If you have got an application written, primarily in C++ code, and you have been asked to understand the code and create architecture diagram, following are some of the biggest challenges:
- Top-down approach: If you have got an understand of functionality at a very high level, it is important to think about some of the components that can map to these functionality. If you have not got enough information on functionality to think about these components, it is always better to ask your customer for a bit more information than what you have. Once you have thought of probable components, you can start hunting for those components in various different folders. If found, you are lucky, and you could start connecting dots from functionality and code perspective.
- Bottoms-up approach: If you may not be able to gather much information on the functionality and you got to get the work done, it is useful to start writing components as you come across. Once you have all the components name written by looking at the class file name, you could start grouping the similar looking components. At the same time, you could also write the functionality delivered by the components-group that can also be called as module.
- Hybrid-approach: This is the best approach I have come across. You know a little about the functionality. You could also start reviewing the code. You should be able to connect/map the components to functionality and come up with different viewpoints of architecture.
Some of the following architectural viewpoints can be created based on above approaches:
- Functional decomposition (may represent business functionality)
- Application architecture representing key modules
Latest posts by Ajitesh Kumar (see all)
- Agentic Reasoning Design Patterns in AI: Examples - October 18, 2024
- LLMs for Adaptive Learning & Personalized Education - October 8, 2024
- Sparse Mixture of Experts (MoE) Models: Examples - October 6, 2024
I found it very helpful. However the differences are not too understandable for me