basic-memory 0.16.1__py3-none-any.whl → 0.17.4__py3-none-any.whl

basic_memory/markdown/entity_parser.py

@@ -23,6 +23,7 @@ from basic_memory.markdown.schemas import (
 )
 from basic_memory.utils import parse_tags
 
+
 md = MarkdownIt().use(observation_plugin).use(relation_plugin)
 
 
@@ -189,35 +190,69 @@ class EntityParser:
         return self.base_path / path
 
     async def parse_file_content(self, absolute_path, file_content):
-        # Parse frontmatter with proper error handling for malformed YAML (issue #185)
+        """Parse markdown content from file stats.
+
+        Delegates to parse_markdown_content() for actual parsing logic.
+        Exists for backwards compatibility with code that passes file paths.
+        """
+        # Extract file stat info for timestamps
+        file_stats = absolute_path.stat()
+
+        # Delegate to parse_markdown_content with timestamps from file stats
+        return await self.parse_markdown_content(
+            file_path=absolute_path,
+            content=file_content,
+            mtime=file_stats.st_mtime,
+            ctime=file_stats.st_ctime,
+        )
+
+    async def parse_markdown_content(
+        self,
+        file_path: Path,
+        content: str,
+        mtime: Optional[float] = None,
+        ctime: Optional[float] = None,
+    ) -> EntityMarkdown:
+        """Parse markdown content without requiring file to exist on disk.
+
+        Useful for parsing content from S3 or other remote sources where the file
+        is not available locally.
+
+        Args:
+            file_path: Path for metadata (doesn't need to exist on disk)
+            content: Markdown content as string
+            mtime: Optional modification time (Unix timestamp)
+            ctime: Optional creation time (Unix timestamp)
+
+        Returns:
+            EntityMarkdown with parsed content
+        """
+        # Strip BOM before parsing (can be present in files from Windows or certain sources)
+        # See issue #452
+        from basic_memory.file_utils import strip_bom
+
+        content = strip_bom(content)
+
+        # Parse frontmatter with proper error handling for malformed YAML
         try:
-            post = frontmatter.loads(file_content)
+            post = frontmatter.loads(content)
         except yaml.YAMLError as e:
-            # Log the YAML parsing error with file context
             logger.warning(
-                f"Failed to parse YAML frontmatter in {absolute_path}: {e}. "
+                f"Failed to parse YAML frontmatter in {file_path}: {e}. "
                 f"Treating file as plain markdown without frontmatter."
             )
-            # Create a post with no frontmatter - treat entire content as markdown
-            post = frontmatter.Post(file_content, metadata={})
-
-        # Extract file stat info
-        file_stats = absolute_path.stat()
+            post = frontmatter.Post(content, metadata={})
 
-        # Normalize frontmatter values to prevent AttributeError on date objects (issue #236)
-        # PyYAML automatically converts date strings like "2025-10-24" to datetime.date objects
-        # This normalization converts them back to ISO format strings to ensure compatibility
-        # with code that expects string values
+        # Normalize frontmatter values
         metadata = normalize_frontmatter_metadata(post.metadata)
 
-        # Ensure required fields have defaults (issue #184, #387)
-        # Handle title - use default if missing, None/null, empty, or string "None"
+        # Ensure required fields have defaults
         title = metadata.get("title")
         if not title or title == "None":
-            metadata["title"] = absolute_path.stem
+            metadata["title"] = file_path.stem
         else:
             metadata["title"] = title
-        # Handle type - use default if missing OR explicitly set to None/null
+
         entity_type = metadata.get("type")
         metadata["type"] = entity_type if entity_type is not None else "note"
 
@@ -225,16 +260,20 @@ class EntityParser:
         if tags:
             metadata["tags"] = tags
 
-        # frontmatter - use metadata with defaults applied
-        entity_frontmatter = EntityFrontmatter(
-            metadata=metadata,
-        )
+        # Parse content for observations and relations
+        entity_frontmatter = EntityFrontmatter(metadata=metadata)
         entity_content = parse(post.content)
+
+        # Use provided timestamps or current time as fallback
+        now = datetime.now().astimezone()
+        created = datetime.fromtimestamp(ctime).astimezone() if ctime else now
+        modified = datetime.fromtimestamp(mtime).astimezone() if mtime else now
+
         return EntityMarkdown(
             frontmatter=entity_frontmatter,
             content=post.content,
             observations=entity_content.observations,
             relations=entity_content.relations,
-            created=datetime.fromtimestamp(file_stats.st_ctime).astimezone(),
-            modified=datetime.fromtimestamp(file_stats.st_mtime).astimezone(),
+            created=created,
+            modified=modified,
         )

basic_memory/markdown/markdown_processor.py

@@ -1,15 +1,19 @@
 from pathlib import Path
-from typing import Optional
+from typing import TYPE_CHECKING, Optional
 from collections import OrderedDict
 
 from frontmatter import Post
 from loguru import logger
 
+
 from basic_memory import file_utils
 from basic_memory.file_utils import dump_frontmatter
 from basic_memory.markdown.entity_parser import EntityParser
 from basic_memory.markdown.schemas import EntityMarkdown, Observation, Relation
 
