Nathan Braswell
|
ece3844c88
|
Quick commit adding br instruction
|
2021-07-24 00:39:45 -04:00 |
|
Nathan Braswell
|
ca25f2ca2b
|
Starting to add control instruction stuff, beginning with the block instruction. No nice labels or anything yet
|
2021-07-22 01:14:51 -04:00 |
|
Nathan Braswell
|
f740dd07e2
|
Implement locals and foldl, which is the direction we should have been folding to not reverse the params/locals anyway
|
2021-07-20 23:36:03 -04:00 |
|
Nathan Braswell
|
e271feed24
|
Implemented $ references for functions and parameters, which are also parsed for real now, along with the result. Setup for locals added, but not it's backend. Added pretty interesting add-dict-to-env function to collections.kp, which has also been added to put collections stuff in one place.
|
2021-07-20 00:37:27 -04:00 |
|
Nathan Braswell
|
835706c97d
|
Add both bare WAT ins and nested WAT ins
|
2021-07-19 00:32:42 -04:00 |
|
Nathan Braswell
|
d6d7af0bc1
|
Super basic and a tiny bit hardcoded WAT-like syntax for generating the array of arrays of arrays of arrays wasm implementation in wasm.kp, allowing us to write psudo-WAT in kraken and output binary wasm with 1 function call
|
2021-07-18 23:42:19 -04:00 |
|
Nathan Braswell
|
d3dd37c60e
|
Ooops, wasn't encoding numbers right at all, that's embarassing
|
2021-04-20 01:49:39 -04:00 |
|
Nathan Braswell
|
f254f11f8d
|
Generating first (in at least wasm3) runnable code! Added type, function, export, and code sections, with instructions and whatnot!
|
2021-04-20 01:36:38 -04:00 |
|
Nathan Braswell
|
e1fd8abd4f
|
Started working on basic wasm encoding/serialization. Added binary_file_writing, generalized arity comparison operators, bitwise operators, parsing of hex numbers, and wasm w/ memory section deserialization
|
2021-04-19 01:39:04 -04:00 |
|
Nathan Braswell
|
ed3b2ce743
|
Finished implementing real Damas-Hindley-Milner type system with let polymorphism!
|
2021-01-24 23:10:27 -05:00 |
|
Nathan Braswell
|
0a056ebef2
|
Limited monomorphic-only Damas-Hindley-Milner implemented in types
|
2021-01-24 02:53:55 -05:00 |
|
Nathan Braswell
|
f0d3be32f6
|
Add lambda to stlc
|
2021-01-18 19:06:28 -05:00 |
|
Nathan Braswell
|
7850866068
|
Start working on STLC again, this one's coming along. Primitives, builtins, and calls, but no lambda yet
|
2021-01-18 02:04:35 -05:00 |
|
Nathan Braswell
|
1f8fb59220
|
Fully functional now, removing set!, set-idx!, array-with-len, and making concat a builtin. Also, added 2 fun rules to new_kraken that allow nesting of new_kraken and k_prime languages
|
2021-01-17 19:57:56 -05:00 |
|
Nathan Braswell
|
7d7b2bd6d5
|
Add error/recover, rep->repl with error catching, and add it to scope so you can do it recursively and debug and whatnot. Also make it take in the grammer to repl with, and fix and commit the new_kraken work from earlier
|
2021-01-14 23:43:50 -05:00 |
|
Nathan Braswell
|
ddd5ce7032
|
Fix some longstanding string bugs and stop cheating with read-string in standard_grammer, implementing both unescape and string-to-int in k'
|
2021-01-04 00:11:15 -05:00 |
|
Nathan Braswell
|
6c0a46099a
|
New kraken with import and #lang passed start symbol
|
2021-01-02 13:55:07 -05:00 |
|
Nathan Braswell
|
4ed9af307f
|
Port method stuff over to new_kraken
|
2020-12-22 19:24:54 -05:00 |
|
Nathan Braswell
|
8d80f38f76
|
Fix a few parsing bugs at both levels, port some of the method demo over to new_kraken
|
2020-12-22 02:40:54 -05:00 |
|
Nathan Braswell
|
5152e1d109
|
Now actually have standard_grammar and with_import in scope let and all other ways in, having implemented let-vrec for mutually recursive vaus and fixing bugs. Tiny placeholder new_kraken definition and test.
|
2020-12-20 02:10:10 -05:00 |
|
Nathan Braswell
|
03b588f1f9
|
Insert standard_grammar into the standard scope via recursion, allowing it to be used in sub-languages
|
2020-12-19 23:05:46 -05:00 |
|
Nathan Braswell
|
85dbaae9de
|
Add vector, quote, and quasiquote syntax to new_kraken, and refactor types to use meta & new syntax
|
2020-10-20 22:59:21 -04:00 |
|
Nathan Braswell
|
371f39c82b
|
First impl of type-systems-as-vau-instead-of-macros in new-new-kraken
|
2020-10-20 22:11:57 -04:00 |
|
Nathan Braswell
|
168589c364
|
Finished generic concat, so let-rec properly works now
|
2020-10-18 20:18:38 -04:00 |
|
Nathan Braswell
|
f9529b02aa
|
Impl function let-rec with Y*
|
2020-10-18 19:11:23 -04:00 |
|
Nathan Braswell
|
d689a59097
|
Experimenting with fungll optimization, implement the okmij.org poly-variadic fix-point combinator for mutual recursion
|
2020-10-17 11:59:54 -04:00 |
|
Nathan Braswell
|
45ecb25bb1
|
Lang with proper scoping
|
2020-10-11 23:24:17 -04:00 |
|
Nathan Braswell
|
844bb0f189
|
Working language mod, though it doesn't have access to anything defined in the outer
|
2020-10-11 22:31:36 -04:00 |
|
Nathan Braswell
|
42c2fcfb66
|
Cheating in the same way, strings!
