automatically normalize audio to -14 LUFS, multithread, use semaphore to

avoid lyrics rate limit

3 files changed, 148 insertions, 50 deletions

diff --git a/main.py b/main.py
index 1982a90..a031dd7 100644
--- a/main.py
+++ b/main.py
@@ -1,5 +1,8 @@
 import argparse
 import subprocess
+import threading
+import time
+from concurrent.futures import ThreadPoolExecutor, as_completed
 from pathlib import Path
 from typing import Iterator
 from tqdm import tqdm
@@ -7,6 +10,8 @@ import syncedlyrics
 from mutagen.flac import FLAC
 import ffmpeg
 
+_lyrics_semaphore = threading.Semaphore(1)
+
 
 def iter_files(base: Path) -> Iterator[Path]:
     iterator = base.rglob("*")
@@ -78,7 +83,8 @@ def get_track_info(path: Path) -> tuple:
     audio = FLAC(str(path))
     title = audio.get("TITLE", [""])
     artist = audio.get("ARTIST", [""])
-    return (title[0], artist[0])
+    album = audio.get("ALBUM", [""])
+    return (title[0], artist[0], album[0])
 
 
 def resize_album_art(path: Path) -> None:
@@ -97,76 +103,125 @@ def resize_album_art(path: Path) -> None:
     audio.save()
 
 
+def normalize_loudness(path: Path, target_lufs: float = -14.0, target_tp: float = -1.0) -> Path:
+    temp_path = path.with_suffix(".tmp.flac")
+    subprocess.run(
+        [
+            "ffmpeg-normalize", str(path),
+            "-o", str(temp_path),
+            "-c:a", "flac",
+            "-t", str(target_lufs),
+            "--true-peak", str(target_tp),
+            "-f",
+        ],
+        capture_output=True,
+        text=True,
+        check=True,
+    )
+    path.unlink()
+    temp_path.rename(path)
+    return path
+
+
 def sanitize_filename(name: str) -> str:
     illegal = r'\/:*?"<>|'
     return "".join(c for c in name if c not in illegal).strip()
 
 
-def main():
-    parser = argparse.ArgumentParser()
-    parser.add_argument("base_dir", type=Path)
-    parser.add_argument("--nolrc", "-n", action="store_true", dest="nolrc")
+def process_file(fp: Path, nolrc: bool) -> str:
+    lines = []
+    log = lines.append
 
-    args = parser.parse_args()
+    log(f"\nProcessing: {fp.name}")
+    title, artist, album = get_track_info(fp)
 
-    base = args.base_dir
+    if not title:
+        log(f"  Warning: TITLE tag is empty, using filename as title")
+        title = fp.stem
+        artist = "UNKNOWN ARTIST"
 
-    files = [p for p in iter_files(base) if p.suffix == ".flac"]
+    new_stem = sanitize_filename(f"{title} - {artist}")
+    target = fp.with_name(new_stem + ".flac")
+    if target.exists() and target.resolve() != fp.resolve():
+        if album:
+            log(f"  Conflict detected, adding album name as differentiator")
+            new_stem = sanitize_filename(f"{title} - {artist} ({album})")
+        else:
+            log(f"  Warning: filename conflict but no album tag, keeping original name")
+            new_stem = fp.stem
+    new_file_name = new_stem + ".flac"
+    if new_file_name != fp.name:
+        rename_file(fp, new_file_name)
+        fp = fp.with_name(new_file_name)
 
-    for fp in tqdm(files, desc="Processing FLAC files", unit="file"):
-        print(f"\nProcessing: {fp.name}")
-        title, artist = get_track_info(fp)
+    issues = get_audio_issues(fp)
+    log(f"  Stats: {issues['sample_rate']}Hz, {issues['bits_per_sample']}-bit, blocksize={issues['max_blocksize']}")
 
-        if not title:
-            print(f"  Warning: TITLE tag is empty, using filename as title")
-            title = fp.stem
-            artist = "UNKNOWN ARTIST"
+    if issues["needs_sample_rate_fix"] or issues["needs_bitdepth_fix"]:
+        reasons = []
+        if issues["needs_sample_rate_fix"]:
+            reasons.append(f"sample rate {issues['sample_rate']}Hz -> 192000Hz")
+        if issues["needs_bitdepth_fix"]:
+            reasons.append(f"bit depth {issues['bits_per_sample']}-bit -> 24-bit")
+        log(f"  Fixing via ffmpeg: {', '.join(reasons)}")
+        fp = fix_with_ffmpeg(fp, issues["needs_sample_rate_fix"], issues["needs_bitdepth_fix"])
 