+if TYPE_CHECKING:  # pragma: no cover
+    from basic_memory.config import BasicMemoryConfig
+
 
 class DirtyFileError(Exception):
     """Raised when attempting to write to a file that has been modified."""
@@ -35,9 +39,14 @@ class MarkdownProcessor:
     3. Track schema changes (that's done by the database)
     """
 
-    def __init__(self, entity_parser: EntityParser):
-        """Initialize processor with base path and parser."""
+    def __init__(
+        self,
+        entity_parser: EntityParser,
+        app_config: Optional["BasicMemoryConfig"] = None,
+    ):
+        """Initialize processor with parser and optional config."""
         self.entity_parser = entity_parser
+        self.app_config = app_config
 
     async def read_file(self, path: Path) -> EntityMarkdown:
         """Read and parse file into EntityMarkdown schema.
@@ -122,7 +131,61 @@ class MarkdownProcessor:
         # Write atomically and return checksum of updated file
         path.parent.mkdir(parents=True, exist_ok=True)
         await file_utils.write_file_atomic(path, final_content)
-        return await file_utils.compute_checksum(final_content)
+
+        # Format file if configured (MarkdownProcessor always handles markdown files)
+        content_for_checksum = final_content
+        if self.app_config:
+            formatted_content = await file_utils.format_file(  # pragma: no cover
+                path, self.app_config, is_markdown=True
+            )
+            if formatted_content is not None:  # pragma: no cover
+                content_for_checksum = formatted_content  # pragma: no cover
+
+        return await file_utils.compute_checksum(content_for_checksum)
+
+    def to_markdown_string(self, markdown: EntityMarkdown) -> str:
+        """Convert EntityMarkdown to markdown string with frontmatter.
+
+        This method handles serialization only - it does not write to files.
+        Use FileService.write_file() to persist the output.
+
+        This enables cloud environments to override file operations via
+        dependency injection while reusing the serialization logic.
+
+        Args:
+            markdown: EntityMarkdown schema to serialize
+
+        Returns:
+            Complete markdown string with frontmatter, content, and structured sections
+        """
+        # Convert frontmatter to dict
+        frontmatter_dict = OrderedDict()
+        frontmatter_dict["title"] = markdown.frontmatter.title
+        frontmatter_dict["type"] = markdown.frontmatter.type
+        frontmatter_dict["permalink"] = markdown.frontmatter.permalink
+
+        metadata = markdown.frontmatter.metadata or {}
+        for k, v in metadata.items():
+            frontmatter_dict[k] = v
+
+        # Start with user content (or minimal title for new files)
+        content = markdown.content or f"# {markdown.frontmatter.title}\n"
+
+        # Add structured sections with proper spacing
+        content = content.rstrip()  # Remove trailing whitespace
+
+        # Add a blank line if we have semantic content
+        if markdown.observations or markdown.relations:
+            content += "\n"
+
+        if markdown.observations:
+            content += self.format_observations(markdown.observations)
+        if markdown.relations:
+            content += self.format_relations(markdown.relations)
+
+        # Create Post object for frontmatter
+        post = Post(content, **frontmatter_dict)
+        return dump_frontmatter(post)
 
     def format_observations(self, observations: list[Observation]) -> str:
         """Format observations section in standard way.

basic_memory/markdown/plugins.py

@@ -30,7 +30,9 @@ def is_observation(token: Token) -> bool:
 
     # Check for proper observation format: [category] content
     match = re.match(r"^\[([^\[\]()]+)\]\s+(.+)", content)
-    has_tags = "#" in content
+    # Check for standalone hashtags (words starting with #)
+    # This excludes # in HTML attributes like color="#4285F4"
+    has_tags = any(part.startswith("#") for part in content.split())
     return bool(match) or has_tags
 
