This chapter will be about getting started with Snap. We will begin by giving some backround and progress through a working scaffolding app. A passing familiarity with Haskell development and the command line is assumed, but a brief rundown is given for those that do not have the background.
This book strives to place example code in your hands rather than lead you through a project. I suggest that you pick an idea such as a simple blog, where you input some data in a form and can render posts at different urls. Then you can use the code examples in the book to build up your project little by little.
The chapters may depend on each other (The Authorization Snaplet references the PostgreSQL chapter for example) and you are encouraged to bounce around while reading the book.
Snap is a web framework for the Haskell language that is most similar to Express or Sinatra. As contrast, Yesod (Another Haskell framework), can be viewed as a more Rails-like framework.
Snap is a simple web development framework for unix systems, written in the Haskell programming language. Snap has a high level of test coverage and is well-documented. Features include:
- A fast HTTP server library
- A sensible and clean monad for web programming
- An HTML-based templating system for generating pages
Snap also contains
snaplets, which are modular units of stateful code that are
usable between applications. For example, there are snaplets for heist,
authentication and Postgres. Snaplets also come with some nice extras such as a unified configuration format.
One of the nice things about the Haskell web framework landscape is that many of the components are replaceable with components from other projects. We will not be going into that in this book but you can read more here
By far the easiest way to install Haskell is the Haskell Platform, which is availible from http://www.haskell.org/platform/. Some package managers also include haskell-platform as well, including homebrew and apt-get. This book is written using the 2013.2.0.0 Haskell Platform, which includes GHC 7.6.3.
It is a good idea to separate Haskell projects from each other if there are multiple Haskell projects on a machine. This makes it easier to manage dependencies and enables a simpler workflow.
There are a few options for creating environments, we will go over
After installing Haskell Platform, we need to upgrade cabal to ~v1.18. Run:
cabal update cabal install cabal-install
Then you will see something like this at the end:
Installing executable(s) in /Library/Haskell/ghc-7.6.3/lib/cabal-install-126.96.36.199/bin Installed cabal-install-188.8.131.52 Updating documentation index /Library/Haskell/doc/index.html
According to the code we just ran, the updated
cabal is in
/Library/Haskell/ghc-7.6.3/lib/cabal-install-184.108.40.206/bin so we need to add that to our path:
Note that this path will be different based on your system.
Now that we have a recent cabal version, we can use
cabal sandbox. For a deeper understanding of
cabal sandbox, refer to this post.
cabal sandbox init
will create a directory and a file in the current directory that will hold packages and other settings for the sandbox.
Now, when we run
cabal install for this project, the executable will be in
.cabal-sandbox/bin/. For example, if we are in
code/heist-app (one of the sample projects that comes with the book) and we created a sandbox in
heist-app, we can install and run the executable without affecting our other projects as such:
cd code/heist-app cabal sandbox init cabal install .cabal-sandbox/bin/heist-app
which will result in
Initializing app @ / Initializing heist @ / ...loaded 7 templates from /heist-app/snaplets/heist/templates Initializing CookieSession @ /sess Initializing JsonFileAuthManager @ /auth Listening on http://0.0.0.0:8000/
Installing Snap is easy because of cabal, Haskell's package manager. Just run:
cabal install snap