).get("duration", 0)
        onsets = d.get("perceptual", {}).get("onset_count", 0)
        density = onsets / max(dur, 0.01)
        tempo = d.get("perceptual", {}).get("tempo", 0)
        if 85 <= tempo <= 150 and 6 <= dur <= 35:
            scored.append((d, density))

    scored.sort(key=lambda x: abs(x[1] - 4.0))
    return [s[0] for s in scored[:count]]


def print_analysis(result_dict: dict) -> None:
    print(f"\n{'='*70}")
    print(f"  {Path(result_dict['file_path']).name}")
    print(f"{'='*70}")
    print(f"  BPM:         {result_dict['bpm']}")
    print(f"  Duration:    {result_dict['duration']:.2f}s")
    print(f"  Bars:        {result_dict['bar_count']}")
    print(f"  Key:         {result_dict['key']} (conf: {result_dict['key_confidence']:.2f})")
    print(f"  Beat grid:   {len(result_dict['beat_grid']['quarter'])} quarters, "
          f"{len(result_dict['beat_grid']['eighth'])} eighths, "
          f"{len(result_dict['beat_grid']['sixteenth'])} sixteenths")

    summary = result_dict["summary"]
    total = summary["kick_count"] + summary["snare_count"] + summary["hihat_count"] + summary["other_count"]
    print(f"\n  Transients:  {total} total")
    print(f"    Kicks:     {summary['kick_count']}")
    print(f"    Snares:    {summary['snare_count']}")
    print(f"    HiHats:    {summary['hihat_count']}")
    print(f"    Other:     {summary['other_count']}")

    transients_by_type = {}
    for t in result_dict["transients"]:
        transients_by_type.setdefault(t["type"], []).append(t)

    for ttype in ["kick", "snare", "hihat", "other"]:
        ts = transients_by_type.get(ttype, [])
        if not ts:
            continue
        print(f"\n  {ttype.upper()} positions (beat positions):")
        positions = [f"{t['beat_pos']:.2f}" for t in ts[:20]]
        line = "    " + "  ".join(positions)
        if len(ts) > 20:
            line += f"  ... +{len(ts)-20} more"
        print(line)

    if result_dict["energy_profile"]:
        print(f"\n  Energy profile (first 16 beats):")
        bars_e = result_dict["energy_profile"][:16]
        max_e = max(bars_e) if bars_e else 1
        for i, e in enumerate(bars_e):
            bar = i // 4
            beat = i % 4
            filled = int((e / max_e) * 30) if max_e > 0 else 0
            print(f"    Bar {bar+1} Beat {beat+1}: {'|' * filled} ({e:.4f})")


def main():
    parser = argparse.ArgumentParser(description="Analyze drumloops forensically")
    parser.add_argument("--count", type=int, default=3, help="Number of drumloops to analyze")
    parser.add_argument("--index", type=str, default=None, help="Path to sample_index.json")
    parser.add_argument("--file", type=str, default=None, help="Analyze a single file instead")
    parser.add_argument("--json", type=str, default=None, help="Save results as JSON")
    args = parser.parse_args()

    index_path = Path(args.index) if args.index else PROJECT / "data" / "sample_index.json"
    results = []

    if args.file:
        print(f"Analyzing: {args.file}")
        analyzer = DrumLoopAnalyzer(args.file)
        result = analyzer.analyze()
        results.append(result.to_dict())
        print_analysis(results[0])
    else:
        drumloops = load_drumloop_paths(index_path, args.count)
        if not drumloops:
            print("No suitable drumloops found.")
            return

        print(f"Selected {len(drumloops)} drumloops for analysis:\n")
        for d in drumloops:
            print(f"  - {d['original_name']} "
                  f"(tempo={d.get('perceptual',{}).get('tempo','?')}, "
                  f"dur={d.get('signal',{}).get('duration',0):.1f}s)")

        for d in drumloops:
            path = d["original_path"]
            print(f"\nAnalyzing: {Path(path).name}...")
            analyzer = DrumLoopAnalyzer(path)
            result = analyzer.analyze()
            results.append(result.to_dict())
            print_analysis(result.to_dict())

    if args.json:
        out_path = Path(args.json)
        out_path.parent.mkdir(parents=True, exist_ok=True)
        with open(out_path, "w", encoding="utf-8") as f:
            json.dump(results, f, indent=2, ensure_ascii=False)
        print(f"\nResults saved to: {out_path}")

    print(f"\n{'='*70}")
    print("DRUMLOOP-FIRST GENERATION APPROACH")
    print(f"{'='*70}")
    print("""
  1. SELECT drumloop -> extract BPM + beat grid + transient map
  2. ALIGN project   -> set REAPER tempo to drumloop BPM
  3. GENERATE bass   -> tresillo pattern in kick-free zones
     - Reggaeton tresillo: notes at 0.0, 0.75, 1.5, 2.0, 2.75, 3.5
     - Place bass between kick transients (margin +/-0.15 beats)
  4. GENERATE chords -> change on downbeats (beat 1 of each bar)
     - Sustain through bar, use i-VI-III-VII progression
     - Match key from drumloop analysis
  5. GENERATE melody -> place on transient-free zones
     - Emphasize chord tones on strong beats
     - Syncopation matches dembow feel
  6. GENERATE vocals -> chops in gaps between drum transients
  7. SELECT samples  -> match drumloop key for compatible tonal samples
""")


if __name__ == "__main__":
    main()