Ruby vs Go – Structuring Data

I’m trying to write a Taskpaper parser in Go.

A very basic Taskpaper file looks like this:

Home:
- Mow the lawn
Don't hit the flowers

Work:
- Fix all the bugs
Seriously, all the bugs

Projects end with a : and tasks begin with a - and any line that doesn’t have that is a note. Projects have tasks and notes and tasks can have notes.

I got totally stuck yesterday on how to structure this data in Go.

If this was Ruby I would do it like this:

class Item
  attr_accessor :parent
  attr_accessor :notes
end

class Project < Item
  attr_accessor :tasks

  def initialize(desc)
    super(desc)
  end
end


class Task < Item
  def initialize(desc)
    super(desc)
  end
end


class Note < Item
  def initialize(desc)
    super(desc)
  end
end

An Item holds generic fields and then there’s a Project, Task and Note classes.

In Go there’s no way to say that a Note struct has just a generic Parent i.e. Parent *Item. To work around this I had to create a ProjectNote and also a TaskNote:

type Item struct {
  Desc string
}

type Project struct {
  Item
  Tasks []Task
  Notes []ProjectNote
  Tags  []ProjectTag
}

type Task struct {
  Item
  Notes  []TaskNote
  Tags   []TaskTag
  Parent *Project
}

type ProjectNote struct {
  Item
  Parent *Project
}

type TaskNote struct {
  Item
  Parent *Task
}

It feels a bit wrong to break these out like this but that’s probably because I’ve been stuck in the OO world of Ruby and Java for too long.