Delete create_metadata.py

create_metadata.py

@@ -1,99 +0,0 @@
-#!/usr/bin/env python3
-"""
-Script to consolidate all transcript files into a single metadata.csv
-- Only filename and transcript
-- Converts Latin punctuation to Japanese punctuation
-- Ensures each transcript ends with 。 (Kuten)
-"""
-
-import csv
-from pathlib import Path
-
-BASE_DIR = Path("/home/phucdb/Documents/uv/cpjd_ver1")
-OUTPUT_FILE = BASE_DIR / "metadata.csv"
-
-def convert_punctuation(text: str) -> str:
-    """Convert Latin punctuation to Japanese punctuation."""
-    # Replace comma with Tōten (、)
-    text = text.replace(",", "、")
-    text = text.replace("，", "、")  # Full-width comma too
-    
-    # Replace period with Kuten (。)
-    text = text.replace(".", "。")
-    text = text.replace("．", "。")  # Full-width period too
-    
-    return text
-
-def ensure_ends_with_kuten(text: str) -> str:
-    """Ensure transcript ends with Japanese period (。)."""
-    text = text.rstrip()
-    if not text.endswith("。"):
-        text += "。"
-    return text
-
-def main():
-    rows = []
-    
-    # Find all dialect folders
-    for folder in sorted(BASE_DIR.iterdir()):
-        if not folder.is_dir():
-            continue
-        
-        # Skip if doesn't start with F_ or M_
-        if not (folder.name.startswith("F_") or folder.name.startswith("M_")):
-            continue
-        
-        dialect = folder.name  # e.g., F_akita, M_osaka
-        transcript_file = folder / "transcript_utf8.txt"
-        
-        if not transcript_file.exists():
-            print(f"Warning: No transcript found in {folder.name}")
-            continue
-        
-        print(f"Processing: {dialect}")
-        
-        # Read transcript file
-        with open(transcript_file, "r", encoding="utf-8") as f:
-            for line in f:
-                line = line.strip()
-                if not line:
-                    continue
-                
-                # Parse format "001: content"
-                if ": " in line:
-                    parts = line.split(": ", 1)
-                    if len(parts) == 2:
-                        audio_id = parts[0].strip()
-                        text = parts[1].strip()
-                        
-                        # Convert punctuation
-                        text = convert_punctuation(text)
-                        
-                        # Ensure ends with 。
-                        text = ensure_ends_with_kuten(text)
-                        
-                        # Create audio filename (just the filename, not full path)
-                        filename = f"{dialect}/wav16k_norm/{audio_id}.wav"
-                        
-                        rows.append({
-                            "filename": filename,
-                            "transcript": text
-                        })
-    
-    # Write to CSV
-    print(f"\nWriting {len(rows)} entries to {OUTPUT_FILE}")
-    
-    with open(OUTPUT_FILE, "w", encoding="utf-8", newline="") as f:
-        writer = csv.DictWriter(f, fieldnames=["filename", "transcript"])
-        writer.writeheader()
-        writer.writerows(rows)
-    
-    print("Done!")
-    
-    # Show sample
-    print("\n--- Sample entries ---")
-    for row in rows[:5]:
-        print(f"{row['filename']}: {row['transcript'][:50]}...")
-
-if __name__ == "__main__":
-    main()