Subjuster | TDD guide for Software Engineers in OOP

A Command-Line tool to adjust your movie subtitle files. Its a common issue that while playing movies audio and subtitle do not sync up. Normally it lags/gains by a few seconds/milliseconds. Using Subjuster You will be able to adjust and generate a new subtitle file.

Table Of Content

Intention / Purposes

I am writing this software and documentation for two of the main reasons

Note: If you feel that this doc/repo needs some modifications to help it better meet its purpose, please feel free to send a Pull Request. I would be very much pleased to merge it after reviewing.
If you feel this doc helpful; please don’t forget to put a Star.

Installation

This gem requires Ruby 2.0+.

  $ gem install subjuster

Usage

  Usage: subjuster [filename.srt] [options]

  Special Case:
  subjuster [fiename.srt] -a-12.23    # for -ve number i.e '-12.23'

  '+ve' number will add time while '-ve' will decrease. 
  i.e. if subtitles appears 2 sec after the audio then use '-2' as adjustment
  ---------------------------------------------------------------------------

      -a, --adjustment [Numeric]       Time adjustment in sec
      -t, --target [Filename]          If Target file name not given then will be '[source_file].modified.srt'
      -h, --help                       Prints this help

Example:

$ /movie/path inception_1080p.srt -t corrected.srt -a-12.34

source                => inception_1080p.srt
adjustment_in_sec     => -12.34

Yeah! successfully adjusted and compiled to file /home/john/corrected.srt

The file is generated in the current working directory.

Steps to TDD

Requirement Gathering

Before you start a software project, there should be a problem somewhere in world in the first place; which will you be solving. You gotta understand everything about the problem as possible.

Next will be, you figuring out ways to solve this problem. You are not supposed to the ultimate solution which is not gonna change ever; It’s not possible, your solution should be robust and always changeable because Requirements always change down the road.


Phase One: UML Drawing and Research

When you’ve thought of solutions, you grab notebook and pen; start making rough sketches(UML diagrams). Think of what your software does and list out separate tasks/actions to perform.

Example(1) of list of tasks you think of:-

  1. Takes input from the users; may be from CLI or via STDIN(Keyboard) Inputs will be filename.srt and number_of_seconds | +ve or -ve
  2. Parse the file supplied into some data-structure like Hash
  3. Modify hash with required adjustments
  4. Export or Generate the adjusted file to FileSystem.

Now, you draw Use Case diagram if possible. Its simple, go through some online blogs and start drawing. After that, you need to draw Class Diagram. This is necessary to figure out what components your software is gonna composed of.

Lets do it: Extract out nouns from the list in Example(1).

  1. UserInput
  2. Parsertoc
  3. Modifier / Adjuster
  4. Exporter / Generator

Rule: According to Single Responsibility Principle(SRP) of Object Oriented Design from SOLID, one class/module/function should not take more than one responsibility.

Use Case diagram of subjuster

We complied to the rules and decided to have 4 modules doing individual tasks and collaborate with each other sending messages. Now we draw Class Diagram of our first thought.

Brief Class Diagram

Class diagram in brief

Question: What is this Subjuster or Subjuster::Core module doing here?

This module is a wrapper to tie your application components together like Namespace. This module will facilitate your application; which we will cover in later topics.

Read this document in Class Diagrams and learn the meaning of arrows and boxes. This will help you in you career as well. Keep in mind that you don’t have to learn UML and gain expertise to learn TDD; you only have to be able to draw boxes and name them. You can use Draw.io to draw UML diagrams; it’s open source and free.

Detail Class Diagram

Class diagram and use case diagram of subjuster

Is it complicated?

If you feel lost, then no worries, forget about the relationship between modules; we will learn later. For now, focus on the boxes and its properties.

Phase Two | Generate a skeleton Ruby gem

We at the end are going to release this project as a standalone gem so that any one willing to use this can get benefit from this.

Following is command to generate a new gem skeleton using bundler.

$ bundle gem [subjuster]

The name of gem can be anything in your case. This command generates a bundler compatible gem skeleton which you can modify and build your idea.

Note: If you are not comfortable with gems then, simply follow the folder structure in this repo. You will get hang of it soon.


Phase Three | Write expectations from each modules in English

Joke about explaining expectations and results

It’s not possible for everybody to figure out which technique are you gonna use to implement that particular function in code before hand. Therefore, think of what you expect from that particular feature in English. You already know what to expect from that module/function, don’t you?. If you are confused then go ask your product owner about the requirements.

Lets write some expectations in pure English:-

1. UserInput
  - Should take `source_filepath` as argument while construction
  - Should take `target_filepath` as argument while construction
  - Should take `no of seconds` to be adjusted as argument while construction
  - Should be able to validate the inputs
    - `source_filepath` should be valid
    - should be a valid `.srt` file
  
2. Parser
  - Should take `user_input` as params
  - Should able to parse the valid `srt` file
  - `parse()`
    - Should return a `hash` of subtitle
      - hash should have `start-time` and `end-time` of every dialog in the hash

3. Modifier / Adjuster
  - Should be able to take the `Hash` containing `srt data-structure`
  - `adjust(no_of_seconds)`
    - Should return the modified version of `Hash` supplied
    - Should be able to adjust the srt `Hash`'s `start-time` and `end-time` by `+2` seconds if `+2` is passed as argument
    - Should not adjust the srt `Hash`'s `start-time` and `end-time` by `-2` seconds if `+2` is passed as argument

4. Exporter / Generator
  - Should accept `Modified Hash` as argument
  - `generate(target_filepath)`
    - Should be able to generate a valid `.srt` file to the path asked

This pure english can easily be transferred to RSpec’s DSL. See this file subjuster_spec.rb. I have already prepared the Pure English version of Test Cases using RSpec’s DSL. Now we need to put assertions to verify the usability of the modules. We will see that in next section.

