Personal tools

How to write a Haskell program

From HaskellWiki

(Difference between revisions)
Jump to: navigation, search
(Deleted section. Link to don's old blog was broken and article was missing on Don's new blog.)
(fix link to "Simple Unit Testing in Haskell")
 
(17 intermediate revisions by 5 users not shown)
Line 12: Line 12:
 
=== Revision control ===
 
=== Revision control ===
   
Use [http://git-scm.com/ git] or [http://darcs.net darcs] unless you have a specific reason not to. Both are lightweight distributed revision control systems (and darcs is written in Haskell). Both have massive market share in the Haskell world. If you want to encourage contributions from other Haskell hackers then git or darcs are the best. Darcs hosting is available on [http://hub.darcs.net/ hub.darcs.net]. For git, [http://github.com/ github] is very popular.
+
Use [http://git-scm.com/ git] or [http://darcs.net darcs] unless you have a specific reason not to. Both are lightweight distributed revision control system. Darcs is written in Haskell. They are the two most popular [http://en.wikipedia.org/wiki/Distributed_version_control_system DVCS]es in the Haskell world. If you want to encourage contributions from other Haskell hackers then git is best. For git, [http://github.com/ GitHub] is very popular. Darcs hosting is available on [http://hub.darcs.net/ hub.darcs.net].
   
 
=== Build system ===
 
=== Build system ===
Line 23: Line 23:
 
You should use [http://haskell.org/cabal/download.html cabal-install] as a front-end for installing your Cabal library. Cabal-install provides commands not only for building libraries but also for installing them from, and uploading them to, Hackage. As a bonus, for almost all programs, it's faster than using Setup.hs scripts directly, since no time is wasted compiling the scripts. (This does not apply for programs that use custom Setup.hs scripts, since those need to be compiled even when using cabal-install.)
 
You should use [http://haskell.org/cabal/download.html cabal-install] as a front-end for installing your Cabal library. Cabal-install provides commands not only for building libraries but also for installing them from, and uploading them to, Hackage. As a bonus, for almost all programs, it's faster than using Setup.hs scripts directly, since no time is wasted compiling the scripts. (This does not apply for programs that use custom Setup.hs scripts, since those need to be compiled even when using cabal-install.)
   
cabal-install is widely available, as part of the [http://haskell.org/platform Haskell Platform], so you can probably assume your users will have it too.
+
cabal-install is widely available, as a [http://www.haskell.org/cabal/download.html binary distribution] or as part of the [http://haskell.org/platform Haskell Platform], so you can probably assume your users will have it too.
   
 
=== Documentation ===
 
=== Documentation ===
   
For libraries, use [http://haskell.org/haddock/ Haddock]. We recommend
+
For libraries, use [http://haskell.org/haddock/ Haddock]. Haddock generates [http://hackage.haskell.org/package/base-4.7.0.0/docs/Prelude.html nice markup], with links to source.
using the version of Haddock that ships with the Haskell Platform. Haddock generates [http://hackage.haskell.org/packages/archive/base/4.3.1.0/doc/html/Prelude.html nice markup], with links to source.
 
   
 
=== Testing ===
 
=== Testing ===
   
You can use [http://hackage.haskell.org/package/QuickCheck QuickCheck] or [http://www.mail-archive.com/haskell@haskell.org/msg19215.html SmallCheck] to test pure code. To test impure code, use [http://hackage.haskell.org/cgi-bin/hackage-scripts/package/HUnit HUnit]. See [http://hackage.haskell.org/packages/archive/hashable/1.1.2.2/hashable.cabal this Cabal file] for an example of how to include tests in your Cabal package.
+
Typical unit/spec based testing, particularly with impure code, can be done with [http://hspec.github.io/ HSpec] and [http://hackage.haskell.org/package/HUnit HUnit].
   
To get started, try [[Introduction to QuickCheck]]. For a slightly more advanced introduction, [http://blog.codersbase.com/2006/09/simple-unit-testing-in-haskell.html Simple Unit Testing in Haskell] is a blog article about creating a testing framework for QuickCheck using some Template Haskell. For HUnit, see [[HUnit 1.0 User's Guide]]
+
You can use [http://hackage.haskell.org/package/QuickCheck QuickCheck] or [http://www.mail-archive.com/haskell@haskell.org/msg19215.html SmallCheck] to test pure code. These libraries work best when you have known invariants in your code's behavior. See [http://hackage.haskell.org/package/hashable-1.1.2.2/hashable.cabal this Cabal file] for an example of how to include tests in your Cabal package.
  +
  +
To get started, try [[Introduction to QuickCheck]]. For a slightly more advanced introduction, [http://blog.codersbase.com/posts/2006-09-01-simple-unit-testing.html Simple Unit Testing in Haskell] is a blog article about creating a testing framework for QuickCheck using some Template Haskell. For HUnit, see [[HUnit 1.0 User's Guide]]
   
 
=== Distribution ===
 
=== Distribution ===
Line 44: Line 44:
 
=== Target Environment ===
 
=== Target Environment ===
   
If at all possible, depend on libraries that are provided by the [http://haskell.org/platform Haskell Platform], and libraries that in turn build against the Haskell Platform. This set of libraries is designed to be widely available, so your end users will be able to build your software.
+
If possible, depend on libraries that are provided by the current stable [http://daniel-diaz.github.io/stackagelist/ Stackage package versions] or the [http://haskell.org/platform Haskell Platform]. Ideally your package should be able to build with users that keep current with Stackage or users of the Platform.
   
