fix(indexing): WebSocket race condition, NaN stats, file streaming

Fixes:
- Add delay before WS close to ensure client receives completion event
- Align stat field names between backend and frontend
- Add current_file to progress events for streaming display
- Preserve completed state when WebSocket disconnects

Root cause: Backend closed WebSocket immediately after sending completion
event, causing race condition where frontend's onclose fired before
onmessage could process the completion.

Closes #154

Author: DevanshuNEU

backend/routes/ws_playground.py

@@ -152,6 +152,9 @@ async def websocket_playground_index(websocket: WebSocket, job_id: str):
                         job_id=job_id[:12],
                         event_type=event_type
                     )
+                    # Small delay to ensure client processes message before close
+                    # This prevents race condition where onclose fires before onmessage
+                    await asyncio.sleep(0.2)
                     break
 
             except json.JSONDecodeError:

backend/services/anonymous_indexer.py

@@ -43,9 +43,9 @@ def to_dict(self) -> dict:
 @dataclass
 class JobStats:
     """Final stats for completed job."""
-    files_indexed: int = 0
-    functions_found: int = 0
-    time_taken_seconds: float = 0
+    files_processed: int = 0
+    functions_indexed: int = 0
+    indexing_time_seconds: float = 0
 
     def to_dict(self) -> dict:
         return asdict(self)
@@ -345,12 +345,18 @@ async def run_indexing_job(
         job_manager.update_status(job_id, JobStatus.PROCESSING)
 
         # Progress callback for real-time updates
-        async def progress_callback(files_processed: int, functions_found: int, total: int):
+        async def progress_callback(
+            files_processed: int,
+            functions_found: int,
+            total: int,
+            current_file: Optional[str] = None
+        ):
             job_manager.update_progress(
                 job_id,
                 files_processed=files_processed,
                 functions_found=functions_found,
-                files_total=total
+                files_total=total,
+                current_file=current_file
             )
 
         # Run indexing with timeout
@@ -370,9 +376,9 @@ async def progress_callback(files_processed: int, functions_found: int, total: i
         # --- Step 3: Mark complete ---
         elapsed = time.time() - start_time
         stats = JobStats(
-            files_indexed=file_count,
-            functions_found=total_functions,
-            time_taken_seconds=round(elapsed, 2)
+            files_processed=file_count,
+            functions_indexed=total_functions,
+            indexing_time_seconds=round(elapsed, 2)
         )
 
         job_manager.update_status(

backend/services/indexer_optimized.py

@@ -672,8 +672,11 @@ async def index_repository_with_progress(
 
             files_processed = min(i + self.FILE_BATCH_SIZE, total_files)
 
-            # Send progress update
-            await progress_callback(files_processed, len(all_functions_data), total_files)
+            # Get the last file in this batch for display
+            current_file = batch[-1].name if batch else None
+            
+            # Send progress update with current file
+            await progress_callback(files_processed, len(all_functions_data), total_files, current_file)
 
             logger.debug("Processing files", 
                         processed=files_processed, 

backend/tests/test_anonymous_indexing.py

@@ -199,9 +199,9 @@ def test_update_status_completed_with_stats(self, job_manager, mock_redis):
         })
 
         stats = JobStats(
-            files_indexed=100,
-            functions_found=500,
-            time_taken_seconds=45.5
+            files_processed=100,
+            functions_indexed=500,
+            indexing_time_seconds=45.5
         )
 
         result = job_manager.update_status(
@@ -259,13 +259,13 @@ def test_job_progress_none_excluded(self):
     def test_job_stats_to_dict(self):
         """JobStats converts to dict correctly."""
         stats = JobStats(
-            files_indexed=100,
-            functions_found=500,
-            time_taken_seconds=45.5
+            files_processed=100,
+            functions_indexed=500,
+            indexing_time_seconds=45.5
         )
         d = stats.to_dict()
-        assert d["files_indexed"] == 100
-        assert d["time_taken_seconds"] == 45.5
+        assert d["files_processed"] == 100
+        assert d["indexing_time_seconds"] == 45.5
 
 
 # =============================================================================
@@ -608,9 +608,9 @@ def test_completed_job_returns_repo_id(self, mock_job_class, client):
             "created_at": "2024-01-01T00:00:00Z",
             "updated_at": "2024-01-01T00:01:00Z",
             "stats": {
-                "files_indexed": 100,
-                "functions_found": 500,
-                "time_taken_seconds": 45.2
+                "files_processed": 100,
+                "functions_indexed": 500,
+                "indexing_time_seconds": 45.2
             }
         }
         mock_job_class.return_value = mock_job_manager
@@ -621,7 +621,7 @@ def test_completed_job_returns_repo_id(self, mock_job_class, client):
         data = response.json()
         assert data["status"] == "completed"
         assert data["repo_id"] == "anon_idx_test123456"
-        assert data["stats"]["files_indexed"] == 100
+        assert data["stats"]["files_processed"] == 100
 
     @patch('routes.playground.AnonymousIndexingJob')
     def test_failed_job_returns_error(self, mock_job_class, client):

backend/tests/test_ws_playground.py

@@ -66,7 +66,7 @@ def test_websocket_handles_already_completed_job(self):
             "job_id": "idx_test123",
             "status": "completed",
             "repo_id": "anon_test123",
-            "stats": {"files_indexed": 100, "functions_found": 500}
+            "stats": {"files_processed": 100, "functions_indexed": 500}
         })
 
         with patch('routes.ws_playground.redis_client', mock_redis):
@@ -172,9 +172,9 @@ def test_completed_event_includes_stats(self):
         job_manager = AnonymousIndexingJob(mock_redis)
 
         stats = JobStats(
-            files_indexed=100,
-            functions_found=500,
-            time_taken_seconds=45.2
+            files_processed=100,
+            functions_indexed=500,
+            indexing_time_seconds=45.2
         )
 
         job_manager.update_status(
@@ -189,7 +189,7 @@ def test_completed_event_includes_stats(self):
 
         assert event_data["type"] == "completed"
         assert event_data["repo_id"] == "anon_test123"
-        assert event_data["stats"]["functions_found"] == 500
+        assert event_data["stats"]["functions_indexed"] == 500
 
     def test_processing_status_skips_duplicate_publish(self):
         """PROCESSING status should not publish (handled by update_progress)."""

frontend/src/hooks/useIndexingWebSocket.ts

@@ -248,8 +248,18 @@ export function useIndexingWebSocket(
     if (jobId) {
       connect(jobId);
     } else {
+      // Only cleanup connection, DON'T reset state!
+      // This preserves completedStats when jobId becomes null after completion
       cleanup();
-      setState(INITIAL_STATE);
+      // Only reset if we were never completed (e.g., user navigated away during indexing)
+      setState(prev => {
+        if (prev.phase === 'completed') {
+          // Keep completed state - just disconnect
+          return { ...prev, connectionState: 'disconnected' };
+        }
+        // Reset if we were mid-indexing (user cancelled, navigated away, etc.)
+        return INITIAL_STATE;
+      });
     }
     return cleanup;
   }, [jobId, connect, cleanup]);