Our great sponsors
mlton | cakeml | |
---|---|---|
9 | 14 | |
916 | 912 | |
1.2% | 2.1% | |
8.3 | 9.8 | |
20 days ago | 7 days ago | |
Standard ML | Standard ML | |
GNU General Public License v3.0 or later | GNU General Public License v3.0 or later |
Stars - the number of stars that a project has on GitHub. Growth - month over month growth in stars.
Activity is a relative number indicating how actively a project is being developed. Recent commits have higher weight than older ones.
For example, an activity of 9.0 indicates that a project is amongst the top 10% of the most actively developed projects that we are tracking.
mlton
-
Flunct: Well-typed, fluent APIs in SML
https://github.com/MLton/mlton/issues/473
Is there sufficient use of MLTon "native" backend out there to consider it mature? or Do people prefer the LLVM or C backend instead in general?
-
Simple JSON parser in c++, rust, ocaml, standard ml
Once I got the parser ready in OCaml, I thought I port it to Standard ML, since it belong to the same ML language family. I was also curious on how well mlton could optimise it. The language lacks custom let bindings, so I resorted to use Result.bind manually. This makes code much less readable and more verbose. The standard library also lacks result type, so I had to come up with my own simple implementation. There's also a lack of any hash map in the standard library, so I just used a list of key-value pairs. This isn't correct, but it's the closest I could get without inventing my own hash map. MLton's compile times are slow. It also lacks interactive REPL. Because of that I used alternative Standard ML implementation for interactive usage: PolyML. Debugging MLton binaries is also pretty hard. gdb doesn't work and there's no bundled debugger. I had to resort to debugging facilities built into PolyML. Valgrind doesn't work with mlton binaries, as it doesn't report any memory allocations. Looks like mlton uses mmap for allocation memory. Surprisingly, performance is not the best. This might be due to heavy usage of my custom Result type and bind calls. Exceptions seem to be a more natural choice for error reporting in Standard ML. I tried to make such a change, but this didn't improve the performance much.
-
old languages compilers
MLton
-
Modules: Overcoming Stockholm and Duning-Kruger
Something I’d highly recommend you do before concluding that SML’s module system is the best is to go through and read the MLton Basis library. MLton uses the module system extremely heavily in its definition of the standard, and I think it’s extremely important to understand what you may be getting yourself into when you add those features, and what you may lose in return.
-
Ante: A low-level functional language
If you’re fine with tracing GC (which depends on the situation, of course), Standard ML is a perfectly boring language (that IIUC predated and inspired Caml) and MLton[1] is a very nice optimizing compiler for it. The language is awkward at times (in particular, the separate sublanguage of modules can be downright unwieldy), and the library has some of the usual blind spots such as nonexistent Unicode support (well, not every language WG is allocated a John Cowan).
Speaking of, Scheme can also be a delightful unexciting static language; consider for example the C-producing implementation Chicken[2]. The pattern-matching / algebraic-datatype story was still rather unsatisfying last I checked, but there are other situations where it shines—it’s complementary to SML in a way.
You’re not going to be writing a kernel or a real-time renderer in either (though I’m certain people have taken that as a challenge), they son’t afford the flashy EDSLs of Tcl, Ruby, or Racket, and I can’t say I can prototype in them like I do in Python or sh+tools, but there is a comfortable middle ground where they fit well. (I hear others use Go in what seem like the same places, but to me it feels so thin and devoid of joy that I can’t really compare.)
The FFI tools in both of the mentioned implementations are excellent, though not quite at the “type in C declarations” level of LuaJIT and D.
[1] http://mlton.org/
[2] https://call-cc.org/
-
Write your own programming language in an hour with Chumsky
Unfortunately, I haven't found a ton of "easily-digestible" and, at the same time, comprehensive guides on compiling functional languages. Generally you'll find a mix of blog posts/class notes/papers covering a single step.
Some resources I like:
- Andrew Kennedy's 2007 paper Compiling with Continuations, Continued [1]. This one is the most clear IMO
- Andrew Appel's Compiling with Continuations book (a bit outdated though... assembly code is for VAX)
- Matt Might's series [2]
- MLton's source and documentation [3]
[1] https://www.microsoft.com/en-us/research/wp-content/uploads/...
