CLAUDE.md

This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.

Project Overview

This is a data mapper for the Open Ownership Register. It converts Open Ownership Register JSONL data into Senzing-compatible JSON format for entity resolution. The data source contains ~18 million entities representing beneficial ownership relationships.

Commands

Install dependencies

python -m venv ./venv
source ./venv/bin/activate
pip install --group all .

Run the mapper

python src/oor_mapper.py -i <input_file.jsonl.gz> -o <output_file.jsonl.gz> [-l <log_file.json>]

Lint

pylint $(git ls-files '*.py' ':!:docs/source/*')
black --check src/
flake8 src/
isort --check src/
mypy src/
bandit -r src/

Architecture

Core Mapper (src/oor_mapper.py)

The mapper class transforms Open Ownership Register statements into Senzing records:

• Statement Types:

  • entityStatement → Organizations with names, addresses, identifiers
  • personStatement → Individuals with names, birth dates, nationalities
  • ownershipOrControlStatement → Relationships linking persons/entities to entities

• Key Methods:

  • map() - Main entry point, routes by statement type
  • map_entity() / map_person() - Convert to Senzing format
  • map_relationship() - Creates REL_POINTER relationships
  • map_addresses() / map_identifiers() - Handle nested data structures

• Data Flow: Input is streamed, records cached by RECORD_ID, relationships merged onto their subject entities, then output as JSONL

Senzing Integration

• Data source: OPEN-OWNERSHIP
• Relationships use REL_ANCHOR_DOMAIN/REL_POINTER_DOMAIN = "OOR"
• Config in src/oor_config_updates.g2c adds relationship date attributes

Code Style

• Line length: 120 characters
• Uses black formatting with isort (black profile)