Better Code Workshop : Sean Parent

Introduction

Engineering is making informed trade-offs to find the best solution given a set of constraints.

My hope is that this workshop is generally applicable, but my experience colors the choices I've made.

Career Background

Demo

Course Outline

• Introduction
• Preface
• Types - Goal: Write complete, expressive, and efficient types
• Algorithms - Goal: No raw loops
• Data Structures - Goal: No incidental data structures
• Runtime Polymorphism - Goal: No raw pointers
• Concurrency - Goal: No raw synchronization primitives
• Relationships - Goal: No contradictions
• Epilogue

Materials

• Slides: https://sean-parent.stlab.cc/notebook
• Exercises: https://github.com/sean-parent/better-code-class

Use of Jupyter w/Xeus-Cling

• Currently limited to C++17
  • I will provide some C++20 examples
• using std; is implied
• Definitions wrapped in a namespace
• Namespaces are often versions so I can refine implementations

namespace v0 {

int f() { return 42; }

} // namespace v0

• Often code is wrapped in a scope so it doesn't interfere with other code
• If the last line doesn't have a semi-colon it is displayed
  • calling display(value); has the same effect

v0::f()

{
    using namespace v0;

    display(f());
}

• Operations can be timed with %%timeit
  • This is not a substitute for benchmarks but gives some information to compare

%%timeit
{
    int a[] = { 9, 8, 7, 6, 5, 4, 3, 2, 1, 0 };
    sort(begin(a), end(a));
}

• I use an instrumented class which prints common operations

{
    instrumented a;
    instrumented b;
    a = b;
    a = move(b);
}