Home

Awesome

edn-ruby

Build Status

© 2012 Relevance Inc

edn-ruby is a Ruby library to read and write EDN (extensible data notation), a subset of Clojure used for transferring data between applications, much like JSON, YAML, or XML.

Installation

Add this line to your application's Gemfile:

gem 'edn'

And then execute:

$ bundle

Or install it yourself as:

$ gem install edn

Note that you might also want to look at edn_turbo which provides a much faster EDN parser (It's written in C) with an interface that is largely compatible with ths gem.

Usage

To read a string of EDN:

EDN.read('[1 2 {:foo "bar"}]')

Alternatively you can pass in an IO instance, for example an open file:

File.open("data.edn") do |f|
  data = EDN.read(f)
  # Do something with data
end

By default EDN.read will throw an execption if you try to read past the end of the data:

EDN.read("")   # Boom!

Alternatively, the EDN.read method takes an optional parameter, which is the value to return when it hits the end of data:

EDN.read("", :nomore)

#=> :nomore

There is no problem using nil as an eof value.

EDN::Reader

You can also do things in a more object oriented way by creating instances of EDN::Reader:

r = EDN::Reader.new('[1 2 3] {:a 1 :b 2}')

r.read #=> [1, 2, 3]
r.read #=> {:a => 1, :b => 2}
r.read #=> RuntimeError: Unexpected end of file

EDN:Reader will also take an IO instance:

r = EDN::Reader.new(open("data.edn"))

r.read  # Read the first form from the file.
r.read  # Read the second form from the file.
r.read  # Read the third from from the file.

You can also iterate through the forms with each:

r = EDN::Reader.new('[1 2 3] {:a 1 :b 2}')

r.each do |form|
  p form
end

#=> [1, 2, 3]
#=> {:a => 1, :b => 2}

Note that in contrast to earlier versions of this gem, EDN::Reader is no longer Enumerable.

Like EDN.read, Reader.read also takes an optional parameter, which is returned when there is no more data:

r = EDN::Reader.new('1 2 3')
r.read(:eof)  # returns 1
r.read(:eof)  # returns 2
r.read(:eof)  # returns 3
r.read(:eof)  # returns :eof

Converting Ruby data to EDN

To convert a data structure to an EDN string:

data.to_edn

By default, this will work for strings, symbols, numbers, arrays, hashes, sets, nil, Time, and boolean values.

Value Translations

Note that EDN uses its own terminology for the types of objects it represents and in some cases those types not map cleanly to Ruby.

In EDN, you have keywords, which look like Ruby symbols and have the same meaning and purpose. These are converted to Ruby symbols.

You also have EDN symbols, which generally reflect variable names, but have several purposes. We parse these and return EDN::Type::Symbol values for them, as they don't map to anything built into Ruby. To create an EDN symbol in Ruby, call EDN::Type::Symbol.new or EDN.symbol with a string argument, or use the convenience unary operator ~ like so: ~"elf/rings".

EDN also has vectors, which map to Ruby arrays, and lists, which are linked lists in Clojure. We map EDN lists to EDN::Type::List values, which are type-compatible with arrays. To create an EDN list in Ruby, call EDN::Type::List.new or EDN.list with all arguments to go in the list. If you have an array, you will use the splat operator, like so: EDN.list(*[1, 2, 3]). You can also use the ~ unary operator like so: ~[1, 2, 3].

EDN also has character types, but Ruby does not. These are converted into one-character strings.

Tagged Values

The interesting part of EDN is the extensible part. Data can be be tagged to coerce interpretation of it to a particular data type. An example of a tagged data element:

#wolf/pack {:alpha "Greybeard" :betas ["Frostpaw" "Blackwind" "Bloodjaw"]}

The tag (#wolf/pack) will tell any consumers of this data to use a data type registered to handle wolf/pack to represent this data.

The rules for tags from the EDN README should be followed. In short, custom tags should have a prefix (the part before the /) designating the user that created them or context they are used in. Non-prefixed tags are reserved for built-in tags.

There are two tags built in by default: #uuid, used for UUIDs, and #inst, used for an instant in time. In edn-ruby, #inst is converted to a Time, and Time values are tagged as #inst. There is not a UUID data type built into Ruby, so #uuid is converted to an instance of EDN::Type::UUID.

Tags that are not registered generate a struct of the type EDN::Type::Unknown with the methods tag and value.

Registering a New Tag For Reading

To register a tag for reading, call the method EDN.register with a tag and one of the following:

Examples:

EDN.register("clinton/uri") do |uri|
  URI(uri)
end

EDN.register("clinton/date", lambda { |date_array| Date.new(*date_array) })

class Dog
  def initialize(name)
    @name = name
  end
end

EDN.register("clinton/dog", Dog)

Writing Tags

Writing tags should be done as part of the class's .to_edn method, like so:

class Dog
  def to_edn
    ["#clinton/dog", @name.to_edn].join(" ")
  end
end

EDN provides a helper method, EDN.tagout:

class Dog
  def to_edn
    EDN.tagout("clinton/dog", @name)
  end
end

This method calls .to_edn on the second argument and joins the arguments appropriately.

Other examples are:

EDN.tagout("wolf/pack", {:alpha=>"Greybeard", :betas=>["Frostpaw", "Blackwind", "Bloodjaw"]})
 => "#wolf/pack {:alpha \"Greybeard\", :betas [\"Frostpaw\" \"Blackwind\" \"Bloodjaw\"]}"

class Range
  def to_edn
    EDN.tagout("ruby/range", [self.begin, self.end, self.exclude_end?])
  end
end

(0..9).to_edn
=> "#ruby/range [0 9 false]"

Metadata

Certain elements of EDN can have metadata. Metadata is a map of values about the element, which must follow specific rules.

Metadata can be expressed in one of the following three ways:

More than one piece of metadata can be applied to an element. Metadata is applied to the next element appearing after it, so in the case of ^:foo ^{:bar false} [1 2], the metadata would be, in total, ^{:foo true, :bar false}. Note that ^:foo is applied to the element [1 2] with the metadata ^{:bar false} applied to it. Because of this, key collisions are resolved right-to-left.

Contributors

Contributing

  1. Fork it
  2. Create your feature branch (git checkout -b my-new-feature)
  3. Commit your changes (git commit -am 'Added some feature')
  4. Push to the branch (git push origin my-new-feature)
  5. Create new Pull Request