File size: 6,612 Bytes
22cc751
 
44402f8
22cc751
44402f8
22cc751
 
 
 
44402f8
22cc751
 
 
 
 
 
 
 
 
 
44402f8
22cc751
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
44402f8
 
 
 
 
 
 
 
 
 
22cc751
 
44402f8
22cc751
 
 
44402f8
22cc751
44402f8
22cc751
 
44402f8
 
22cc751
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
44402f8
 
22cc751
 
44402f8
22cc751
44402f8
 
22cc751
 
 
 
44402f8
 
 
22cc751
 
 
 
 
 
44402f8
 
 
 
 
 
22cc751
44402f8
 
 
22cc751
44402f8
 
 
 
 
 
 
22cc751
 
 
 
 
 
 
44402f8
 
 
 
 
 
22cc751
 
 
44402f8
 
22cc751
 
 
 
 
44402f8
 
 
 
 
 
 
 
22cc751
 
44402f8
22cc751
 
 
 
 
 
 
 
 
 
 
 
 
44402f8
22cc751
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
#!/usr/bin/env python3
"""
ํ•œ ์žฅ ์ด์ƒ์˜ ์ด๋ฏธ์ง€์— ๋Œ€ํ•ด /analyze v1 ๊ณผ ๋™์ผํ•œ ํŒŒ์ดํ”„๋ผ์ธ์„ ๋Œ๋ฆฌ๊ณ 
`<stem>_analyze_response.json`, `<stem>_analyze_summary.txt` ๋ฅผ ์”๋‹ˆ๋‹ค.
์—ฌ๋Ÿฌ ์žฅ์ด๋ฉด ์—…๋กœ๋“œ ์ˆœ์„œ์™€ ๊ฐ™์ด ํ•œ ์š”์ฒญ์œผ๋กœ ๋ณ‘ํ•ฉํ•ฉ๋‹ˆ๋‹ค.

ํ”„๋กœ์ ํŠธ ๋ฃจํŠธ์—์„œ:

  python dump_sample_analyze_artifacts.py sample_oneline_0.png
  python dump_sample_analyze_artifacts.py a.png b.png c.png
  python dump_sample_analyze_artifacts.py sample_imgs/sample_c.jpeg --return-debug
"""

from __future__ import annotations

import argparse
import json
import sys
import uuid
from pathlib import Path
from typing import Any, Dict, List

import cv2

ROOT = Path(__file__).resolve().parent
SAMPLE_IMGS = ROOT / "sample_imgs"
ALLOWED_EXT = {".jpg", ".jpeg", ".png", ".webp"}

if str(ROOT) not in sys.path:
    sys.path.insert(0, str(ROOT))

import main as stella_main  # noqa: E402


def _resolve_image_path(name_or_path: str) -> Path:
    raw = Path(name_or_path)
    if raw.is_file():
        return raw.resolve()
    candidate = SAMPLE_IMGS / raw.name
    if candidate.is_file():
        return candidate.resolve()
    raise FileNotFoundError(
        f"์ด๋ฏธ์ง€๋ฅผ ์ฐพ์„ ์ˆ˜ ์—†์Šต๋‹ˆ๋‹ค: {name_or_path!r} "
        f"(๋ฃจํŠธ ์ƒ๋Œ€ ๊ฒฝ๋กœ ๋˜๋Š” {SAMPLE_IMGS.name}/ ์•ˆ์˜ ํŒŒ์ผ ์ด๋ฆ„์„ ์ฃผ์„ธ์š”.)"
    )


def _load_decoded(path: Path) -> Dict[str, Any]:
    ext = path.suffix.lower()
    if ext not in ALLOWED_EXT:
        raise ValueError(f"์ง€์›ํ•˜์ง€ ์•Š๋Š” ํ™•์žฅ์ž์ž…๋‹ˆ๋‹ค: {ext} (ํ—ˆ์šฉ: {sorted(ALLOWED_EXT)})")
    image = cv2.imread(str(path), cv2.IMREAD_COLOR)
    if image is None:
        raise ValueError(f"์ด๋ฏธ์ง€๋ฅผ ๋””์ฝ”๋”ฉํ•  ์ˆ˜ ์—†์Šต๋‹ˆ๋‹ค: {path}")
    h, w = image.shape[:2]
    return {
        "filename": path.name,
        "bytes": int(path.stat().st_size),
        "width": int(w),
        "height": int(h),
        "image": image,
    }


def _default_score_context() -> Dict[str, Any]:
    return {
        "clef": "treble",
        "key_signature": {"fifths": -1},
        "time_signature": "4/4",
        "tempo_bpm_reference": None,
        "divisions": 4,
    }


def _build_analyze_response(
    decoded: List[Dict[str, Any]],
    score_context: Dict[str, Any],
    *,
    return_debug: bool,
) -> Dict[str, Any]:
    body = stella_main.build_analyze_response_v1(
        decoded,
        score_context,
        return_debug=return_debug,
    )
    body["request_id"] = str(uuid.uuid4())
    return body


