fix: remove whitespace from blank line

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <noreply@anthropic.com>

.github/workflows/build-and-publish.yml

@@ -5,7 +5,6 @@ on:
     branches: [ main ]
   pull_request:
     branches: [ main ]
-  workflow_dispatch:
 
 jobs:
   build:

.github/workflows/build-reusable.yml

@@ -28,7 +28,7 @@ jobs:
 
       - name: Install ruff
         run: |
-          uv tool install ruff
+          uv tool install ruff==0.12.7
 
       - name: Run ruff check
         run: |
@@ -54,36 +54,16 @@ jobs:
             python: '3.12'
           - os: ubuntu-22.04
             python: '3.13'
-          - os: macos-14
+          - os: macos-latest
             python: '3.9'
-          - os: macos-14
+          - os: macos-latest
             python: '3.10'
-          - os: macos-14
+          - os: macos-latest
             python: '3.11'
-          - os: macos-14
+          - os: macos-latest
             python: '3.12'
-          - os: macos-14
+          - os: macos-latest
             python: '3.13'
-          - os: macos-15
-            python: '3.9'
-          - os: macos-15
-            python: '3.10'
-          - os: macos-15
-            python: '3.11'
-          - os: macos-15
-            python: '3.12'
-          - os: macos-15
-            python: '3.13'
-          - os: macos-13
-            python: '3.9'
-          - os: macos-13
-            python: '3.10'
-          - os: macos-13
-            python: '3.11'
-          - os: macos-13
-            python: '3.12'
-          # Note: macos-13 + Python 3.13 excluded due to PyTorch compatibility
-          # (PyTorch 2.5+ supports Python 3.13 but not Intel Mac x86_64)
     runs-on: ${{ matrix.os }}
 
     steps:
@@ -129,70 +109,41 @@ jobs:
             uv pip install --system delocate
           fi
 
-      - name: Set macOS environment variables
-        if: runner.os == 'macOS'
-        run: |
-          # Use brew --prefix to automatically detect Homebrew installation path
-          HOMEBREW_PREFIX=$(brew --prefix)
-          echo "HOMEBREW_PREFIX=${HOMEBREW_PREFIX}" >> $GITHUB_ENV
-          echo "OpenMP_ROOT=${HOMEBREW_PREFIX}/opt/libomp" >> $GITHUB_ENV
-
-          # Set CMAKE_PREFIX_PATH to let CMake find all packages automatically
-          echo "CMAKE_PREFIX_PATH=${HOMEBREW_PREFIX}" >> $GITHUB_ENV
-
-          # Set compiler flags for OpenMP (required for both backends)
-          echo "LDFLAGS=-L${HOMEBREW_PREFIX}/opt/libomp/lib" >> $GITHUB_ENV
-          echo "CPPFLAGS=-I${HOMEBREW_PREFIX}/opt/libomp/include" >> $GITHUB_ENV
-
       - name: Build packages
         run: |
-          # Build core (platform independent)
+          # Build core (platform independent) on all platforms for consistency
           cd packages/leann-core
           uv build
           cd ../..
 
           # Build HNSW backend
           cd packages/leann-backend-hnsw
-          if [[ "${{ matrix.os }}" == macos-* ]]; then
-            # Use system clang for better compatibility
+          if [ "${{ matrix.os }}" == "macos-latest" ]; then
+            # Use system clang instead of homebrew LLVM for better compatibility
             export CC=clang
             export CXX=clang++
-            # Homebrew libraries on each macOS version require matching minimum version
-            if [[ "${{ matrix.os }}" == "macos-13" ]]; then
-              export MACOSX_DEPLOYMENT_TARGET=13.0
-            elif [[ "${{ matrix.os }}" == "macos-14" ]]; then
-              export MACOSX_DEPLOYMENT_TARGET=14.0
-            elif [[ "${{ matrix.os }}" == "macos-15" ]]; then
-              export MACOSX_DEPLOYMENT_TARGET=15.0
-            fi
-            uv build --wheel --python ${{ matrix.python }} --find-links ${GITHUB_WORKSPACE}/packages/leann-core/dist
+            export MACOSX_DEPLOYMENT_TARGET=11.0
+            uv build --wheel --python python
           else
-            uv build --wheel --python ${{ matrix.python }} --find-links ${GITHUB_WORKSPACE}/packages/leann-core/dist
+            uv build --wheel --python python
           fi
           cd ../..
 
           # Build DiskANN backend
           cd packages/leann-backend-diskann
-          if [[ "${{ matrix.os }}" == macos-* ]]; then
-            # Use system clang for better compatibility
+          if [ "${{ matrix.os }}" == "macos-latest" ]; then
+            # Use system clang instead of homebrew LLVM for better compatibility
             export CC=clang
             export CXX=clang++
-            # DiskANN requires macOS 13.3+ for sgesdd_ LAPACK function
-            # But Homebrew libraries on each macOS version require matching minimum version
-            if [[ "${{ matrix.os }}" == "macos-13" ]]; then
-              export MACOSX_DEPLOYMENT_TARGET=13.3
-            elif [[ "${{ matrix.os }}" == "macos-14" ]]; then
-              export MACOSX_DEPLOYMENT_TARGET=14.0
-            elif [[ "${{ matrix.os }}" == "macos-15" ]]; then
-              export MACOSX_DEPLOYMENT_TARGET=15.0
-            fi
-            uv build --wheel --python ${{ matrix.python }} --find-links ${GITHUB_WORKSPACE}/packages/leann-core/dist
+            # sgesdd_ is only available on macOS 13.3+
+            export MACOSX_DEPLOYMENT_TARGET=13.3
+            uv build --wheel --python python
           else
-            uv build --wheel --python ${{ matrix.python }} --find-links ${GITHUB_WORKSPACE}/packages/leann-core/dist
+            uv build --wheel --python python
           fi
           cd ../..
 
-          # Build meta package (platform independent)
+          # Build meta package (platform independent) on all platforms
           cd packages/leann
           uv build
           cd ../..
@@ -209,10 +160,15 @@ jobs:
           fi
           cd ../..
 
-          # Repair DiskANN wheel
+          # Repair DiskANN wheel - use show first to debug
           cd packages/leann-backend-diskann
           if [ -d dist ]; then
+            echo "Checking DiskANN wheel contents before repair:"
+            unzip -l dist/*.whl | grep -E "\.so|\.pyd|_diskannpy" || echo "No .so files found"
+            auditwheel show dist/*.whl || echo "auditwheel show failed"
             auditwheel repair dist/*.whl -w dist_repaired
+            echo "Checking DiskANN wheel contents after repair:"
+            unzip -l dist_repaired/*.whl | grep -E "\.so|\.pyd|_diskannpy" || echo "No .so files found after repair"
             rm -rf dist
             mv dist_repaired dist
           fi
@@ -221,24 +177,10 @@ jobs:
       - name: Repair wheels (macOS)
         if: runner.os == 'macOS'
         run: |
-          # Determine deployment target based on runner OS
-          # Must match the Homebrew libraries for each macOS version
-          if [[ "${{ matrix.os }}" == "macos-13" ]]; then
-            HNSW_TARGET="13.0"
-            DISKANN_TARGET="13.3"
-          elif [[ "${{ matrix.os }}" == "macos-14" ]]; then
-            HNSW_TARGET="14.0"
-            DISKANN_TARGET="14.0"
-          elif [[ "${{ matrix.os }}" == "macos-15" ]]; then
-            HNSW_TARGET="15.0"
-            DISKANN_TARGET="15.0"
-          fi
-
           # Repair HNSW wheel
           cd packages/leann-backend-hnsw
           if [ -d dist ]; then
-            export MACOSX_DEPLOYMENT_TARGET=$HNSW_TARGET
-            delocate-wheel -w dist_repaired -v --require-target-macos-version $HNSW_TARGET dist/*.whl
+            delocate-wheel -w dist_repaired -v dist/*.whl
             rm -rf dist
             mv dist_repaired dist
           fi
@@ -247,8 +189,7 @@ jobs:
           # Repair DiskANN wheel
           cd packages/leann-backend-diskann
           if [ -d dist ]; then
-            export MACOSX_DEPLOYMENT_TARGET=$DISKANN_TARGET
-            delocate-wheel -w dist_repaired -v --require-target-macos-version $DISKANN_TARGET dist/*.whl
+            delocate-wheel -w dist_repaired -v dist/*.whl
             rm -rf dist
             mv dist_repaired dist
           fi
@@ -259,34 +200,44 @@ jobs:
           echo "📦 Built packages:"
           find packages/*/dist -name "*.whl" -o -name "*.tar.gz" | sort
 
-
       - name: Install built packages for testing
         run: |
           # Create a virtual environment with the correct Python version
-          uv venv --python ${{ matrix.python }}
+          uv venv --python python${{ matrix.python }}
           source .venv/bin/activate || source .venv/Scripts/activate
 
-          # Install packages using --find-links to prioritize local builds
-          uv pip install --find-links packages/leann-core/dist --find-links packages/leann-backend-hnsw/dist --find-links packages/leann-backend-diskann/dist packages/leann-core/dist/*.whl || uv pip install --find-links packages/leann-core/dist packages/leann-core/dist/*.tar.gz
-          uv pip install --find-links packages/leann-core/dist packages/leann-backend-hnsw/dist/*.whl
-          uv pip install --find-links packages/leann-core/dist packages/leann-backend-diskann/dist/*.whl
-          uv pip install packages/leann/dist/*.whl || uv pip install packages/leann/dist/*.tar.gz
+          # Install the built wheels directly to ensure we use locally built packages
+          # Use only locally built wheels on all platforms for full consistency
+          FIND_LINKS="--find-links packages/leann-core/dist --find-links packages/leann/dist"
+          FIND_LINKS="$FIND_LINKS --find-links packages/leann-backend-hnsw/dist --find-links packages/leann-backend-diskann/dist"
+
+          uv pip install leann-core leann leann-backend-hnsw leann-backend-diskann \
+            $FIND_LINKS --force-reinstall
 
           # Install test dependencies using extras
           uv pip install -e ".[test]"
 
+          # Debug: Check if _diskannpy module is installed correctly
+          echo "Checking installed DiskANN module structure:"
+          python -c "import leann_backend_diskann; print('leann_backend_diskann location:', leann_backend_diskann.__file__)" || echo "Failed to import leann_backend_diskann"
+          python -c "from leann_backend_diskann import _diskannpy; print('_diskannpy imported successfully')" || echo "Failed to import _diskannpy"
+          ls -la $(python -c "import leann_backend_diskann; import os; print(os.path.dirname(leann_backend_diskann.__file__))" 2>/dev/null) 2>/dev/null || echo "Failed to list module directory"
+
       - name: Run tests with pytest
         env:
-          CI: true
+          CI: true  # Mark as CI environment to skip memory-intensive tests
           OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
           HF_HUB_DISABLE_SYMLINKS: 1
           TOKENIZERS_PARALLELISM: false
-          PYTORCH_ENABLE_MPS_FALLBACK: 0
-          OMP_NUM_THREADS: 1
-          MKL_NUM_THREADS: 1
+          PYTORCH_ENABLE_MPS_FALLBACK: 0  # Disable MPS on macOS CI to avoid memory issues
+          OMP_NUM_THREADS: 1  # Disable OpenMP parallelism to avoid libomp crashes
+          MKL_NUM_THREADS: 1  # Single thread for MKL operations
         run: |
+          # Activate virtual environment
           source .venv/bin/activate || source .venv/Scripts/activate
-          pytest tests/ -v --tb=short
+
+          # Run all tests
+          pytest tests/
 
       - name: Run sanity checks (optional)
         run: |

README.md

@@ -3,11 +3,10 @@
 </p>
 
 <p align="center">
-  <img src="https://img.shields.io/badge/Python-3.9%20%7C%203.10%20%7C%203.11%20%7C%203.12%20%7C%203.13-blue.svg" alt="Python Versions">
-  <img src="https://github.com/yichuan-w/LEANN/actions/workflows/build-and-publish.yml/badge.svg" alt="CI Status">
-  <img src="https://img.shields.io/badge/Platform-Ubuntu%20%7C%20macOS%20(ARM64%2FIntel)-lightgrey" alt="Platform">
+  <img src="https://img.shields.io/badge/Python-3.9%2B-blue.svg" alt="Python 3.9+">
   <img src="https://img.shields.io/badge/License-MIT-green.svg" alt="MIT License">
-  <img src="https://img.shields.io/badge/MCP-Native%20Integration-blue" alt="MCP Integration">
+  <img src="https://img.shields.io/badge/Platform-Linux%20%7C%20macOS-lightgrey" alt="Platform">
+  <img src="https://img.shields.io/badge/MCP-Native%20Integration-blue?style=flat-square" alt="MCP Integration">
 </p>
 
 <h2 align="center" tabindex="-1" class="heading-element" dir="auto">
@@ -98,7 +97,6 @@ uv sync
 
 </details>
 
-
 ## Quick Start
 
 Our declarative API makes RAG as easy as writing a config file.
@@ -190,7 +188,7 @@ All RAG examples share these common parameters. **Interactive mode** is availabl
 --force-rebuild         # Force rebuild index even if it exists
 
 # Embedding Parameters
---embedding-model MODEL  # e.g., facebook/contriever, text-embedding-3-small, nomic-embed-text,mlx-community/Qwen3-Embedding-0.6B-8bit or nomic-embed-text
+--embedding-model MODEL  # e.g., facebook/contriever, text-embedding-3-small, nomic-embed-text, or mlx-community/multilingual-e5-base-mlx
 --embedding-mode MODE    # sentence-transformers, openai, mlx, or ollama
 
