openstage
Typed Python models for legislative data. openstage provides a common data model for procedures, events, and documents with built-in support for multilingual text and multi-scheme identifiers.
Note: openstage is both the project name and the Python package name (import openstage).
Guiding principle
Legislative data across systems shares common structures: procedures contain events, events produce documents, text appears in multiple languages, and entities carry identifiers from several naming schemes. openstage captures these patterns in typed Pydantic models that work across legislative systems.
The EU is the primary implemented case. EU-specific models extend the base types with typed fields for procedure types, legislative phases, institutional actors, and other domain concepts drawn from the Common Data Model (CDM).
Usage
Most researchers download a pre-compiled dataset and load it directly into typed models. No collection, no parsing, no extra dependencies.
Power users can collect raw data from EUR-Lex and Cellar using openstage-infrastructure, parse it with openbasement, and map the results into the same typed models using the adapter layer.
Both paths produce the same model objects.
Package layers
| Layer | What it does | Dependencies |
|---|---|---|
models/ |
Typed Pydantic models with multilingual text and multi-scheme identifiers | pydantic |
models/eu/ |
EU case models extending base types with domain-specific fields | pydantic |
adapters/ |
Map external data (openbasement dicts) into models | pydantic |
collections |
ProcedureList for filtering and cross-procedure analytics |
-- |
dataset |
Dataset for loading and saving packaged procedure collections |
-- |
Part of the openstage project
- openstage (this package): Data models and adapters
- openbasement: Template-based RDF extraction from EU Cellar data
- openstage-infrastructure: Workflow orchestration, storage, and publishing
Next steps
- Getting started: Install, load data, explore models.
- Base models: Entity hierarchy, multilingual text, identifiers.
- EU models: EU-specific fields, controlled vocabularies, codebook.
- Datasets: Loading, saving, filtering, and analyzing procedure collections.
- Fields and codebooks: Field metadata system and codebook generation.