A purely functional programming language with first class types
Go to file
Edwin Brady 26cdfc7830 Make records work in parameter blocks
This involves making 'unelab' aware of nested names so that it can
remove the parameters from names in the current block. It's a bit of a
hacky solution, but it is also the easiest one.
Ideally we'd build the getter types directly, rather than using unelab,
but that's one to save for another time.
Fixes #1482
2021-07-10 18:12:44 +01:00
.github [ ci ] just dump everything into a single file already (#1645) 2021-06-29 19:42:48 +01:00
benchmark [ refactor ] Remove Data.Strings module (#1607) 2021-06-28 13:48:37 +01:00
bootstrap Update bootstrap scheme 2021-06-23 18:16:13 +01:00
docs Add comment about import from the contrib library (#1625) 2021-06-28 16:23:01 +01:00
icons Add icons 2020-05-20 18:48:55 +01:00
libs Define and implement Relation interfaces (#1472) 2021-07-09 09:06:27 +01:00
lint Lint utility 2021-01-16 10:00:03 +00:00
nix Nix: make nixpkgs options overridable 2021-06-30 15:33:37 +01:00
Release Don't need benchmarks in the release (#1600) 2021-06-28 19:07:24 +01:00
samples Use pragma once instead of include guard 2021-06-28 12:05:22 +01:00
src Make records work in parameter blocks 2021-07-10 18:12:44 +01:00
support [ fix ] Chez channels (#1596) 2021-07-02 13:13:50 +01:00
tests Make records work in parameter blocks 2021-07-10 18:12:44 +01:00
.editorconfig [ install ] Check if 'realpath' exists for Chez and Racket backends (#1210) 2021-04-06 15:42:04 +01:00
.gitattributes Mark bootstrap code as generated 2021-06-30 22:11:54 +01:00
.gitignore [ test ] cosmetic changes & retest capability (#1394) 2021-05-11 09:46:21 +01:00
bootstrap-stage1-chez.sh [ install ] Check if 'realpath' exists for Chez and Racket backends (#1210) 2021-04-06 15:42:04 +01:00
bootstrap-stage1-racket.sh [ install ] Check if 'realpath' exists for Chez and Racket backends (#1210) 2021-04-06 15:42:04 +01:00
bootstrap-stage2.sh [ install ] Check if 'realpath' exists for Chez and Racket backends (#1210) 2021-04-06 15:42:04 +01:00
CHANGELOG.md [ fix ] Chez channels (#1596) 2021-07-02 13:13:50 +01:00
config.mk Stop pointing at the Void 2021-06-18 11:06:45 +01:00
CONTRIBUTING.md fixed whitespace for *.md and .rst files 2021-01-22 15:08:49 +00:00
CONTRIBUTORS [ fix ] Chez channels (#1596) 2021-07-02 13:13:50 +01:00
default.nix Nix fixes (#1623) 2021-06-28 11:47:47 +01:00
flake.lock Added new nix functionalities (#1154) 2021-03-15 14:21:50 +00:00
flake.nix Nix: make nixpkgs options overridable 2021-06-30 15:33:37 +01:00
idris2.ipkg Support for incremental compilation with Chez 2021-06-27 15:40:23 +01:00
idris2api.ipkg [ improvement ] VMCode (#1662) 2021-07-07 17:06:59 +01:00
INSTALL.md Nix fixes (#1623) 2021-06-28 11:47:47 +01:00
LICENSE Add licence and changelog and update REAMDE 2020-05-20 11:31:48 +01:00
Makefile Insert @generated markers into generated launcher scripts (#1586) 2021-06-28 16:44:33 +01:00
README.md Contributors -> Contributions 2021-06-12 21:38:14 +01:00

Idris 2

Documentation Status Windows Status Ubuntu Status macOS Status Nix Status

Idris 2 is a purely functional programming language with first class types.

For full installation instructions, see INSTALL.md. Briefly, if you have Chez Scheme installed, with the executable name chez, type:

  • make bootstrap SCHEME=chez
  • make install

You may need to change chez to be the local name of your Chez Scheme. This is often one of scheme, chezscheme or chezscheme9.5 (depending on the version). On a modern desktop machine, this process (including tests) should take less than 5 minutes.

Idris 2 is mostly backwards compatible with Idris 1, with some minor exceptions. The most notable user visible differences, which might cause Idris 1 programs to fail to type check, are:

  • Unbound implicit arguments are always erased, so it is a type error to attempt to pattern match on one.

  • Simplified resolution of ambiguous names, which might mean you need to explicitly disambiguate more often. As a general rule, Idris 2 will be able to disambiguate between names which have different concrete return types (such as data constructors), or which have different concrete argument types (such as record projections). It may struggle to resolve ambiguities if one name requires an interface to be resolved.

  • The cong function now takes its congruence explicitly as its first argument.

  • Minor differences in the meaning of export modifiers private, export, and public export, which now refer to visibility of names from other namespaces rather than visibility from other files.

  • Module names must match the filename in which they are defined (unless the module's name is "Main").

  • Anything which uses a %language pragma in Idris 1 is likely to be different. Notably, elaborator reflection will exist, but most likely in a slightly different form because the internal details of the elaborator are different.

  • The Prelude is much smaller (and easier to replace with an alternative). Command-line option --no-prelude can be used to not implicitly import Prelude.

  • let x = val in e no longer computes with x in e, instead being essentially equivalent to (\x => e) val. This is to make the behaviour of let consistent in the presence of case and with (where it is hard to push the computation inside the case/with efficiently). Instead, you can define functions locally with let, which do have computational force, as follows:

    let x : ?
        x = val in
        e
    

Watch this space for more details and the rationale for the changes, as I get around to writing it...

Summary of new features:

  • A core language based on "Quantitative Type Theory" which allows explicit annotation of erased types, and linear types.
  • let bindings are now more expressive, and can be used to define pattern matching functions locally.
  • Names which are in scope in a type are also always in scope in the body of the corresponding definition.
  • Better inference. Holes are global to a source file, rather than local to a definition, meaning that some holes can be left in function types to be inferred by the type checker. This also gives better inference for the types of case expressions, and means fewer annotations are needed in interface declarations.
  • Better type checker implementation which minimises the need for compile time evaluation.
  • New Chez Scheme based back end which both compiles and runs faster than the default Idris 1 back end. (Also, optionally, Racket and Gambit can be used as targets).
  • Everything works faster :).

A significant change in the implementation is that there is an intermediate language TTImp, which is essentially a desugared Idris, and is cleanly separated from the high level language which means it is potentially usable as a core language for other high level syntaxes.

Javascript

The javascript codegen uses the new BigInt, hence nodejs 10.4 or higher is required.

Editor Plugins

The wiki lists the current plugins available for common text editors and their features.

Things still missing

  • Disambiguation via 'with'
  • Cumulativity (so we currently have Type : Type! Bear that in mind when you think you've proved something :))
  • 'rewrite' doesn't yet work on dependent types
  • Parts of the ide-mode, particularly syntax highlighting
  • Documentation strings and HTML documentation generation
  • ':search' and ':apropos' at the REPL
  • Metaprogramming (reflection, partial evaluation)

Contributions wanted

If you want to learn about Idris more, contributing to the compiler could be one way to do so. Just select one good first issue and ask about it on the Discord channel.

Talks

Edwin Brady Tells Us What's New in Idris 2 (Berlin Functional Programming Group)

Scheme Workshop Keynote (ACM SIGPLAN)

Idris 2 - Type-driven Development of Idris (Curry On - London 2019)

Idris 2: Type-driven Development of Idris (Code Mesh LDN 18)