 
@@ -160,7 +162,7 @@ def parse_inline_relations(content: str) -> List[Dict[str, Any]]:
 
         target = content[start + 2 : end].strip()
         if target:
-            relations.append({"type": "links to", "target": target, "context": None})
+            relations.append({"type": "links_to", "target": target, "context": None})
 
         start = end + 2
 

basic_memory/markdown/utils.py

@@ -3,6 +3,7 @@
 from pathlib import Path
 from typing import Any, Optional
 
+
 from frontmatter import Post
 
 from basic_memory.file_utils import has_frontmatter, remove_frontmatter, parse_frontmatter
@@ -12,7 +13,10 @@ from basic_memory.models import Observation as ObservationModel
 
 
 def entity_model_from_markdown(
-    file_path: Path, markdown: EntityMarkdown, entity: Optional[Entity] = None
+    file_path: Path,
+    markdown: EntityMarkdown,
+    entity: Optional[Entity] = None,
+    project_id: Optional[int] = None,
 ) -> Entity:
     """
     Convert markdown entity to model. Does not include relations.
@@ -21,6 +25,7 @@ def entity_model_from_markdown(
         file_path: Path to the markdown file
         markdown: Parsed markdown entity
         entity: Optional existing entity to update
+        project_id: Project ID for new observations (uses entity.project_id if not provided)
 
     Returns:
         Entity model populated from markdown
@@ -50,9 +55,13 @@ def entity_model_from_markdown(
     metadata = markdown.frontmatter.metadata or {}
     model.entity_metadata = {k: str(v) for k, v in metadata.items() if v is not None}
 
+    # Get project_id from entity if not provided
+    obs_project_id = project_id or (model.project_id if hasattr(model, "project_id") else None)
+
     # Convert observations
     model.observations = [
         ObservationModel(
+            project_id=obs_project_id,
             content=obs.content,
             category=obs.category,
             context=obs.context,

basic_memory/mcp/async_client.py

@@ -95,6 +95,7 @@ async def get_client() -> AsyncIterator[AsyncClient]:
                 yield client
         else:
             # Local mode: ASGI transport for in-process calls
+            # Note: ASGI transport does NOT trigger FastAPI lifespan, so no special handling needed
             logger.info("Creating ASGI client for local Basic Memory API")
             async with AsyncClient(
                 transport=ASGITransport(app=fastapi_app), base_url="http://test", timeout=timeout

basic_memory/mcp/clients/__init__.py

@@ -0,0 +1,28 @@
+"""Typed internal API clients for MCP tools.
+
+These clients encapsulate API paths, error handling, and response validation.
+MCP tools become thin adapters that call these clients and format results.
+
+Usage:
+    from basic_memory.mcp.clients import KnowledgeClient, SearchClient
+
+    async with get_client() as http_client:
+        knowledge = KnowledgeClient(http_client, project_id)
+        entity = await knowledge.create_entity(entity_data)
+"""
+
+from basic_memory.mcp.clients.knowledge import KnowledgeClient
+from basic_memory.mcp.clients.search import SearchClient
+from basic_memory.mcp.clients.memory import MemoryClient
+from basic_memory.mcp.clients.directory import DirectoryClient
+from basic_memory.mcp.clients.resource import ResourceClient
+from basic_memory.mcp.clients.project import ProjectClient
+
+__all__ = [
+    "KnowledgeClient",
+    "SearchClient",
+    "MemoryClient",
+    "DirectoryClient",
+    "ResourceClient",
+    "ProjectClient",
+]

basic_memory/mcp/clients/directory.py

@@ -0,0 +1,70 @@
+"""Typed client for directory API operations.
+
+Encapsulates all /v2/projects/{project_id}/directory/* endpoints.
+"""
+
+from typing import Optional, Any
+
+from httpx import AsyncClient
+
+from basic_memory.mcp.tools.utils import call_get
+
+
+class DirectoryClient:
+    """Typed client for directory listing operations.
+
+    Centralizes:
+    - API path construction for /v2/projects/{project_id}/directory/*
+    - Response validation
+    - Consistent error handling through call_* utilities
+
+    Usage:
+        async with get_client() as http_client:
+            client = DirectoryClient(http_client, project_id)
+            nodes = await client.list("/", depth=2)
+    """
+
+    def __init__(self, http_client: AsyncClient, project_id: str):
+        """Initialize the directory client.
+
+        Args:
+            http_client: HTTPX AsyncClient for making requests
+            project_id: Project external_id (UUID) for API calls
+        """
+        self.http_client = http_client
+        self.project_id = project_id
+        self._base_path = f"/v2/projects/{project_id}/directory"
+
+    async def list(
+        self,
+        dir_name: str = "/",
+        *,
+        depth: int = 1,
+        file_name_glob: Optional[str] = None,
+    ) -> list[dict[str, Any]]:
+        """List directory contents.
+
+        Args:
+            dir_name: Directory path to list (default: root)
+            depth: How deep to traverse (default: 1)
+            file_name_glob: Optional glob pattern to filter files
+
+        Returns:
+            List of directory nodes with their contents
+
+        Raises:
+            ToolError: If the request fails
+        """
+        params: dict = {
+            "dir_name": dir_name,
+            "depth": depth,
+        }
+        if file_name_glob:
+            params["file_name_glob"] = file_name_glob
+
+        response = await call_get(
+            self.http_client,
+            f"{self._base_path}/list",
+            params=params,
+        )
+        return response.json()