 
== Structure of a simple project ==
 
== Structure of a simple project ==
Line 55: Line 55:
 
* haq.cabal -- the cabal build description
 
* haq.cabal -- the cabal build description
 
* Setup.hs -- build script itself
 
* Setup.hs -- build script itself
* _darcs -- revision control
+
* .git -- revision control
 
* README -- info
 
* README -- info
 
* LICENSE -- license
 
* LICENSE -- license
Line 62: Line 62:
 
modules. See [[Structure of a Haskell project]] for an example of a larger project's directory structure.
 
modules. See [[Structure of a Haskell project]] for an example of a larger project's directory structure.
   
Here is a transcript that shows how you'd create a minimal darcs and cabalised
+
Here is a transcript that shows how you'd create a minimal git and cabalised
 
Haskell project for the cool new Haskell program "haq", build it,
 
Haskell project for the cool new Haskell program "haq", build it,
 
install it and release.
 
install it and release.
Line 105: Line 105:
   
 
<code>
 
<code>
$ darcs init
+
$ git init
$ darcs add Haq.hs
+
$ git add Haq.hs
$ darcs record
+
$ git commit -am "my commit message"
addfile ./Haq.hs
 
Shall I record this change? (1/?) [ynWsfqadjkc], or ? for help: y
 
hunk ./Haq.hs 1
 
+--
 
+-- Copyright (c) 2006 Don Stewart - http://www.cse.unsw.edu.au/~dons/
 
+-- GPL version 2 or later (see http://www.gnu.org/copyleft/gpl.html)
 
+--
 
+import System.Environment
 
+
 
+-- | 'main' runs the main program
 
+main :: IO ()
 
+main = getArgs >>= print . haqify . head
 
+
 
+haqify s = "Haq! " ++ s
 
Shall I record this change? (2/?) [ynWsfqadjkc], or ? for help: y
 
What is the patch name? Import haq source
 
Do you want to add a long comment? [yn]n
 
Finished recording patch 'Import haq source'
 
 
</code>
 
</code>
   
And we can see that darcs is now running the show:
+
=== Add a build system ===
   
  +
Create a .cabal file describing how to build your project:
 
<code>
 
<code>
$ ls
+
$ cabal init
Haq.hs _darcs
 
 
</code>
 
</code>
   
=== Add a build system ===
+
that will ask a few questions about your project and generate a file similar to the example.
  +
<code>
  +
-- Initial scratch.cabal generated by cabal init. For further
  +
-- documentation, see http://haskell.org/cabal/users-guide/
   
Create a .cabal file describing how to build your project:
+
name: haq
  +
version: 0.1.0.0
  +
description: Super cool mega lambdas
  +
license: GPL
  +
license-file: LICENSE
  +
author: Don Stewart
  +
maintainer: dons@cse.unsw.edu.au
  +
build-type: Simple
  +
cabal-version: >=1.10
   
<code>
+
executable haq
$ cat > haq.cabal
+
main-is: Haq.hs
Name: haq
+
build-depends: base >=4.5 && <4.8
Version: 0.0
+
default-language: Haskell2010
Description: Super cool mega lambdas
 
License: GPL
 
License-file: LICENSE
 
Author: Don Stewart
 
Maintainer: dons@cse.unsw.edu.au
 
Build-Type: Simple
 
Cabal-Version: >=1.2
 
 
Executable haq
 
Main-is: Haq.hs
 
Build-Depends: base >= 3 && < 5
 
 
</code>
 
</code>
   
(If your package uses other packages, e.g. <tt>haskell98</tt>, you'll need to add them to the <tt>Build-Depends:</tt> field as a comma separated list.)
+
(If your package uses other packages, e.g. <tt>text</tt>, you'll need to add them to the <tt>build-depends:</tt> field as a comma separated list.)
Add a <tt>Setup.hs</tt> that will actually do the building:
 
   
<haskell>
+
Cabal will also generate a <tt>Setup.hs</tt> file that will do the actual building. You will rarely need to modify it.
$ cat > Setup.hs
 
import Distribution.Simple
 
main = defaultMain
 
</haskell>
 
Cabal allows either <tt>Setup.hs</tt> or <tt>Setup.lhs</tt>.
 
   
Now would also be a good time to add a LICENSE file and a README file. Examples are in the tarball for HNop.
+
If you specifed a known license, it will also add a LICENSE file.
  +
  +
You might like to add a README file to tell what your project is about.
   
 
Record your changes:
 
Record your changes:
   
 
<code>
 
<code>
$ darcs add haq.cabal Setup.hs LICENSE README
+
$ git add haq.cabal Setup.hs LICENSE README
$ darcs record --all
+
$ git commit -am "Add a build system"
What is the patch name? Add a build system
 
Do you want to add a long comment? [yn]n
 
Finished recording patch 'Add a build system'
 
 
</code>
 
</code>
   
 
=== Build your project ===
 
=== Build your project ===
   
Now build it! There are two methods of accessing Cabal functionality: through your Setup.hs script or through cabal-install. In most cases, cabal-install is now the preferred method.
+
Now build it! There are two methods of accessing Cabal functionality: through your Setup.hs script or through cabal-install. cabal-install is now the preferred method.
   
Building using cabal-install:
+
Building using cabal-install and sandboxes, always use sandboxes in your projects unless you are an expert and know what you are doing! Demonstrated here:
   
 
<code>
 
<code>
$ cabal install --prefix=$HOME --user
+
$ cabal sandbox init
  +
$ cabal install -j
 
</code>
 
</code>
   
Building using the traditional Setup.hs method:
+
This will install your newly minted haq program in $PROJECT_DIR/.cabal-sandbox/bin.
 
