Better Software - November 2007 - (Page 14)

Code Craft “A precursor to generic programming, functional programming is especially powerful in constructing succinct, flexible solutions that process sequences of data regardless of type.” def union(s,t): return reduce(lambda sofar, x: sofar + ([] if x in s else [x]), t, s) Listing 5 def intersection(s,t): return reduce(lambda sofar,x: sofar + ([x] if x in s else []), t, []) Listing 6 If there are no functions in the list funs, then compose returns the identity function. Otherwise, it returns a series of nested invocations on the free parameter x—just what we wanted, and all in one line! An example that validates compose is shown in listing 8, which prints the first ten odd numbers by applying 2x+1 to tion as a parameter but return a computed value, not a function. In an ambitious effort to convince you of the power of functional programming (and of Python!), the next example illustrates both. Consider how you would formulate the composition of an arbitrary number of unary functions. Given the list of functions [f,g,h], we want to create a new function that, when given the argument x, computes f(g(h(x)))—but for any number of functions. Since reduce traverses sequences left to right, we’ll need to reverse the list of functions so that h is applied first. The accumulator function should take the current accumulated result as its first parameter, the next function in the list as a second parameter, and apply the latter to the former as the new accumulated result. The return value should itself be a function that takes a single parameter (x above). All the above verbiage can be reduced to the one-line function shown in listing 7. the list [0,1,…,9]. Summary The object paradigm is just one of many that a software developer should have in his arsenal. A precursor to generic programming, functional programming is especially powerful in constructing succinct, flexible solutions that process sequences of data regardless of type. I have shown examples in Python, but many languages support functional programming. Languages—such as C++, Python, and the increasingly popular D—that support multiple styles of programming tend to get you to an acceptable software solution more directly than those that don’t. {end} Chuck Allison developed software for twenty years before becoming a professor of computer science at Utah Valley State College. He was senior editor of the C/C++ Users Journal and is def compose(funs): founding editor of The C++ return lambda x: reduce(lambda z,f: f(z), list(reversed(funs)), x) Source. Chuck is also the author of two C++ books and gives onListing 7 site training in C++, Python, and design patterns. def add1(stuff): return [x+1 for x in stuff] def mult2(stuff): return [x*2 for x in stuff] c = compose([add1, mult2]) print c(range(10)) # [1, 3, 5, 7, 9, 11, 13, 15, 17, 19] Which paradigms do you regularly use? Have you noticed that the language you use influences the solutions you create? M Listing 8 Follow the link on the StickyMinds.com homepage to join the conversation. 14 BETTER SOFTWARE NOVEMBER 2007 www.StickyMinds.com http://www.StickyMinds.com http://www.StickyMinds.com

Table of Contents Feed for the Digital Edition of Better Software - November 2007

Better Software - November 2007 Contents MarkYour Calendar Technically Speaking What’s Happening @StickyMinds.com Code Craft Test Connection Management Chronicles The Measure of a Management System Behind the Scenes A Story About User Stories and Test-Driven Development Product Announcements The Last Word Ad Index

Better Software - November 2007

For optimal viewing of this digital publication, please enable JavaScript and then refresh the page. If you would like to try to load the digital publication without using Flash Player detection, please click here.