Rattus: A modal FRP language
This library implements the Rattus programming language as an embedded DSL. To this end the library provides a GHC plugin that performs the additional checks that are necessary for Rattus. What follows is a brief introduction to the language and its usage. A more detailed introduction can be found in this paper.
Rattus is a functional reactive programming (FRP) language that uses modal types to ensure operational properties that are crucial for reactive programs: productivity (in each computation step, the program makes progress), causality (output depends only on current and earlier input), and no implicit space leaks (programs do not implicitly retain memory over time).
To ensure these properties, Rattus uses the type modality
O
to express the concept of time at the type
level. Intuitively speaking, a value of type O a
represents a computation that will produce a value of type
a
in the next time step. Additionally, the language also
features the Box
type modality. A value of type Box a
is a time-independent computation that can be executed at
any time to produce a value of type a
.
For example, the type of streams is defined as
data Str a = a ::: (O (Str a))
So the head of the stream is available now, but its tail
is only available in the next time step. Writing a map
function for this type of streams, requires us to use the
Box
modality:
map :: Box (a -> b) -> Str a -> Str b map f (x ::: xs) = unbox f x ::: delay (map f (adv xs))
This makes sure that the function f
that we give to
map
is available at any time in the future.
The core of the language is defined in the module
Rattus.Primitives. Note that the operations on O
and
Box
have non-standard typing rules. Therefore, this
library provides a compiler plugin that checks these
non-standard typing rules. To write Rattus programs, one
must enable this plugin via the GHC option
-fplugin=Rattus.Plugin
, e.g. by including the following
line in the source file:
{-# OPTIONS -fplugin=Rattus.Plugin #-}
In addition, one must annotate the functions that are written in Rattus:
{-# ANN myFunction Rattus #-}
Or annotate the whole module as a Rattus module:
{-# ANN module Rattus #-}
Below is a minimal Rattus program using the Rattus.Stream module for programming with streams:
{-# OPTIONS -fplugin=Rattus.Plugin #-} import Rattus import Rattus.Stream {-# ANN sums Rattus #-} sums :: Str Int -> Str Int sums = scan (box (+)) 0
The source code of the Rattus.Stream module provides more examples on how to program in Rattus.
Modules
[Index] [Quick Jump]
Downloads
- Rattus-0.5.1.1.tar.gz [browse] (Cabal source package)
- Package description (as included in the package)
Maintainer's Corner
For package maintainers and hackage trustees
Candidates
Versions [RSS] | 0.1.0.0, 0.1.1.0, 0.2, 0.3, 0.3.1, 0.4, 0.5, 0.5.0.1, 0.5.1, 0.5.1.1 |
---|---|
Change log | CHANGELOG.md |
Dependencies | base (>=4.12 && <5), containers (>=0.5 && <0.8), ghc (>=8.6 && <9.9), simple-affine-space (>=0.2.1 && <0.3), transformers (>=0.5 && <0.7) [details] |
License | BSD-3-Clause |
Copyright | Copyright (C) 2020 Patrick Bahr |
Author | Patrick Bahr |
Maintainer | Patrick Bahr <paba@itu.dk> |
Category | FRP |
Home page | https://github.com/pa-ba/Rattus |
Bug tracker | https://github.com/pa-ba/Rattus/issues |
Uploaded | by PatrickBahr at 2023-10-16T13:56:42Z |
Distributions | LTSHaskell:0.5.1.1, NixOS:0.5.1.1, Stackage:0.5.1.1 |
Downloads | 1727 total (36 in the last 30 days) |
Rating | 2.25 (votes: 2) [estimated by Bayesian average] |
Your Rating | |
Status | Docs available [build log] Last success reported on 2023-10-16 [all 1 reports] |