<code>
 
$ runhaskell Setup configure --prefix=$HOME --user
 
$ runhaskell Setup build
 
$ runhaskell Setup install
 
</code>
 
 
This will install your newly minted haq program in $HOME/bin.
 
   
 
=== Run it ===
 
=== Run it ===
Line 155: Line 156:
 
And now you can run your cool project:
 
And now you can run your cool project:
 
<code>
 
<code>
$ haq me
+
$ .cabal-sandbox/bin/haq me
 
"Haq! me"
 
"Haq! me"
 
</code>
 
</code>
   
You can also run it in-place, even if you skip the install phase:
+
Since our program doesn't rely on any other assets we can just copy it to a directory in our path if we want to be able to use it from anywhere without referencing the sandbox directory. Like so:
  +
 
<code>
 
<code>
$ dist/build/haq/haq you
+
$ sudo cp .cabal-sandbox/bin/haq /usr/local/bin/haq
"Haq! you"
+
$ haq me
  +
"Haq! me"
 
</code>
 
</code>
   
Line 174: Line 175:
 
</code>
 
</code>
   
Traditional method:
+
Using cabal install if you planned to upload your haq package to Hackage:
 
<code>
 
<code>
$ runhaskell Setup haddock
+
$ cabal haddock --hyperlink-source --html-location='http://hackage.haskell.org/package/haq/docs' --contents-location='http://hackage.haskell.org/package/haq'
 
</code>
 
</code>
   
Line 224: Line 225:
 
</haskell>
 
</haskell>
   
=== Add some automated testing: QuickCheck ===
+
=== Add some tests ===
   
==== QuickCheck v1 ====
+
==== HSpec ====
   
We'll use QuickCheck to specify a simple property of our Haq.hs code. Create a tests module, Tests.hs, with some QuickCheck boilerplate:
+
If you'd like to use HSpec:
   
<haskell>
+
To your cabal, add:
$ cat > Tests.hs
 
import Char
 
import List
 
import Test.QuickCheck
 
import Text.Printf
 
   
main = mapM_ (\(s,a) -> printf "%-25s: " s >> a) tests
+
<code>
  +
test-suite tests
  +
ghc-options: -Wall
  +
default-extensions: OverloadedStrings
  +
type: exitcode-stdio-1.0
  +
main-is: HSpecTests.hs
  +
build-depends: base,
  +
haq,
  +
hspec >= 1.8
  +
default-language: Haskell2010
  +
</code>
   
instance Arbitrary Char where
+
To enable tests and install HSpec (and any other needed dependencies):
arbitrary = choose ('\0', '\128')
+
coarbitrary c = variant (ord c `rem` 4)
+
<code>
</haskell>
+
$ cabal install --enable-tests
  +
</code>
   
Now let's write a simple property:
+
Then for your actual file containing the test code:
   
 
<haskell>
 
<haskell>
$ cat >> Tests.hs
+
$ cat > HSpecTests.hs
-- reversing twice a finite list, is the same as identity
+
module Main where
prop_reversereverse s = (reverse . reverse) s == id s
 
where _ = s :: [Int]
 
   
-- and add this to the tests list
+
import Haq
tests = [("reverse.reverse/id", test prop_reversereverse)]
+
import Test.Hspec
</haskell>
 
   
We can now run this test, and have QuickCheck generate the test data:
+
main :: IO ()
  +
main = hspec $ do
   
<code>
+
describe "Validate haqify function" $ do
$ runhaskell Tests.hs
+
it "haqify is supposed to prefix Haq! to things" $ do
reverse.reverse/id : OK, passed 100 tests.
+
haqify "me" `shouldBe` "Haq! me"
</code>
 
 
Let's add a test for the 'haqify' function:
 
 
<haskell>
 
-- Dropping the "Haq! " string is the same as identity
 
prop_haq s = drop (length "Haq! ") (haqify s) == id s
 
where haqify s = "Haq! " ++ s
 
   
tests = [("reverse.reverse/id", test prop_reversereverse)
 
,("drop.haq/id", test prop_haq)]
 
 
</haskell>
 
</haskell>
   
and let's test that:
+
Execute the tests with:
   
 
<code>
 
<code>
$ runhaskell Tests.hs
+
$ cabal test
reverse.reverse/id : OK, passed 100 tests.
 
drop.haq/id : OK, passed 100 tests.
 
 
</code>
 
</code>
   
Great!
+
==== QuickCheck ====
   
==== QuickCheck v2 ====
+
If you're using QuickCheck:
 
If you're using version 2 of QuickCheck, the code in the previous section needs some minor modifications:
 
   
 
<haskell>
 
<haskell>
$ cat > Tests.hs
+
$ cat > QuickTests.hs
import Char
+
import List
+
import Data.Char
  +
import Data.List
  +
import Haq
 
import Test.QuickCheck
 
import Test.QuickCheck
 
import Text.Printf
 
import Text.Printf
+
 
main = mapM_ (\(s,a) -> printf "%-25s: " s >> a) tests
 
main = mapM_ (\(s,a) -> printf "%-25s: " s >> a) tests
   
Line 298: Line 297:
   
 
Success!
 
Success!
 
=== Running the test suite from darcs ===
 
 
We can arrange for darcs to run the test suite on every commit that is run with the flag --test:
 
 
<code>
 
$ darcs setpref test "runhaskell Tests.hs"
 
Changing value of test from '' to 'runhaskell Tests.hs'
 
</code>
 
 
will run the full set of QuickChecks.
 