 # LLM Parameters (Text generation models)
@@ -468,7 +466,7 @@ leann --help
 ### Usage Examples
 
 ```bash
-# build from a specific directory, and my_docs is the index name(Here you can also build from multiple dict or multiple files)
+# build from a specific directory, and my_docs is the index name
 leann build my-docs --docs ./your_documents
 
 # Search your documents
@@ -611,9 +609,8 @@ We welcome more contributors! Feel free to open issues or submit PRs.
 
 This work is done at [**Berkeley Sky Computing Lab**](https://sky.cs.berkeley.edu/).
 
-## Star History
+---
 
-[![Star History Chart](https://api.star-history.com/svg?repos=yichuan-w/LEANN&type=Date)](https://www.star-history.com/#yichuan-w/LEANN&Date)
 <p align="center">
   <strong>⭐ Star us on GitHub if Leann is useful for your research or applications!</strong>
 </p>

apps/base_rag_example.py

@@ -178,9 +178,6 @@ class BaseRAGExample(ABC):
             config["host"] = args.llm_host
         elif args.llm == "hf":
             config["model"] = args.llm_model or "Qwen/Qwen2.5-1.5B-Instruct"
-        elif args.llm == "simulated":
-            # Simulated LLM doesn't need additional configuration
-            pass
 
         return config
 

docs/configuration-guide.md

@@ -236,15 +236,9 @@ python apps/document_rag.py --query "What are the main techniques LEANN explores
 
 3. **Use MLX on Apple Silicon** (optional optimization):
    ```bash
-   --embedding-mode mlx --embedding-model mlx-community/Qwen3-Embedding-0.6B-8bit
+   --embedding-mode mlx --embedding-model mlx-community/multilingual-e5-base-mlx
    ```
-    MLX might not be the best choice, as we tested and found that it only offers 1.3x acceleration compared to HF, so maybe using ollama is a better choice for embedding generation
 
-4. **Use Ollama**
-   ```bash
-   --embedding-mode ollama --embedding-model nomic-embed-text
-   ```
-   To discover additional embedding models in ollama, check out https://ollama.com/search?c=embedding or read more about embedding models at https://ollama.com/blog/embedding-models, please do check the model size that works best for you
 ### If Search Quality is Poor
 
 1. **Increase retrieval count**:

packages/leann-backend-diskann/CMakeLists.txt

@@ -0,0 +1,8 @@
+# packages/leann-backend-diskann/CMakeLists.txt (simplified version)
+
+cmake_minimum_required(VERSION 3.20)
+project(leann_backend_diskann_wrapper)
+
+# Tell CMake to directly enter the DiskANN submodule and execute its own CMakeLists.txt
+# DiskANN will handle everything itself, including compiling Python bindings
+add_subdirectory(src/third_party/DiskANN)

packages/leann-backend-diskann/leann_backend_diskann/diskann_backend.py

@@ -22,11 +22,6 @@ logger = logging.getLogger(__name__)
 @contextlib.contextmanager
 def suppress_cpp_output_if_needed():
     """Suppress C++ stdout/stderr based on LEANN_LOG_LEVEL"""
-    # In CI we avoid fiddling with low-level file descriptors to prevent aborts
-    if os.getenv("CI") == "true":
-        yield
-        return
-
     log_level = os.getenv("LEANN_LOG_LEVEL", "WARNING").upper()
 
     # Only suppress if log level is WARNING or higher (ERROR, CRITICAL)
@@ -464,3 +459,25 @@ class DiskannSearcher(BaseSearcher):
         string_labels = [[str(int_label) for int_label in batch_labels] for batch_labels in labels]
 
         return {"labels": string_labels, "distances": distances}
+
+    def cleanup(self):
+        """Cleanup DiskANN-specific resources including C++ index."""
+        # Call parent cleanup first
+        super().cleanup()
+
+        # Delete the C++ index to trigger destructors
+        try:
+            if hasattr(self, "_index") and self._index is not None:
+                del self._index
+                self._index = None
+                self._current_zmq_port = None
+        except Exception:
+            pass
+
+        # Force garbage collection to ensure C++ objects are destroyed
+        try:
+            import gc
+
+            gc.collect()
+        except Exception:
+            pass

packages/leann-backend-diskann/leann_backend_diskann/diskann_embedding_server.py

@@ -103,9 +103,8 @@ def create_diskann_embedding_server(
         socket.bind(f"tcp://*:{zmq_port}")
         logger.info(f"DiskANN ZMQ REP server listening on port {zmq_port}")
 
-        socket.setsockopt(zmq.RCVTIMEO, 1000)
-        socket.setsockopt(zmq.SNDTIMEO, 1000)
-        socket.setsockopt(zmq.LINGER, 0)
+        socket.setsockopt(zmq.RCVTIMEO, 300000)
+        socket.setsockopt(zmq.SNDTIMEO, 300000)
 
         while True:
             try:
@@ -222,217 +221,30 @@ def create_diskann_embedding_server(
                 traceback.print_exc()
                 raise
 
-    def zmq_server_thread_with_shutdown(shutdown_event):
-        """ZMQ server thread that respects shutdown signal.
-
-        This creates its own REP socket, binds to zmq_port, and periodically
-        checks shutdown_event using recv timeouts to exit cleanly.
-        """
-        logger.info("DiskANN ZMQ server thread started with shutdown support")
-
-        context = zmq.Context()
-        rep_socket = context.socket(zmq.REP)
-        rep_socket.bind(f"tcp://*:{zmq_port}")
-        logger.info(f"DiskANN ZMQ REP server listening on port {zmq_port}")
-
-        # Set receive timeout so we can check shutdown_event periodically
-        rep_socket.setsockopt(zmq.RCVTIMEO, 1000)  # 1 second timeout
-        rep_socket.setsockopt(zmq.SNDTIMEO, 1000)
-        rep_socket.setsockopt(zmq.LINGER, 0)
-
-        try:
-            while not shutdown_event.is_set():
-                try:
-                    e2e_start = time.time()
-                    # REP socket receives single-part messages
-                    message = rep_socket.recv()
-
-                    # Check for empty messages - REP socket requires response to every request
-                    if not message:
-                        logger.warning("Received empty message, sending empty response")
-                        rep_socket.send(b"")
-                        continue
-
-                    # Try protobuf first (same logic as original)
-                    texts = []
-                    is_text_request = False
-
-                    try:
-                        req_proto = embedding_pb2.NodeEmbeddingRequest()
-                        req_proto.ParseFromString(message)
-                        node_ids = list(req_proto.node_ids)
-
-                        # Look up texts by node IDs
-                        for nid in node_ids:
-                            try:
-                                passage_data = passages.get_passage(str(nid))
-                                txt = passage_data["text"]
-                                if not txt:
-                                    raise RuntimeError(f"FATAL: Empty text for passage ID {nid}")
-                                texts.append(txt)
-                            except KeyError:
-                                raise RuntimeError(f"FATAL: Passage with ID {nid} not found")
-
-                        logger.info(f"ZMQ received protobuf request for {len(node_ids)} node IDs")
-                    except Exception:
-                        # Fallback to msgpack for text requests
-                        try:
-                            import msgpack
-
-                            request = msgpack.unpackb(message)
-                            if isinstance(request, list) and all(
-                                isinstance(item, str) for item in request
-                            ):
-                                texts = request
-                                is_text_request = True
-                                logger.info(
-                                    f"ZMQ received msgpack text request for {len(texts)} texts"
-                                )
-                            else:
-                                raise ValueError("Not a valid msgpack text request")
-                        except Exception:
-                            logger.error("Both protobuf and msgpack parsing failed!")
-                            # Send error response
-                            resp_proto = embedding_pb2.NodeEmbeddingResponse()
-                            rep_socket.send(resp_proto.SerializeToString())
-                            continue
-
-                    # Process the request
-                    embeddings = compute_embeddings(texts, model_name, mode=embedding_mode)
-                    logger.info(f"Computed embeddings shape: {embeddings.shape}")
-
-                    # Validation
-                    if np.isnan(embeddings).any() or np.isinf(embeddings).any():
-                        logger.error("NaN or Inf detected in embeddings!")
-                        # Send error response
-                        if is_text_request:
-                            import msgpack
-
-                            response_data = msgpack.packb([])
-                        else:
-                            resp_proto = embedding_pb2.NodeEmbeddingResponse()
-                            response_data = resp_proto.SerializeToString()
-                        rep_socket.send(response_data)
-                        continue
-
-                    # Prepare response based on request type
-                    if is_text_request:
-                        # For direct text requests, return msgpack
-                        import msgpack
-
-                        response_data = msgpack.packb(embeddings.tolist())
-                    else:
-                        # For protobuf requests, return protobuf
-                        resp_proto = embedding_pb2.NodeEmbeddingResponse()
-                        hidden_contiguous = np.ascontiguousarray(embeddings, dtype=np.float32)
-
-                        resp_proto.embeddings_data = hidden_contiguous.tobytes()
-                        resp_proto.dimensions.append(hidden_contiguous.shape[0])
-                        resp_proto.dimensions.append(hidden_contiguous.shape[1])
-
-                        response_data = resp_proto.SerializeToString()
-
-                    # Send response back to the client
-                    rep_socket.send(response_data)
-
-                    e2e_end = time.time()
-                    logger.info(f"⏱️  ZMQ E2E time: {e2e_end - e2e_start:.6f}s")
-
-                except zmq.Again:
-                    # Timeout - check shutdown_event and continue
-                    continue
-                except Exception as e:
-                    if not shutdown_event.is_set():
-                        logger.error(f"Error in ZMQ server loop: {e}")
-                        try:
-                            # Send error response for REP socket
-                            resp_proto = embedding_pb2.NodeEmbeddingResponse()
-                            rep_socket.send(resp_proto.SerializeToString())
-                        except Exception:
-                            pass
-                    else:
-                        logger.info("Shutdown in progress, ignoring ZMQ error")
-                        break
-        finally:
-            try:
-                rep_socket.close(0)
-            except Exception:
-                pass
-            try:
-                context.term()
-            except Exception:
-                pass
-
-        logger.info("DiskANN ZMQ server thread exiting gracefully")
-
-    # Add shutdown coordination
-    shutdown_event = threading.Event()
-
-    def shutdown_zmq_server():
-        """Gracefully shutdown ZMQ server."""
-        logger.info("Initiating graceful shutdown...")
-        shutdown_event.set()
-
-        if zmq_thread.is_alive():
-            logger.info("Waiting for ZMQ thread to finish...")
-            zmq_thread.join(timeout=5)
-            if zmq_thread.is_alive():
-                logger.warning("ZMQ thread did not finish in time")
-
-        # Clean up ZMQ resources
-        try:
-            # Note: socket and context are cleaned up by thread exit
-            logger.info("ZMQ resources cleaned up")
-        except Exception as e:
-            logger.warning(f"Error cleaning ZMQ resources: {e}")
-
-        # Clean up other resources
-        try:
-            import gc
-
-            gc.collect()
-            logger.info("Additional resources cleaned up")
-        except Exception as e:
-            logger.warning(f"Error cleaning additional resources: {e}")
-
-        logger.info("Graceful shutdown completed")
-        sys.exit(0)
-
-    # Register signal handlers within this function scope
-    import signal
-
-    def signal_handler(sig, frame):
-        logger.info(f"Received signal {sig}, shutting down gracefully...")
-        shutdown_zmq_server()
-
-    signal.signal(signal.SIGTERM, signal_handler)
-    signal.signal(signal.SIGINT, signal_handler)
-
-    # Start ZMQ thread (NOT daemon!)
-    zmq_thread = threading.Thread(
-        target=lambda: zmq_server_thread_with_shutdown(shutdown_event),
-        daemon=False,  # Not daemon - we want to wait for it
-    )
+    zmq_thread = threading.Thread(target=zmq_server_thread, daemon=True)
     zmq_thread.start()
     logger.info(f"Started DiskANN ZMQ server thread on port {zmq_port}")
 
     # Keep the main thread alive
     try:
-        while not shutdown_event.is_set():
-            time.sleep(0.1)  # Check shutdown more frequently
+        while True:
+            time.sleep(1)
     except KeyboardInterrupt:
         logger.info("DiskANN Server shutting down...")
-        shutdown_zmq_server()
         return
 
-    # If we reach here, shutdown was triggered by signal
-    logger.info("Main loop exited, process should be shutting down")
-
 
 if __name__ == "__main__":
+    import signal
     import sys
 
-    # Signal handlers are now registered within create_diskann_embedding_server
+    def signal_handler(sig, frame):
+        logger.info(f"Received signal {sig}, shutting down gracefully...")
+        sys.exit(0)
+
+    # Register signal handlers for graceful shutdown
+    signal.signal(signal.SIGTERM, signal_handler)
+    signal.signal(signal.SIGINT, signal_handler)
 
     parser = argparse.ArgumentParser(description="DiskANN Embedding service")
     parser.add_argument("--zmq-port", type=int, default=5555, help="ZMQ port to run on")

packages/leann-backend-diskann/leann_backend_diskann/graph_partition_simple.py

@@ -0,0 +1,137 @@
+#!/usr/bin/env python3
+"""
+Simplified Graph Partition Module for LEANN DiskANN Backend
+
+This module provides a simple Python interface for graph partitioning
+that directly calls the existing executables.
+"""
+
+import os
+import subprocess
+import tempfile
+from pathlib import Path
+from typing import Optional
+
+
+def partition_graph_simple(
+    index_prefix_path: str, output_dir: Optional[str] = None, **kwargs
+) -> tuple[str, str]:
+    """
+    Simple function to partition a graph index.
+
+    Args:
+        index_prefix_path: Path to the index prefix (e.g., "/path/to/index")
+        output_dir: Output directory (defaults to parent of index_prefix_path)
+        **kwargs: Additional parameters for graph partitioning
+
+    Returns:
+        Tuple of (disk_graph_index_path, partition_bin_path)
+    """
+    # Set default parameters
+    params = {
+        "gp_times": 10,
+        "lock_nums": 10,
+        "cut": 100,
+        "scale_factor": 1,
+        "data_type": "float",
+        "thread_nums": 10,
+        **kwargs,
+    }
+
+    # Determine output directory
+    if output_dir is None:
+        output_dir = str(Path(index_prefix_path).parent)
+
+    # Find the graph_partition directory
+    current_file = Path(__file__)
+    graph_partition_dir = current_file.parent.parent / "third_party" / "DiskANN" / "graph_partition"
+
+    if not graph_partition_dir.exists():
+        raise RuntimeError(f"Graph partition directory not found: {graph_partition_dir}")
+
+    # Find input index file
+    old_index_file = f"{index_prefix_path}_disk_beam_search.index"
+    if not os.path.exists(old_index_file):
+        old_index_file = f"{index_prefix_path}_disk.index"
+
+    if not os.path.exists(old_index_file):
+        raise RuntimeError(f"Index file not found: {old_index_file}")
+
+    # Create temporary directory for processing
+    with tempfile.TemporaryDirectory() as temp_dir:
+        temp_data_dir = Path(temp_dir) / "data"
+        temp_data_dir.mkdir(parents=True, exist_ok=True)
+
+        # Set up paths for temporary files
+        graph_path = temp_data_dir / "starling" / "_M_R_L_B" / "GRAPH"
+        graph_gp_path = (
+            graph_path
+            / f"GP_TIMES_{params['gp_times']}_LOCK_{params['lock_nums']}_GP_USE_FREQ0_CUT{params['cut']}_SCALE{params['scale_factor']}"
+        )
+        graph_gp_path.mkdir(parents=True, exist_ok=True)
+
+        # Run the build script with our parameters
+        cmd = [str(graph_partition_dir / "build.sh"), "release", "split_graph", index_prefix_path]
+
+        # Set environment variables for parameters
+        env = os.environ.copy()
+        env.update(
+            {
+                "GP_TIMES": str(params["gp_times"]),
+                "GP_LOCK_NUMS": str(params["lock_nums"]),
+                "GP_CUT": str(params["cut"]),
+                "GP_SCALE_F": str(params["scale_factor"]),
+                "DATA_TYPE": params["data_type"],
+                "GP_T": str(params["thread_nums"]),
+            }
+        )
+
+        print(f"Running graph partition with command: {' '.join(cmd)}")
+        print(f"Working directory: {graph_partition_dir}")
+
+        # Run the command
+        result = subprocess.run(
+            cmd, env=env, capture_output=True, text=True, cwd=graph_partition_dir
+        )
+
+        if result.returncode != 0:
+            print(f"Command failed with return code {result.returncode}")
+            print(f"stdout: {result.stdout}")
+            print(f"stderr: {result.stderr}")
+            raise RuntimeError(
+                f"Graph partitioning failed with return code {result.returncode}.\n"
+                f"stdout: {result.stdout}\n"
+                f"stderr: {result.stderr}"
+            )
+
+        # Check if output files were created
+        disk_graph_path = Path(output_dir) / "_disk_graph.index"
+        partition_bin_path = Path(output_dir) / "_partition.bin"
+
+        if not disk_graph_path.exists():
+            raise RuntimeError(f"Expected output file not found: {disk_graph_path}")
+
+        if not partition_bin_path.exists():
+            raise RuntimeError(f"Expected output file not found: {partition_bin_path}")
+
+        print("✅ Partitioning completed successfully!")
+        print(f"   Disk graph index: {disk_graph_path}")
+        print(f"   Partition binary: {partition_bin_path}")
