更新支持pg下载多版本歌曲

+#!/usr/bin/env python3
+import argparse
+import csv
+import os
+import re
+import sys
+from pathlib import Path
+from urllib.parse import urlparse
+
+from dotenv import load_dotenv
+from qcloud_cos import CosConfig, CosS3Client
+
+load_dotenv(Path(__file__).resolve().parents[1] / '.env')
+
+AUDIO_EXTS = {'.mp3', '.wav', '.flac', '.m4a', '.aac', '.ogg', '.wma', '.ape', '.alac'}
+
+
+def normalize_key(raw_url: str) -> str:
+    if not raw_url:
+        return ''
+    raw_url = raw_url.strip()
+    if re.match(r'^https?://', raw_url, re.I):
+        return urlparse(raw_url).path.lstrip('/')
+    return raw_url.lstrip('/')
+
+
+def ext_ok(path: str) -> bool:
+    return Path(path).suffix.lower() in AUDIO_EXTS
+
+
+def main():
+    ap = argparse.ArgumentParser()
+    ap.add_argument('--manifest', default='output_selection_budgeted/selected_files.csv')
+    ap.add_argument('--output-dir', default='downloads')
+    ap.add_argument('--types', default='1,7,8,11,16')
+    ap.add_argument('--song-limit', type=int, default=3)
+    ap.add_argument('--start-song-offset', type=int, default=0)
+    ap.add_argument('--overwrite', action='store_true')
+    ap.add_argument('--fail-log', default='download_failures.csv')
+    args = ap.parse_args()
+
+    region = os.environ['COS_REGION']
+    secret_id = os.environ['COS_SECRET_ID']
+    secret_key = os.environ['COS_SECRET_KEY']
+    bucket = os.environ['COS_BUCKET']
+
+    config = CosConfig(Region=region, SecretId=secret_id, SecretKey=secret_key)
+    client = CosS3Client(config)
+    wanted_types = {int(x) for x in args.types.split(',') if x.strip()}
+
+    out_dir = Path(args.output_dir)
+    out_dir.mkdir(parents=True, exist_ok=True)
+    fail_log = Path(args.fail_log)
+    fail_exists = fail_log.exists()
+    fail_f = fail_log.open('a', encoding='utf-8', newline='')
+    fail_w = csv.writer(fail_f)
+    if not fail_exists:
+        fail_w.writerow(['song_id', 'type', 'url', 'error'])
+
+    downloaded = skipped = errors = 0
+    active_song_ids = []
+    current_song = None
+    started = False
+    song_counter = 0
+
+    with open(args.manifest, 'r', encoding='utf-8', newline='') as f:
+        reader = csv.DictReader(f)
+        for row in reader:
+            sid = row['song_id']
+            if sid != current_song:
+                current_song = sid
+                if song_counter < args.start_song_offset:
+                    song_counter += 1
+                    continue
+                if args.song_limit and started and len(active_song_ids) >= args.song_limit and sid not in set(active_song_ids):
+                    break
+                if sid not in active_song_ids:
+                    active_song_ids.append(sid)
+                    started = True
+            if sid not in active_song_ids:
+                continue
+            try:
+                t = int(row['type'])
+            except Exception:
+                skipped += 1
+                continue
+            if t not in wanted_types:
+                skipped += 1
+                continue
+            key = normalize_key(row['url'])
+            if not key or not ext_ok(key):
+                skipped += 1
+                continue
+            target_dir = out_dir / sid / f'type_{t}'
+            target_dir.mkdir(parents=True, exist_ok=True)
+            target = target_dir / Path(key).name
+            if target.exists() and not args.overwrite:
+                skipped += 1
+                continue
+            try:
+                client.download_file(Bucket=bucket, Key=key, DestFilePath=str(target))
+                downloaded += 1
+                if downloaded % 50 == 0:
+                    print({'downloaded': downloaded, 'songs': len(active_song_ids), 'last': str(target)})
+            except Exception as e:
+                errors += 1
+                fail_w.writerow([sid, t, row['url'], str(e)])
+                fail_f.flush()
+                print({'error_song': sid, 'type': t, 'key': key, 'error': str(e)}, file=sys.stderr)
+
+    fail_f.close()
+    print({
+        'downloaded_files': downloaded,
+        'skipped_rows': skipped,
+        'error_files': errors,
+        'song_count': len(active_song_ids),
+        'output_dir': str(out_dir.resolve()),
+    })
+
+
+if __name__ == '__main__':
+    main()