-        new_file_name = sanitize_filename(f"{title} - {artist}") + ".flac"
-        if new_file_name != fp.name:
-            rename_file(fp, new_file_name)
-            fp = fp.with_name(new_file_name)
+    log("  Normalizing loudness to -14 LUFS")
+    fp = normalize_loudness(fp)
 
-        issues = get_audio_issues(fp)
-        print(f"  Stats: {issues['sample_rate']}Hz, {issues['bits_per_sample']}-bit, blocksize={issues['max_blocksize']}")
+    post_blocksize = getattr(FLAC(str(fp)).info, "max_blocksize", 4096)
+    if post_blocksize != 4096:
+        log(f"  Fixing blocksize -> 4096 via flac CLI")
+        fp = fix_blocksize(fp)
 
-        if issues["needs_sample_rate_fix"] or issues["needs_bitdepth_fix"]:
-            reasons = []
-            if issues["needs_sample_rate_fix"]:
-                reasons.append(f"sample rate {issues['sample_rate']}Hz -> 192000Hz")
-            if issues["needs_bitdepth_fix"]:
-                reasons.append(f"bit depth {issues['bits_per_sample']}-bit -> 24-bit")
-            print(f"  Fixing via ffmpeg: {', '.join(reasons)}")
-            fp = fix_with_ffmpeg(fp, issues["needs_sample_rate_fix"], issues["needs_bitdepth_fix"])
+    log("  Resizing album art to 500x500")
+    resize_album_art(fp)
 
-            post_info = FLAC(str(fp)).info
-            if getattr(post_info, "max_blocksize", 4096) != 4096:
-                issues["needs_blocksize_fix"] = True
+    if nolrc:
+        return "\n".join(lines)
 
-        if issues["needs_blocksize_fix"]:
-            print(f"  Fixing blocksize -> 4096 via flac CLI")
-            fp = fix_blocksize(fp)
+    if not title or not artist:
+        log(f"  Skipping LRC for {fp.name} (missing title or artist tag)")
+        return "\n".join(lines)
 
-        print("  Resizing album art to 500x500")
-        resize_album_art(fp)
+    lrc_path = fp.with_suffix(".lrc")
+    if lrc_path.exists():
+        log(f"  Skipping LRC for {fp.name} (already exists)")
+        return "\n".join(lines)
 
-        if args.nolrc:
-            continue
+    log(f"  Fetching LRC for: {title} - {artist}")
+    with _lyrics_semaphore:
+        lrc = syncedlyrics.search(f"{title} {artist}", providers=["Lrclib", "Megalobiz", "NetEase"])
+        time.sleep(0.5)
 
-        if not title or not artist:
-            print(f"  Skipping LRC for {fp.name} (missing title or artist tag)")
-            continue
+    with open(lrc_path, "w", encoding="utf-8") as f:
+        f.write(lrc if lrc else "")
 
-        lrc_path = fp.with_suffix(".lrc")
-        if lrc_path.exists():
-            print(f"  Skipping LRC for {fp.name} (already exists)")
-            continue
+    return "\n".join(lines)
 
-        print(f"  Fetching LRC for: {title} - {artist}")
-        lrc = syncedlyrics.search(f"{title} {artist}", providers=["Lrclib", "Megalobiz", "NetEase"])
 
-        with open(lrc_path, "w", encoding="utf-8") as f:
-            f.write(lrc if lrc else "")
+def main():
+    parser = argparse.ArgumentParser()
+    parser.add_argument("base_dir", type=Path)
+    parser.add_argument("--nolrc", action="store_true", dest="nolrc")
+    parser.add_argument("-n", "--workers", type=int, default=1, dest="workers",
+                        help="Number of parallel workers (default: 1)")
+    args = parser.parse_args()
+
+    files = [p for p in iter_files(args.base_dir) if p.suffix == ".flac"]
+
+    with ThreadPoolExecutor(max_workers=args.workers) as executor:
+        futures = {executor.submit(process_file, fp, args.nolrc): fp for fp in files}
+        with tqdm(total=len(files), desc="Processing FLAC files", unit="file") as pbar:
+            for future in as_completed(futures):
+                try:
+                    result = future.result()
+                    tqdm.write(result)
+                except Exception as e:
+                    fp = futures[future]
+                    tqdm.write(f"\n  ERROR processing {fp.name}: {e}")
+                finally:
+                    pbar.update(1)
 
 
 if __name__ == "__main__":
diff --git a/pyproject.toml b/pyproject.toml
index d6449eb..9364074 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -5,6 +5,7 @@ description = "Media tool for retagging FLACs for Snowsky Echo/Mini"
 readme = "README.md"
 requires-python = ">=3.13"
 dependencies = [
+    "ffmpeg-normalize>=1.37.6",
     "ffmpeg-python>=0.2.0",
     "mutagen>=1.47.0",
     "pillow>=12.2.0",
diff --git a/uv.lock b/uv.lock
index eec6aed..2503869 100644
--- a/uv.lock
+++ b/uv.lock
@@ -91,6 +91,46 @@ wheels = [
 ]
 
