Table-Driven Methods

[[deleted user]]

Table-driven techniques can be so powerful because of greater clarity and easier maintenance, along with providing an opportunity for runtime and post-deployment change. I am surprised, though, that this chapter leaves out hash tables (except for one sentence that briefly mentions Java HashMap), yet covers most other basic table types. Surprisingly, I searched the book and found no mention of this fundamental data structure!

[Aleksander]

Sometimes the best solution to the problem is a combination of both object-oriented and table-driven approaches. Also, recently I read an article written by Andrew Koenig (his Accelerated C++ book is highly recommended for the beginners by C++ object-oriented community) about misusing arrays. Some of that information can be applied to misusing tables…

http://drdobbs.com/blogs/cpp/229401975

… and further..

http://drdobbs.com/blogs/cpp/229402321

I guess before considering using table-driven approach and writing search algorithms the better idea is considering using standard or well tested algorithms and utilities.

Also I was surprised by not seeing a reference to Big O notation in discussion of searching algorithms :) jk

[Erik]

Good chapter. I like the code samples and real world usage ideas. I could handle a few more from the book too.

I've used some of these approaches to logging. Code often starts out simple (if-else or switch) and then I refactor it in to a table model as it grows.