Rename test.py to extract_rar.py

extract_rar.py

@@ -0,0 +1,128 @@
+#!/usr/bin/env python3
+"""
+Helper script to extract kiu_segment_metadata.json from RAR file
+
+USAGE:
+  1. Run locally: python extract_rar.py
+  2. Upload the extracted JSON to your HF Space at: /data/kiu_segment_metadata/kiu_segment_metadata.json
+  
+OR add to your Space's Dockerfile:
+  RUN apt-get update && apt-get install -y unrar
+"""
+
+import os
+import shutil
+
+try:
+    import rarfile
+except ImportError:
+    print("Installing rarfile...")
+    os.system("pip install rarfile")
+    import rarfile
+
+RAR_FILE = "kiu_segment_metadata.rar"
+OUTPUT_DIR = "extracted"
+OUTPUT_FILE = "kiu_segment_metadata.json"
+
+def extract_with_unrar():
+    """Try using system unrar command"""
+    print("Attempting extraction with system unrar...")
+    os.makedirs(OUTPUT_DIR, exist_ok=True)
+    
+    result = os.system(f"unrar x -y {RAR_FILE} {OUTPUT_DIR}/")
+    
+    if result == 0:
+        # Find the extracted JSON
+        for root, dirs, files in os.walk(OUTPUT_DIR):
+            for file in files:
+                if file.endswith('.json') and 'segment_metadata' in file:
+                    src = os.path.join(root, file)
+                    dst = OUTPUT_FILE
+                    shutil.copy(src, dst)
+                    print(f"✅ Extracted to: {dst}")
+                    print(f"📤 Upload this file to your HF Space")
+                    return True
+    
+    return False
+
+def extract_with_rarfile():
+    """Try using rarfile library"""
+    print("Attempting extraction with rarfile library...")
+    
+    try:
+        with rarfile.RarFile(RAR_FILE) as rf:
+            print(f"Files in RAR: {[f.filename for f in rf.infolist()]}")
+            
+            # Find JSON file
+            json_file = None
+            for file_info in rf.infolist():
+                if file_info.filename.endswith('.json') and 'segment_metadata' in file_info.filename:
+                    json_file = file_info.filename
+                    break
+            
+            if not json_file:
+                print("❌ No JSON file found in RAR")
+                return False
+            
+            print(f"Found: {json_file}")
+            
+            # Extract
+            os.makedirs(OUTPUT_DIR, exist_ok=True)
+            rf.extract(json_file, OUTPUT_DIR)
+            
+            # Copy to output location
+            src = os.path.join(OUTPUT_DIR, json_file)
+            shutil.copy(src, OUTPUT_FILE)
+            
+            print(f"✅ Extracted to: {OUTPUT_FILE}")
+            print(f"📤 Upload this file to your HF Space at: /data/kiu_segment_metadata/kiu_segment_metadata.json")
+            return True
+            
+    except rarfile.RarCannotExec:
+        print("❌ unrar tool not found")
+        return False
+    except Exception as e:
+        print(f"❌ Error: {e}")
+        return False
+
+def main():
+    if not os.path.exists(RAR_FILE):
+        print(f"❌ {RAR_FILE} not found in current directory")
+        print(f"Current directory: {os.getcwd()}")
+        print(f"Files: {os.listdir('.')}")
+        return
+    
+    print(f"Found {RAR_FILE}")
+    print(f"Size: {os.path.getsize(RAR_FILE):,} bytes")
+    print()
+    
+    # Try rarfile first
+    if extract_with_rarfile():
+        return
+    
+    print()
+    print("Trying system unrar command...")
+    if extract_with_unrar():
+        return
+    
+    print()
+    print("="*60)
+    print("❌ EXTRACTION FAILED")
+    print("="*60)
+    print()
+    print("SOLUTIONS:")
+    print()
+    print("1. Install unrar on your system:")
+    print("   Ubuntu/Debian: sudo apt-get install unrar")
+    print("   Mac: brew install unrar")
+    print("   Windows: Download from https://www.rarlab.com/")
+    print()
+    print("2. Extract manually using WinRAR/7-Zip")
+    print(f"   Then upload kiu_segment_metadata.json to your HF Space")
+    print()
+    print("3. For HF Spaces, add to Dockerfile:")
+    print("   RUN apt-get update && apt-get install -y unrar")
+    print()
+
+if __name__ == "__main__":
+    main()

test.py

@@ -1,35 +0,0 @@
-# test_download.py
-import os
-from huggingface_hub import hf_hub_download
-
-HF_TOKEN = os.getenv("HF_TOKEN")
-REPO_ID = "alyex/karnak_data"
-
-print("Testing download...")
-
-try:
-    # Try with xet=True
-    print("Trying with xet=True...")
-    path = hf_hub_download(
-        repo_id=REPO_ID,
-        filename="kiu_segment_metadata.rar",
-        token=HF_TOKEN,
-        repo_type="dataset",
-        xet=True
-    )
-    print(f"✅ Success! File at: {path}")
-except Exception as e:
-    print(f"❌ xet=True failed: {e}")
-    
-    try:
-        # Try without xet
-        print("\nTrying without xet...")
-        path = hf_hub_download(
-            repo_id=REPO_ID,
-            filename="kiu_segment_metadata.rar",
-            token=HF_TOKEN,
-            repo_type="dataset"
-        )
-        print(f"✅ Success! File at: {path}")
-    except Exception as e2:
-        print(f"❌ Standard download failed: {e2}")