Layer Details¶

Layer 1: Presentation Layer¶

The presentation layer forms the contact point with the user and handles command processing, logging, and configuration.

cli.py — Command Line Interface¶

Main entrypoint registered as crunch_uml console script. Provides three subcommands:

crunch_uml [-v] [-d] [-w] [-db_url URL] [-sch SCHEMA] {import,transform,export} ...

Argument	Description
`-v / --verbose`	INFO-level logging
`-d / --debug`	DEBUG-level logging
`-db_url`	Database connection string
`-sch / --schema_name`	Name of the working schema

The main() function validates arguments, configures logging, and delegates to the appropriate registry based on the command.

const.py — Constants¶

Contains database URL defaults, XML namespaces (XMI 2.0.1, UML 2.0.1), command constants, EA Repository mappers, tag profiles, and language configuration.

Planned Extensions¶

REST API Interface

FastAPI or Flask-based web interface for controlling import/transform/export via HTTP.

Configuration Module

Overarching interface for managing transformation pipelines, including monitoring and reproducibility.

Layer 2: Orchestration Layer¶

The orchestration layer manages the registry pattern and plugin framework.

Registry Pattern¶

class Registry:
    @classmethod
    def register(name, descr="")   # Decorator for registration

    @classmethod
    def entries()                   # List of registered names

    @classmethod
    def getinstance(name)           # Instantiate registered class

    @classmethod
    def getDescription(name)        # Get description

Three subclasses implement this pattern:

Registry	Count	Function
`ParserRegistry`	7	Maps input types to parser classes
`RendererRegistry`	11	Maps output types to renderer classes
`TransformerRegistry`	2+	Maps transformation types to transformer classes

Plugin Framework¶

Dynamic loading of custom transformation plugins via CLI arguments --plugin_file_name and --plugin_class_name. Plugins extend the Plugin base class.

Layer 3: Processing Layers¶

See the detailed pages:

Layer 4: Persistence Layer¶

See Persistence and Data Model.

Layer 5: External Systems¶

crunch_uml supports multiple databases via SQLAlchemy:

graph LR
    DB["Database class<br/>SQLAlchemy 2.0"] --> SQLite[("SQLite<br/>default")]
    DB --> PG[("PostgreSQL<br/>production")]
    DB --> MySQL[("MySQL / MariaDB")]
    DB --> EA[("Enterprise Architect<br/>ODBC")]
    DB --> FS["File System<br/>XMI, JSON, CSV, XLSX"]
    DB --> SNOW[("Snowflake<br/><i>Planned</i>")]
    DB --> AZURE[("AzureDB<br/><i>Planned</i>")]

    style SNOW fill:#e1d5e7,stroke:#9673a6,stroke-dasharray: 5 5
    style AZURE fill:#e1d5e7,stroke:#9673a6,stroke-dasharray: 5 5

Layer 6: Helper Modules¶

Module	Function
`util.py`	URL validation, GUID generation, date parsing, Dutch pluralization
`lang.py`	Translation wrapper via `translators` library with retry logic
`exceptions.py`	Custom exception classes
`templates/`	Jinja2 templates (GGM Markdown, JSON Schema, DDAS, SQLAlchemy)
`json_datatypes.json`	Datatype mapping configuration