def _format_event_table(events: List[Dict[str, Any]]) -> str:
    header = f"{'idx':>3}  {'onset':>5}  {'dur':>3}   {'step':>4}  {'oct':>3}  {'alt':>3}  type"
    sep = "-" * 50
    lines = [header, sep]
    for i, ev in enumerate(events, start=1):
        onset = ev.get("onset_div", "")
        dur = ev.get("duration_div", "")
        typ = ev.get("type", "")
        if typ == "rest":
            lines.append(f"{i:>3}  {onset:>5}  {dur:>3}   {'':>4}  {'':>3}  {'':>3}  {typ}")
        else:
            step = ev.get("step", "")
            oct_ = ev.get("octave", "")
            alt = ev.get("alter", "")
            lines.append(f"{i:>3}  {onset:>5}  {dur:>3}   {str(step):>4}  {str(oct_):>3}  {str(alt):>3}  {typ}")
    return "\n".join(lines)


def build_summary_text(response: Dict[str, Any]) -> str:
    meta = response["meta"]
    tl = response["timeline"]
    sc = response["score_context"]
    mel = response["melody"]
    lines: List[str] = [
        f"request_id: {response['request_id']}",
        f"source: {json.dumps(response.get('source'), ensure_ascii=False)}",
        f"warnings: {json.dumps(response['warnings'], ensure_ascii=False)}",
        f"score_context: {json.dumps(sc, ensure_ascii=False)}",
        f"segment_map: {json.dumps(response.get('segment_map'), ensure_ascii=False)}",
        f"meta.preprocess: {json.dumps(meta.get('preprocess'), ensure_ascii=False)}",
        f"pipeline_mode: {meta.get('pipeline_mode')}",
        f"timeline: {json.dumps({'divisions': tl.get('divisions'), 'time_signature': tl.get('time_signature')}, ensure_ascii=False)}",
        "",
        f"--- melody ({mel.get('voice_id')}) reduction={mel.get('reduction_rule')} events={len(mel.get('events', []))} ---",
        _format_event_table(mel.get("events", [])),
        "",
    ]
    while lines and lines[-1] == "":
        lines.pop()
    return "\n".join(lines) + "\n"


def _output_stem(paths: List[Path]) -> str:
    if len(paths) == 1:
        return paths[0].stem
    return "__".join(p.stem for p in paths)


def main() -> None:
    parser = argparse.ArgumentParser(
        description="์ด๋ฏธ์ง€ 1์žฅ ์ด์ƒ์— ๋Œ€ํ•ด analyze v1 ํŒŒ์ดํ”„๋ผ์ธ ๊ฒฐ๊ณผ๋ฅผ JSONยท์š”์•ฝ ํ…์ŠคํŠธ๋กœ ์ €์žฅํ•ฉ๋‹ˆ๋‹ค."
    )
    parser.add_argument(
        "filenames",
        nargs="+",
        metavar="IMAGE",
        help=(
            f"์ด๋ฏธ์ง€ ๊ฒฝ๋กœ 1๊ฐœ ์ด์ƒ (์ˆœ์„œ = ์•…๋ณด ์ด์–ด์ง). "
            f"ํŒŒ์ผ ์ด๋ฆ„๋งŒ ์ฃผ๋ฉด {SAMPLE_IMGS.name}/ ์—์„œ ์ฐพ์Šต๋‹ˆ๋‹ค."
        ),
    )
    parser.add_argument(
        "--out-dir",
        type=Path,
        default=ROOT / "artifacts",
        help=f"์ถœ๋ ฅ ๋””๋ ‰ํ„ฐ๋ฆฌ (๊ธฐ๋ณธ: {ROOT.name}/artifacts)",
    )
    parser.add_argument(
        "--score-context-json",
        type=str,
        default=None,
        help="JSON ๋ฌธ์ž์—ด (๊ธฐ๋ณธ: treble, fifths=-1, 4/4)",
    )
    parser.add_argument("--return-debug", action="store_true")
    args = parser.parse_args()

    paths = [_resolve_image_path(name) for name in args.filenames]
    stem = _output_stem(paths)
    out_dir = args.out_dir.resolve()
    out_dir.mkdir(parents=True, exist_ok=True)
    json_path = out_dir / f"{stem}_analyze_response.json"
    txt_path = out_dir / f"{stem}_analyze_summary.txt"

    if args.score_context_json:
        score_ctx = json.loads(args.score_context_json)
        if not isinstance(score_ctx, dict):
            raise ValueError("score_context JSON must be an object")
    else:
        score_ctx = _default_score_context()

    decoded = [_load_decoded(p) for p in paths]
    response = _build_analyze_response(
        decoded,
        score_ctx,
        return_debug=args.return_debug,
    )

    json_path.write_text(json.dumps(response, ensure_ascii=False, indent=2), encoding="utf-8")
    txt_path.write_text(build_summary_text(response), encoding="utf-8")

    print(f"Wrote {json_path}")
    print(f"Wrote {txt_path}")


if __name__ == "__main__":
    try:
        main()
    except (FileNotFoundError, ValueError, json.JSONDecodeError, stella_main.SingleStaffAnalyzeError) as exc:
        print(f"error: {exc}", file=sys.stderr)
        sys.exit(1)