Add script to extract source scripts

tools/extract.py

@@ -0,0 +1,68 @@
+#!/usr/bin/env python3
+
+# $1: source ep number
+
+import json
+import os
+from pathlib import Path
+import subprocess
+import sys
+
+
+this_dir = Path(__file__).parent
+src_dir = this_dir.parent / "[Chyuu-PAS] Shoujo☆Kageki Revue Starlight [WEB 1080p EAC3]"
+target_dir_template = this_dir.parent / "Folge {ep_num:02d}" / "english"
+sub_template = target_dir_template / "[Chyuu-PAS] SKRS - {ep_num:02d}.ass"
+font_dir_template = target_dir_template / "cpas_fonts"
+chapters_template = target_dir_template.parent / "chapters.xml"
+
+
+# def mkv(prog, *args, **kwargs):
+#     cmd = [f'mkv{prog}', *args]
+#     print("running", cmd)
+#     return subprocess.run(cmd, check=True, capture_output=True, text=True, **kwargs)
+
+
+def mkvidentify(path):
+    cmd = ['mkvmerge', "-J", str(path)]
+    result = subprocess.run(cmd, check=True, capture_output=True, text=True)
+    return json.loads(result.stdout)
+
+
+def mkvextract(path, *args):
+    cmd = ['mkvextract', str(path), *args]
+    print("\n[CMD]", cmd)
+    return subprocess.run(cmd, check=True, stdout=sys.stdout, stderr=sys.stderr)
+
+
+def main():
+    for ep_num in map(int, sys.argv[1:]):
+        src_path = next(src_dir.glob(f"* - {ep_num:02d}*"))
+        print("[SRC]", src_path)
+        identify = mkvidentify(src_path)
+
+        # extract subs, chapters, attachments
+        for track in identify['tracks']:
+            if track['codec'] == 'SubStationAlpha':
+                subs_id = track['id']
+                os.makedirs(str(target_dir_template).format(ep_num=ep_num), exist_ok=True)
+                mkvextract(src_path, 'tracks',
+                           f"{subs_id}:{str(sub_template).format(ep_num=ep_num)}")
+                break
+        else:
+            print("No subtitle track found")
+
+        font_dir = str(font_dir_template).format(ep_num=ep_num)
+        os.makedirs(font_dir, exist_ok=True)
+        attach_args = [
+            f"{attach['id']}:{str(font_dir_template / attach['file_name']).format(ep_num=ep_num)}"
+            for attach in identify['attachments']
+        ]
+        mkvextract(src_path, 'attachments', *attach_args)
+
+        mkvextract(src_path, 'chapters', str(chapters_template).format(ep_num=ep_num))
+        print()
+
+
+if __name__ == '__main__':
+    sys.exit(main())