spatial-memory-mcp 1.6.1__py3-none-any.whl

spatial_memory/services/utility.py

@@ -0,0 +1,409 @@
+"""Utility service for database management operations.
+
+This service provides the application layer for utility operations:
+- stats: Get database statistics and health metrics
+- namespaces: List namespaces with memory counts
+- delete_namespace: Delete all memories in a namespace
+- rename_namespace: Rename namespace (move all memories)
+- hybrid_recall: Combined vector + FTS search
+
+The service uses dependency injection for repository and embedding services.
+"""
+
+from __future__ import annotations
+
+import logging
+from typing import TYPE_CHECKING
+
+from spatial_memory.core.errors import (
+    NamespaceNotFoundError,
+    NamespaceOperationError,
+    ValidationError,
+)
+from spatial_memory.core.models import (
+    DeleteNamespaceResult,
+    HybridMemoryMatch,
+    HybridRecallResult,
+    IndexInfo,
+    NamespaceInfo,
+    NamespacesResult,
+    RenameNamespaceResult,
+    StatsResult,
+    UtilityConfig,
+)
+from spatial_memory.core.validation import validate_namespace
+
+logger = logging.getLogger(__name__)
+
+if TYPE_CHECKING:
+    from spatial_memory.ports.repositories import (
+        EmbeddingServiceProtocol,
+        MemoryRepositoryProtocol,
+    )
+
+
+class UtilityService:
+    """Service for database utility operations.
+
+    Uses Clean Architecture - depends on protocol interfaces, not implementations.
+    Provides database statistics, namespace management, and hybrid search.
+    """
+
+    def __init__(
+        self,
+        repository: MemoryRepositoryProtocol,
+        embeddings: EmbeddingServiceProtocol,
+        config: UtilityConfig | None = None,
+    ) -> None:
+        """Initialize the utility service.
+
+        Args:
+            repository: Repository for memory storage.
+            embeddings: Service for generating embeddings.
+            config: Optional configuration (uses defaults if not provided).
+        """
+        self._repo = repository
+        self._embeddings = embeddings
+        self._config = config or UtilityConfig()
+
+    def stats(
+        self,
+        namespace: str | None = None,
+        include_index_details: bool = True,
+    ) -> StatsResult:
+        """Get comprehensive database statistics.
+
+        Retrieves statistics about the memory database including total counts,
+        storage size, index information, and health metrics.
+
+        Args:
+            namespace: Filter statistics to a specific namespace.
+                If None, returns statistics for all namespaces.
+            include_index_details: Include detailed index statistics.
+
+        Returns:
+            StatsResult with database information.
+
+        Raises:
+            ValidationError: If namespace is invalid.
+            NamespaceOperationError: If stats retrieval fails.
+        """
+        # Validate namespace if provided
+        if namespace is not None:
+            namespace = validate_namespace(namespace)
+
+        try:
+            # Get stats from repository
+            raw_stats = self._repo.get_stats(namespace=namespace)
+
+            # Transform indices to IndexInfo objects
+            indices: list[IndexInfo] = []
+            if include_index_details:
+                for idx_data in raw_stats.get("indices", []):
+                    indices.append(
+                        IndexInfo(
+                            name=idx_data.get("name", "unknown"),
+                            index_type=idx_data.get("index_type", "unknown"),
+                            column=idx_data.get("column", "unknown"),
+                            num_indexed_rows=idx_data.get("num_indexed_rows", 0),
+                            status=idx_data.get("status", "unknown"),
+                        )
+                    )
+
+            # Get namespace breakdown from raw_stats
+            memories_by_namespace: dict[str, int] = raw_stats.get("namespaces", {})
+
+            # Calculate estimated vector bytes (dims * 4 bytes per float * num memories)
+            total_memories = raw_stats.get("total_memories", 0)
+            embedding_dims = self._embeddings.dimensions
+            estimated_vector_bytes = total_memories * embedding_dims * 4
+
+            return StatsResult(
+                total_memories=total_memories,
+                memories_by_namespace=memories_by_namespace,
+                storage_bytes=raw_stats.get("storage_bytes", 0),
+                storage_mb=raw_stats.get("storage_mb", 0.0),
+                estimated_vector_bytes=estimated_vector_bytes,
+                has_vector_index=raw_stats.get("has_vector_index", False),
+                has_fts_index=raw_stats.get("has_fts_index", False),
+                indices=indices,
+                num_fragments=raw_stats.get("num_fragments", 0),
+                needs_compaction=raw_stats.get("needs_compaction", False),
+                table_version=raw_stats.get("table_version", 1),
+                oldest_memory_date=raw_stats.get("oldest_memory_date"),
+                newest_memory_date=raw_stats.get("newest_memory_date"),
+                avg_content_length=raw_stats.get("avg_content_length"),
+            )
+
+        except ValidationError:
+            raise
+        except Exception as e:
+            raise NamespaceOperationError(f"Failed to get stats: {e}") from e
+
+    def namespaces(
+        self,
+        include_stats: bool = True,
+    ) -> NamespacesResult:
+        """List all namespaces with optional statistics.
