Geoff Bones

19 March 2014

Code kata 4: Word ladder

Naming things can be very hard, yet a well-chosen name can really help a reader understand how your code works. This code kata is an old favorite, one that we’ve done before, but this time we’ve added an emphasis on naming things well.

The kata

A word ladder is a sequence of words (each of which forms a rung on the ladder) where each word is formed from its predecessor in the sequence by a simple transformation. The possible transformations are substitution (changing a single letter for any other in the alphabet), rearrangement (swapping places of two adjacent letters in the word, inserting a letter, and deleting a letter.

Thus, if we take move and pave as our start and end words, then the following are two valid word ladders formed by substitution only:

1. move
2. more
3. core
4. care
5. cave
6. pave

and

1. move
2. cove
3. cave
4. pave

The task

Write a program that, given a start word and an end word, will compute a corresponding word ladder. Your program should prefer shorter ladders over longer ones.

Start off by using only substitution to form the ladder. If you have time at the end, add the ability to include rearrangements as well.

You can use any suitable word list, by the way, as your source of valid words.

Constraints

We’re going to do this kata by TDD, but with an added step to focus the task on clear naming. So, split out the TDD steps to look like this:

1. Red: Write a failing test.
2. Green: Implement the minimal, sensible functionality to pass the test.
3. Refactor until you’re proud of your code.
4. Rename any classes, members or variables that don’t explicitly state their purpose.