This is an Elixir-based SDK for Prismic.io Mostly based on https://github.com/prismicio/ruby-kit and https://github.com/prismicio/javascript-kit

• The primary authors of this library are:
  1. Coburn Berry (coburncoburn)
  • API
  • Cache
  • Predicates
  2. David Wu (sudostack)
  • Fragments
  • Parser
• TODOs:
  • Add caching on api entrypoint request ( Prismic.api/2 )
  • Add optional caching on document requests ( submitting search form )
  • Support authentication in the API

def deps do
  [
      {:prismic, git: "https://github.com/TheRealReal/prismic-elixir", branch: "master"}
  ]
end

set repo url in your project's config

  config :prismic,
    repo_url: "https://micro.cdn.prismic.io/api"

Prismic.V2.API.new()
|> SearchForm.from_api()
|> SearchForm.set_ref(<ref>) # master ref / versioned ref
|> SearchForm.set_query_predicates([Predicate.at("document.id", <id>)])
|> SearchForm.submit()
|> Map.fetch!(:results)
|> Parser.parse_documents() # %Prismic.Document{id: ..., uid: ..., href: ..., fragments...}

The default HTTP Client is Poison. It is possible to use any http client that implements the Prismic.HTTPClient behaviour .

Then, set the HTTPClient Module in config or within

    Application.put_env(:prismic, :http_client_module, MyApp.HTTPClient)

The default Cache is an Agent . It is possible to use any cache that implements the Prismic.Cache behaviour .

Then, set the Cache Module in config or within

    Application.put_env(:prismic, :cache_module, MyApp.Cache)

Only the first leg of the request ( api entrypoint ) is cached. Actual document queries are not cached. It is probably better to just cache calls to the library itself rather than worrying about caching in the library.