DataReader

The DataReader gem is used to provide a standard mechanism for providing a YAML data path and loading data from it. DataReader is mainly used as a support gem that can be included by other libraries that need this functionality.

To get the latest stable release, add this line to your application's Gemfile:

gem 'data_reader'

To get the latest code:

gem 'data_reader', git: 'https://github.com/jeffnyman/data_reader'

After doing one of the above, execute the following command:

$ bundle

You can also install DataReader just as you would any other gem:

$ gem install data_reader

The basic idea of DataReader is simple: you set a data path and DataReader will load data files from that path. But there are some nuances that it's worth discussing in this documentation.

You can include the DataReader in a class or module.

require "data_reader"

class Testing
  include DataReader
end

This will provide DataReader functionality on any instance of the class where DataReader is mixed in.

DataReader does not set defaults for anything. It provides a data_path variable that you can set. It also provides a data_contents variable that will be populated with the result of a data file that gets loaded from any specified data path.

Consider the following file and directory setup:

project_dir\
  combined\
    invalid.yml
    conditions.yml
    stars.yml

  provision\
    invalid.yml

  stardates\
    conditions.yml

  warp\
    stars.yml

This is in fact the structure that is provided as part of the examples directory with this repository.

Within the project_dir you could create a file called script.rb to see how DataReader works. Put the above class in place in that file and then add this:

test = Testing.new

puts test.data_path
puts test.data_contents

This would print nothing for either of those values, showing that they have no default values. You could now do this:

test.data_path = 'warp'

puts test.data_path

Here you are setting the data_path to a directory called warp. The puts statement after that simply confirms that this was set. This has now set the data path for DataReader. Here the "data path" indicates where DataReader will look for data files. Thus you could load any file that is in that directory:

test.load 'stars.yml'

Loading causes the data from the file to be put into a data_contents attribute.

puts test.data_contents

The puts call for the data_contents will show you the contents of the stars.yml file.

You could set the data contents on the class instance if you wanted to:

class Testing
  include DataReader

  def data
    @data_contents
  end
end

Now you can access the data via:

puts test.data

The reason this might be useful is because the data contents may change, such as if you read different files at different times, but this way you refer to the relevant contents via one variable.

Note that you can change the data path on the fly if you need to. For example:

test.data_path = 'stardates'

test.load 'conditions.yml'

This would set the data path to the stardates directory and then the load file would grab the contents of the conditions.yml file.

You could have specified the data_path as a method on the class instead, like this:

class Testing
  include DataReader

  def data_path
    'provision'
  end
end

Then you don't have to set the path specifically as we've been doing.

Note that if you are setting the data_path on the class, the idea is that you want this to be the data path. So it can't be reassigned. To see that, try the above and the have the script logic as such:

test.data_path = 'provision'
test.load 'invalid.yml'

This would lead to an error because while you have set the data_path, that will not override what has been set on the class. So the upshot is that if you define a data_path as a method on the class, that's what will be used even if you re-define the data_path on a specific instance of that class.

You may want to make sure that a default data path is always available should a data path not have been specifically set. You can do that as follows:

class Testing
  include DataReader

  def default_data_path
    'provision'
  end
end

Keep in mind that DataReader will always favor whatever it has stored in data_path. The default_data_path can be used for a fallback. So, with the default data path specifed as above, consider this:

test = Testing.new

test.load 'invalid.yml'
puts test.data_contents

This owuld work just fine. But if you were to set the data path, that overrides the default:

test.data_path = 'warp'
test.load 'invalid.yml'

Here I've set the data_path but I'm still trying to load invalid.yml (which is in the provision directory). But the data_path, since it's set, overrides that. The upshot is that a specific data path overrides the default.

If you want to be able to revert to the default, you need to set the data_path to nil.

users: &users
  admin:
  - username: admin
  	password: admin

This would fail to load based on the &users part.

You can load multiple YAML files. The load method takes a list of comma separated names of files that are in that same directory. So if you were to place all the above example YAML files in one directory, such as the combined directory shown above, you could do this:

test.data_path = 'combined'

test.load 'stars.yml, conditions.yml, invalid.yml'

When loading in multiple files, the data_contents will hold the contents of all the files in the list.

You don't have to use the data_context value. For example, you could do this:

stars = test.load 'stars.yml'
invalid = test.load 'invalid.yml'

In this case, the appropriate data would be stored in each variable. Do note that data_context will always contain the last data read by the load method. So in the above case, data_context would contain the contents of invalid.yml even if you never intended to use that variable.

You can set environment variables in YAML files. To do this you have to use ERB, like this:

<%= ENV['XYZZY'] %>

To handle this, DataReader parses any values with ERB before it parses the YAML itself. Here's an example YAML file:

config:
  current:
    server: test
    user: jeff_nyman
    browser: <%= ENV['BROWSER'] %>

Now let's say I loaded up this file and looked at the data source:

test.load 'config.yml'
puts test.data_contents

Assuming the BROWSER environment variable was set, the data_contents variable would look as follows:

{
    "config" => {
        "current" => {
             "server" => "test",
               "user" => "jeff_nyman",
            "browser" => "chrome"
        }
    }
}

The support for ERB allows for custom method calls. One that is included with DataReader is include_data. First consider the directory structure:

included\
  included_nested.yml
  included.yml
  with_includes.yml
  with_nested_includes.yml

In the with_includes.yml file, there is a line like this:

<%= include_data("included.yml") %>

Now you can do this:

test.data_path = 'included'

test.load 'with_includes.yml'

This will load up with_includes.yml and, because of the include_data call would attempt to load the file included.yml.

In this case, the value of data_contents would contain both data sets, first the data from with_includes.yml and then the data from included.yml.

Note, however, that DataReader will attempt to load this from the same location as with_includes.yml. You can absolute or relative paths as part of the call, as such:

<%= include_data("../warp/stars.yml") %>

You can also extend, rather than include, DataReader. This means you deal with the class rather than an instance of it. For example:

require "data_reader"

class Testing
  extend DataReader
end

Testing.data_path = 'config'

puts Testing.data_path

Testing.load 'config.yml'

puts Testing.data_source

Note that you can provide methods as you did in the include case, but make sure they are defined on self. For example:

class Testing
  extend DataReader

  def self.data_path
    'config'
  end
end

If you were using default_data_path, likewise just make sure you prepend self to it.

After checking out the repo, run bin/setup to install dependencies. Then, run rake spec:all to run the tests. You can also run bin/console for an interactive prompt that will allow you to experiment.

The default rake command will run all tests as well as a RuboCop analysis.

To install this gem onto your local machine, run bundle exec rake install.

Bug reports and pull requests are welcome on GitHub at https://github.com/jeffnyman/data_reader. The testing ecosystem of Ruby is very large and this project is intended to be a welcoming arena for collaboration on yet another testing tool. As such, contributors are very much welcome but are expected to adhere to the Contributor Covenant code of conduct.

To contribute to DataReader:

1. Fork the project.
2. Create your feature branch. (git checkout -b my-new-feature)
3. Commit your changes. (git commit -am 'new feature')
4. Push the branch. (git push origin my-new-feature)
5. Create a new pull request.

• Jeff Nyman

This code is based upon the YmlReader gem. I wanted to give myself room to make a more generic version that may not be focused only on YAML files. More importantly, I wanted to clean up the implementation and documentation a bit.

DataReader is distributed under the MIT license. See the LICENSE file for details.