basic_memory/mcp/clients/knowledge.py

@@ -0,0 +1,176 @@
+"""Typed client for knowledge/entity API operations.
+
+Encapsulates all /v2/projects/{project_id}/knowledge/* endpoints.
+"""
+
+from typing import Any
+
+from httpx import AsyncClient
+
+from basic_memory.mcp.tools.utils import call_get, call_post, call_put, call_patch, call_delete
+from basic_memory.schemas.response import EntityResponse, DeleteEntitiesResponse
+
+
+class KnowledgeClient:
+    """Typed client for knowledge graph entity operations.
+
+    Centralizes:
+    - API path construction for /v2/projects/{project_id}/knowledge/*
+    - Response validation via Pydantic models
+    - Consistent error handling through call_* utilities
+
+    Usage:
+        async with get_client() as http_client:
+            client = KnowledgeClient(http_client, project_id)
+            entity = await client.create_entity(entity_data)
+    """
+
+    def __init__(self, http_client: AsyncClient, project_id: str):
+        """Initialize the knowledge client.
+
+        Args:
+            http_client: HTTPX AsyncClient for making requests
+            project_id: Project external_id (UUID) for API calls
+        """
+        self.http_client = http_client
+        self.project_id = project_id
+        self._base_path = f"/v2/projects/{project_id}/knowledge"
+
+    # --- Entity CRUD Operations ---
+
+    async def create_entity(self, entity_data: dict[str, Any]) -> EntityResponse:
+        """Create a new entity.
+
+        Args:
+            entity_data: Entity data including title, content, folder, etc.
+
+        Returns:
+            EntityResponse with created entity details
+
+        Raises:
+            ToolError: If the request fails
+        """
+        response = await call_post(
+            self.http_client,
+            f"{self._base_path}/entities",
+            json=entity_data,
+        )
+        return EntityResponse.model_validate(response.json())
+
+    async def update_entity(self, entity_id: str, entity_data: dict[str, Any]) -> EntityResponse:
+        """Update an existing entity (full replacement).
+
+        Args:
+            entity_id: Entity external_id (UUID)
+            entity_data: Complete entity data for replacement
+
+        Returns:
+            EntityResponse with updated entity details
+
+        Raises:
+            ToolError: If the request fails
+        """
+        response = await call_put(
+            self.http_client,
+            f"{self._base_path}/entities/{entity_id}",
+            json=entity_data,
+        )
+        return EntityResponse.model_validate(response.json())
+
+    async def get_entity(self, entity_id: str) -> EntityResponse:
+        """Get an entity by ID.
+
+        Args:
+            entity_id: Entity external_id (UUID)
+
+        Returns:
+            EntityResponse with entity details
+
+        Raises:
+            ToolError: If the entity is not found or request fails
+        """
+        response = await call_get(
+            self.http_client,
+            f"{self._base_path}/entities/{entity_id}",
+        )
+        return EntityResponse.model_validate(response.json())
+
+    async def patch_entity(self, entity_id: str, patch_data: dict[str, Any]) -> EntityResponse:
+        """Partially update an entity.
+
+        Args:
+            entity_id: Entity external_id (UUID)
+            patch_data: Partial entity data to update
+
+        Returns:
+            EntityResponse with updated entity details
+
+        Raises:
+            ToolError: If the request fails
+        """
+        response = await call_patch(
+            self.http_client,
+            f"{self._base_path}/entities/{entity_id}",
+            json=patch_data,
+        )
+        return EntityResponse.model_validate(response.json())
+
+    async def delete_entity(self, entity_id: str) -> DeleteEntitiesResponse:
+        """Delete an entity.
+
+        Args:
+            entity_id: Entity external_id (UUID)
+
+        Returns:
+            DeleteEntitiesResponse confirming deletion
+
+        Raises:
+            ToolError: If the entity is not found or request fails
+        """
+        response = await call_delete(
+            self.http_client,
+            f"{self._base_path}/entities/{entity_id}",
+        )
+        return DeleteEntitiesResponse.model_validate(response.json())
+
+    async def move_entity(self, entity_id: str, destination_path: str) -> EntityResponse:
+        """Move an entity to a new location.
+
+        Args:
+            entity_id: Entity external_id (UUID)
+            destination_path: New file path for the entity
+
+        Returns:
+            EntityResponse with updated entity details
+
+        Raises:
+            ToolError: If the request fails
+        """
+        response = await call_put(
+            self.http_client,
+            f"{self._base_path}/entities/{entity_id}/move",
+            json={"destination_path": destination_path},
+        )
+        return EntityResponse.model_validate(response.json())
+
+    # --- Resolution ---
+
+    async def resolve_entity(self, identifier: str) -> str:
+        """Resolve a string identifier to an entity external_id.
+
+        Args:
+            identifier: The identifier to resolve (permalink, title, or path)
+
+        Returns:
+            The resolved entity external_id (UUID)
+
+        Raises:
+            ToolError: If the identifier cannot be resolved
+        """
+        response = await call_post(
+            self.http_client,
+            f"{self._base_path}/resolve",
+            json={"identifier": identifier},
+        )
+        data = response.json()
+        return data["external_id"]