+
+        return str(disk_graph_path), str(partition_bin_path)
+
+
+# Example usage
+if __name__ == "__main__":
+    try:
+        disk_graph_path, partition_bin_path = partition_graph_simple(
+            "/Users/yichuan/Desktop/release2/leann/diskannbuild/test_doc_files",
+            gp_times=5,
+            lock_nums=5,
+            cut=50,
+        )
+        print("Success! Output files:")
+        print(f"  - {disk_graph_path}")
+        print(f"  - {partition_bin_path}")
+    except Exception as e:
+        print(f"Error: {e}")

packages/leann-backend-diskann/pyproject.toml

@@ -4,8 +4,8 @@ build-backend = "scikit_build_core.build"
 
 [project]
 name = "leann-backend-diskann"
-version = "0.2.8"
-dependencies = ["leann-core==0.2.8", "numpy", "protobuf>=3.19.0"]
+version = "0.2.5"
+dependencies = ["leann-core==0.2.5", "numpy", "protobuf>=3.19.0"]
 
 [tool.scikit-build]
 # Key: simplified CMake path
@@ -17,5 +17,3 @@ editable.mode = "redirect"
 cmake.build-type = "Release"
 build.verbose = true
 build.tool-args = ["-j8"]
-# Let CMake find packages via Homebrew prefix
-cmake.define = {CMAKE_PREFIX_PATH = {env = "CMAKE_PREFIX_PATH"}, OpenMP_ROOT = {env = "OpenMP_ROOT"}}

packages/leann-backend-hnsw/CMakeLists.txt

@@ -5,20 +5,11 @@ set(CMAKE_CXX_COMPILER_WORKS 1)
 
 # Set OpenMP path for macOS
 if(APPLE)
-    # Detect Homebrew installation path (Apple Silicon vs Intel)
-    if(EXISTS "/opt/homebrew/opt/libomp")
-        set(HOMEBREW_PREFIX "/opt/homebrew")
-    elseif(EXISTS "/usr/local/opt/libomp")
-        set(HOMEBREW_PREFIX "/usr/local")
-    else()
-        message(FATAL_ERROR "Could not find libomp installation. Please install with: brew install libomp")
-    endif()
-
-    set(OpenMP_C_FLAGS "-Xpreprocessor -fopenmp -I${HOMEBREW_PREFIX}/opt/libomp/include")
-    set(OpenMP_CXX_FLAGS "-Xpreprocessor -fopenmp -I${HOMEBREW_PREFIX}/opt/libomp/include")
+    set(OpenMP_C_FLAGS "-Xpreprocessor -fopenmp -I/opt/homebrew/opt/libomp/include")
+    set(OpenMP_CXX_FLAGS "-Xpreprocessor -fopenmp -I/opt/homebrew/opt/libomp/include")
     set(OpenMP_C_LIB_NAMES "omp")
     set(OpenMP_CXX_LIB_NAMES "omp")
-    set(OpenMP_omp_LIBRARY "${HOMEBREW_PREFIX}/opt/libomp/lib/libomp.dylib")
+    set(OpenMP_omp_LIBRARY "/opt/homebrew/opt/libomp/lib/libomp.dylib")
 
     # Force use of system libc++ to avoid version mismatch
     set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -stdlib=libc++")

packages/leann-backend-hnsw/leann_backend_hnsw/convert_to_csr.py

@@ -1,6 +1,5 @@
 import argparse
 import gc  # Import garbage collector interface
-import logging
 import os
 import struct
 import sys
@@ -8,12 +7,6 @@ import time
 
 import numpy as np
 
-# Set up logging to avoid print buffer issues
-logger = logging.getLogger(__name__)
-LOG_LEVEL = os.getenv("LEANN_LOG_LEVEL", "WARNING").upper()
-log_level = getattr(logging, LOG_LEVEL, logging.WARNING)
-logger.setLevel(log_level)
-
 # --- FourCCs (add more if needed) ---
 INDEX_HNSW_FLAT_FOURCC = int.from_bytes(b"IHNf", "little")
 # Add other HNSW fourccs if you expect different storage types inside HNSW
@@ -250,8 +243,6 @@ def convert_hnsw_graph_to_csr(input_filename, output_filename, prune_embeddings=
         output_filename: Output CSR index file
         prune_embeddings: Whether to prune embedding storage (write NULL storage marker)
     """
-    # Keep prints simple; rely on CI runner to flush output as needed
-
     print(f"Starting conversion: {input_filename} -> {output_filename}")
     start_time = time.time()
     original_hnsw_data = {}

packages/leann-backend-hnsw/leann_backend_hnsw/hnsw_embedding_server.py

@@ -82,317 +82,188 @@ def create_hnsw_embedding_server(
     with open(passages_file) as f:
         meta = json.load(f)
 
-    # Let PassageManager handle path resolution uniformly. It supports fallback order:
-    # 1) path/index_path; 2) *_relative; 3) standard siblings next to meta
+    # Let PassageManager handle path resolution uniformly
     passages = PassageManager(meta["passage_sources"], metadata_file_path=passages_file)
-    # Dimension from metadata for shaping responses
-    try:
-        embedding_dim: int = int(meta.get("dimensions", 0))
-    except Exception:
-        embedding_dim = 0
     logger.info(
         f"Loaded PassageManager with {len(passages.global_offset_map)} passages from metadata"
     )
 
-    # (legacy ZMQ thread removed; using shutdown-capable server only)
-
-    def zmq_server_thread_with_shutdown(shutdown_event):
-        """ZMQ server thread that respects shutdown signal.
-
-        Creates its own REP socket bound to zmq_port and polls with timeouts
-        to allow graceful shutdown.
-        """
-        logger.info("ZMQ server thread started with shutdown support")
-
+    def zmq_server_thread():
+        """ZMQ server thread"""
         context = zmq.Context()
-        rep_socket = context.socket(zmq.REP)
-        rep_socket.bind(f"tcp://*:{zmq_port}")
-        logger.info(f"HNSW ZMQ REP server listening on port {zmq_port}")
-        rep_socket.setsockopt(zmq.RCVTIMEO, 1000)
-        # Keep sends from blocking during shutdown; fail fast and drop on close
-        rep_socket.setsockopt(zmq.SNDTIMEO, 1000)
-        rep_socket.setsockopt(zmq.LINGER, 0)
+        socket = context.socket(zmq.REP)
+        socket.bind(f"tcp://*:{zmq_port}")
+        logger.info(f"HNSW ZMQ server listening on port {zmq_port}")
 
-        # Track last request type/length for shape-correct fallbacks
-        last_request_type = "unknown"  # 'text' | 'distance' | 'embedding' | 'unknown'
-        last_request_length = 0
+        socket.setsockopt(zmq.RCVTIMEO, 300000)
+        socket.setsockopt(zmq.SNDTIMEO, 300000)
 
-        try:
-            while not shutdown_event.is_set():
-                try:
-                    e2e_start = time.time()
-                    logger.debug("🔍 Waiting for ZMQ message...")
-                    request_bytes = rep_socket.recv()
+        while True:
+            try:
+                message_bytes = socket.recv()
+                logger.debug(f"Received ZMQ request of size {len(message_bytes)} bytes")
 
-                    # Rest of the processing logic (same as original)
-                    request = msgpack.unpackb(request_bytes)
+                e2e_start = time.time()
+                request_payload = msgpack.unpackb(message_bytes)
 
-                    if len(request) == 1 and request[0] == "__QUERY_MODEL__":
-                        response_bytes = msgpack.packb([model_name])
-                        rep_socket.send(response_bytes)
-                        continue
+                # Handle direct text embedding request
+                if isinstance(request_payload, list) and len(request_payload) > 0:
+                    # Check if this is a direct text request (list of strings)
+                    if all(isinstance(item, str) for item in request_payload):
+                        logger.info(
+                            f"Processing direct text embedding request for {len(request_payload)} texts in {embedding_mode} mode"
+                        )
 
-                    # Handle direct text embedding request
-                    if (
-                        isinstance(request, list)
-                        and request
-                        and all(isinstance(item, str) for item in request)
-                    ):
-                        last_request_type = "text"
-                        last_request_length = len(request)
-                        embeddings = compute_embeddings(request, model_name, mode=embedding_mode)
-                        rep_socket.send(msgpack.packb(embeddings.tolist()))
+                        # Use unified embedding computation (now with model caching)
+                        embeddings = compute_embeddings(
+                            request_payload, model_name, mode=embedding_mode
+                        )
+
+                        response = embeddings.tolist()
+                        socket.send(msgpack.packb(response))
                         e2e_end = time.time()
                         logger.info(f"⏱️  Text embedding E2E time: {e2e_end - e2e_start:.6f}s")
                         continue
 
-                    # Handle distance calculation request: [[ids], [query_vector]]
-                    if (
-                        isinstance(request, list)
-                        and len(request) == 2
-                        and isinstance(request[0], list)
-                        and isinstance(request[1], list)
-                    ):
-                        node_ids = request[0]
-                        # Handle nested [[ids]] shape defensively
-                        if len(node_ids) == 1 and isinstance(node_ids[0], list):
-                            node_ids = node_ids[0]
-                        query_vector = np.array(request[1], dtype=np.float32)
-                        last_request_type = "distance"
-                        last_request_length = len(node_ids)
+                # Handle distance calculation requests
+                if (
+                    isinstance(request_payload, list)
+                    and len(request_payload) == 2
+                    and isinstance(request_payload[0], list)
+                    and isinstance(request_payload[1], list)
+                ):
+                    node_ids = request_payload[0]
+                    query_vector = np.array(request_payload[1], dtype=np.float32)
 
-                        logger.debug("Distance calculation request received")
-                        logger.debug(f"    Node IDs: {node_ids}")
-                        logger.debug(f"    Query vector dim: {len(query_vector)}")
+                    logger.debug("Distance calculation request received")
+                    logger.debug(f"    Node IDs: {node_ids}")
+                    logger.debug(f"    Query vector dim: {len(query_vector)}")
 
-                        # Gather texts for found ids
-                        texts: list[str] = []
-                        found_indices: list[int] = []
-                        for idx, nid in enumerate(node_ids):
-                            try:
-                                passage_data = passages.get_passage(str(nid))
-                                txt = passage_data.get("text", "")
-                                if isinstance(txt, str) and len(txt) > 0:
-                                    texts.append(txt)
-                                    found_indices.append(idx)
-                                else:
-                                    logger.error(f"Empty text for passage ID {nid}")
-                            except KeyError:
-                                logger.error(f"Passage ID {nid} not found")
-                            except Exception as e:
-                                logger.error(f"Exception looking up passage ID {nid}: {e}")
-
-                        # Prepare full-length response with large sentinel values
-                        large_distance = 1e9
-                        response_distances = [large_distance] * len(node_ids)
-
-                        if texts:
-                            try:
-                                embeddings = compute_embeddings(
-                                    texts, model_name, mode=embedding_mode
-                                )
-                                logger.info(
-                                    f"Computed embeddings for {len(texts)} texts, shape: {embeddings.shape}"
-                                )
-                                if distance_metric == "l2":
-                                    partial = np.sum(
-                                        np.square(embeddings - query_vector.reshape(1, -1)), axis=1
-                                    )
-                                else:  # mips or cosine
-                                    partial = -np.dot(embeddings, query_vector)
-
-                                for pos, dval in zip(found_indices, partial.flatten().tolist()):
-                                    response_distances[pos] = float(dval)
-                            except Exception as e:
-                                logger.error(f"Distance computation error, using sentinels: {e}")
-
-                        # Send response in expected shape [[distances]]
-                        rep_socket.send(msgpack.packb([response_distances], use_single_float=True))
-                        e2e_end = time.time()
-                        logger.info(f"⏱️  Distance calculation E2E time: {e2e_end - e2e_start:.6f}s")
-                        continue
-
-                    # Fallback: treat as embedding-by-id request
-                    if (
-                        isinstance(request, list)
-                        and len(request) == 1
-                        and isinstance(request[0], list)
-                    ):
-                        node_ids = request[0]
-                    elif isinstance(request, list):
-                        node_ids = request
-                    else:
-                        node_ids = []
-                    last_request_type = "embedding"
-                    last_request_length = len(node_ids)
-                    logger.info(f"ZMQ received {len(node_ids)} node IDs for embedding fetch")
-
-                    # Preallocate zero-filled flat data for robustness
-                    if embedding_dim <= 0:
-                        dims = [0, 0]
-                        flat_data: list[float] = []
-                    else:
-                        dims = [len(node_ids), embedding_dim]
-                        flat_data = [0.0] * (dims[0] * dims[1])
-
-                    # Collect texts for found ids
-                    texts: list[str] = []
-                    found_indices: list[int] = []
-                    for idx, nid in enumerate(node_ids):
+                    # Get embeddings for node IDs
+                    texts = []
+                    for nid in node_ids:
                         try:
                             passage_data = passages.get_passage(str(nid))
-                            txt = passage_data.get("text", "")
-                            if isinstance(txt, str) and len(txt) > 0:
-                                texts.append(txt)
-                                found_indices.append(idx)
-                            else:
-                                logger.error(f"Empty text for passage ID {nid}")
+                            txt = passage_data["text"]
+                            texts.append(txt)
                         except KeyError:
-                            logger.error(f"Passage with ID {nid} not found")
+                            logger.error(f"Passage ID {nid} not found")
+                            raise RuntimeError(f"FATAL: Passage with ID {nid} not found")
                         except Exception as e:
                             logger.error(f"Exception looking up passage ID {nid}: {e}")
+                            raise
 
-                    if texts:
-                        try:
-                            embeddings = compute_embeddings(texts, model_name, mode=embedding_mode)
-                            logger.info(
-                                f"Computed embeddings for {len(texts)} texts, shape: {embeddings.shape}"
-                            )
+                    # Process embeddings
+                    embeddings = compute_embeddings(texts, model_name, mode=embedding_mode)
+                    logger.info(
+                        f"Computed embeddings for {len(texts)} texts, shape: {embeddings.shape}"
+                    )
 
-                            if np.isnan(embeddings).any() or np.isinf(embeddings).any():
-                                logger.error(
-                                    f"NaN or Inf detected in embeddings! Requested IDs: {node_ids[:5]}..."
-                                )
-                                dims = [0, embedding_dim]
-                                flat_data = []
-                            else:
-                                emb_f32 = np.ascontiguousarray(embeddings, dtype=np.float32)
-                                flat = emb_f32.flatten().tolist()
-                                for j, pos in enumerate(found_indices):
-                                    start = pos * embedding_dim
-                                    end = start + embedding_dim
-                                    if end <= len(flat_data):
-                                        flat_data[start:end] = flat[
-                                            j * embedding_dim : (j + 1) * embedding_dim
-                                        ]
-                        except Exception as e:
-                            logger.error(f"Embedding computation error, returning zeros: {e}")
+                    # Calculate distances
+                    if distance_metric == "l2":
+                        distances = np.sum(
+                            np.square(embeddings - query_vector.reshape(1, -1)), axis=1
+                        )
+                    else:  # mips or cosine
+                        distances = -np.dot(embeddings, query_vector)
 
-                    response_payload = [dims, flat_data]
-                    response_bytes = msgpack.packb(response_payload, use_single_float=True)
+                    response_payload = distances.flatten().tolist()
+                    response_bytes = msgpack.packb([response_payload], use_single_float=True)
+                    logger.debug(f"Sending distance response with {len(distances)} distances")
 
-                    rep_socket.send(response_bytes)
+                    socket.send(response_bytes)
                     e2e_end = time.time()
-                    logger.info(f"⏱️  ZMQ E2E time: {e2e_end - e2e_start:.6f}s")
-
-                except zmq.Again:
-                    # Timeout - check shutdown_event and continue
+                    logger.info(f"⏱️  Distance calculation E2E time: {e2e_end - e2e_start:.6f}s")
                     continue
-                except Exception as e:
-                    if not shutdown_event.is_set():
-                        logger.error(f"Error in ZMQ server loop: {e}")
-                        # Shape-correct fallback
-                        try:
-                            if last_request_type == "distance":
-                                large_distance = 1e9
-                                fallback_len = max(0, int(last_request_length))
-                                safe = [[large_distance] * fallback_len]
-                            elif last_request_type == "embedding":
-                                bsz = max(0, int(last_request_length))