|
2020-10-11 20:01:18 -04:00 |
|
Nathan Braswell
|
0820113c67
|
Fix error swallowing grammer errors for custom grammers on read-string, start constructing standard_grammar
|
2020-10-11 19:42:00 -04:00 |
|
Nathan Braswell
|
e6319d18f7
|
Add all-k_prime grammar argument to read-string, enabling all-in-k_prime grammer changing
|
2020-09-20 13:52:56 -04:00 |
|
Nathan Braswell
|
23d2c166cd
|
Basic provide & with_import, fix a rep bug where all but first execution was in root_env instead of standard env
|
2020-09-19 12:41:52 -04:00 |
|
Nathan Braswell
|
e8899fc093
|
Better hermetic scoping
|
2020-09-19 00:04:09 -04:00 |
|
Nathan Braswell
|
1d97366933
|
Implement user-interatction part of k_prime in new_kraken.kp, moved from load_test.kp
|
2020-09-17 23:28:45 -04:00 |
|
Nathan Braswell
|
62e991d0dd
|
Y combinator works, as well as vY (the vau version, which is different b/c applicative order), and prototype for new load style removing set works
|
2020-09-16 00:07:49 -04:00 |
|
Nathan Braswell
|
23910ba51e
|
Update web version, add a sharable try.html
|
2020-09-13 19:03:44 -04:00 |
|
Nathan Braswell
|
aa75269608
|
Working string interpolation!
|
2020-09-13 18:25:30 -04:00 |
|
Nathan Braswell
|
59c6bceb4d
|
work in progress string interpolation
|
2020-09-13 18:05:54 -04:00 |
|
Nathan Braswell
|
33d28fe1fa
|
Added {} bodies, C style function call, and Kraken-style fun
|
2020-09-12 21:34:23 -04:00 |
|
Nathan Braswell
|
54cffb4185
|
massage new constructor to be a function and add methods to both it and the new syntax
|
2020-09-08 23:30:56 -04:00 |
|
Nathan Braswell
|
ba64276630
|
Begin object/struct syntax & semantics
|
2020-09-08 00:25:41 -04:00 |
|
Nathan Braswell
|
3cb5c8d827
|
Add TCO option to BuiltinCombinator and convert cond and eval to use this, then rewrite self-hosted do so that it too is TCO. This allows us to self-host cond (which we did) so without worring about stack space for large arrays
|
2020-09-07 15:41:27 -04:00 |
|
Nathan Braswell
|
c0eca02e43
|
Update website
|
2020-09-06 22:10:59 -04:00 |
|
Nathan Braswell
|
0d07d22995
|
Move to wrap and unwrap
|
2020-09-06 12:19:19 -04:00 |
|
Nathan Braswell
|
8e47cb2ec2
|
Move to prelude and add enough to run bf (with modifications)
|
2020-08-29 00:33:04 -04:00 |
|
Nathan Braswell
|
241b4ca3f4
|
Implemented quasiquoting as a vau!
|
2020-08-28 00:38:16 -04:00 |
|
Nathan Braswell
|
a5f15f70f8
|
Fixed the eval issue for builtins mentioned last time, and undid lambda fix. Fixed it by unevaling in appropriate spots by wrapping in a call to quote, or equilivant. Map at least should be rewritten a language feature, for sure.
|
2020-08-20 22:57:01 -04:00 |
|
Nathan Braswell
|
94a24cb30f
|
Somehow forgot to commit rc all this time
|
2020-08-20 18:37:43 -04:00 |
|
Nathan Braswell
|
d69e77ee33
|
Added vector syntatic sugar [], set-idx!, and neatend up some of the self-implementation with the ' and [] syntatic sugar. Realized that currently most builtins pass arguments evaluated such that functions that they are passed really have to be vau instead of lambda. This includes both add_grammer_rule and more common ones like map, etc. I think I might have to undo my lambda fix in recognition of this. Anyway, either I have to pass arguments unevaluated or I need to swap to wrap/unwrap instead of my current lambda impl. It would be more like Kernel, and might make optimization etc easier, but I'm a tad worried about unwrap being exposed to the user / not forcing the more general optimization based on partial evaluation on myself (maybe doing this even makes optimization harder or less elegant?)
|
2020-08-14 22:37:03 -04:00 |
|
Nathan Braswell
|
8ea030547a
|
Finally commit earlier work now that I've fixed the bug where you couldn't use new syntax inside of functions because the inner closure struct was being corrupted and giving a null env. Now I use a nil env
|
2020-08-14 00:37:55 -04:00 |
|