The mcr_sas2yaml Module

This module defines an introspector for SAS 7BDAT files related to Medicare data.

The SASIntrospector class crawls a directory for SAS files matching a pattern, extracts metadata (column info), and generates data model definitions in YAML format. These models are written to a centralized registry.

Typical use case: building data models from Medicare SAS files (e.g., 1999–2010).

See Also: - SASIntrospector (this class) - MedicareSAS (superclass) - Introspector - MedicareRegistry

class SASIntrospector(registry_path: str, root_dir: str = '.')

This class traverses a file path looking for SAS .sas7bdat files, extracts their schema using Introspector, and creates a structured data model serialized to a YAML registry.

In addition to field-level metadata, this introspector: - Attempts to identify common fields such as bene_id, state, zip, and year - Automatically generates a year column if missing - Marks special fields as indexed - Adds virtual key fields for file and record identifiers

Inherits from:

MedicareSAS: For file traversal and handling utilities MedicareRegistry: For interacting with the data model YAML registry

Initializes the SASIntrospector with the given registry path and SAS root directory.

Args:

registry_path (str): Path to the YAML registry file. root_dir (str): Base directory for SAS 7BDAT files.

classmethod process(registry_path: str, pattern: str, root_dir: str = '.')

Entry point that initializes and runs the introspector.

Args:

registry_path (str): Path to output YAML registry file. pattern (str): Glob-like pattern to match .sas7bdat files. root_dir (str): Root directory to start searching. Default is current directory.

classmethod matches(s: str, candidates: List[str])

Determines whether a string matches any string or wildcard pattern in candidates.

Args:

s (str): String to match. candidates (List[str]): List of exact names or patterns (may include *).

Returns:

bool: True if s matches any candidate.

handle(table: str, file_path: str, file_type: str, year: int)

Handles metadata extraction for a single .sas7bdat file.

Args:

table (str): Target table name to use in the registry. file_path (str): File path to the SAS data file. file_type (str): Type of file (e.g., ‘denominator’). year (int): Associated year of data.

add_sas_table(table: str, file_path: str, index_all: bool, year: int)

Extracts schema from a SAS file and registers columns into the YAML registry.

• Uses introspection to extract columns and attach metadata.

• Detects and indexes key columns (e.g., bene_id, state, year).

• Auto-generates a ‘year’ column if missing using a virtual GENERATED column.

• Indexes all columns if index_all is True (e.g., for denominator files).

• Adds FILE and RECORD fields to simulate full uniqueness using a compound PK.

Args:

table (str): Name of the table in the registry. file_path (str): Path to the SAS file. index_all (bool): Whether all fields should be indexed. year (int): Year to use when generating missing year columns.

Raises:

ValueError: If duplicate key fields are detected or mandatory fields are missing.