 [[package]]
+name = "colorlog"
+version = "6.7.0"
+source = { registry = "https://pypi.org/simple" }
+dependencies = [
+    { name = "colorama", marker = "sys_platform == 'win32'" },
+]
+sdist = { url = "https://files.pythonhosted.org/packages/78/6b/4e5481ddcdb9c255b2715f54c863629f1543e97bc8c309d1c5c131ad14f2/colorlog-6.7.0.tar.gz", hash = "sha256:bd94bd21c1e13fac7bd3153f4bc3a7dc0eb0974b8bc2fdf1a989e474f6e582e5", size = 29920, upload-time = "2022-08-29T14:51:27.945Z" }
+wheels = [
+    { url = "https://files.pythonhosted.org/packages/58/43/a363c213224448f9e194d626221123ce00e3fb3d87c0c22aed52b620bdd1/colorlog-6.7.0-py2.py3-none-any.whl", hash = "sha256:0d33ca236784a1ba3ff9c532d4964126d8a2c44f1f0cb1d2b0728196f512f662", size = 11286, upload-time = "2022-08-29T14:51:26.426Z" },
+]
+
+[[package]]
+name = "ffmpeg-normalize"
+version = "1.37.6"
+source = { registry = "https://pypi.org/simple" }
+dependencies = [
+    { name = "colorama", marker = "sys_platform == 'win32'" },
+    { name = "colorlog" },
+    { name = "ffmpeg-progress-yield" },
+    { name = "mutagen" },
+    { name = "tqdm" },
+]
+sdist = { url = "https://files.pythonhosted.org/packages/c8/1a/0d27abae69de51b27dae48892f2d325dda3f80cc7def8ac216058b27c20c/ffmpeg_normalize-1.37.6.tar.gz", hash = "sha256:cec7d6a9d1b2108f0e4f8fcbd304f192676a708e8d11faf948073bf8eec562ee", size = 34800, upload-time = "2026-04-13T11:59:59.226Z" }
+wheels = [
+    { url = "https://files.pythonhosted.org/packages/8e/a6/63f2d450cc11275a5e9e67e4734e9cd8f85ab3bf5b4af50f4c304424735a/ffmpeg_normalize-1.37.6-py3-none-any.whl", hash = "sha256:3e346c6f5bc2fc5411d22b52d5e33b994cd690ece657cd68d58911dfcbd5bd6b", size = 41075, upload-time = "2026-04-13T11:59:57.881Z" },
+]
+
+[[package]]
+name = "ffmpeg-progress-yield"
+version = "1.1.3"
+source = { registry = "https://pypi.org/simple" }
+dependencies = [
+    { name = "tqdm" },
+]
+sdist = { url = "https://files.pythonhosted.org/packages/36/eb/f4836c4c30cef5f26cb0d77fa16938303665f87c1e0f64c2812d00706ecf/ffmpeg_progress_yield-1.1.3.tar.gz", hash = "sha256:79bf782a6d6bf9be64bea1d3b7a0f777f11057705935128a225970db74dedf85", size = 10012, upload-time = "2026-03-30T15:38:45.453Z" }
+wheels = [
+    { url = "https://files.pythonhosted.org/packages/fd/c7/c42ce9353ae6b70bd920856a97736ac81e068901fb537126365e6735f1c8/ffmpeg_progress_yield-1.1.3-py3-none-any.whl", hash = "sha256:4eb4948a72414608c7b05c8061aff85ff47e31e33f629abc22d08d0b03196c62", size = 12709, upload-time = "2026-03-30T15:38:44.575Z" },
+]
+
+[[package]]
 name = "ffmpeg-python"
 version = "0.2.0"
 source = { registry = "https://pypi.org/simple" }
@@ -107,6 +147,7 @@ name = "fiio-snowsky-flac-media"
 version = "0.1.0"
 source = { virtual = "." }
 dependencies = [
+    { name = "ffmpeg-normalize" },
     { name = "ffmpeg-python" },
     { name = "mutagen" },
     { name = "pillow" },
@@ -116,6 +157,7 @@ dependencies = [
 
 [package.metadata]
 requires-dist = [
+    { name = "ffmpeg-normalize", specifier = ">=1.37.6" },
     { name = "ffmpeg-python", specifier = ">=0.2.0" },
     { name = "mutagen", specifier = ">=1.47.0" },
     { name = "pillow", specifier = ">=12.2.0" },