If your test requires it, you may need to ensure other things are built too -- for example:<code>darcs setpref test "alex Tokens.x;happy Grammar.y;runhaskell Tests.hs"</code>.
 
You will encounter that this way a darcs patch is also accepted if a QuickCheck test fails.
 
You have two choices to [http://www.haskell.org/pipermail/haskell-cafe/2007-October/033834.html work around] this:
 
* Use <hask>quickCheck'</hask> from the package QuickCheck-2 and call <hask>exitWithFailure</hask> if it return <hask>False</hask>.
 
* Keep the test program as it is, and implement the failure on the shell level:
 
: <code>runhaskell Tests.hs | tee test.log && if grep Falsifiable test.log >/dev/null; then exit 1; fi</code>
 
 
Let's commit a new patch:
 
 
<code>
 
$ darcs add Tests.hs
 
$ darcs record --all --test
 
What is the patch name? Add testsuite
 
Do you want to add a long comment? [yn]n
 
Running test...
 
reverse.reverse/id : OK, passed 100 tests.
 
drop.haq/id : OK, passed 100 tests.
 
Test ran successfully.
 
Looks like a good patch.
 
Finished recording patch 'Add testsuite'
 
</code>
 
 
Excellent: now, patches must pass the test suite before they can be committed provided the --test flag is passed.
 
   
 
=== Tag the stable version, create a tarball, and sell it! ===
 
=== Tag the stable version, create a tarball, and sell it! ===
Line 338: Line 303:
   
 
<code>
 
<code>
$ darcs tag
+
$ git tag 0.0
What is the version name? 0.0
 
Finished tagging patch 'TAG 0.0'
 
 
</code>
 
</code>
   
 
==== Create a tarball ====
 
==== Create a tarball ====
You can do this using either Cabal or darcs, or even an explicit <tt>tar</tt> command.
+
You can do this using either Cabal or an explicit <tt>tar</tt> command.
   
 
===== Using Cabal =====
 
===== Using Cabal =====
Line 364: Line 329:
   
 
to the .cabal file to have everything included.
 
to the .cabal file to have everything included.
 
===== Using darcs =====
 
 
Alternatively, you can use darcs:
 
<code>
 
$ darcs dist -d haq-0.0
 
Created dist as haq-0.0.tar.gz
 
</code>
 
 
And you're all set up!
 
   
 
==== Check that your source package is complete ====
 
==== Check that your source package is complete ====
Line 391: Line 346:
 
==== Upload your package to Hackage ====
 
==== Upload your package to Hackage ====
   
Whichever of the above methods you've used to create your package, you can upload it to the Hackage package collection via a [http://hackage.haskell.org/packages/upload.html web interface].
+
Whichever of the above methods you've used to create your package, you can upload it to the Hackage package collection via a [http://hackage.haskell.org/upload web interface].
 
You may wish to use the package checking interface there first, and fix things it warns about, before uploading your package.
 
You may wish to use the package checking interface there first, and fix things it warns about, before uploading your package.
  +
  +
==== Upload your package documentation to Hackage ====
  +
  +
Assuming you built your documentation like this:
  +
  +
<code>
  +
$ cabal haddock --hyperlink-source --html-location='http://hackage.haskell.org/package/haq/docs' --contents-location='http://hackage.haskell.org/package/haq'
  +
</code>
  +
  +
You can force upload (so you don't have to wait on the server to build it for you) your documentation for your package like so:
  +
  +
<code>
  +
$ cp -R ./dist/doc/html/haq/ haq-0.0-docs
  +
$ tar cvzf --format=ustar -f haq-0.0-docs.tar.gz haq-0.0-docs
  +
$ curl -X PUT -H 'Content-Type: application/x-tar' -H 'Content-Encoding: gzip' --data-binary '@haq-0.0-docs.tar.gz' 'https://$USERNAME:$PASSWORD@hackage.haskell.org/package/haq-0.0/docs'
  +
</code>
   
 
=== Summary ===
 
=== Summary ===
Line 399: Line 370:
   
 
$ ls
 
$ ls
Haq.hs Tests.hs dist haq.cabal
+
Haq.hs HSpecTests.hs dist haq.cabal
Setup.hs _darcs haq-0.0.tar.gz
+
Setup.hs .git haq-0.0.tar.gz QuickTests.hs
   
 
== Libraries ==
 
== Libraries ==
Line 483: Line 454:
 
You can also set up Cabal to run configure scripts, among other features. For more information consult the
 
You can also set up Cabal to run configure scripts, among other features. For more information consult the
 
[http://www.haskell.org/cabal/users-guide/ Cabal user guide].
 
[http://www.haskell.org/cabal/users-guide/ Cabal user guide].
 
== Automation ==
 
 
A tool to automatically populate a new cabal project is available:
 
 
cabal init
 
 
Usage is:
 
 
<code>
 
$ cabal init
 
Package name [default "haq"]?
 
Package version [default "0.1"]?
 
Please choose a license:
 
1) GPL
 
2) GPL-2
 
3) GPL-3
 
4) LGPL
 
5) LGPL-2.1
 
6) LGPL-3
 
* 7) BSD3
 
8) BSD4
 
9) MIT
 
10) PublicDomain
 
11) AllRightsReserved
 
12) OtherLicense
 
13) Other (specify)
 
Your choice [default "BSD3"]?
 
