Awesome

Welcome to HyperHaskell,

… the strongly hyped Haskell interpreter.

HyperHaskell is a graphical interpreter for the programming language Haskell. You use worksheets to enter expressions and evaluate them. Results are displayed graphically using HTML.

HyperHaskell is intended to be easy to install. It is cross-platform and should run on Linux, Mac and Windows. Internally, it uses the GHC API to interpret Haskell programs, and the graphical front-end is built on the cross-platform Electron framework.

HyperHaskell's main attraction is a Display class that supersedes the good old Show class. The result looks like this:

Releases

The latest release of HyperHaskell is Level α (version 0.2.3.0).

Installation

Overview

HyperHaskell is intended to be easy to install. The easiest way to install it is to download the binary distribution. This is explained in the next subsection. However, there is a pitfall which you have to know about, and which requires knowledge of the installation structure.

A HyperHaskell installation consists of two parts:

The graphical front-end.

Currently written in HTML and JavaScript, packaged with the Electron framework.
The interpreter back-end.

Consists of an executable hyper-haskell-server, written in Haskell using the GHC API, and a library (module) Hyper for visualizing and pretty printing Haskell values.

Both parts depend on several different Haskell packages. Unfortunately, the versions of the packages used to compile the executable and to compile the library have to be exactly the same.

This is why, at the moment, the front-end does not come with the back-end executable included. Instead, the user is asked to install the hyper-haskell-server back-end into his or her own database of Haskell packages, and then tell the front-end about it. This way, the user is free to use different package or compiler versions.

Installation of the binary distribution

Installation from the binary distribution follows the structure explained above.

Download the graphical front-end from the latest release and unpack it.

At the moment, binaries for macOS and Windows (thanks to Nicholas Silvestrin) are provided. Since Heinrich Apfelmus only has access to a macOS machine, help is appreciated!
Install the back-end server
1. Make sure that you have a working installation of the GHC Haskell compiler.
2. Install the back-end with Cabal by executing
```
 cabal install hyper hyper-haskell-server
```
  It is also recommended (but not necessary) that you install the additional support for other popular Haskell packages, e.g. the Diagrams library by additionally executing
```
 cabal install hyper-extra
```
3. Now you can start the front-end application and create a new worksheet, or open an existing one. Make sure that the "Interpreter Back-end" in the "Settings" section of the worksheet is set to "cabal". (The path field does not matter in this case.)
  
  It is also possible to use Stack by using stack install, but that is not fully explained here, only to some extent below.

That's it! Happy hyper!

Run from source

When developing HyperHaskell itself, it is also possible to run it from source. Follow these steps:

Download and install Electron

The whole thing is currently developed and tested with Electron version 10.1.5.

(If you use the npm package manager, you can install it in your home directory with cd ~ && npm install electron@10.5.1. On Debian-based Linux distributions, Electron currently requires the nodejs-legacy package.)
Make sure that you have a working installation of
- the GHC Haskell compiler
- the stack utility
(See the Haskell homepage for more on how to obtain these.)
You also need the make utility, which should be standard on any UN*X platform. Edit the file named Makefile and tell it where to find the Electron executable

On macOS: Typically,
```
 ELECTRON=/Applications/Electron.app/Contents/MacOS/Electron
```
On Linux: Typically,
```
 ELECTRON=/usr/local/bin/electron
```
On Windows: You can locate electron.exe and double-click it. Then simply drop the hyper-haskell\app folder onto the lower pane of the window. Alternatively, from the terminal invoke what is suggested in the upper portion of the Electron window, i.e.
```
 <path-to-electron>\electron.exe hyper-haskell\app
```
Go into the root directory of this repository and type make run.
```
 $ cd hyper-haskell
 $ make run
```
This will call the stack utility to build the server back-end, and finally run the front-end.
Use the File menu to open one of the example worksheets from the worksheets folder. Voilà!

You can also create a new worksheet, but note that you have to set the back-end path in the "Settings" section. The path is relative to the directory where the worksheet was saved. For instance, if you run a worksheet from the worksheets directory, the path ../haskell/stack.yaml will point to the right hyper-haskell-server executable. Screenshot:

Note that for this setting, the stack utility has to be in your path. You can also set an explicit path for this utility in the "Preferences…" menu item.

Packaging

The normal way to obtain HyperHaskell is to download the application bundle in binary form. This section describes how to generate this from source.

We use the electron-packager utility. To install it, you need to use the npm package manager and execute the following commands:

cd ~
npm install electron-packager
npm install electron@10.1.5 --save-dev

To create an application bundle and compress it in a zip-file, use the following commands:

On macOS:
```
make pkg-darwin
make zip-darwin
```
On Windows: You need the 7zip utility in your path.
```
make pkg-win32
nake zip-win32
```
On Linux: not implemented yet

Contributors

Many thanks to everyone who contributed, provided feedback or simply found a nice application for HyperHaskell! In particular, many thanks to:

Moritz Angermann, Simon Jakobi, Rodney Lorrimar, Karshan Sharma, Nicholas Silvestrin, and many others.

The project was started by Heinrich Apfelmus.