-                                dim = max(0, int(embedding_dim))
-                                safe = (
-                                    [[bsz, dim], [0.0] * (bsz * dim)] if dim > 0 else [[0, 0], []]
-                                )
-                            elif last_request_type == "text":
-                                safe = []  # direct text embeddings expectation is a flat list
-                            else:
-                                safe = [[0, int(embedding_dim) if embedding_dim > 0 else 0], []]
-                            rep_socket.send(msgpack.packb(safe, use_single_float=True))
-                        except Exception:
-                            pass
-                    else:
-                        logger.info("Shutdown in progress, ignoring ZMQ error")
-                        break
-        finally:
-            try:
-                rep_socket.close(0)
-            except Exception:
-                pass
-            try:
-                context.term()
-            except Exception:
-                pass
 
-        logger.info("ZMQ server thread exiting gracefully")
+                # Standard embedding request (passage ID lookup)
+                if (
+                    not isinstance(request_payload, list)
+                    or len(request_payload) != 1
+                    or not isinstance(request_payload[0], list)
+                ):
+                    logger.error(
+                        f"Invalid MessagePack request format. Expected [[ids...]] or [texts...], got: {type(request_payload)}"
+                    )
+                    socket.send(msgpack.packb([[], []]))
+                    continue
 
-    # Add shutdown coordination
-    shutdown_event = threading.Event()
+                node_ids = request_payload[0]
+                logger.debug(f"Request for {len(node_ids)} node embeddings")
 
-    def shutdown_zmq_server():
-        """Gracefully shutdown ZMQ server."""
-        logger.info("Initiating graceful shutdown...")
-        shutdown_event.set()
+                # Look up texts by node IDs
+                texts = []
+                for nid in node_ids:
+                    try:
+                        passage_data = passages.get_passage(str(nid))
+                        txt = passage_data["text"]
+                        if not txt:
+                            raise RuntimeError(f"FATAL: Empty text for passage ID {nid}")
+                        texts.append(txt)
+                    except KeyError:
+                        raise RuntimeError(f"FATAL: Passage with ID {nid} not found")
+                    except Exception as e:
+                        logger.error(f"Exception looking up passage ID {nid}: {e}")
+                        raise
 
-        if zmq_thread.is_alive():
-            logger.info("Waiting for ZMQ thread to finish...")
-            zmq_thread.join(timeout=5)
-            if zmq_thread.is_alive():
-                logger.warning("ZMQ thread did not finish in time")
+                # Process embeddings
+                embeddings = compute_embeddings(texts, model_name, mode=embedding_mode)
+                logger.info(
+                    f"Computed embeddings for {len(texts)} texts, shape: {embeddings.shape}"
+                )
 
-        # Clean up ZMQ resources
-        try:
-            # Note: socket and context are cleaned up by thread exit
-            logger.info("ZMQ resources cleaned up")
-        except Exception as e:
-            logger.warning(f"Error cleaning ZMQ resources: {e}")
+                # Serialization and response
+                if np.isnan(embeddings).any() or np.isinf(embeddings).any():
+                    logger.error(
+                        f"NaN or Inf detected in embeddings! Requested IDs: {node_ids[:5]}..."
+                    )
+                    raise AssertionError()
 
-        # Clean up other resources
-        try:
-            import gc
+                hidden_contiguous_f32 = np.ascontiguousarray(embeddings, dtype=np.float32)
+                response_payload = [
+                    list(hidden_contiguous_f32.shape),
+                    hidden_contiguous_f32.flatten().tolist(),
+                ]
+                response_bytes = msgpack.packb(response_payload, use_single_float=True)
 
-            gc.collect()
-            logger.info("Additional resources cleaned up")
-        except Exception as e:
-            logger.warning(f"Error cleaning additional resources: {e}")
+                socket.send(response_bytes)
+                e2e_end = time.time()
+                logger.info(f"⏱️  ZMQ E2E time: {e2e_end - e2e_start:.6f}s")
 
-        logger.info("Graceful shutdown completed")
-        sys.exit(0)
+            except zmq.Again:
+                logger.debug("ZMQ socket timeout, continuing to listen")
+                continue
+            except Exception as e:
+                logger.error(f"Error in ZMQ server loop: {e}")
+                import traceback
 
-    # Register signal handlers within this function scope
-    import signal
+                traceback.print_exc()
+                socket.send(msgpack.packb([[], []]))
 
-    def signal_handler(sig, frame):
-        logger.info(f"Received signal {sig}, shutting down gracefully...")
-        shutdown_zmq_server()
-
-    signal.signal(signal.SIGTERM, signal_handler)
-    signal.signal(signal.SIGINT, signal_handler)
-
-    # Pass shutdown_event to ZMQ thread
-    zmq_thread = threading.Thread(
-        target=lambda: zmq_server_thread_with_shutdown(shutdown_event),
-        daemon=False,  # Not daemon - we want to wait for it
-    )
+    zmq_thread = threading.Thread(target=zmq_server_thread, daemon=True)
     zmq_thread.start()
     logger.info(f"Started HNSW ZMQ server thread on port {zmq_port}")
 
     # Keep the main thread alive
     try:
-        while not shutdown_event.is_set():
-            time.sleep(0.1)  # Check shutdown more frequently
+        while True:
+            time.sleep(1)
     except KeyboardInterrupt:
         logger.info("HNSW Server shutting down...")
-        shutdown_zmq_server()
         return
 
-    # If we reach here, shutdown was triggered by signal
-    logger.info("Main loop exited, process should be shutting down")
-
 
 if __name__ == "__main__":
+    import signal
     import sys
 
-    # Signal handlers are now registered within create_hnsw_embedding_server
+    def signal_handler(sig, frame):
+        logger.info(f"Received signal {sig}, shutting down gracefully...")
+        sys.exit(0)
+
+    # Register signal handlers for graceful shutdown
+    signal.signal(signal.SIGTERM, signal_handler)
+    signal.signal(signal.SIGINT, signal_handler)
 
     parser = argparse.ArgumentParser(description="HNSW Embedding service")
     parser.add_argument("--zmq-port", type=int, default=5555, help="ZMQ port to run on")

packages/leann-backend-hnsw/pyproject.toml

@@ -6,10 +6,10 @@ build-backend = "scikit_build_core.build"
 
 [project]
 name = "leann-backend-hnsw"
-version = "0.2.8"
+version = "0.2.5"
 description = "Custom-built HNSW (Faiss) backend for the Leann toolkit."
 dependencies = [
-    "leann-core==0.2.8",
+    "leann-core==0.2.5",
     "numpy",
     "pyzmq>=23.0.0",
     "msgpack>=1.0.0",
@@ -22,8 +22,6 @@ cmake.build-type = "Release"
 build.verbose = true
 build.tool-args = ["-j8"]
 
-# CMake definitions to optimize compilation and find Homebrew packages
+# CMake definitions to optimize compilation
 [tool.scikit-build.cmake.define]
 CMAKE_BUILD_PARALLEL_LEVEL = "8"
-CMAKE_PREFIX_PATH = {env = "CMAKE_PREFIX_PATH"}
-OpenMP_ROOT = {env = "OpenMP_ROOT"}

packages/leann-core/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "leann-core"
-version = "0.2.8"
+version = "0.2.5"
 description = "Core API and plugin system for LEANN"
 readme = "README.md"
 requires-python = ">=3.9"
@@ -31,10 +31,8 @@ dependencies = [
     "PyPDF2>=3.0.0",
     "pymupdf>=1.23.0",
     "pdfplumber>=0.10.0",
-    "nbconvert>=7.0.0",  # For .ipynb file support
-    "gitignore-parser>=0.1.12",  # For proper .gitignore handling
-    "mlx>=0.26.3; sys_platform == 'darwin' and platform_machine == 'arm64'",
-    "mlx-lm>=0.26.0; sys_platform == 'darwin' and platform_machine == 'arm64'",
+    "mlx>=0.26.3; sys_platform == 'darwin'",
+    "mlx-lm>=0.26.0; sys_platform == 'darwin'",
 ]
 
 [project.optional-dependencies]

packages/leann-core/src/leann/api.py

@@ -87,21 +87,26 @@ def compute_embeddings_via_server(chunks: list[str], model_name: str, port: int)
     # Connect to embedding server
     context = zmq.Context()
     socket = context.socket(zmq.REQ)
+    socket.setsockopt(zmq.LINGER, 0)  # Don't block on close
+    socket.setsockopt(zmq.RCVTIMEO, 300000)
+    socket.setsockopt(zmq.SNDTIMEO, 300000)
+    socket.setsockopt(zmq.IMMEDIATE, 1)
     socket.connect(f"tcp://localhost:{port}")
 
-    # Send chunks to server for embedding computation
-    request = chunks
-    socket.send(msgpack.packb(request))
+    try:
+        # Send chunks to server for embedding computation
+        request = chunks
+        socket.send(msgpack.packb(request))
 
-    # Receive embeddings from server
-    response = socket.recv()
-    embeddings_list = msgpack.unpackb(response)
+        # Receive embeddings from server
+        response = socket.recv()
+        embeddings_list = msgpack.unpackb(response)
 
-    # Convert back to numpy array
-    embeddings = np.array(embeddings_list, dtype=np.float32)
-
-    socket.close()
-    context.term()
+        # Convert back to numpy array
+        embeddings = np.array(embeddings_list, dtype=np.float32)
+    finally:
+        socket.close()
+        # Don't call context.term() - this was causing hangs
 
     return embeddings
 
@@ -122,55 +127,31 @@ class PassageManager:
         self.passage_files = {}
         self.global_offset_map = {}  # Combined map for fast lookup
 
-        # Derive index base name for standard sibling fallbacks, e.g., <index_name>.passages.*
-        index_name_base = None
-        if metadata_file_path:
-            meta_name = Path(metadata_file_path).name
-            if meta_name.endswith(".meta.json"):
-                index_name_base = meta_name[: -len(".meta.json")]
-
         for source in passage_sources:
             assert source["type"] == "jsonl", "only jsonl is supported"
-            passage_file = source.get("path", "")
-            index_file = source.get("index_path", "")  # .idx file
+            passage_file = source["path"]
+            index_file = source["index_path"]  # .idx file
 
             # Fix path resolution - relative paths should be relative to metadata file directory
-            def _resolve_candidates(
-                primary: str,
-                relative_key: str,
-                default_name: Optional[str],
-                source_dict: dict[str, Any],
-            ) -> list[Path]:
-                candidates: list[Path] = []
-                # 1) Primary as-is (absolute or relative)
-                if primary:
-                    p = Path(primary)
-                    candidates.append(p if p.is_absolute() else (Path.cwd() / p))
-                # 2) metadata-relative explicit relative key
-                if metadata_file_path and source_dict.get(relative_key):
-                    candidates.append(Path(metadata_file_path).parent / source_dict[relative_key])
-                # 3) metadata-relative standard sibling filename
-                if metadata_file_path and default_name:
-                    candidates.append(Path(metadata_file_path).parent / default_name)
-                return candidates
-
-            # Build candidate lists and pick first existing; otherwise keep last candidate for error message
-            idx_default = f"{index_name_base}.passages.idx" if index_name_base else None
-            idx_candidates = _resolve_candidates(
-                index_file, "index_path_relative", idx_default, source
-            )
-            pas_default = f"{index_name_base}.passages.jsonl" if index_name_base else None
-            pas_candidates = _resolve_candidates(passage_file, "path_relative", pas_default, source)
-
-            def _pick_existing(cands: list[Path]) -> str:
-                for c in cands:
-                    if c.exists():
-                        return str(c.resolve())
-                # Fallback to last candidate (best guess) even if not exists; will error below
-                return str(cands[-1].resolve()) if cands else ""
-
-            index_file = _pick_existing(idx_candidates)
-            passage_file = _pick_existing(pas_candidates)
+            if not Path(index_file).is_absolute():
+                if metadata_file_path:
+                    # Resolve relative to metadata file directory
+                    metadata_dir = Path(metadata_file_path).parent
+                    logger.debug(
+                        f"PassageManager: Resolving relative paths from metadata_dir: {metadata_dir}"
+                    )
+                    index_file = str((metadata_dir / index_file).resolve())
+                    passage_file = str((metadata_dir / passage_file).resolve())
+                    logger.debug(f"PassageManager: Resolved index_file: {index_file}")
+                else:
+                    # Fallback to current directory resolution (legacy behavior)
+                    logger.warning(
+                        "PassageManager: No metadata_file_path provided, using fallback resolution from cwd"
+                    )
+                    logger.debug(f"PassageManager: Current working directory: {Path.cwd()}")
+                    index_file = str(Path(index_file).resolve())
+                    passage_file = str(Path(passage_file).resolve())
+                    logger.debug(f"PassageManager: Fallback resolved index_file: {index_file}")
 
             if not Path(index_file).exists():
                 raise FileNotFoundError(f"Passage index file not found: {index_file}")
@@ -356,12 +337,8 @@ class LeannBuilder:
             "passage_sources": [
                 {
                     "type": "jsonl",
-                    # Preserve existing relative file names (backward-compatible)
-                    "path": passages_file.name,
-                    "index_path": offset_file.name,
-                    # Add optional redundant relative keys for remote build portability (non-breaking)
-                    "path_relative": passages_file.name,
-                    "index_path_relative": offset_file.name,
+                    "path": passages_file.name,  # Use relative path (just filename)
+                    "index_path": offset_file.name,  # Use relative path (just filename)
                 }
             ],
         }
@@ -476,12 +453,8 @@ class LeannBuilder:
             "passage_sources": [
                 {
                     "type": "jsonl",
-                    # Preserve existing relative file names (backward-compatible)
-                    "path": passages_file.name,
-                    "index_path": offset_file.name,
-                    # Add optional redundant relative keys for remote build portability (non-breaking)
-                    "path_relative": passages_file.name,
-                    "index_path_relative": offset_file.name,
+                    "path": passages_file.name,  # Use relative path (just filename)
+                    "index_path": offset_file.name,  # Use relative path (just filename)
                 }
             ],
             "built_from_precomputed_embeddings": True,
@@ -604,7 +577,6 @@ class LeannSearcher:
         enriched_results = []
         if "labels" in results and "distances" in results:
             logger.info(f"  Processing {len(results['labels'][0])} passage IDs:")
-            # Python 3.9 does not support zip(strict=...); lengths are expected to match
             for i, (string_id, dist) in enumerate(
                 zip(results["labels"][0], results["distances"][0])
             ):
@@ -632,25 +604,17 @@ class LeannSearcher:
                     )
                 except KeyError:
                     RED = "\033[91m"
-                    RESET = "\033[0m"
                     logger.error(
                         f"   {RED}✗{RESET} [{i + 1:2d}] ID: '{string_id}' -> {RED}ERROR: Passage not found!{RESET}"
                     )
 
-        # Define color codes outside the loop for final message
-        GREEN = "\033[92m"
-        RESET = "\033[0m"
         logger.info(f"  {GREEN}✓ Final enriched results: {len(enriched_results)} passages{RESET}")
         return enriched_results
 
     def cleanup(self):
-        """Explicitly cleanup embedding server resources.
-
-        This method should be called after you're done using the searcher,
-        especially in test environments or batch processing scenarios.
-        """
-        if hasattr(self.backend_impl, "embedding_server_manager"):
-            self.backend_impl.embedding_server_manager.stop_server()
+        """Cleanup embedding server and other resources."""
+        if hasattr(self.backend_impl, "cleanup"):
+            self.backend_impl.cleanup()
 
 
 class LeannChat:
@@ -721,12 +685,3 @@ class LeannChat:
             except (KeyboardInterrupt, EOFError):
                 print("\nGoodbye!")
                 break
-
-    def cleanup(self):
-        """Explicitly cleanup embedding server resources.
-
-        This method should be called after you're done using the chat interface,
-        especially in test environments or batch processing scenarios.
-        """
-        if hasattr(self.searcher, "cleanup"):
-            self.searcher.cleanup()

packages/leann-core/src/leann/cli.py

@@ -1,11 +1,10 @@
 import argparse
 import asyncio
 from pathlib import Path
-from typing import Union
+from typing import Optional
 
 from llama_index.core import SimpleDirectoryReader
 from llama_index.core.node_parser import SentenceSplitter
-from tqdm import tqdm
 
 from .api import LeannBuilder, LeannChat, LeannSearcher
 
@@ -76,14 +75,11 @@ class LeannCLI:
             formatter_class=argparse.RawDescriptionHelpFormatter,
             epilog="""
 Examples:
-  leann build my-docs --docs ./documents                                  # Build index from directory
-  leann build my-code --docs ./src ./tests ./config                      # Build index from multiple directories
-  leann build my-files --docs ./file1.py ./file2.txt ./docs/             # Build index from files and directories
-  leann build my-mixed --docs ./readme.md ./src/ ./config.json           # Build index from mixed files/dirs
-  leann build my-ppts --docs ./ --file-types .pptx,.pdf                  # Index only PowerPoint and PDF files
-  leann search my-docs "query"                                           # Search in my-docs index