Author name? Henry Laxen
 
Maintainer email? nadine.and.henry@pobox.com
 
Project homepage/repo URL? http://somewhere.com/haq/
 
Project synopsis? A wonderful little module
 
Project category:
 
1) Codec
 
2) Concurrency
 
3) Control
 
4) Data
 
5) Database
 
6) Development
 
7) Distribution
 
8) Game
 
9) Graphics
 
10) Language
 
11) Math
 
12) Network
 
13) Sound
 
14) System
 
15) Testing
 
16) Text
 
17) Web
 
18) Other (specify)
 
Your choice? 3
 
What does the package build:
 
1) Library
 
2) Executable
 
Your choice? 1
 
Generating LICENSE...
 
Generating Setup.hs...
 
Generating haq.cabal...
 
 
You may want to edit the .cabal file and add a Description field.
 
</code>
 
   
 
== Licenses ==
 
== Licenses ==
Line 562: Line 471:
 
== Releases ==
 
== Releases ==
   
It's important to release your code as stable, tagged tarballs. Don't
+
It's important to release your code as stable, tagged tarballs. Don't just rely on your commit hashes for tracking history, tag your released versions and milestones!
just [http://jackunrue.blogspot.com/2006/11/don-do-releases.html rely on darcs for distribution].
 
   
* '''darcs dist''' generates tarballs directly from a darcs repository
+
* '''git archive''' generates tarballs directly from a git repository based on the provided tag.
   
 
For example:
 
For example:
Line 570: Line 479:
 
$ cd fps
 
$ cd fps
 
$ ls
 
$ ls
Data LICENSE README Setup.hs TODO _darcs cbits dist fps.cabal tests
+
Data LICENSE README Setup.hs TODO .git cbits dist fps.cabal tests
$ darcs dist -d fps-0.8
+
$ git tag v0.8
Created dist as fps-0.8.tar.gz
+
$ git archive --format=tar.gz v0.8 > fps-0.8.tar.gz
   
 
You can now just post your fps-0.8.tar.gz
 
You can now just post your fps-0.8.tar.gz
   
You can also have darcs do the equivalent of 'daily snapshots' for you by using a post-hook.
 
   
put the following in _darcs/prefs/defaults:
+
* Tag each release using '''git tag'''. For example:
apply posthook darcs dist
 
apply run-posthook
 
 
Advice:
 
* Tag each release using '''darcs tag'''. For example:
 
 
$ darcs tag 0.8
 
Finished tagging patch 'TAG 0.8'
 
 
Then people can <tt>darcs pull --partial -t 0.8</tt>, to get just the tagged version (and not the entire history).
 
   
 
== Hosting ==
 
== Hosting ==
   
Hosting for repos is available from the Haskell community server:
+
Hosting for repos is available from [http://github.com Github] and the [http://community.haskell.org/ Haskell community server]
 
http://community.haskell.org/
 
 
A Darcs repository can be published simply by making it available from a
 
web page.
 
   
 
There is also a (minimal) Github equivalent for Darcs at [http://hub.darcs.net/ hub.darcs.net].
 
There is also a (minimal) Github equivalent for Darcs at [http://hub.darcs.net/ hub.darcs.net].
Line 589: Line 497:
   
 
Create a web page documenting your project! An easy way to do this is to
 
Create a web page documenting your project! An easy way to do this is to
add a project specific page to [[Haskell|the Haskell wiki]]
+
add a project specific page to [[HaskellWiki:Contributing|the Haskell wiki]]
   
 
== The user experience ==
 
== The user experience ==
Line 655: Line 563:
 
* Add your library or application to the [[Libraries and tools]] page, under the relevant category, so people can find it.
 
* Add your library or application to the [[Libraries and tools]] page, under the relevant category, so people can find it.
   
* If your release is a Cabal package, add it to the [http://hackage.haskell.org/packages/hackage.html Hackage database] (Haskell's CPAN wanna-be).
+
* If your release is a Cabal package, add it to the [http://hackage.haskell.org/packages/hackage.html Hackage database]
  +
  +
* If your library stays up to date with its dependencies, add it to the vetted packages on [https://github.com/fpco/stackage Stackage].
   
 
=== Blog about it ===
 
=== Blog about it ===

Latest revision as of 07:44, 4 September 2014

A developers' guide to creating a new Haskell project or program, and working in the Haskell developer ecosystem.

Note: for learning the Haskell language itself we recommend these resources.

Contents

[edit] 1 Recommended tools

Almost all new Haskell projects use the following tools. Each is intrinsically useful, but using a set of common tools also helps everyone by increasing productivity, and you're more likely to get patches.

[edit] 1.1 Revision control

Use git or darcs unless you have a specific reason not to. Both are lightweight distributed revision control system. Darcs is written in Haskell. They are the two most popular DVCSes in the Haskell world. If you want to encourage contributions from other Haskell hackers then git is best. For git, GitHub is very popular. Darcs hosting is available on hub.darcs.net.

[edit] 1.2 Build system

(thumbnail)
Built with Cabal

Use Cabal. You should read at least the start of section 2 of the Cabal User's Guide.

You should use cabal-install as a front-end for installing your Cabal library. Cabal-install provides commands not only for building libraries but also for installing them from, and uploading them to, Hackage. As a bonus, for almost all programs, it's faster than using Setup.hs scripts directly, since no time is wasted compiling the scripts. (This does not apply for programs that use custom Setup.hs scripts, since those need to be compiled even when using cabal-install.)

cabal-install is widely available, as a binary distribution or as part of the Haskell Platform, so you can probably assume your users will have it too.