+
+        Args:
+            include_stats: Include memory counts and date ranges per namespace.
+
+        Returns:
+            NamespacesResult with namespace list and totals.
+
+        Raises:
+            NamespaceOperationError: If namespace listing fails.
+        """
+        try:
+            # Get list of namespaces from repository
+            namespace_names = self._repo.get_namespaces()
+
+            namespaces: list[NamespaceInfo] = []
+            total_memories = 0
+
+            for ns_name in namespace_names:
+                if include_stats:
+                    # Get detailed stats for each namespace
+                    try:
+                        ns_stats = self._repo.get_namespace_stats(ns_name)
+                        memory_count = ns_stats.get("memory_count", 0)
+                        namespaces.append(
+                            NamespaceInfo(
+                                name=ns_name,
+                                memory_count=memory_count,
+                                oldest_memory=ns_stats.get("oldest_memory"),
+                                newest_memory=ns_stats.get("newest_memory"),
+                            )
+                        )
+                        total_memories += memory_count
+                    except Exception as e:
+                        logger.warning(f"Failed to get stats for namespace {ns_name}: {e}")
+                        # Still include namespace with zero count
+                        namespaces.append(
+                            NamespaceInfo(
+                                name=ns_name,
+                                memory_count=0,
+                            )
+                        )
+                else:
+                    # Just include name without stats
+                    namespaces.append(
+                        NamespaceInfo(
+                            name=ns_name,
+                            memory_count=0,
+                        )
+                    )
+
+            # If we didn't include stats, get total from count
+            if not include_stats:
+                total_memories = self._repo.count()
+
+            return NamespacesResult(
+                namespaces=namespaces,
+                total_namespaces=len(namespaces),
+                total_memories=total_memories,
+            )
+
+        except Exception as e:
+            raise NamespaceOperationError(f"Failed to list namespaces: {e}") from e
+
+    def delete_namespace(
+        self,
+        namespace: str,
+        confirm: bool = False,
+        dry_run: bool = True,
+    ) -> DeleteNamespaceResult:
+        """Delete all memories in a namespace.
+
+        DESTRUCTIVE OPERATION - requires explicit confirmation unless dry_run.
+
+        Args:
+            namespace: Namespace to delete.
+            confirm: Set to True to confirm deletion (required for non-dry-run).
+            dry_run: Preview deletion without executing (default True).
+
+        Returns:
+            DeleteNamespaceResult with deletion details.
+
+        Raises:
+            ValidationError: If namespace is invalid or confirmation missing.
+            NamespaceOperationError: If deletion fails.
+        """
+        # Validate namespace
+        namespace = validate_namespace(namespace)
+
+        try:
+            # Get count of memories that would be deleted
+            memory_count = self._repo.count(namespace=namespace)
+
+            # If dry run, just return preview
+            if dry_run:
+                return DeleteNamespaceResult(
+                    namespace=namespace,
+                    memories_deleted=memory_count,
+                    success=True,
+                    message=f"DRY RUN: Would delete {memory_count} memories from namespace '{namespace}'",
+                    dry_run=True,
+                )
+
+            # Not a dry run - require confirmation
+            if not confirm:
+                raise ValidationError(
+                    f"Deletion of {memory_count} memories in namespace '{namespace}' "
+                    "requires confirm=True. Use dry_run=True to preview first."
+                )
+
+            # Perform actual deletion
+            deleted_count = self._repo.delete_by_namespace(namespace)
+
+            return DeleteNamespaceResult(
+                namespace=namespace,
+                memories_deleted=deleted_count,
+                success=True,
+                message=f"Successfully deleted {deleted_count} memories from namespace '{namespace}'",
+                dry_run=False,
+            )
+
+        except ValidationError:
+            raise
+        except Exception as e:
+            raise NamespaceOperationError(
+                f"Failed to delete namespace '{namespace}': {e}"
+            ) from e
+
+    def rename_namespace(
+        self,
+        old_namespace: str,
+        new_namespace: str,
+    ) -> RenameNamespaceResult:
+        """Rename all memories from one namespace to another.
