Frontmatter MCP

"""Query schema module for DuckDB table column information.""" from collections import defaultdict from typing import Any, NotRequired, TypedDict class ColumnInfo(TypedDict): """Column information for a single field.""" type: str nullable: bool examples: NotRequired[list[Any]] # Type alias for table schema Schema = dict[str, ColumnInfo] def create_base_schema(records: list[dict[str, Any]], max_samples: int = 5) -> Schema: """Create base schema with path and frontmatter columns. Extracts column information from parsed frontmatter records. All values are treated as strings or arrays in DuckDB queries. Args: records: List of parsed frontmatter records. max_samples: Maximum number of sample values to include. Returns: Schema dict with path and frontmatter columns. """ schema: Schema = {} property_values: dict[str, list[Any]] = defaultdict(list) for record in records: for key, value in record.items(): property_values[key].append(value) total_files = len(records) for prop, values in property_values.items(): non_null_values = [v for v in values if v is not None] count = len(non_null_values) # path is never nullable if prop == "path": nullable = False else: nullable = count < total_files # Detect if values are arrays is_array = any(isinstance(v, list) for v in non_null_values) # Collect unique sample values seen: set[str] = set() samples: list[Any] = [] for v in non_null_values: key = str(v) if key not in seen and len(samples) < max_samples: seen.add(key) samples.append(v) schema[prop] = ColumnInfo( type="array" if is_array else "string", nullable=nullable, examples=samples, ) return schema