How taghubnet/frepple works
Main Components
• Core Planning Engine: Represented by the C++ source files found under the “src” directory (such as src/model and src/solver) and header files under “include”. This engine embodies the planning algorithms, business logic, and optimization routines (covering scheduling, load planning, operation planning, and bottleneck management).
• Python/Django Integration Layer: Found under “contrib/django”. This part provides the web frontend (administrative UI, reporting and REST API endpoints) and the binding to the planning engine. Here, modules for input (data ingestion), output (reports, dashboards) and connectors (e.g., for interfacing with external ERP systems) are defined.
• External Connectors and Integrations: Components like the Odoo and Openbravo integrations (located under “contrib/odoo” and “contrib/django/freppledb”) are essential to interface with ERP systems. They handle data exchange such as importing parameters or exporting schedules.
• Packaging/Deployment and Installer Modules: Parts in “contrib/debian”, “contrib/rpm”, “contrib/vc”, and “contrib/installer” indicate that the system has tailored deployment processes and installers (Windows, Linux packages). This should be shown as a separate component for distribution and deployment.
• Documentation and Testing: Although not directly part of runtime architecture, the “doc” and “test” directories highlight extensive technical documentation and a comprehensive testing framework. (You may optionally include these in an “Auxiliary” or “Support” layer to indicate continuous integration and quality assurance aspects.)
Relationships and Data Flows
• That the Django-based UI (presentation layer) interacts with the Core Planning Engine. This is often achieved via the Python extension (located in “src/pythonextension.cpp” and corresponding bindings).
• That external modules (such as the ERP connectors) both receive inputs from and output results back to the planning engine.
• Packaging and installer components as pipelines that take the combined application (engine + integration UI) and produce deployment artifacts (e.g., deb, rpm, Windows installer).
Architectural Patterns and Technologies
• Layered Architecture: Visually separate the system into layers – the Business Logic (C++ engine), the Integration/Interface Layer (Python/Django application), and the Deployment/Packaging/Support Layer.
• Hybrid Integration: Note how the system employs language interoperability where high-performance routines run in C++ while user interaction and external communication are managed by Django.
• Extensibility: With multiple integration modules (ERP connectors, additional extension modules as seen in “doc/extension-modules”), highlight how the architecture supports plug-ins and extension points.