basic_memory/mcp/clients/memory.py

@@ -0,0 +1,120 @@
+"""Typed client for memory/context API operations.
+
+Encapsulates all /v2/projects/{project_id}/memory/* endpoints.
+"""
+
+from typing import Optional
+
+from httpx import AsyncClient
+
+from basic_memory.mcp.tools.utils import call_get
+from basic_memory.schemas.memory import GraphContext
+
+
+class MemoryClient:
+    """Typed client for memory context operations.
+
+    Centralizes:
+    - API path construction for /v2/projects/{project_id}/memory/*
+    - Response validation via Pydantic models
+    - Consistent error handling through call_* utilities
+
+    Usage:
+        async with get_client() as http_client:
+            client = MemoryClient(http_client, project_id)
+            context = await client.build_context("memory://specs/search")
+    """
+
+    def __init__(self, http_client: AsyncClient, project_id: str):
+        """Initialize the memory client.
+
+        Args:
+            http_client: HTTPX AsyncClient for making requests
+            project_id: Project external_id (UUID) for API calls
+        """
+        self.http_client = http_client
+        self.project_id = project_id
+        self._base_path = f"/v2/projects/{project_id}/memory"
+
+    async def build_context(
+        self,
+        path: str,
+        *,
+        depth: int = 1,
+        timeframe: Optional[str] = None,
+        page: int = 1,
+        page_size: int = 10,
+        max_related: int = 10,
+    ) -> GraphContext:
+        """Build context from a memory path.
+
+        Args:
+            path: The path to build context for (without memory:// prefix)
+            depth: How deep to traverse relations
+            timeframe: Time filter (e.g., "7d", "1 week")
+            page: Page number (1-indexed)
+            page_size: Results per page
+            max_related: Maximum related items per result
+
+        Returns:
+            GraphContext with hierarchical results
+
+        Raises:
+            ToolError: If the request fails
+        """
+        params: dict = {
+            "depth": depth,
+            "page": page,
+            "page_size": page_size,
+            "max_related": max_related,
+        }
+        if timeframe:
+            params["timeframe"] = timeframe
+
+        response = await call_get(
+            self.http_client,
+            f"{self._base_path}/{path}",
+            params=params,
+        )
+        return GraphContext.model_validate(response.json())
+
+    async def recent(
+        self,
+        *,
+        timeframe: str = "7d",
+        depth: int = 1,
+        types: Optional[list[str]] = None,
+        page: int = 1,
+        page_size: int = 10,
+    ) -> GraphContext:
+        """Get recent activity.
+
+        Args:
+            timeframe: Time filter (e.g., "7d", "1 week", "2 days ago")
+            depth: How deep to traverse relations
+            types: Filter by item types
+            page: Page number (1-indexed)
+            page_size: Results per page
+
+        Returns:
+            GraphContext with recent activity
+
+        Raises:
+            ToolError: If the request fails
+        """
+        params: dict = {
+            "timeframe": timeframe,
+            "depth": depth,
+            "page": page,
+            "page_size": page_size,
+        }
+        if types:
+            # Join types as comma-separated string if provided
+            params["type"] = ",".join(types) if isinstance(types, list) else types
+
+        response = await call_get(
+            self.http_client,
+            f"{self._base_path}/recent",
+            params=params,
+        )
+        return GraphContext.model_validate(response.json())