[edit] 1.3 Documentation

For libraries, use Haddock. Haddock generates nice markup, with links to source.

[edit] 1.4 Testing

Typical unit/spec based testing, particularly with impure code, can be done with HSpec and HUnit.

You can use QuickCheck or SmallCheck to test pure code. These libraries work best when you have known invariants in your code's behavior. See this Cabal file for an example of how to include tests in your Cabal package.

To get started, try Introduction to QuickCheck. For a slightly more advanced introduction, Simple Unit Testing in Haskell is a blog article about creating a testing framework for QuickCheck using some Template Haskell. For HUnit, see HUnit 1.0 User's Guide

[edit] 1.5 Distribution

The standard mechanism for distributing Haskell libraries and applications is Hackage. Hackage can host your cabalised tarball releases, and link to any library dependencies your code has. Users will find and install your packages via "cabal install", and your package will be integrated into Haskell search engines, like hoogle

[edit] 1.6 Target Environment

If possible, depend on libraries that are provided by the current stable Stackage package versions or the Haskell Platform. Ideally your package should be able to build with users that keep current with Stackage or users of the Platform.

[edit] 2 Structure of a simple project

The basic structure of a new Haskell project can be adopted from HNop, the minimal Haskell project. It consists of the following files, for the mythical project "haq".

  • Haq.hs -- the main haskell source file
  • haq.cabal -- the cabal build description
  • Setup.hs -- build script itself
  • .git -- revision control
  • README -- info
  • LICENSE -- license

Of course, you can elaborate on this, with subdirectories and multiple modules. See Structure of a Haskell project for an example of a larger project's directory structure.

Here is a transcript that shows how you'd create a minimal git and cabalised Haskell project for the cool new Haskell program "haq", build it, install it and release.

Note: The new tool "cabal init" automates all this for you, but you should understand all the parts even so.

We will now walk through the creation of the infrastructure for a simple Haskell executable. Advice for libraries follows after.

[edit] 2.1 Create a directory

Create somewhere for the source:

$ mkdir haq
$ cd haq

[edit] 2.2 Write some Haskell source

Write your program:

$ cat > Haq.hs
--
-- Copyright (c) 2006 Don Stewart - http://www.cse.unsw.edu.au/~dons/
-- GPL version 2 or later (see http://www.gnu.org/copyleft/gpl.html)
--
import System.Environment
 
-- | 'main' runs the main program
main :: IO ()
main = getArgs >>= print . haqify . head
 
haqify s = "Haq! " ++ s

[edit] 2.3 Stick it in version control

Place the source under revision control (you may need to enter your e-mail address first, to identify you as maintainer of this source):

$ git init
$ git add Haq.hs 
$ git commit -am "my commit message"

[edit] 2.4 Add a build system

Create a .cabal file describing how to build your project:

$ cabal init

that will ask a few questions about your project and generate a file similar to the example. -- Initial scratch.cabal generated by cabal init. For further -- documentation, see http://haskell.org/cabal/users-guide/

name: haq version: 0.1.0.0 description: Super cool mega lambdas license: GPL license-file: LICENSE author: Don Stewart maintainer: dons@cse.unsw.edu.au build-type: Simple cabal-version: >=1.10

executable haq

 main-is:           Haq.hs 
 build-depends:     base >=4.5 && <4.8
 default-language:  Haskell2010

(If your package uses other packages, e.g. text, you'll need to add them to the build-depends: field as a comma separated list.)

Cabal will also generate a Setup.hs file that will do the actual building. You will rarely need to modify it.

If you specifed a known license, it will also add a LICENSE file.

You might like to add a README file to tell what your project is about.

Record your changes:

$ git add haq.cabal Setup.hs LICENSE README
$ git commit -am "Add a build system"

[edit] 2.5 Build your project

Now build it! There are two methods of accessing Cabal functionality: through your Setup.hs script or through cabal-install. cabal-install is now the preferred method.

Building using cabal-install and sandboxes, always use sandboxes in your projects unless you are an expert and know what you are doing! Demonstrated here:

$ cabal sandbox init
$ cabal install -j

This will install your newly minted haq program in $PROJECT_DIR/.cabal-sandbox/bin.

[edit] 2.6 Run it

And now you can run your cool project:

$ .cabal-sandbox/bin/haq me
"Haq! me"

Since our program doesn't rely on any other assets we can just copy it to a directory in our path if we want to be able to use it from anywhere without referencing the sandbox directory. Like so:

$ sudo cp .cabal-sandbox/bin/haq /usr/local/bin/haq
$ haq me
"Haq! me"

[edit] 2.7 Build some haddock documentation

Generate some API documentation into dist/doc/*

Using cabal install:

$ cabal haddock

Using cabal install if you planned to upload your haq package to Hackage: $ cabal haddock --hyperlink-source --html-location='http://hackage.haskell.org/package/haq/docs' --contents-location='http://hackage.haskell.org/package/haq'

which generates files in dist/doc/ including:

$ w3m -dump dist/doc/html/haq/Main.html

haq Contents Index
Main
Synopsis
main :: IO ()
Documentation
main :: IO ()
main runs the main program
Produced by Haddock version 0.7

No output? Make sure you have actually installed haddock. It is a separate program, not something that comes with Cabal. Note that the stylized comment in the source gets picked up by Haddock.

[edit] 2.8 (Optional) Improve your code: HLint

HLint can be a valuable tool for improving your coding style, particularly if you're new to Haskell. Let's run it now.

