#PyOhio_talk_2013

Notes and slides for talk on Code Reading Python

###Writing small programs is no longer a problem, how to transition to longer ones?

• learn from others, check out open source projects
• better yet, look at what interests you and contribute
• you'll learn:
  • how to partition programs, (e.g. modules, objects, utilities)
  • see other's pattern usage:
    • Architectural patterns, (e.g. MVC, event-driven)
    • Design patterns
    • even simpler ones that never got their own name

###How to wrap your mind around big programs

• There is only one book on this, "Code Reading"
  • We will pull out what we can from this, but it's examples are not in Python ...
• Oddly enough, a lot of techniques here are inspired by an book on teaching reading for content: "Content Area Reading"
  • you can't read code, "cover to cover" and get much of a handle on it
  • you need strategies, (remember SQ3R?) our teachers tried ...
    • use, (and build-up), your knowledge object system design to speculate on:
      • what objects would you expect to find in this project?
        • what real world things exist in the domain of the project, are they modelled?
      • what responsibilites are needed, which object would have each?
    • use, (and build-up), your knowledge of re-factoring to speculate on what you'll find
    • use, (and build-up), naming guidelines to specualate on the objects and methods you'll find
  • once have done a bit of specualation, read the code to see how accurate they were, (whose ideas are better?)
  • you need to go "multimodal"
    • pictures
    • sounds, (hey, talk out loud!)
    • motion, (hey, play is good!)
  • you need ways to check your understanding
  • you need ways to "lift the hood"

###We will illustrate this by looking at the open source program, "Plover"

• What is this program supposed to do?
• Speculate how you would do that
• Read with an eye to how they did it & how that compares to your idea
• check out the module structure, speculate from their names what they might do
• any Architectural, (i.e. App-wide), pattern being used?
• check out classes, their variables and methods
  • do the names make sense?
• Mnemonic tools
  • using pictures
  • give your memory a hand (Anki)
• Analysis tools
  • outliners: Leo, VOom
  • grep
• Under the hood tools
  • Ipython
  • Winpdb
• Probing for understanding with Tests