Update app.py

app.py

@@ -1,7 +1,7 @@
 import gradio as gr
 import sqlite3
 import pandas as pd
-from huggingface_hub import hf_hub_download, HfApi, HfFolder
+from huggingface_hub import hf_hub_download, HfApi
 import os
 import time
 import shutil
@@ -18,17 +18,15 @@ LOCAL_DB_PATH = "/tmp/conceptnet-indexed.db"
 
 print(f"🌍 Filtering to: {', '.join([l.upper() for l in TARGET_LANGUAGES])}")
 
-# Get HF token
-HF_TOKEN = os.environ.get("HF_TOKEN") or os.environ.get("HUGGINGFACE_TOKEN")
-if not HF_TOKEN:
-    try:
-        HF_TOKEN = HfFolder.get_token()
-    except:
-        pass
+# Get HF token (multiple methods for compatibility)
+HF_TOKEN = os.environ.get("HF_TOKEN") or os.environ.get("HUGGINGFACE_TOKEN") or os.environ.get("HF_API_TOKEN")
 
 if not HF_TOKEN:
     print("⚠️  WARNING: No HF_TOKEN found!")
     print("   Add HF_TOKEN in Space settings to enable checkpointing")
+    print("   Go to: Settings > Variables and secrets > HF_TOKEN")
+else:
+    print(f"✅ HF_TOKEN found (length: {len(HF_TOKEN)})")
 
 # Original database
 ORIGINAL_REPO_ID = "ysenarath/conceptnet-sqlite"
@@ -163,6 +161,8 @@ def update_remote_progress(completed_indices, analyzed_tables=None, database_upl
         
     except Exception as e:
         log_progress(f"Failed to update progress: {e}", "ERROR")
+        import traceback
+        traceback.print_exc()
         return False
 
 def upload_database_checkpoint(message=""):
@@ -180,7 +180,8 @@ def upload_database_checkpoint(message=""):
         
         db_size = os.path.getsize(LOCAL_DB_PATH) / (2**30)
         log_progress(f"Uploading database checkpoint ({db_size:.2f} GB)...", "CHECKPOINT")
-        log_progress(f"  {message}", "INFO")
+        if message:
+            log_progress(f"  {message}", "INFO")
         log_progress(f"  This may take 5-10 minutes...", "INFO")
         
         start = time.time()
@@ -195,7 +196,8 @@ def upload_database_checkpoint(message=""):
         )
         
         elapsed = time.time() - start
-        log_progress(f"Database uploaded in {elapsed:.1f}s ({db_size*8/elapsed:.1f} Mbps)", "SUCCESS")
+        speed_mbps = (db_size * 8) / elapsed if elapsed > 0 else 0
+        log_progress(f"Database uploaded in {elapsed:.1f}s ({speed_mbps:.1f} Mbps)", "SUCCESS")
         
         return True
         
@@ -257,8 +259,8 @@ def create_indexed_database():
             )
             shutil.copy2(remote_db, LOCAL_DB_PATH)
             log_progress("Downloaded partial database", "SUCCESS")
-        except:
-            log_progress("No partial database, starting from original", "INFO")
+        except Exception as e:
+            log_progress(f"No partial database: {e}", "INFO")
     
     if not os.path.exists(LOCAL_DB_PATH):
         # Download and copy original
@@ -292,6 +294,7 @@ def create_indexed_database():
     cursor.execute("PRAGMA journal_mode = WAL")
     cursor.execute("PRAGMA synchronous = NORMAL")
     cursor.execute("PRAGMA cache_size = -512000")
+    cursor.execute("PRAGMA temp_store = MEMORY")
     
     # PHASE 1: Create Indices
     log_progress("="*60, "INFO")
@@ -331,7 +334,7 @@ def create_indexed_database():
             )
             
             # Upload checkpoint
-            upload_database_checkpoint(f"Checkpoint: {idx_name} created")
+            upload_database_checkpoint(f"Checkpoint: {idx_name} created ({i}/{len(indices_to_create)})")
             
         except Exception as e:
             log_progress(f"Failed to create {idx_name}: {e}", "ERROR")
@@ -356,10 +359,13 @@ def create_indexed_database():
         
         log_progress(f"[{i}/{len(tables)}] Analyzing table: {table}", "INFO")
         
-        # Get table size for progress estimation
-        cursor.execute(f"SELECT COUNT(*) FROM {table}")
-        row_count = cursor.fetchone()[0]
-        log_progress(f"  Table has {row_count:,} rows", "INFO")
+        # Get table size
+        try:
+            cursor.execute(f"SELECT COUNT(*) FROM {table}")
+            row_count = cursor.fetchone()[0]
+            log_progress(f"  Table has {row_count:,} rows", "INFO")
+        except:
+            log_progress(f"  Could not count rows", "WARN")
         
         start = time.time()