-  leann ask my-docs "question"                                           # Ask my-docs index
-  leann list                                                             # List all stored indexes
+  leann build my-docs --docs ./documents                    # Build index named my-docs
+  leann build my-ppts --docs ./ --file-types .pptx,.pdf    # Index only PowerPoint and PDF files
+  leann search my-docs "query"                             # Search in my-docs index
+  leann ask my-docs "question"                             # Ask my-docs index
+  leann list                                              # List all stored indexes
             """,
         )
 
@@ -91,15 +87,9 @@ Examples:
 
         # Build command
         build_parser = subparsers.add_parser("build", help="Build document index")
+        build_parser.add_argument("index_name", help="Index name")
         build_parser.add_argument(
-            "index_name", nargs="?", help="Index name (default: current directory name)"
-        )
-        build_parser.add_argument(
-            "--docs",
-            type=str,
-            nargs="+",
-            default=["."],
-            help="Documents directories and/or files (default: current directory)",
+            "--docs", type=str, default=".", help="Documents directory (default: current directory)"
         )
         build_parser.add_argument(
             "--backend", type=str, default="hnsw", choices=["hnsw", "diskann"]
@@ -212,63 +202,6 @@ Examples:
         with open(global_registry, "w") as f:
             json.dump(projects, f, indent=2)
 
-    def _build_gitignore_parser(self, docs_dir: str):
-        """Build gitignore parser using gitignore-parser library."""
-        from gitignore_parser import parse_gitignore
-
-        # Try to parse the root .gitignore
-        gitignore_path = Path(docs_dir) / ".gitignore"
-
-        if gitignore_path.exists():
-            try:
-                # gitignore-parser automatically handles all subdirectory .gitignore files!
-                matches = parse_gitignore(str(gitignore_path))
-                print(f"📋 Loaded .gitignore from {docs_dir} (includes all subdirectories)")
-                return matches
-            except Exception as e:
-                print(f"Warning: Could not parse .gitignore: {e}")
-        else:
-            print("📋 No .gitignore found")
-
-        # Fallback: basic pattern matching for essential files
-        essential_patterns = {".git", ".DS_Store", "__pycache__", "node_modules", ".venv", "venv"}
-
-        def basic_matches(file_path):
-            path_parts = Path(file_path).parts
-            return any(part in essential_patterns for part in path_parts)
-
-        return basic_matches
-
-    def _should_exclude_file(self, relative_path: Path, gitignore_matches) -> bool:
-        """Check if a file should be excluded using gitignore parser."""
-        return gitignore_matches(str(relative_path))
-
-    def _is_git_submodule(self, path: Path) -> bool:
-        """Check if a path is a git submodule."""
-        try:
-            # Find the git repo root
-            current_dir = Path.cwd()
-            while current_dir != current_dir.parent:
-                if (current_dir / ".git").exists():
-                    gitmodules_path = current_dir / ".gitmodules"
-                    if gitmodules_path.exists():
-                        # Read .gitmodules to check if this path is a submodule
-                        gitmodules_content = gitmodules_path.read_text()
-                        # Convert path to relative to git root
-                        try:
-                            relative_path = path.resolve().relative_to(current_dir)
-                            # Check if this path appears in .gitmodules
-                            return f"path = {relative_path}" in gitmodules_content
-                        except ValueError:
-                            # Path is not under git root
-                            return False
-                    break
-                current_dir = current_dir.parent
-            return False
-        except Exception:
-            # If anything goes wrong, assume it's not a submodule
-            return False
-
     def list_indexes(self):
         print("Stored LEANN indexes:")
 
@@ -298,9 +231,7 @@ Examples:
             valid_projects.append(current_path)
 
         if not valid_projects:
-            print(
-                "No indexes found. Use 'leann build <name> --docs <dir> [<dir2> ...]' to create one."
-            )
+            print("No indexes found. Use 'leann build <name> --docs <dir>' to create one.")
             return
 
         total_indexes = 0
@@ -347,88 +278,41 @@ Examples:
                     print(f'  leann search {example_name} "your query"')
                     print(f"  leann ask {example_name} --interactive")
 
-    def load_documents(
-        self, docs_paths: Union[str, list], custom_file_types: Union[str, None] = None
-    ):
-        # Handle both single path (string) and multiple paths (list) for backward compatibility
-        if isinstance(docs_paths, str):
-            docs_paths = [docs_paths]
-
-        # Separate files and directories
-        files = []
-        directories = []
-        for path in docs_paths:
-            path_obj = Path(path)
-            if path_obj.is_file():
-                files.append(str(path_obj))
-            elif path_obj.is_dir():
-                # Check if this is a git submodule - if so, skip it
-                if self._is_git_submodule(path_obj):
-                    print(f"⚠️  Skipping git submodule: {path}")
-                    continue
-                directories.append(str(path_obj))
-            else:
-                print(f"⚠️  Warning: Path '{path}' does not exist, skipping...")
-                continue
-
-        # Print summary of what we're processing
-        total_items = len(files) + len(directories)
-        items_desc = []
-        if files:
-            items_desc.append(f"{len(files)} file{'s' if len(files) > 1 else ''}")
-        if directories:
-            items_desc.append(
-                f"{len(directories)} director{'ies' if len(directories) > 1 else 'y'}"
-            )
-
-        print(f"Loading documents from {' and '.join(items_desc)} ({total_items} total):")
-        if files:
-            print(f"  📄 Files: {', '.join([Path(f).name for f in files])}")
-        if directories:
-            print(f"  📁 Directories: {', '.join(directories)}")
-
+    def load_documents(self, docs_dir: str, custom_file_types: Optional[str] = None):
+        print(f"Loading documents from {docs_dir}...")
         if custom_file_types:
             print(f"Using custom file types: {custom_file_types}")
 
-        all_documents = []
+        # Try to use better PDF parsers first
+        documents = []
+        docs_path = Path(docs_dir)
 
-        # First, process individual files if any
-        if files:
-            print(f"\n🔄 Processing {len(files)} individual file{'s' if len(files) > 1 else ''}...")
+        for file_path in docs_path.rglob("*.pdf"):
+            print(f"Processing PDF: {file_path}")
 
-            # Load individual files using SimpleDirectoryReader with input_files
-            # Note: We skip gitignore filtering for explicitly specified files
-            try:
-                # Group files by their parent directory for efficient loading
-                from collections import defaultdict
+            # Try PyMuPDF first (best quality)
+            text = extract_pdf_text_with_pymupdf(str(file_path))
+            if text is None:
+                # Try pdfplumber
+                text = extract_pdf_text_with_pdfplumber(str(file_path))
 
-                files_by_dir = defaultdict(list)
-                for file_path in files:
-                    parent_dir = str(Path(file_path).parent)
-                    files_by_dir[parent_dir].append(file_path)
+            if text:
+                # Create a simple document structure
+                from llama_index.core import Document
 
-                # Load files from each parent directory
-                for parent_dir, file_list in files_by_dir.items():
-                    print(
-                        f"  Loading {len(file_list)} file{'s' if len(file_list) > 1 else ''} from {parent_dir}"
-                    )
-                    try:
-                        file_docs = SimpleDirectoryReader(
-                            parent_dir,
-                            input_files=file_list,
-                            filename_as_id=True,
-                        ).load_data()
-                        all_documents.extend(file_docs)
-                        print(
-                            f"    ✅ Loaded {len(file_docs)} document{'s' if len(file_docs) > 1 else ''}"
-                        )
-                    except Exception as e:
-                        print(f"    ❌ Warning: Could not load files from {parent_dir}: {e}")
+                doc = Document(text=text, metadata={"source": str(file_path)})
+                documents.append(doc)
+            else:
+                # Fallback to default reader
+                print(f"Using default reader for {file_path}")
+                default_docs = SimpleDirectoryReader(
+                    str(file_path.parent),
+                    filename_as_id=True,
+                    required_exts=[file_path.suffix],
+                ).load_data()
+                documents.extend(default_docs)
 
-            except Exception as e:
-                print(f"❌ Error processing individual files: {e}")
-
-        # Define file extensions to process
+        # Load other file types with default reader
         if custom_file_types:
             # Parse custom file types from comma-separated string
             code_extensions = [ext.strip() for ext in custom_file_types.split(",") if ext.strip()]
@@ -490,106 +374,20 @@ Examples:
                 ".py",
                 ".jl",
             ]
-
-        # Process each directory
-        if directories:
-            print(
-                f"\n🔄 Processing {len(directories)} director{'ies' if len(directories) > 1 else 'y'}..."
-            )
-
-        for docs_dir in directories:
-            print(f"Processing directory: {docs_dir}")
-            # Build gitignore parser for each directory
-            gitignore_matches = self._build_gitignore_parser(docs_dir)
-
-            # Try to use better PDF parsers first, but only if PDFs are requested
-            documents = []
-            docs_path = Path(docs_dir)
-
-            # Check if we should process PDFs
-            should_process_pdfs = custom_file_types is None or ".pdf" in custom_file_types
-
-            if should_process_pdfs:
-                for file_path in docs_path.rglob("*.pdf"):
-                    # Check if file matches any exclude pattern
-                    try:
-                        relative_path = file_path.relative_to(docs_path)
-                        if self._should_exclude_file(relative_path, gitignore_matches):
-                            continue
-                    except ValueError:
-                        # Skip files that can't be made relative to docs_path
-                        print(f"⚠️  Skipping file outside directory scope: {file_path}")
-                        continue
-
-                    print(f"Processing PDF: {file_path}")
-
-                    # Try PyMuPDF first (best quality)
-                    text = extract_pdf_text_with_pymupdf(str(file_path))
-                    if text is None:
-                        # Try pdfplumber
-                        text = extract_pdf_text_with_pdfplumber(str(file_path))
-
-                    if text:
-                        # Create a simple document structure
-                        from llama_index.core import Document
-
-                        doc = Document(text=text, metadata={"source": str(file_path)})
-                        documents.append(doc)
-                    else:
-                        # Fallback to default reader
-                        print(f"Using default reader for {file_path}")
-                        try:
-                            default_docs = SimpleDirectoryReader(
-                                str(file_path.parent),
-                                filename_as_id=True,
-                                required_exts=[file_path.suffix],
-                            ).load_data()
-                            documents.extend(default_docs)
-                        except Exception as e:
-                            print(f"Warning: Could not process {file_path}: {e}")
-
-            # Load other file types with default reader
-            try:
-                # Create a custom file filter function using our PathSpec
-                def file_filter(
-                    file_path: str, docs_dir=docs_dir, gitignore_matches=gitignore_matches
-                ) -> bool:
-                    """Return True if file should be included (not excluded)"""
-                    try:
-                        docs_path_obj = Path(docs_dir)
-                        file_path_obj = Path(file_path)
-                        relative_path = file_path_obj.relative_to(docs_path_obj)
-                        return not self._should_exclude_file(relative_path, gitignore_matches)
-                    except (ValueError, OSError):
-                        return True  # Include files that can't be processed
-
-                other_docs = SimpleDirectoryReader(
-                    docs_dir,
-                    recursive=True,
-                    encoding="utf-8",
-                    required_exts=code_extensions,
-                    file_extractor={},  # Use default extractors
-                    filename_as_id=True,
-                ).load_data(show_progress=True)
-
-                # Filter documents after loading based on gitignore rules
-                filtered_docs = []
-                for doc in other_docs:
-                    file_path = doc.metadata.get("file_path", "")
-                    if file_filter(file_path):
-                        filtered_docs.append(doc)
-
-                documents.extend(filtered_docs)
-            except ValueError as e:
-                if "No files found" in str(e):
-                    print(f"No additional files found for other supported types in {docs_dir}.")
-                else:
-                    raise e
-
-            all_documents.extend(documents)
-            print(f"Loaded {len(documents)} documents from {docs_dir}")
-
-        documents = all_documents
+        # Try to load other file types, but don't fail if none are found
+        try:
+            other_docs = SimpleDirectoryReader(
+                docs_dir,
+                recursive=True,
+                encoding="utf-8",
+                required_exts=code_extensions,
+            ).load_data(show_progress=True)
+            documents.extend(other_docs)
+        except ValueError as e:
+            if "No files found" in str(e):
+                print("No additional files found for other supported types.")
+            else:
+                raise e
 
         all_texts = []
 
@@ -640,9 +438,7 @@ Examples:
             ".jl",
         }
 
-        print("start chunking documents")
-        # Add progress bar for document chunking
-        for doc in tqdm(documents, desc="Chunking documents", unit="doc"):
+        for doc in documents:
             # Check if this is a code file based on source path
             source_path = doc.metadata.get("source", "")
             is_code_file = any(source_path.endswith(ext) for ext in code_file_exts)
@@ -658,36 +454,18 @@ Examples:
         return all_texts
 
     async def build_index(self, args):
-        docs_paths = args.docs
-        # Use current directory name if index_name not provided
-        if args.index_name:
-            index_name = args.index_name
-        else:
-            index_name = Path.cwd().name
-            print(f"Using current directory name as index: '{index_name}'")
-
+        docs_dir = args.docs
+        index_name = args.index_name
         index_dir = self.indexes_dir / index_name
         index_path = self.get_index_path(index_name)
 
-        # Display all paths being indexed with file/directory distinction
-        files = [p for p in docs_paths if Path(p).is_file()]
-        directories = [p for p in docs_paths if Path(p).is_dir()]
-
-        print(f"📂 Indexing {len(docs_paths)} path{'s' if len(docs_paths) > 1 else ''}:")
-        if files:
-            print(f"  📄 Files ({len(files)}):")
-            for i, file_path in enumerate(files, 1):
-                print(f"    {i}. {Path(file_path).resolve()}")
-        if directories:
-            print(f"  📁 Directories ({len(directories)}):")
-            for i, dir_path in enumerate(directories, 1):
-                print(f"    {i}. {Path(dir_path).resolve()}")
+        print(f"📂 Indexing: {Path(docs_dir).resolve()}")
 
         if index_dir.exists() and not args.force:
             print(f"Index '{index_name}' already exists. Use --force to rebuild.")
             return
 
-        all_texts = self.load_documents(docs_paths, args.file_types)
+        all_texts = self.load_documents(docs_dir, args.file_types)
         if not all_texts:
             print("No documents found")
             return
@@ -723,7 +501,7 @@ Examples:
 
         if not self.index_exists(index_name):
             print(
-                f"Index '{index_name}' not found. Use 'leann build {index_name} --docs <dir> [<dir2> ...]' to create it."
+                f"Index '{index_name}' not found. Use 'leann build {index_name} --docs <dir>' to create it."
             )
             return
 
@@ -750,7 +528,7 @@ Examples:
 
         if not self.index_exists(index_name):
             print(
-                f"Index '{index_name}' not found. Use 'leann build {index_name} --docs <dir> [<dir2> ...]' to create it."
+                f"Index '{index_name}' not found. Use 'leann build {index_name} --docs <dir>' to create it."
             )
             return
 

packages/leann-core/src/leann/embedding_compute.py

@@ -6,6 +6,7 @@ Preserves all optimization parameters to ensure performance
 
 import logging
 import os
+from concurrent.futures import ThreadPoolExecutor, as_completed
 from typing import Any
 
 import numpy as np
@@ -373,9 +374,7 @@ def compute_embeddings_ollama(
     texts: list[str], model_name: str, is_build: bool = False, host: str = "http://localhost:11434"
 ) -> np.ndarray:
     """
-    Compute embeddings using Ollama API with simplified batch processing.
-
-    Uses batch size of 32 for MPS/CPU and 128 for CUDA to optimize performance.
