How spanag/eden works
Main Components
a. Core Components:
• NeuroML Parser/Processor (based on eden/NeuroML.cpp and eden/NeuroML.h)
• LEMS Expression Engine (eden/neuroml/LEMS_Expr.*)
• Code Generator (implied by README's mention of runtime code generation)
• Parallel Execution Engine (supports both OpenMP and optional MPI)
• Memory Management System (eden/MMMallocator.h)
b. External Interfaces:
• Command Line Interface
• Python Package Interface (testing/python_package)
• File I/O System (for model inputs and simulation outputs)
[Input Layer]
• NeuroML Files / LEMS Files
• ↓
[Processing Layer]
• NeuroML Parser → LEMS Expression Engine
• ↓
[Generation Layer]
• Code Generator
• ↓
[Execution Layer]
• Parallel Execution Engine (OpenMP/MPI)
• ↓
[Output Layer]
• Simulation Results
Important Elements to Highlight
• Input/Output components (rectangles)
• Processing modules (rounded rectangles)
• Execution components (hexagons)
b. Parallel processing capabilities:
• A parallel execution pool representation
• MPI communication paths if applicable
• OpenMP thread distribution
Integration Points
• Both CLI and Python API entry points
• Third-party dependencies (pugixml, cJSON)
• Compiler integration point (crucial runtime dependency)
Specific Implementation Details
• NeuroML Parser and LEMS Expression Engine
• Code Generator and Compiler
• Execution Engine and Memory Manager
Technical Notes
• Runtime code generation and compilation workflow
• Parallel execution capabilities (both OpenMP and MPI)
• Memory management strategy
• File I/O patterns for model input and result output
Implementation-Specific Details
Specialized components:
• Custom memory allocator (MMMallocator)
• LEMS core components integration
• Platform-specific compilation paths
• Python package wrapper structure