Python Cookbook, 2nd Edition

Chapter 16. Programs About Programs

Introduction

Credit: Paul F. Dubois, Ph.D., Program for Climate Model Diagnosis and Intercomparison, Lawrence Livermore National Laboratory

This chapter was originally meant to cover mainly topics such as lexing, parsing, and code generation—the classic issues of programs that are about programs. It turns out, however, that Pythonistas did not post many recipes about such tasks, focusing more on highly Python-specific topics such as program introspection, dynamic importing, and generation of functions by closure. Many of those recipes, we decided, were more properly located in various other chapters—on shortcuts, debugging, object oriented programming, algorithms, metaprogramming, and specific areas such as the handling of text, files, and persistence Therefore, you will find those topics covered in other chapters. In this chapter, we included only those recipes that are still best described as programs about programs. Of these, probably the most important one is that about currying, the creation of new functions by predetermining some arguments of other functions.

This arrangement doesn’t mean that the classic issues aren’t important! Python has extensive facilities related to lexing and parsing, as well as a large number of user-contributed modules related to parsing standard languages, which reduces the need for doing your own programming. If Pythonistas are not using these tools, then, in this one area, they are doing more work than they need ...