+    Compute embeddings using Ollama API.
 
     Args:
         texts: List of texts to compute embeddings for
@@ -439,19 +438,12 @@ def compute_embeddings_ollama(
             if any(emb in base_name for emb in ["embed", "bge", "minilm", "e5"]):
                 embedding_models.append(model)
 
-        # Check if model exists (handle versioned names) and resolve to full name
-        resolved_model_name = None
-        for name in model_names:
-            # Exact match
-            if model_name == name:
-                resolved_model_name = name
-                break
-            # Match without version tag (use the versioned name)
-            elif model_name == name.split(":")[0]:
-                resolved_model_name = name
-                break
+        # Check if model exists (handle versioned names)
+        model_found = any(
+            model_name == name.split(":")[0] or model_name == name for name in model_names
+        )
 
-        if not resolved_model_name:
+        if not model_found:
             error_msg = f"❌ Model '{model_name}' not found in local Ollama.\n\n"
 
             # Suggest pulling the model
@@ -473,11 +465,6 @@ def compute_embeddings_ollama(
             error_msg += "\n📚 Browse more: https://ollama.com/library"
             raise ValueError(error_msg)
 
-        # Use the resolved model name for all subsequent operations
-        if resolved_model_name != model_name:
-            logger.info(f"Resolved model name '{model_name}' to '{resolved_model_name}'")
-        model_name = resolved_model_name
-
         # Verify the model supports embeddings by testing it
         try:
             test_response = requests.post(
@@ -498,148 +485,138 @@ def compute_embeddings_ollama(
     except requests.exceptions.RequestException as e:
         logger.warning(f"Could not verify model existence: {e}")
 
-    # Determine batch size based on device availability
-    # Check for CUDA/MPS availability using torch if available
-    batch_size = 32  # Default for MPS/CPU
-    try:
-        import torch
+    # Process embeddings with optimized concurrent processing
+    import requests
 
-        if torch.cuda.is_available():
-            batch_size = 128  # CUDA gets larger batch size
-        elif hasattr(torch.backends, "mps") and torch.backends.mps.is_available():
-            batch_size = 32  # MPS gets smaller batch size
-    except ImportError:
-        # If torch is not available, use conservative batch size
-        batch_size = 32
+    def get_single_embedding(text_idx_tuple):
+        """Helper function to get embedding for a single text."""
+        text, idx = text_idx_tuple
+        max_retries = 3
+        retry_count = 0
 
-    logger.info(f"Using batch size: {batch_size}")
+        # Truncate very long texts to avoid API issues
+        truncated_text = text[:8000] if len(text) > 8000 else text
 
-    def get_batch_embeddings(batch_texts):
-        """Get embeddings for a batch of texts."""
-        all_embeddings = []
-        failed_indices = []
+        while retry_count < max_retries:
+            try:
+                response = requests.post(
+                    f"{host}/api/embeddings",
+                    json={"model": model_name, "prompt": truncated_text},
+                    timeout=30,
+                )
+                response.raise_for_status()
 
-        for i, text in enumerate(batch_texts):
-            max_retries = 3
-            retry_count = 0
+                result = response.json()
+                embedding = result.get("embedding")
 
-            # Truncate very long texts to avoid API issues
-            truncated_text = text[:8000] if len(text) > 8000 else text
-            while retry_count < max_retries:
-                try:
-                    response = requests.post(
-                        f"{host}/api/embeddings",
-                        json={"model": model_name, "prompt": truncated_text},
-                        timeout=30,
+                if embedding is None:
+                    raise ValueError(f"No embedding returned for text {idx}")
+
+                return idx, embedding
+
+            except requests.exceptions.Timeout:
+                retry_count += 1
+                if retry_count >= max_retries:
+                    logger.warning(f"Timeout for text {idx} after {max_retries} retries")
+                    return idx, None
+
+            except Exception as e:
+                if retry_count >= max_retries - 1:
+                    logger.error(f"Failed to get embedding for text {idx}: {e}")
+                    return idx, None
+                retry_count += 1
+
+        return idx, None
+
+    # Determine if we should use concurrent processing
+    use_concurrent = (
+        len(texts) > 5 and not is_build
+    )  # Don't use concurrent in build mode to avoid overwhelming
+    max_workers = min(4, len(texts))  # Limit concurrent requests to avoid overwhelming Ollama
+
+    all_embeddings = [None] * len(texts)  # Pre-allocate list to maintain order
+    failed_indices = []
+
+    if use_concurrent:
+        logger.info(
+            f"Using concurrent processing with {max_workers} workers for {len(texts)} texts"
+        )
+
+        with ThreadPoolExecutor(max_workers=max_workers) as executor:
+            # Submit all tasks
+            future_to_idx = {
+                executor.submit(get_single_embedding, (text, idx)): idx
+                for idx, text in enumerate(texts)
+            }
+
+            # Add progress bar for concurrent processing
+            try:
+                if is_build or len(texts) > 10:
+                    from tqdm import tqdm
+
+                    futures_iterator = tqdm(
+                        as_completed(future_to_idx),
+                        total=len(texts),
+                        desc="Computing Ollama embeddings",
                     )
-                    response.raise_for_status()
-
-                    result = response.json()
-                    embedding = result.get("embedding")
-
-                    if embedding is None:
-                        raise ValueError(f"No embedding returned for text {i}")
-
-                    if not isinstance(embedding, list) or len(embedding) == 0:
-                        raise ValueError(f"Invalid embedding format for text {i}")
-
-                    all_embeddings.append(embedding)
-                    break
-
-                except requests.exceptions.Timeout:
-                    retry_count += 1
-                    if retry_count >= max_retries:
-                        logger.warning(f"Timeout for text {i} after {max_retries} retries")
-                        failed_indices.append(i)
-                        all_embeddings.append(None)
-                        break
+                else:
+                    futures_iterator = as_completed(future_to_idx)
+            except ImportError:
+                futures_iterator = as_completed(future_to_idx)
 
+            # Collect results as they complete
+            for future in futures_iterator:
+                try:
+                    idx, embedding = future.result()
+                    if embedding is not None:
+                        all_embeddings[idx] = embedding
+                    else:
+                        failed_indices.append(idx)
                 except Exception as e:
-                    retry_count += 1
-                    if retry_count >= max_retries:
-                        logger.error(f"Failed to get embedding for text {i}: {e}")
-                        failed_indices.append(i)
-                        all_embeddings.append(None)
-                        break
-        return all_embeddings, failed_indices
+                    idx = future_to_idx[future]
+                    logger.error(f"Exception for text {idx}: {e}")
+                    failed_indices.append(idx)
 
-    # Process texts in batches
-    all_embeddings = []
-    all_failed_indices = []
-
-    # Setup progress bar if needed
-    show_progress = is_build or len(texts) > 10
-    try:
-        if show_progress:
-            from tqdm import tqdm
-    except ImportError:
-        show_progress = False
-
-    # Process batches
-    num_batches = (len(texts) + batch_size - 1) // batch_size
-
-    if show_progress:
-        batch_iterator = tqdm(range(num_batches), desc="Computing Ollama embeddings")
     else:
-        batch_iterator = range(num_batches)
+        # Sequential processing with progress bar
+        show_progress = is_build or len(texts) > 10
 
-    for batch_idx in batch_iterator:
-        start_idx = batch_idx * batch_size
-        end_idx = min(start_idx + batch_size, len(texts))
-        batch_texts = texts[start_idx:end_idx]
+        try:
+            if show_progress:
+                from tqdm import tqdm
 
-        batch_embeddings, batch_failed = get_batch_embeddings(batch_texts)
+                iterator = tqdm(
+                    enumerate(texts), total=len(texts), desc="Computing Ollama embeddings"
+                )
+            else:
+                iterator = enumerate(texts)
+        except ImportError:
+            iterator = enumerate(texts)
 
-        # Adjust failed indices to global indices
-        global_failed = [start_idx + idx for idx in batch_failed]
-        all_failed_indices.extend(global_failed)
-        all_embeddings.extend(batch_embeddings)
+        for idx, text in iterator:
+            result_idx, embedding = get_single_embedding((text, idx))
+            if embedding is not None:
+                all_embeddings[idx] = embedding
+            else:
+                failed_indices.append(idx)
 
     # Handle failed embeddings
-    if all_failed_indices:
-        if len(all_failed_indices) == len(texts):
+    if failed_indices:
+        if len(failed_indices) == len(texts):
             raise RuntimeError("Failed to compute any embeddings")
 
-        logger.warning(
-            f"Failed to compute embeddings for {len(all_failed_indices)}/{len(texts)} texts"
-        )
+        logger.warning(f"Failed to compute embeddings for {len(failed_indices)}/{len(texts)} texts")
 
         # Use zero embeddings as fallback for failed ones
         valid_embedding = next((e for e in all_embeddings if e is not None), None)
         if valid_embedding:
             embedding_dim = len(valid_embedding)
-            for i, embedding in enumerate(all_embeddings):
-                if embedding is None:
-                    all_embeddings[i] = [0.0] * embedding_dim
+            for idx in failed_indices:
+                all_embeddings[idx] = [0.0] * embedding_dim
 
-    # Remove None values
+    # Remove None values and convert to numpy array
     all_embeddings = [e for e in all_embeddings if e is not None]
 
-    if not all_embeddings:
-        raise RuntimeError("No valid embeddings were computed")
-
-    # Validate embedding dimensions
-    expected_dim = len(all_embeddings[0])
-    inconsistent_dims = []
-    for i, embedding in enumerate(all_embeddings):
-        if len(embedding) != expected_dim:
-            inconsistent_dims.append((i, len(embedding)))
-
-    if inconsistent_dims:
-        error_msg = f"Ollama returned inconsistent embedding dimensions. Expected {expected_dim}, but got:\n"
-        for idx, dim in inconsistent_dims[:10]:  # Show first 10 inconsistent ones
-            error_msg += f"  - Text {idx}: {dim} dimensions\n"
-        if len(inconsistent_dims) > 10:
-            error_msg += f"  ... and {len(inconsistent_dims) - 10} more\n"
-        error_msg += f"\nThis is likely an Ollama API bug with model '{model_name}'. Please try:\n"
-        error_msg += "1. Restart Ollama service: 'ollama serve'\n"
-        error_msg += f"2. Re-pull the model: 'ollama pull {model_name}'\n"
-        error_msg += (
-            "3. Use sentence-transformers instead: --embedding-mode sentence-transformers\n"
-        )
-        error_msg += "4. Report this issue to Ollama: https://github.com/ollama/ollama/issues"
-        raise ValueError(error_msg)
-
     # Convert to numpy array and normalize
     embeddings = np.array(all_embeddings, dtype=np.float32)
 

packages/leann-core/src/leann/embedding_server_manager.py

@@ -1,6 +1,7 @@
 import atexit
 import logging
 import os
+import signal
 import socket
 import subprocess
 import sys
@@ -8,7 +9,7 @@ import time
 from pathlib import Path
 from typing import Optional
 
-# Lightweight, self-contained server manager with no cross-process inspection
+import psutil
 
 # Set up logging based on environment variable
 LOG_LEVEL = os.getenv("LEANN_LOG_LEVEL", "WARNING").upper()
@@ -43,7 +44,130 @@ def _check_port(port: int) -> bool:
         return s.connect_ex(("localhost", port)) == 0
 
 
-# Note: All cross-process scanning helpers removed for simplicity
+def _check_process_matches_config(
+    port: int, expected_model: str, expected_passages_file: str
+) -> bool:
+    """
+    Check if the process using the port matches our expected model and passages file.
+    Returns True if matches, False otherwise.
+    """
+    try:
+        for proc in psutil.process_iter(["pid", "cmdline"]):
+            if not _is_process_listening_on_port(proc, port):
+                continue
+
+            cmdline = proc.info["cmdline"]
+            if not cmdline:
+                continue
+
+            return _check_cmdline_matches_config(
+                cmdline, port, expected_model, expected_passages_file
+            )
+
+        logger.debug(f"No process found listening on port {port}")
+        return False
+
+    except Exception as e:
+        logger.warning(f"Could not check process on port {port}: {e}")
+        return False
+
+
+def _is_process_listening_on_port(proc, port: int) -> bool:
+    """Check if a process is listening on the given port."""
+    try:
+        connections = proc.net_connections()
+        for conn in connections:
+            if conn.laddr.port == port and conn.status == psutil.CONN_LISTEN:
+                return True
+        return False
+    except (psutil.NoSuchProcess, psutil.AccessDenied, psutil.ZombieProcess):
+        return False
+
+
+def _check_cmdline_matches_config(
+    cmdline: list, port: int, expected_model: str, expected_passages_file: str
+) -> bool:
+    """Check if command line matches our expected configuration."""
+    cmdline_str = " ".join(cmdline)
+    logger.debug(f"Found process on port {port}: {cmdline_str}")
+
+    # Check if it's our embedding server
+    is_embedding_server = any(
+        server_type in cmdline_str
+        for server_type in [
+            "embedding_server",
+            "leann_backend_diskann.embedding_server",
+            "leann_backend_hnsw.hnsw_embedding_server",
+        ]
+    )
+
+    if not is_embedding_server:
+        logger.debug(f"Process on port {port} is not our embedding server")
+        return False
+
+    # Check model name
+    model_matches = _check_model_in_cmdline(cmdline, expected_model)
+
+    # Check passages file if provided
+    passages_matches = _check_passages_in_cmdline(cmdline, expected_passages_file)
+
+    result = model_matches and passages_matches
+    logger.debug(
+        f"model_matches: {model_matches}, passages_matches: {passages_matches}, overall: {result}"
+    )
+    return result
+
+
+def _check_model_in_cmdline(cmdline: list, expected_model: str) -> bool:
+    """Check if the command line contains the expected model."""
+    if "--model-name" not in cmdline:
+        return False
+
+    model_idx = cmdline.index("--model-name")
+    if model_idx + 1 >= len(cmdline):
+        return False
+
+    actual_model = cmdline[model_idx + 1]
+    return actual_model == expected_model
+
+
+def _check_passages_in_cmdline(cmdline: list, expected_passages_file: str) -> bool:
+    """Check if the command line contains the expected passages file."""
+    if "--passages-file" not in cmdline:
+        return False  # Expected but not found
+
+    passages_idx = cmdline.index("--passages-file")
+    if passages_idx + 1 >= len(cmdline):
+        return False
+
+    actual_passages = cmdline[passages_idx + 1]
+    expected_path = Path(expected_passages_file).resolve()
+    actual_path = Path(actual_passages).resolve()
+    return actual_path == expected_path
+
+
+def _find_compatible_port_or_next_available(
+    start_port: int, model_name: str, passages_file: str, max_attempts: int = 100
+) -> tuple[int, bool]:
+    """
+    Find a port that either has a compatible server or is available.
+    Returns (port, is_compatible) where is_compatible indicates if we found a matching server.
+    """
+    for port in range(start_port, start_port + max_attempts):
+        if not _check_port(port):
+            # Port is available
+            return port, False
+
+        # Port is in use, check if it's compatible
+        if _check_process_matches_config(port, model_name, passages_file):
+            logger.info(f"Found compatible server on port {port}")
+            return port, True
+        else:
+            logger.info(f"Port {port} has incompatible server, trying next port...")
+
+    raise RuntimeError(
+        f"Could not find compatible or available port in range {start_port}-{start_port + max_attempts}"
+    )
 
 
 class EmbeddingServerManager:
@@ -62,16 +186,7 @@ class EmbeddingServerManager:
         self.backend_module_name = backend_module_name
         self.server_process: Optional[subprocess.Popen] = None
         self.server_port: Optional[int] = None
-        # Track last-started config for in-process reuse only
-        self._server_config: Optional[dict] = None
         self._atexit_registered = False
-        # Also register a weakref finalizer to ensure cleanup when manager is GC'ed
-        try:
-            import weakref
-
-            self._finalizer = weakref.finalize(self, self._finalize_process)
-        except Exception:
-            self._finalizer = None
 
     def start_server(
         self,
@@ -81,24 +196,26 @@ class EmbeddingServerManager:
         **kwargs,
     ) -> tuple[bool, int]:
         """Start the embedding server."""
-        # passages_file may be present in kwargs for server CLI, but we don't need it here
+        passages_file = kwargs.get("passages_file")
 
-        # If this manager already has a live server, just reuse it
-        if self.server_process and self.server_process.poll() is None and self.server_port:
-            logger.info("Reusing in-process server")
-            return True, self.server_port
+        # Check if we have a compatible server already running
+        if self._has_compatible_running_server(model_name, passages_file):
+            logger.info("Found compatible running server!")