[2] https://matt.might.net/articles/closure-conversion/
[3] http://mlton.org/
-
Why are imperative programs considered faster than their functional counterparts?
More broadly, they can be fast even without such extensions if they aggressively pursue optimization opportunities afforded by static typing, like MLton for example, but that also impacts compilation performance negatively.
- Coalton: How to Have Our (Typed) Cake and (Safely) Eat It Too, in Common Lisp
-
Are there any efficient key-value map/dictionary implementations in SML?
https://github.com/MLton/mlton/blob/master/lib/mlton/basic/hash-set.sig https://github.com/MLton/mlton/blob/master/lib/mlton/basic/hash-table.sig
cakeml
-
The Deep Link Equating Math Proofs and Computer Programs
If I understand what you are asking about correctly, then I do think you are mistaken.
As a sibling comment observed, you would be proving something about a program, but proving things about programs is both possible and done.
This ranges from things like CakeML (https://cakeml.org/) and CompCert (compilers with verified correctness proofs of their optimizations) to something simple like absence of runtime type errors in statically strongly soundly-typed languages.
Of note is that you are proving properties of your program, not proving them perfect in every way. The properties of your program that you prove can vary wildly in both difficulty and usefulness. A sufficiently advanced formally verified compiler like CakeML can transfer a high-level proof about your source code to a corresponding proof about the behavior of the generated machine-executable code.
-
The future of Clang-based tooling
> A single IR with multiple passes is a good way to build a compiler
https://mlir.llvm.org/, which is using, is largely claiming the opposite. Most passes more naturally are not "a -> a", but "a -> b". data structures and data structures work hand in hand, it is very nice to produce "evidence" for what is done in the output data structure.
This is why https://cakeml.org/, which "can't cheat" with partial functions, has so many IRs!
Using just a single IR was historically done for cost-control, the idea being that having many IRs was a disaster in repetitive boilerplate. MLIR seeks to solve that exact problem!
- CakeML – A Verified Implementation of ML
-
Tools for Verifying a Language and its Semantics
You may want to look at [CakeML](https://cakeml.org) done in HOL4, there is also a nice proof pearl about a more .. minimalistic verified bootstrapped compiler also in HOL4.
-
old languages compilers
CakeML
-
Is there a formally-proven real-time language/computing env. or operating system?
There is also Cake ML which is a formally verified functional programming language compiler and runtime.
-
CakeML: A Verified Implementation of ML
There is also a CakeML -> Standard ML compiler though it seems to have been built to translate benchmarks and sort of old so I'm not sure how comprehensive it is: https://github.com/CakeML/cakeml/tree/master/unverified/front-end
-
The λ-Cube
> One guess is that lisps cope with being minimal through use of macros and metaprogramming, it's difficult for a typed language to support that level of metaprogramming while maintaining the various guarantees that one wants from such a system.
Difficult, but certainly not impossible [0].
[0] https://cakeml.org/
-
Two Mechanisations of WebAssembly 1.0
If this interests you, I'd highly recommend checking out CompCert (docs here) and CakeML.
- VLISP: A Verified Implementation of Scheme [pdf]
What are some alternatives?
LunarML - The Standard ML compiler that produces Lua/JavaScript
Daikon - Dynamic detection of likely invariants
typed-racket - Typed Racket
hardware - Verilog development and verification project for HOL4
tao - A statically-typed functional language with generics, typeclasses, sum types, pattern-matching, first-class functions, currying, algebraic effects, associated types, good diagnostics, etc.
mpl - The MaPLe compiler for efficient and scalable parallel functional programming
seL4 - The seL4 microkernel
CompCert - The CompCert formally-verified C compiler
sml-parseq - parallel sequences library in Standard ML
Checker Framework - Pluggable type-checking for Java
smlfmt - A custom parser/auto-formatter for Standard ML
checkedc - Checked C is an extension to C that lets programmers write C code that is guaranteed by the compiler to be type-safe. The goal is to let people easily make their existing C code type-safe and eliminate entire classes of errors. Checked C does not address use-after-free errors. This repo has a wiki for Checked C, sample code, the specification, and test code.