$ hlint .
./Haq.hs:11:1: Warning: Eta reduce
Found:
  haqify s = "Haq! " ++ s
Why not:
  haqify = ("Haq! " ++)

The existing code will work, but let's follow that suggestion. Open Haq.hs in your favourite editor and change the line:

    where haqify s = "Haq! " ++ s

to:

    where haqify = ("Haq! " ++)

[edit] 2.9 Add some tests

[edit] 2.9.1 HSpec

If you'd like to use HSpec:

To your cabal, add:

test-suite tests

 ghc-options: -Wall
 default-extensions:  OverloadedStrings
 type: exitcode-stdio-1.0
 main-is: HSpecTests.hs
 build-depends:       base,
                      haq,
                      hspec                >= 1.8
 default-language:    Haskell2010

To enable tests and install HSpec (and any other needed dependencies):

$ cabal install --enable-tests

Then for your actual file containing the test code:

$ cat > HSpecTests.hs
module Main where
 
import Haq
import Test.Hspec
 
main :: IO ()
main = hspec $ do
 
  describe "Validate haqify function" $ do
    it "haqify is supposed to prefix Haq! to things" $ do
      haqify "me" `shouldBe` "Haq! me"

Execute the tests with:

$ cabal test

[edit] 2.9.2 QuickCheck

If you're using QuickCheck:

$ cat > QuickTests.hs
 
import Data.Char
import Data.List
import Haq
import Test.QuickCheck
import Text.Printf
 
main  = mapM_ (\(s,a) -> printf "%-25s: " s >> a) tests
 
-- reversing twice a finite list, is the same as identity
prop_reversereverse s = (reverse . reverse) s == id s
    where _ = s :: [Int]
 
-- Dropping the "Haq! " string is the same as identity
prop_haq s = drop (length "Haq! ") (haqify s) == id s
    where haqify s = "Haq! " ++ s
 
tests  = [("reverse.reverse/id", quickCheck prop_reversereverse)
        ,("drop.haq/id",        quickCheck prop_haq)]

To run the test:

$ runhaskell Tests.hs
reverse.reverse/id       : +++ OK, passed 100 tests.
drop.haq/id              : +++ OK, passed 100 tests.

Success!

[edit] 2.10 Tag the stable version, create a tarball, and sell it!

Tag the stable version:

$ git tag 0.0

[edit] 2.10.1 Create a tarball

You can do this using either Cabal or an explicit tar command.

[edit] 2.10.1.1 Using Cabal

Since the code is cabalised, we can create a tarball with cabal-install directly (you can also use runhaskell Setup.hs sdist, but you need tar on your system [1]):

$ cabal sdist
Building source dist for haq-0.0...
Source tarball created: dist/haq-0.0.tar.gz

This has the advantage that Cabal will do a bit more checking, and ensure that the tarball has the structure that HackageDB expects. Note that it does require the LICENSE file to exist. It packages up the files needed to build the project; to include other files (such as Test.hs in the above example, and our README), we need to add:

extra-source-files: Tests.hs README

to the .cabal file to have everything included.

[edit] 2.10.2 Check that your source package is complete

Just to make sure everything works, try building the source package in some temporary directory:

$ tar xzf haq-0.0.tar.gz
$ cd haq-0.0
$ cabal configure
$ cabal build

and for packages containing libraries,

$ cabal haddock

[edit] 2.10.3 Upload your package to Hackage

Whichever of the above methods you've used to create your package, you can upload it to the Hackage package collection via a web interface. You may wish to use the package checking interface there first, and fix things it warns about, before uploading your package.

[edit] 2.10.4 Upload your package documentation to Hackage

Assuming you built your documentation like this:

$ cabal haddock --hyperlink-source --html-location='http://hackage.haskell.org/package/haq/docs' --contents-location='http://hackage.haskell.org/package/haq'

You can force upload (so you don't have to wait on the server to build it for you) your documentation for your package like so:

$ cp -R ./dist/doc/html/haq/ haq-0.0-docs $ tar cvzf --format=ustar -f haq-0.0-docs.tar.gz haq-0.0-docs $ curl -X PUT -H 'Content-Type: application/x-tar' -H 'Content-Encoding: gzip' --data-binary '@haq-0.0-docs.tar.gz' 'https://$USERNAME:$PASSWORD@hackage.haskell.org/package/haq-0.0/docs'

[edit] 2.11 Summary

The following files were created:

   $ ls
   Haq.hs           HSpecTests.hs         dist             haq.cabal
   Setup.hs         .git                  haq-0.0.tar.gz   QuickTests.hs

[edit] 3 Libraries

The process for creating a Haskell library is almost identical. The differences are as follows, for the hypothetical "ltree" library:

[edit] 3.1 Hierarchical source

The source should live under a directory path that fits into the existing module layout guide. So we would create the following directory structure, for the module Data.LTree:

   $ mkdir Data
   $ cat > Data/LTree.hs 
   module Data.LTree where

So our Data.LTree module lives in Data/LTree.hs

[edit] 3.2 The Cabal file

Cabal files for libraries list the publically visible modules, and have no executable section:

   $ cat > ltree.cabal 
   Name:                ltree
   Version:             0.1
   Description:         Lambda tree implementation
   License:             BSD3
   License-file:        LICENSE
   Author:              Don Stewart
   Maintainer:          dons@cse.unsw.edu.au
   Build-Type:          Simple
   Cabal-Version:       >=1.2
   
   Library
     Build-Depends:     base >= 3 && < 5
     Exposed-modules:   Data.LTree
     ghc-options:       -Wall

We can thus build our library:

   $ cabal configure --prefix=$HOME --user
   $ cabal build    
   Preprocessing library ltree-0.1...
   Building ltree-0.1...
   [1 of 1] Compiling Data.LTree       ( Data/LTree.hs, dist/build/Data/LTree.o )
   /usr/bin/ar: creating dist/build/libHSltree-0.1.a

and our library has been created as a object archive. Now install it:

   $ cabal install
   Installing: /home/dons/lib/ltree-0.1/ghc-6.6 & /home/dons/bin ltree-0.1...
   Registering ltree-0.1...
   Reading package info from ".installed-pkg-config" ... done.
   Saving old package config file... done.
   Writing new package config file... done.