+
+        Atomically updates the namespace field for all memories belonging
+        to the source namespace.
+
+        Args:
+            old_namespace: The current namespace name (source).
+            new_namespace: The new namespace name (target).
+
+        Returns:
+            RenameNamespaceResult with rename details.
+
+        Raises:
+            ValidationError: If namespace names are invalid or the same.
+            NamespaceNotFoundError: If old_namespace doesn't exist.
+            NamespaceOperationError: If rename fails.
+        """
+        # Validate both namespaces
+        old_namespace = validate_namespace(old_namespace)
+        new_namespace = validate_namespace(new_namespace)
+
+        # Check they are different
+        if old_namespace == new_namespace:
+            raise ValidationError(
+                f"Cannot rename namespace to same name: '{old_namespace}'"
+            )
+
+        try:
+            # Call repository to perform rename
+            renamed_count = self._repo.rename_namespace(old_namespace, new_namespace)
+
+            return RenameNamespaceResult(
+                old_namespace=old_namespace,
+                new_namespace=new_namespace,
+                memories_renamed=renamed_count,
+                success=True,
+                message=f"Successfully renamed {renamed_count} memories from '{old_namespace}' to '{new_namespace}'",
+            )
+
+        except NamespaceNotFoundError:
+            raise
+        except ValidationError:
+            raise
+        except Exception as e:
+            raise NamespaceOperationError(
+                f"Failed to rename namespace '{old_namespace}' to '{new_namespace}': {e}"
+            ) from e
+
+    def hybrid_recall(
+        self,
+        query: str,
+        alpha: float = 0.5,
+        limit: int = 5,
+        namespace: str | None = None,
+        min_similarity: float = 0.0,
+    ) -> HybridRecallResult:
+        """Search using combined vector similarity and keyword matching.
+
+        Performs hybrid search combining semantic similarity (vector search)
+        and keyword matching (full-text search). Alpha parameter controls
+        the balance: 1.0 = pure vector, 0.0 = pure keyword, 0.5 = balanced.
+
+        Args:
+            query: Search query text.
+            alpha: Balance between vector (1.0) and keyword (0.0) search.
+            limit: Maximum number of results.
+            namespace: Filter to specific namespace.
+            min_similarity: Minimum similarity threshold (0-1).
+
+        Returns:
+            HybridRecallResult with matching memories.
+
+        Raises:
+            ValidationError: If input validation fails.
+        """
+        # Validate query
+        if not query or not query.strip():
+            raise ValidationError("Query cannot be empty")
+
+        # Validate alpha
+        if alpha < self._config.hybrid_min_alpha or alpha > self._config.hybrid_max_alpha:
+            raise ValidationError(
+                f"Alpha must be between {self._config.hybrid_min_alpha} "
+                f"and {self._config.hybrid_max_alpha}, got {alpha}"
+            )
+
+        # Validate namespace if provided
+        if namespace is not None:
+            namespace = validate_namespace(namespace)
+
+        # Generate query embedding
+        query_vector = self._embeddings.embed(query)
+
+        # Perform hybrid search
+        results = self._repo.hybrid_search(
+            query_vector=query_vector,
+            query_text=query,
+            limit=limit,
+            namespace=namespace,
+            alpha=alpha,
+        )
+
+        # Transform results to HybridMemoryMatch and filter by min_similarity
+        memories: list[HybridMemoryMatch] = []
+        for result in results:
+            if result.similarity >= min_similarity:
+                memories.append(
+                    HybridMemoryMatch(
+                        id=result.id,
+                        content=result.content,
+                        similarity=result.similarity,
+                        namespace=result.namespace,
+                        tags=list(result.tags),
+                        importance=result.importance,
+                        created_at=result.created_at,
+                        metadata=dict(result.metadata),
+                        # These may be populated if repository provides them
+                        vector_score=getattr(result, "vector_score", None),
+                        fts_score=getattr(result, "fts_score", None),
+                        combined_score=result.similarity,
+                    )
+                )
+
+        return HybridRecallResult(
+            query=query,
+            alpha=alpha,
+            memories=memories,
+            total=len(memories),
+            search_type="hybrid",
+        )

spatial_memory/tools/__init__.py

@@ -0,0 +1,5 @@
+"""MCP Tool definitions for Spatial Memory Server."""
+
+from spatial_memory.tools.definitions import TOOLS
+
+__all__ = ["TOOLS"]