+            return True, port
 
         # For Colab environment, use a different strategy
         if _is_colab_environment():
             logger.info("Detected Colab environment, using alternative startup strategy")
             return self._start_server_colab(port, model_name, embedding_mode, **kwargs)
 
-        # Always pick a fresh available port
-        try:
-            actual_port = _get_available_port(port)
-        except RuntimeError:
-            logger.error("No available ports found")
-            return False, port
+        # Find a compatible port or next available
+        actual_port, is_compatible = _find_compatible_port_or_next_available(
+            port, model_name, passages_file
+        )
+
+        if is_compatible:
+            logger.info(f"Found compatible server on port {actual_port}")
+            return True, actual_port
 
         # Start a new server
         return self._start_new_server(actual_port, model_name, embedding_mode, **kwargs)
@@ -131,7 +248,17 @@ class EmbeddingServerManager:
             logger.error(f"Failed to start embedding server in Colab: {e}")
             return False, actual_port
 
-    # Note: No compatibility check needed; manager is per-searcher and configs are stable per instance
+    def _has_compatible_running_server(self, model_name: str, passages_file: str) -> bool:
+        """Check if we have a compatible running server."""
+        if not (self.server_process and self.server_process.poll() is None and self.server_port):
+            return False
+
+        if _check_process_matches_config(self.server_port, model_name, passages_file):
+            logger.info(f"Existing server process (PID {self.server_process.pid}) is compatible")
+            return True
+
+        logger.info("Existing server process is incompatible. Should start a new server.")
+        return False
 
     def _start_new_server(
         self, port: int, model_name: str, embedding_mode: str, **kwargs
@@ -178,61 +305,23 @@ class EmbeddingServerManager:
         project_root = Path(__file__).parent.parent.parent.parent.parent
         logger.info(f"Command: {' '.join(command)}")
 
-        # In CI environment, redirect stdout to avoid buffer deadlock but keep stderr for debugging
-        # Embedding servers use many print statements that can fill stdout buffers
-        is_ci = os.environ.get("CI") == "true"
-        if is_ci:
-            stdout_target = subprocess.DEVNULL
-            stderr_target = None  # Keep stderr for error debugging in CI
-            logger.info(
-                "CI environment detected, redirecting embedding server stdout to DEVNULL, keeping stderr"
-            )
-        else:
-            stdout_target = None  # Direct to console for visible logs
-            stderr_target = None  # Direct to console for visible logs
-
-        # Start embedding server subprocess
+        # Let server output go directly to console
+        # The server will respect LEANN_LOG_LEVEL environment variable
         self.server_process = subprocess.Popen(
             command,
             cwd=project_root,
-            stdout=stdout_target,
-            stderr=stderr_target,
+            stdout=None,  # Direct to console
+            stderr=None,  # Direct to console
+            start_new_session=True,  # Create new process group for better cleanup
         )
         self.server_port = port
-        # Record config for in-process reuse
-        try:
-            self._server_config = {
-                "model_name": command[command.index("--model-name") + 1]
-                if "--model-name" in command
-                else "",
-                "passages_file": command[command.index("--passages-file") + 1]
-                if "--passages-file" in command
-                else "",
-                "embedding_mode": command[command.index("--embedding-mode") + 1]
-                if "--embedding-mode" in command
-                else "sentence-transformers",
-            }
-        except Exception:
-            self._server_config = {
-                "model_name": "",
-                "passages_file": "",
-                "embedding_mode": "sentence-transformers",
-            }
         logger.info(f"Server process started with PID: {self.server_process.pid}")
 
         # Register atexit callback only when we actually start a process
         if not self._atexit_registered:
-            # Always attempt best-effort finalize at interpreter exit
-            atexit.register(self._finalize_process)
+            # Use a lambda to avoid issues with bound methods
+            atexit.register(lambda: self.stop_server() if self.server_process else None)
             self._atexit_registered = True
-        # Touch finalizer so it knows there is a live process
-        if getattr(self, "_finalizer", None) is not None and not self._finalizer.alive:
-            try:
-                import weakref
-
-                self._finalizer = weakref.finalize(self, self._finalize_process)
-            except Exception:
-                pass
 
     def _wait_for_server_ready(self, port: int) -> tuple[bool, int]:
         """Wait for the server to be ready."""
@@ -257,28 +346,37 @@ class EmbeddingServerManager:
         if not self.server_process:
             return
 
-        if self.server_process and self.server_process.poll() is not None:
+        if self.server_process.poll() is not None:
             # Process already terminated
             self.server_process = None
-            self.server_port = None
-            self._server_config = None
             return
 
         logger.info(
             f"Terminating server process (PID: {self.server_process.pid}) for backend {self.backend_module_name}..."
         )
 
-        # Use simple termination - our improved server shutdown should handle this properly
-        self.server_process.terminate()
+        # Try terminating the whole process group first
+        try:
+            pgid = os.getpgid(self.server_process.pid)
+            os.killpg(pgid, signal.SIGTERM)
+        except Exception:
+            # Fallback to terminating just the process
+            self.server_process.terminate()
 
         try:
-            self.server_process.wait(timeout=5)  # Give more time for graceful shutdown
-            logger.info(f"Server process {self.server_process.pid} terminated gracefully.")
+            self.server_process.wait(timeout=3)
+            logger.info(f"Server process {self.server_process.pid} terminated.")
         except subprocess.TimeoutExpired:
             logger.warning(
-                f"Server process {self.server_process.pid} did not terminate within 5 seconds, force killing..."
+                f"Server process {self.server_process.pid} did not terminate gracefully within 3 seconds, killing it."
             )
-            self.server_process.kill()
+            # Try killing the whole process group
+            try:
+                pgid = os.getpgid(self.server_process.pid)
+                os.killpg(pgid, signal.SIGKILL)
+            except Exception:
+                # Fallback to killing just the process
+                self.server_process.kill()
             try:
                 self.server_process.wait(timeout=2)
                 logger.info(f"Server process {self.server_process.pid} killed successfully.")
@@ -286,33 +384,20 @@ class EmbeddingServerManager:
                 logger.error(
                     f"Failed to kill server process {self.server_process.pid} - it may be hung"
                 )
+                # Don't hang indefinitely
 
-        # Clean up process resources with timeout to avoid CI hang
+        # Clean up process resources to prevent resource tracker warnings
         try:
-            # Use shorter timeout in CI environments
-            is_ci = os.environ.get("CI") == "true"
-            timeout = 3 if is_ci else 10
-            self.server_process.wait(timeout=timeout)
-            logger.info(f"Server process {self.server_process.pid} cleanup completed")
+            self.server_process.wait(timeout=1)  # Give it one final chance with timeout
         except subprocess.TimeoutExpired:
-            logger.warning(f"Process cleanup timeout after {timeout}s, proceeding anyway")
-        except Exception as e:
-            logger.warning(f"Error during process cleanup: {e}")
-        finally:
-            self.server_process = None
-            self.server_port = None
-            self._server_config = None
-
-    def _finalize_process(self) -> None:
-        """Best-effort cleanup used by weakref.finalize/atexit."""
-        try:
-            self.stop_server()
+            logger.warning(
+                f"Process {self.server_process.pid} still hanging after all kill attempts"
+            )
+            # Don't wait indefinitely - just abandon it
         except Exception:
             pass
 
-    def _adopt_existing_server(self, *args, **kwargs) -> None:
-        # Removed: cross-process adoption no longer supported
-        return
+        self.server_process = None
 
     def _launch_server_process_colab(self, command: list, port: int) -> None:
         """Launch the server process with Colab-specific settings."""
@@ -328,16 +413,10 @@ class EmbeddingServerManager:
         self.server_port = port
         logger.info(f"Colab server process started with PID: {self.server_process.pid}")
 
-        # Register atexit callback (unified)
+        # Register atexit callback
         if not self._atexit_registered:
-            atexit.register(self._finalize_process)
+            atexit.register(lambda: self.stop_server() if self.server_process else None)
             self._atexit_registered = True
-        # Record config for in-process reuse is best-effort in Colab mode
-        self._server_config = {
-            "model_name": "",
-            "passages_file": "",
-            "embedding_mode": "sentence-transformers",
-        }
 
     def _wait_for_server_ready_colab(self, port: int) -> tuple[bool, int]:
         """Wait for the server to be ready with Colab-specific timeout."""

packages/leann-core/src/leann/mcp.py

@@ -25,61 +25,32 @@ def handle_request(request):
                 "tools": [
                     {
                         "name": "leann_search",
-                        "description": """🔍 Search code using natural language - like having a coding assistant who knows your entire codebase!
-
-🎯 **Perfect for**:
-- "How does authentication work?" → finds auth-related code
-- "Error handling patterns" → locates try-catch blocks and error logic
-- "Database connection setup" → finds DB initialization code
-- "API endpoint definitions" → locates route handlers
-- "Configuration management" → finds config files and usage
-
-💡 **Pro tip**: Use this before making any changes to understand existing patterns and conventions.""",
+                        "description": "Search LEANN index",
                         "inputSchema": {
                             "type": "object",
                             "properties": {
-                                "index_name": {
-                                    "type": "string",
-                                    "description": "Name of the LEANN index to search. Use 'leann_list' first to see available indexes.",
-                                },
-                                "query": {
-                                    "type": "string",
-                                    "description": "Search query - can be natural language (e.g., 'how to handle errors') or technical terms (e.g., 'async function definition')",
-                                },
-                                "top_k": {
-                                    "type": "integer",
-                                    "default": 5,
-                                    "minimum": 1,
-                                    "maximum": 20,
-                                    "description": "Number of search results to return. Use 5-10 for focused results, 15-20 for comprehensive exploration.",
-                                },
-                                "complexity": {
-                                    "type": "integer",
-                                    "default": 32,
-                                    "minimum": 16,
-                                    "maximum": 128,
-                                    "description": "Search complexity level. Use 16-32 for fast searches (recommended), 64+ for higher precision when needed.",
-                                },
+                                "index_name": {"type": "string"},
+                                "query": {"type": "string"},
+                                "top_k": {"type": "integer", "default": 5},
                             },
                             "required": ["index_name", "query"],
                         },
                     },
                     {
-                        "name": "leann_status",
-                        "description": "📊 Check the health and stats of your code indexes - like a medical checkup for your codebase knowledge!",
+                        "name": "leann_ask",
+                        "description": "Ask question using LEANN RAG",
                         "inputSchema": {
                             "type": "object",
                             "properties": {
-                                "index_name": {
-                                    "type": "string",
-                                    "description": "Optional: Name of specific index to check. If not provided, shows status of all indexes.",
-                                }
+                                "index_name": {"type": "string"},
+                                "question": {"type": "string"},
                             },
+                            "required": ["index_name", "question"],
                         },
                     },
                     {
                         "name": "leann_list",
-                        "description": "📋 Show all your indexed codebases - your personal code library! Use this to see what's available for search.",
+                        "description": "List all LEANN indexes",
                         "inputSchema": {"type": "object", "properties": {}},
                     },
                 ]
@@ -92,40 +63,19 @@ def handle_request(request):
 
         try:
             if tool_name == "leann_search":
-                # Validate required parameters
-                if not args.get("index_name") or not args.get("query"):
-                    return {
-                        "jsonrpc": "2.0",
-                        "id": request.get("id"),
-                        "result": {
-                            "content": [
-                                {
-                                    "type": "text",
-                                    "text": "Error: Both index_name and query are required",
-                                }
-                            ]
-                        },
-                    }
-
-                # Build simplified command
                 cmd = [
                     "leann",
                     "search",
                     args["index_name"],
                     args["query"],
+                    "--recompute-embeddings",
                     f"--top-k={args.get('top_k', 5)}",
-                    f"--complexity={args.get('complexity', 32)}",
                 ]
                 result = subprocess.run(cmd, capture_output=True, text=True)
 
-            elif tool_name == "leann_status":
-                if args.get("index_name"):
-                    # Check specific index status - for now, we'll use leann list and filter
-                    result = subprocess.run(["leann", "list"], capture_output=True, text=True)
-                    # We could enhance this to show more detailed status per index
-                else:
-                    # Show all indexes status
-                    result = subprocess.run(["leann", "list"], capture_output=True, text=True)
+            elif tool_name == "leann_ask":
+                cmd = f'echo "{args["question"]}" | leann ask {args["index_name"]} --recompute-embeddings --llm ollama --model qwen3:8b'
+                result = subprocess.run(cmd, shell=True, capture_output=True, text=True)
 
             elif tool_name == "leann_list":
                 result = subprocess.run(["leann", "list"], capture_output=True, text=True)

packages/leann-core/src/leann/searcher_base.py

@@ -132,10 +132,15 @@ class BaseSearcher(LeannBackendSearcherInterface, ABC):
         import msgpack
         import zmq
 
+        context = None
+        socket = None
         try:
             context = zmq.Context()
             socket = context.socket(zmq.REQ)
-            socket.setsockopt(zmq.RCVTIMEO, 30000)  # 30 second timeout
+            socket.setsockopt(zmq.LINGER, 0)  # Don't block on close
+            socket.setsockopt(zmq.RCVTIMEO, 300000)
+            socket.setsockopt(zmq.SNDTIMEO, 300000)
+            socket.setsockopt(zmq.IMMEDIATE, 1)
             socket.connect(f"tcp://localhost:{zmq_port}")
 
             # Send embedding request
@@ -147,9 +152,6 @@ class BaseSearcher(LeannBackendSearcherInterface, ABC):
             response_bytes = socket.recv()
             response = msgpack.unpackb(response_bytes)
 
-            socket.close()
-            context.term()
-
             # Convert response to numpy array
             if isinstance(response, list) and len(response) > 0:
                 return np.array(response, dtype=np.float32)
@@ -158,6 +160,10 @@ class BaseSearcher(LeannBackendSearcherInterface, ABC):
 
         except Exception as e:
             raise RuntimeError(f"Failed to compute embeddings via server: {e}")
+        finally:
+            if socket:
+                socket.close()
+            # Don't call context.term() - this was causing hangs
 
     @abstractmethod
     def search(
@@ -191,7 +197,15 @@ class BaseSearcher(LeannBackendSearcherInterface, ABC):
         """
         pass
 
-    def __del__(self):
-        """Ensures the embedding server is stopped when the searcher is destroyed."""
+    def cleanup(self):
+        """Cleanup resources including embedding server."""
         if hasattr(self, "embedding_server_manager"):
             self.embedding_server_manager.stop_server()
+
+    def __del__(self):
+        """Ensures resources are cleaned up when the searcher is destroyed."""
+        try:
+            self.cleanup()
+        except Exception:
+            # Ignore errors during destruction
+            pass

packages/leann-mcp/README.md

@@ -45,42 +45,6 @@ leann build my-project --docs ./
 claude
 ```
 
