Obsidian MCP Server

"""Find orphaned notes in the vault.""" import logging import re from typing import List, Dict, Any, Optional, Literal from pathlib import Path from datetime import datetime, timedelta from ..utils.filesystem import get_vault from ..utils import is_markdown_file from .link_management import get_backlinks, get_outgoing_links logger = logging.getLogger(__name__) async def find_orphaned_notes( orphan_type: str = "no_backlinks", exclude_folders: Optional[List[str]] = None, min_age_days: Optional[int] = None, ctx=None ) -> Dict[str, Any]: """ Find orphaned notes based on specified criteria. Args: orphan_type: Type of orphaned notes to find exclude_folders: Folders to exclude from search min_age_days: Only include notes older than X days ctx: MCP context for progress reporting Returns: Dictionary containing orphaned notes and statistics """ vault = get_vault() # Default exclusions if not specified if exclude_folders is None: exclude_folders = ["Templates", "Archive", "Daily"] if ctx: ctx.info(f"Finding orphaned notes of type: {orphan_type}") # Get all notes all_notes = await vault.list_notes(recursive=True) orphaned_notes = [] # Calculate date threshold if min_age_days is specified date_threshold = None if min_age_days is not None: date_threshold = datetime.now() - timedelta(days=min_age_days) # Process each note total_notes = len(all_notes) for i, note_info in enumerate(all_notes): path = note_info["path"] # Skip excluded folders if any(path.startswith(folder + "/") or path.startswith(folder + "\\") for folder in exclude_folders): continue # Skip if file is in root and we're excluding root level files if "/" not in path and "\\" not in path: # Root level file - you might want to handle these differently pass try: # Read note to get full information note = await vault.read_note(path) # Check age if threshold is set if date_threshold and note.metadata.modified: # Handle both timezone-aware and naive datetime strings mod_str = note.metadata.modified if mod_str.endswith('Z'): mod_str = mod_str[:-1] + '+00:00' try: # Try to parse with timezone mod_time = datetime.fromisoformat(mod_str) except ValueError: # If that fails, parse as naive and assume local timezone mod_time = datetime.fromisoformat(mod_str.split('+')[0].split('.')[0]) # Make date_threshold timezone-naive for comparison if mod_time.tzinfo is not None: mod_time = mod_time.replace(tzinfo=None) if mod_time > date_threshold: continue # Skip recent notes # Check orphan criteria is_orphaned = False orphan_reason = "" if orphan_type == "no_backlinks": # Get backlinks for this note backlinks_result = await get_backlinks(path) backlinks = backlinks_result.get("backlinks", []) if not backlinks or len(backlinks) == 0: is_orphaned = True orphan_reason = "No incoming links" elif orphan_type == "no_links": # Check both incoming and outgoing links backlinks_result = await get_backlinks(path) backlinks = backlinks_result.get("backlinks", []) outgoing_result = await get_outgoing_links(path) outgoing = outgoing_result.get("outgoing_links", []) if (not backlinks or len(backlinks) == 0) and (not outgoing or len(outgoing) == 0): is_orphaned = True orphan_reason = "No incoming or outgoing links" elif orphan_type == "no_tags": # Check if note has any tags if not note.metadata.tags or len(note.metadata.tags) == 0: is_orphaned = True orphan_reason = "No tags" elif orphan_type == "no_metadata": # Check if note has any frontmatter properties (beyond basic ones) frontmatter = note.metadata.frontmatter # Remove system properties user_properties = {k: v for k, v in frontmatter.items() if k not in ["tags", "aliases", "cssclass"]} if not user_properties: is_orphaned = True orphan_reason = "No metadata properties" elif orphan_type == "isolated": # Multiple criteria: no links AND no tags backlinks_result = await get_backlinks(path) backlinks = backlinks_result.get("backlinks", []) outgoing_result = await get_outgoing_links(path) outgoing = outgoing_result.get("outgoing_links", []) has_no_links = (not backlinks or len(backlinks) == 0) and (not outgoing or len(outgoing) == 0) has_no_tags = not note.metadata.tags or len(note.metadata.tags) == 0 if has_no_links and has_no_tags: is_orphaned = True orphan_reason = "No links and no tags" if is_orphaned: # Calculate size and word count content = note.content size_bytes = len(content.encode('utf-8')) word_count = len(content.split()) orphaned_notes.append({ "path": path, "reason": orphan_reason, "modified": note.metadata.modified.isoformat() if note.metadata.modified else None, "size": size_bytes, "word_count": word_count }) except Exception as e: logger.warning(f"Error processing note {path}: {e}") continue # Progress reporting if ctx and (i + 1) % 50 == 0: ctx.info(f"Processed {i + 1}/{total_notes} notes...") # Sort by modified date (oldest first) orphaned_notes.sort(key=lambda x: x.get("modified", ""), reverse=False) # Prepare summary statistics stats = { "total_notes_scanned": total_notes, "excluded_folders": exclude_folders, "orphan_type": orphan_type, "min_age_days": min_age_days } if min_age_days: stats["date_threshold"] = date_threshold.isoformat() return { "orphaned_notes": orphaned_notes, "count": len(orphaned_notes), "stats": stats }