And we're done! To try it out, first make sure that your working directory is anything but the source directory of your library:

   $ cd ..

And then use your new library from, for example, ghci:

   $ ghci -package ltree
   Prelude> :m + Data.LTree
   Prelude Data.LTree> 

The new library is in scope, and ready to go.

[edit] 3.3 More complex build systems

For larger projects, you may want to store source trees in subdirectories. This can be done simply by creating a directory -- for example, "src" -- into which you will put your src tree.

To have Cabal find this code, you add the following line to your Cabal file:

   hs-source-dirs: src

You can also set up Cabal to run configure scripts, among other features. For more information consult the Cabal user guide.

[edit] 4 Licenses

Code for the common base library package must be BSD licensed. Otherwise, it is entirely up to you as the author. Choose a licence (inspired by this). Check the licences of things you use (both other Haskell packages and C libraries), since these may impose conditions you must follow. Use the same licence as related projects, where possible. The Haskell community is split into 2 camps, roughly: those who release everything under BSD, and (L)GPLers. Some Haskellers recommend avoiding LGPL, due to cross-module optimisation issues. Like many licensing questions, this advice is controversial. Several Haskell projects (wxHaskell, HaXml, etc) use the LGPL with an extra permissive clause which gets round the cross-module optimisation problem.

[edit] 5 Releases

It's important to release your code as stable, tagged tarballs. Don't just rely on your commit hashes for tracking history, tag your released versions and milestones!

  • git archive generates tarballs directly from a git repository based on the provided tag.

For example:

$ cd fps
$ ls       
Data      LICENSE   README    Setup.hs  TODO      .git    cbits dist      fps.cabal tests
$ git tag v0.8
$ git archive --format=tar.gz v0.8 > fps-0.8.tar.gz

You can now just post your fps-0.8.tar.gz


  • Tag each release using git tag. For example:

[edit] 6 Hosting

Hosting for repos is available from Github and the Haskell community server

There is also a (minimal) Github equivalent for Darcs at hub.darcs.net.

[edit] 7 Web page

Create a web page documenting your project! An easy way to do this is to add a project specific page to the Haskell wiki

[edit] 8 The user experience

When developing a new Haskell library, it is important to remember how the user expects to be able to build and use a library.

[edit] 8.1 Introductory information and build guide

A typical library user expects to:

  1. Visit Haskell.org
  2. Find the library/program they are looking for:
    1. if not found, try mailing list;
    2. if it is hidden, try improving the documentation on haskell.org;
    3. if it does not exist, try contributing code and documentation)
  3. Download
  4. Build and install
  5. Enjoy

Each of these steps can pose potential road blocks, and code authors can do a lot to help code users avoid such blocks. Steps 1..2 may be easy enough, and many coders and users are mainly concerned with step 5. Steps 3..4 are the ones that often get in the way. In particular, the following questions should have clear answers:

  • Which is the latest version?
  • What state is it in?
  • What are its aims?
  • Where is the documentation?
  • Which is the right version for given OS and Haskell implementation?
  • How is it packaged, and what tools are needed to get and unpack it?
  • How is it installed, and what tools are needed to install it?
  • How do we handle dependencies?
  • How do we provide/acquire the knowledge and tool-chains needed?

The best place to answer these questions is a README file, distributed with the library or application, and often accompanied with similar text on a more extensive web page.

[edit] 8.2 Tutorials

Generated haddock documentation is usually not enough to help new programmers learn how to use a library. You must also provide accompanying examples, and even tutorials about the library.

Please consider providing example code for your library or application. The code should be type-correct and well-commented.

[edit] 9 Program structure

Monad transformers are very useful for programming in the large, encapsulating state, and controlling side effects. To learn more about this approach, try Monad Transformers Step by Step.

[edit] 10 Publicity

The best code in the world is meaningless if nobody knows about it. The process to follow once you've tagged and released your code is:

[edit] 10.1 Join the community

If you haven't already, join the community. The best way to do this is to subscribe to at least haskell-cafe@ and haskell@ mailing lists. Joining the #haskell IRC channel is also an excellent idea.

[edit] 10.2 Announce your project on haskell@

Most important: announce your project releases to the haskell@haskell.org mailing list. Tag your email subject line with "ANNOUNCE: ...". This ensure it will then make it into the Haskell Weekly News. To be doubly sure, you can email the release text to the HWN editor.

[edit] 10.3 Add your code to the public collections

  • Add your library or application to the Libraries and tools page, under the relevant category, so people can find it.
  • If your library stays up to date with its dependencies, add it to the vetted packages on Stackage.

[edit] 10.4 Blog about it

Blog about it! Blog about your new code on Planet Haskell. Write about your project in your blog, then email the Planet Haskell maintainer (ibid on #haskell) the RSS feed url for your blog