Nabo

Dead simple blog engine written in Elixir.

Nabo is designed to be a simple, fast, extendable blog engine. You can integrate Nabo to any component in your application like Phoenix, Plug.

It does not include routers or html generators, but focuses on doing one thing and does it well: manages your blog posts.

Installation

Add :nabo to your mix.exs.

By default Nabo uses Jason and Earmark for its JSON and Markdown parsers.

def deps() do
  [
    {:nabo, "~> 1.0.2", runtime: false},
    {:jason, "~> 1.0.0", runtime: false}, # optional
    {:earmark, "~> 1.2.4", runtime: false} # optional
  ]
end

Documentation

Full documentation can be found on HexDocs and you can view an example of how to integrate Nabo to Phoenix here.

Getting started

To start using Nabo, first you need to create your own repo. Use this mix task:

mix nabo.gen.repo MyApp.Repo

Nabo assumes your posts are in priv/posts, if they are not, you can change the :root option in the generated repo.

defmodule MyWeb.Repo do
  use Nabo.Repo, root: "path/to/posts"
end

Nabo provides nabo.gen.post mix task to generate posts.

mix nabo.gen.post my-first-blog-post

All posts in Nabo should follow this format:

{
  "title": "Welcome to Nabo",
  "slug": "welcome",
  "published_at": "2017-01-01T00:00:00Z"
}
---
This is the excerpt of the post in markdown
---
This is the *body* of the post in markdown

Phoenix integration

defmodule MyWeb.PostController do
  use MyWeb.Web, :controller

  def index(conn, _params) do
    posts = MyWeb.Repo.all()
    render conn, "index.html", posts: posts
  end

  def show(conn, %{"id" => slug}) do
    {:ok, post} = MyWeb.Repo.get(slug)
    #or post = MyWeb.Repo.get!(slug) This will raise if no post was found
    render conn, "show.html", post: post
  end
end

Then in your template.

# index.html.eex
<div class="posts">
  <%= for post <- posts do %>
  <div class="post">
    <h3><%= post.title %></h3>
    <div class="excerpt"><%= post.excerpt_html %></div>
  </div>
  <% end %>
</div>

# show.html.eex
<div class="post">
  <h1><%= post.title %></h1>
  <div class="body"><%= post.body_html %></div>
</div>

Plug integration

defmodule MyWeb.Router do
  use Plug.Router
  import Plug.Conn

  plug :match
  plug :dispatch

  get "/posts" do
    posts = MyWeb.Repo.all()

    body =
      posts
      |> Enum.map(& %{title: &1.title, slug: &1.slug})
      |> Poison.encode!

    conn
    |> put_resp_header("content-type", "application/json;charset=utf-8")
    |> send_resp(200, body)
    |> halt()
  end

  get "/posts/:slug" do
    case MyWeb.Repo.get(slug) do
      {:ok, %Nabo.Post{title: title, body_html: body_html}} ->
        body = %{title: title, body: body_html} |> Poison.encode!
        conn
        |> put_resp_header("content-type", "application/json;charset=utf-8")
        |> send_resp(200, body)
        |> halt()
      {:error, reason} -> send_resp(conn, 404, reason) |> halt()
    end
  end
end

Extending Nabo

By default Nabo uses JSON for post metadata and Markdown for post excerpt and body, but if you prefer them in other format, you are encourage to write your own parsers.

defmodule MyBBCodeParser do
  @behaviour Nabo.Parser

  def parse(binary, options) do
    BBCode.to_html(binary, options)
  end
end

Then configure your parser to the repo.

defmodule MyRepo do
  use Nabo.Repo,
      root: "path/to/posts",
      compiler: [
        body_parser: {MyBBCodeParser, []}
      ]
end

And that's just it!

See the documentation for more Nabo.Repo options.

Q&A

How does Nabo work?

Nabo parses all posts in the configured repo in compile-time, keeps them in the module, and returns them when you ask in run-time.

Why this post engine does not support generating static html like Jekyll and such?

Nabo is not meant to replace Jekyll since Jekyll has done their job really well (trust me the previous version of my blog was built with Jekyll). But Nabo takes another approach and is designed to integrate with other components in your application. If you want to build a blog with more controllable functions like comments or traffic count, then Nabo might fit your needs. If you need a minimal version of static html blog that can play well with Github page, then Jekyll would probably be your choice.

Why Nabo does not support template, router, controller or server?

As mentioned, Nabo is designed to be simple and integrate-able with other components in your application so it does not presume your needs.

If you want routing or templating, you can easily integrate Nabo to Phoenix or Plug.

How customizable and extendable is Nabo?

Nabo uses JSON for metadata and Markdown for post excerpt and body by default, that is quite an easy decision. But they are optional.

You can control the format of all the three components mentioned above, even the delimiter between them. Nothing is going to stop you to have your own XML metadata, Markdown for excerpt, and BBCode for post body.

Websites built with Nabo

• hqc.io
• quan-cam.com

Development and issue report

If you have any issues or ideas, feel free to write to https://github.com/qcam/nabo/issues.