Glimpse of Specs

describe UserInput do  
  it 'Should take `source_filepath` as argument while construction'
  it 'Should take `target_filepath` as argument while construction'
  it 'Should take `no of seconds` to be adjusted as argument while construction'
  
  context 'Should be able to validate the inputs' do
    it '`source_filepath` should be valid'
    it 'should be a valid `.srt` file'
  end
end

Defining TDD in simple language

It’s like signing a contract with everything needed to be done specified in formal language. This is needed to define the Job Done. We write test-cases before implementing a feature; which will let us know when that particular feature/function is done, so that we can move on to next function. It will help you stick with the features client/owner demanded. Often developers lose track of what they should be doing in that iteration.


Phase Four | Write Failing Expectations

Firstly, we do some cleaning; we separate out specs related to individual module to corresponding spec files, like, user_input_spec.rb will contain specs only related to UserInput module and so forth.

Now, we use our knowledge of Ruby and little bit of RSpec DSL. We pick one module at a time, because we are on war. We write expectations that we know it should fail because they are not yet implemented. I am implying syntax errors as well by failing test. It can be Class not defined kinda error though. The errors are intended and expected.

In the beginning we pick the simplest module i.e. UserInput. Its legit to go in sequence.

In the file user_input_spec.rb we pick the top most expectation i.e.

  it 'Should take `source_filepath` as argument while construction'

Now after putting an expectation it looks like:

  it 'Should take `source_filepath` as argument while construction' do
    source_filepath = '/tmp/source_file.srt'
    expect(UserInput.new(source: source_filepath).source_filepath).to eq(source_filepath)
  end

While running the command rspec spec/ I encounter this error

  NameError:
    uninitialized constant `UserInput`

  Finished in 0.00022 seconds (files took 0.40604 seconds to load)
  0 examples, 0 failures, 1 error occurred outside of examples

Error is Good.

To comply with our requirements, there should be a class UserInput in the first place. This technique is also called Error Driven Development. Errors and failure will guide you to the destination; a better tomorrow.

Little improvisation needed here; We need to use Subjuster namespace to bind modules in a box. Lets create a file called lib/subjuster/user_input.rb where our class UserInput will be defined. And, in lib/subjuster.rb we will require the file so that RSpec could load the file in test-suites. Also will use Subjuster::UserInput instead of just UserInput.


Phase Five | Write Production Code to Pass Failing Expectations

Next thing you do is, define the class UserInput.

Rule: You deal with one RED/Failing test at a time; you take that examples from Red –> Green; then only you touch next example/test-case.

# lib/subjuster/user_input.rb
  module Subjuster
    class UserInput
    end
  end

Now see the execution result:-

  $ rspec spec/user_input_spec.rb

  ArgumentError:
    wrong number of arguments (given 1, expected 0)
  # ./spec/user_input_spec.rb:6:in `initialize'
  # ./spec/user_input_spec.rb:6:in `new'
  # ./spec/user_input_spec.rb:6:in `block (2 levels) in <top (required)>'

It means we are advancing, we surpassed our first error and now in second. It says, may be we have not defined the initialize function with proper arguments. So, we obey the suggestion.

  module Subjuster
    class UserInput
      def initialize(source:)
      end
    end
  end

Now see the execution result:-

  $ rspec spec/user_input_spec.rb

  NoMethodError:
    undefined method `source_filepath` for #<Subjuster::UserInput:0x0056444cadabd8>
  # ./spec/user_input_spec.rb:6:in `block (2 levels) in <top (required)>'

Now, we need to define the getter method source_filepath

  module Subjuster
    class UserInput
      attr_reader :source_filepath

      def initialize(source:)
      end
    end
  end

Now see the execution result:-

  $ rspec spec/user_input_spec.rb

  Failure/Error: expect(Subjuster::UserInput.new(source: source_filepath).source_filepath).to eq(source_filepath)

    expected: "/tmp/source_file.srt"
         got: nil

    (compared using ==)
  # ./spec/user_input_spec.rb:6:in `block (2 levels) in <top (required)>'

It means, still the mill/function source_filepath is not functional. Now lets set instance var @source_filepath. Then, may be it will work.

  attr_reader :source_filepath
  def initialize(source:)
    @source_filepath = source
  end

Now see the execution result:-

  $ rspec spec/user_input_spec.rb

  Finished in 0.00242 seconds (files took 0.26666 seconds to load)
  5 examples, 0 failures, 4 pending

Hurrah! we did it. We made the Red test-case to GREEN.

Now, you can repeat the same process for Parser.
This process should be repeated till you are done with all the features. Please see the individual files in spec/ and lib/subjuster as references.

If still you need assistance then see the documentation in /tdd_for_parser.md.

Phase-5 Write_commits_in_steps_for_Parser

Phase-6_Write_commits_in_steps_for_Adjuster

Phase-7_write_steps_for_generator

Phase-8_write_failing_test_and_make_it_green_for_SubjusterCore

Some rules