-## 🚀 Advanced Usage Examples
-
-### Index Entire Git Repository
-```bash
-# Index all tracked files in your git repository, note right now we will skip submodules, but we can add it back easily if you want
-leann build my-repo --docs $(git ls-files) --embedding-mode sentence-transformers --embedding-model all-MiniLM-L6-v2 --backend hnsw
-
-# Index only specific file types from git
-leann build my-python-code --docs $(git ls-files "*.py") --embedding-mode sentence-transformers --embedding-model all-MiniLM-L6-v2 --backend hnsw
-```
-
-### Multiple Directories and Files
-```bash
-# Index multiple directories
-leann build my-codebase --docs ./src ./tests ./docs ./config --embedding-mode sentence-transformers --embedding-model all-MiniLM-L6-v2 --backend hnsw
-
-# Mix files and directories
-leann build my-project --docs ./README.md ./src/ ./package.json ./docs/ --embedding-mode sentence-transformers --embedding-model all-MiniLM-L6-v2 --backend hnsw
-
-# Specific files only
-leann build my-configs --docs ./tsconfig.json ./package.json ./webpack.config.js --embedding-mode sentence-transformers --embedding-model all-MiniLM-L6-v2 --backend hnsw
-```
-
-### Advanced Git Integration
-```bash
-# Index recently modified files
-leann build recent-changes --docs $(git diff --name-only HEAD~10..HEAD) --embedding-mode sentence-transformers --embedding-model all-MiniLM-L6-v2 --backend hnsw
-
-# Index files matching pattern
-leann build frontend --docs $(git ls-files "*.tsx" "*.ts" "*.jsx" "*.js") --embedding-mode sentence-transformers --embedding-model all-MiniLM-L6-v2 --backend hnsw
-
-# Index documentation and config files
-leann build docs-and-configs --docs $(git ls-files "*.md" "*.yml" "*.yaml" "*.json" "*.toml") --embedding-mode sentence-transformers --embedding-model all-MiniLM-L6-v2 --backend hnsw
-```
-
-
 **Try this in Claude Code:**
 ```
 Help me understand this codebase. List available indexes and search for authentication patterns.

packages/leann/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "leann"
-version = "0.2.8"
+version = "0.2.5"
 description = "LEANN - The smallest vector index in the world. RAG Everything with LEANN!"
 readme = "README.md"
 requires-python = ">=3.9"

pyproject.toml

@@ -40,13 +40,10 @@ dependencies = [
     # Other dependencies
     "ipykernel==6.29.5",
     "msgpack>=1.1.1",
-    "mlx>=0.26.3; sys_platform == 'darwin' and platform_machine == 'arm64'",
-    "mlx-lm>=0.26.0; sys_platform == 'darwin' and platform_machine == 'arm64'",
+    "mlx>=0.26.3; sys_platform == 'darwin'",
+    "mlx-lm>=0.26.0; sys_platform == 'darwin'",
     "psutil>=5.8.0",
     "pybind11>=3.0.0",
-    "pathspec>=0.12.1",
-    "nbconvert>=7.16.6",
-    "gitignore-parser>=0.1.12",
 ]
 
 [project.optional-dependencies]
@@ -63,7 +60,7 @@ dev = [
 
 test = [
     "pytest>=7.0",
-    "pytest-timeout>=2.0",
+    "pytest-timeout>=2.0",  # Simple timeout protection for CI
     "llama-index-core>=0.12.0",
     "llama-index-readers-file>=0.4.0",
     "python-dotenv>=1.0.0",
@@ -155,7 +152,7 @@ markers = [
     "slow: marks tests as slow (deselect with '-m \"not slow\"')",
     "openai: marks tests that require OpenAI API key",
 ]
-timeout = 300  # Reduced from 600s (10min) to 300s (5min) for CI safety
+timeout = 300  # Simple timeout for CI safety (5 minutes)
 addopts = [
     "-v",
     "--tb=short",

tests/conftest.py

@@ -0,0 +1,41 @@
+"""Pytest configuration and fixtures for LEANN tests."""
+
+import os
+
+import pytest
+
+
+@pytest.fixture(autouse=True)
+def test_environment():
+    """Set up test environment variables."""
+    # Mark as test environment to skip memory-intensive operations
+    os.environ["CI"] = "true"
+    yield
+
+
+@pytest.fixture(scope="session", autouse=True)
+def cleanup_session():
+    """Session-level cleanup to ensure no hanging processes."""
+    yield
+
+    # Basic cleanup after all tests
+    try:
+        import os
+
+        import psutil
+
+        current_process = psutil.Process(os.getpid())
+        children = current_process.children(recursive=True)
+
+        for child in children:
+            try:
+                child.terminate()
+            except psutil.NoSuchProcess:
+                pass
+
+        # Give them time to terminate gracefully
+        psutil.wait_procs(children, timeout=3)
+
+    except Exception:
+        # Don't fail tests due to cleanup errors
+        pass

tests/test_basic.py

@@ -64,9 +64,6 @@ def test_backend_basic(backend_name):
         assert isinstance(results[0], SearchResult)
         assert "topic 2" in results[0].text or "document" in results[0].text
 
-        # Ensure cleanup to avoid hanging background servers
-        searcher.cleanup()
-
 
 @pytest.mark.skipif(
     os.environ.get("CI") == "true", reason="Skip model tests in CI to avoid MPS memory issues"
@@ -93,5 +90,3 @@ def test_large_index():
         searcher = LeannSearcher(index_path)
         results = searcher.search(["word10 word20"], top_k=10)
         assert len(results[0]) == 10
-        # Cleanup
-        searcher.cleanup()

tests/test_document_rag.py

@@ -58,9 +58,6 @@ def test_document_rag_simulated(test_data_dir):
 
 
 @pytest.mark.skipif(not os.environ.get("OPENAI_API_KEY"), reason="OpenAI API key not available")
-@pytest.mark.skipif(
-    os.environ.get("CI") == "true", reason="Skip OpenAI tests in CI to avoid API costs"
-)
 def test_document_rag_openai(test_data_dir):
     """Test document_rag with OpenAI embeddings."""
     with tempfile.TemporaryDirectory() as temp_dir:

tests/test_readme_examples.py

@@ -16,9 +16,6 @@ def test_readme_basic_example(backend_name):
     # Skip on macOS CI due to MPS environment issues with all-MiniLM-L6-v2
     if os.environ.get("CI") == "true" and platform.system() == "Darwin":
         pytest.skip("Skipping on macOS CI due to MPS environment issues with all-MiniLM-L6-v2")
-    # Skip DiskANN on CI (Linux runners) due to C++ extension memory/hardware constraints
-    if os.environ.get("CI") == "true" and backend_name == "diskann":
-        pytest.skip("Skip DiskANN tests in CI due to resource constraints and instability")
 
     # This is the exact code from README (with smaller model for CI)
     from leann import LeannBuilder, LeannChat, LeannSearcher
@@ -62,9 +59,6 @@ def test_readme_basic_example(backend_name):
         # The second text about banana-crocodile should be more relevant
         assert "banana" in results[0].text or "crocodile" in results[0].text
 
-        # Ensure we cleanup background embedding server
-        searcher.cleanup()
-
         # Chat with your data (using simulated LLM to avoid external dependencies)
         chat = LeannChat(INDEX_PATH, llm_config={"type": "simulated"})
         response = chat.ask("How much storage does LEANN save?", top_k=1)
@@ -72,8 +66,6 @@ def test_readme_basic_example(backend_name):
         # Verify chat works
         assert isinstance(response, str)
         assert len(response) > 0
-        # Cleanup chat resources
-        chat.cleanup()
 
 
 def test_readme_imports():