Ravi Programming Language

https://github.com/dibyendumajumdar/ravi/workflows/build/badge.svg

Ravi is a dialect of Lua with limited optional static typing and features a JIT compiler powered by MIR as well as support for AOT compilation to native code. The name Ravi comes from the Sanskrit word for the Sun. Interestingly a precursor to Lua was Sol which had support for static types; Sol means the Sun in Portugese.

Lua is perfect as a small embeddable dynamic language so why a derivative? Ravi extends Lua with static typing for improved performance when JIT compilation is enabled. However, the static typing is optional and therefore Lua programs are also valid Ravi programs.

There are other attempts to add static typing to Lua - e.g. Typed Lua but these efforts are mostly about adding static type checks in the language while leaving the VM unmodified. The Typed Lua effort is very similar to the approach taken by Typescript in the JavaScript world. The static typing is to aid programming in the large - the code is eventually translated to standard Lua and executed in the unmodified Lua VM.

My motivation is somewhat different - I want to enhance the VM to support more efficient operations when types are known. Type information can be exploited by JIT compilation technology to improve performance. At the same time, I want to keep the language safe and therefore usable by non-expert programmers.

Of course there is the fantastic LuaJIT implementation. Ravi has a different goal compared to LuaJIT. Ravi prioritizes ease of maintenance and support, language safety, and compatibility with Lua 5.3, over maximum performance. For more detailed comparison please refer to the documentation links below.

Features

Articles about Ravi

Documentation

Lua Goodies

Lua 5.4 Position Statement

Lua 5.4 relationship to Ravi is as follows:

  • Generational GC - back-ported to Ravi.

  • New random number generator - back-ported to Ravi.

  • Multiple user values can be associated with userdata - under consideration.

  • <const> variables - not planned.

  • <close> variables - Ravi has 'defer' statement which is the better option in my opinion, hence no plans to support <close> variables.

  • Interpreter performance improvements - these are beneficial to Lua interpreter but not to the JIT backends, hence not much point in back-porting.

  • Table implementation changes - under consideration.

  • String to number coercion is now part of string library metamethods - back-ported to Ravi.

  • utf8 library accepts codepoints up to 2^31 - back-ported to Ravi.

  • Removal of compatibility layers for 5.1, and 5.2 - not implemented as Ravi continues to provide these layers as per Lua 5.3.

Compatibility with Lua 5.3

Ravi should be able to run all Lua 5.3 programs in interpreted mode, but following should be noted:

  • Ravi supports optional typing and enhanced types such as arrays (see the documentation). Programs using these features cannot be run by standard Lua. However all types in Ravi can be passed to Lua functions; operations on Ravi arrays within Lua code will be subject to restrictions as described in the section above on arrays.

  • Values crossing from Lua to Ravi will be subjected to typechecks should these values be assigned to typed variables.

  • Upvalues cannot subvert the static typing of local variables (issue #26) when types are annotated.

  • Certain Lua limits are reduced due to changed byte code structure. These are described below.

  • Ravi uses an extended bytecode which means it is not compatible with Lua 5.x bytecode.

  • Ravi incorporates the new Generational GC from Lua 5.4, hence the GC interface has changed.

Limit name

Lua value

Ravi value

MAXUPVAL

255

125

LUAI_MAXCCALLS

200

125

MAXREGS

255

125

MAXVARS

200

125

MAXARGLINE

250

120

When JIT compilation is enabled there are following additional constraints:

  • Ravi will only execute JITed code from the main Lua thread; any secondary threads (coroutines) execute in interpreter mode.

  • In JITed code tailcalls are implemented as regular calls so unlike the interpreter VM which supports infinite tail recursion JIT compiled code only supports tail recursion to a depth of about 110 (issue #17)

  • Debug api and hooks are not supported in JIT mode

History

  • 2015
  • 2016
  • 2017
    • Embedded C compiler using dmrC project (C JIT compiler) (now discontinued)

    • Additional type-annotations

  • 2018
  • 2019
    • New language feature - defer statement

    • New JIT backend MIR.

  • 2020
  • 2021
  • Current Priorities
    • Improve Embedded C support with more validation

    • Improve tests and documentation overall

    • Ensure new compiler is production grade (i.e. always generates correct code)

License

MIT License