How do you manage things?

[audinue]

How do you manage C++ classes for reusable-code purpose?

Since performance is the top priority for me, there is duplicated class names for their special purpose, for instance:
- Node, with single link.
- Node, with double link.
- Node, with multiple link (a.k.a. Vertex).

As well as String:
- String for appending (back only).
- String for appending (front/back), and insertion.

Thanks in advance.

[CornedBee]

Nothing wrong with having different classes with different trade-offs. If you can write these trade-offs as template policies, it might increase your implementation code reuse, but that's a different issue.

The important thing here is concepts. For example, define what it means to be a string. Define the essential operations that all strings must support, and exactly how these things are done. Then you can write template functions that can take any string - as long as it conforms to these requirements.

The STL is founded on this principle. Compare deque and vector - they're both models of the Random Access Container concept, but with different trade-offs.