Commit graph

5234 commits

Author SHA1 Message Date
Sebastian Ullrich
9476ee2f54 fix: benchmarks CI 2020-05-14 14:47:54 +02:00
Sebastian Ullrich
10253e89ea chore: move bin/ and .oleans into build directory 2020-05-14 14:47:54 +02:00
Sebastian Ullrich
872d5fc7ba feat: stop compiling Lean code as C++, remove --cpp option
Since we don't use static initializers, really the only difference between using `clang` and `clang++` is the default
inclusion of the C++ standard library.
2020-05-14 14:45:33 +02:00
Sebastian Ullrich
053d4bab1c chore: factor out and unify common test behavior; retrieve lean from PATH
`./test_single.sh foo.lean yes` is now `./test_single.sh -i foo.lean`
2020-05-14 14:38:52 +02:00
Sebastian Ullrich
2325f5d358 chore: ignore more test output files 2020-05-14 14:38:52 +02:00
Sebastian Ullrich
bc40796729 chore: remove fail/ tests
Checking for *any* failure is never a good idea. Use `$f.expected.ret` instead.
2020-05-14 14:38:52 +02:00
Sebastian Ullrich
76a97ea4fc feat: infer module name from cwd instead of LEAN_PATH, also make build system less specific to Init/ 2020-05-14 14:38:52 +02:00
Leonardo de Moura
ebc0663b3f chore: fix tests 2020-05-12 15:02:03 -07:00
Sebastian Ullrich
386c706f3e feat: basic parenthesizer 2020-05-04 14:28:36 -07:00
Sebastian Ullrich
5107403d24 feat: detect stack overflows on all platforms and threads 2020-05-04 11:11:11 +02:00
Sebastian Ullrich
e7920bcdb5 chore: remove test special case 2020-05-04 11:11:11 +02:00
Leonardo de Moura
5437ce9106 chore: workaround bugs exposed by previous commit 2020-04-23 11:32:39 -07:00
Sebastian Ullrich
8f67db0101 refactor: never implicitly ignore monadic results
Also change `do e; f` to desugar to `e *> f` so that it is affected as well
2020-04-23 11:09:59 -07:00
Marc Huisinga
aa4fc12dff feat: translate and refactor Gabriel Ebner's JSON parser
implement the JSON parser from https://github.com/gebner/lean4/tree/server by @gebner using new Lean 4 syntax and fix some minor bugs.
2020-04-15 11:29:18 -07:00
Leonardo de Moura
7322345768 fix: private names
@Kha I tried to fix a few issues with private names. The new test
tries to cover them. If you have more, please create an issue.

1- Scoping. A private declaration should shadow one in a previous scope.
2- We should not be able to define the same `private` in the same
module more than once.
   ```
   private def x := 10
   private def x := "hello" -- should produce error here
   ```
3- Dot-notation should work with private declarations in the module
where they were defined.
4- The following should work
   ```
   namespace N
   private def x := 10
   end N
   #check N.x
   ```
5- The following should **not** work
   ```
   def y := 10
   private def y := "hello" -- produce error

   private def z := 10
   def z := "hello" -- produce error
   ```
   BTW, I am happy to change this behavior. I just mimicked C's
   behavior for `static`.

It is not clear whether the following should work or not.
```
    namespace N
      private def b := 10
    end N
    open N
    #check b
```
2020-04-09 16:28:21 -07:00
Leonardo de Moura
360cebf680 fix: induction tactic must check expected major premise type
`whnfUntil` now returns `Option Expr`.
2020-04-09 11:18:02 -07:00
Leonardo de Moura
9a5bed0f94 fix: missing ! at whnfUntil 2020-04-09 10:49:20 -07:00
Leonardo de Moura
7a5fcfae8b feat: add [recursor] attribute to Iff.elim 2020-04-09 10:45:27 -07:00
Leonardo de Moura
40336131bb feat: do not switch to term mode in the RHS of induction and cases tactics 2020-04-08 17:45:06 -07:00
Daniel Selsam
7c1c0dc01e test: example that requires the instance arg reordering 2020-04-08 16:56:14 -07:00
Leonardo de Moura
e22af8d1ef feat: add FloatArray
cc @dselsam
2020-04-07 18:05:54 -07:00
Leonardo de Moura
2a1d9a0285 test: simple demo mixing C/C++ and Lean
@dselsam @kha
I did not have to create a new shared library.
The main limitation of this approach is that the new `extern`
functions are only available in compile code. That is, we cannot use
them with `#eval`.
2020-04-06 16:16:49 -07:00
Leonardo de Moura
b8ea55c989 feat: expose some transcendental functions from the C library
cc @dselsam
2020-04-06 15:22:25 -07:00
Leonardo de Moura
705530b62b fix: remove DecidableEq for Float
We cannot implement `DecidableEq Float` using C equality for
`double`. Reason: the C implementation is not even reflexive.
If we need `DecidableEq Float`, we will need to provide our own
implementation (i.e., a wrapper around the one provided by the
hardware). In this commit, we implement `HasBeq Float` instead.

cc @dselsam
2020-04-06 14:10:18 -07:00
Leonardo de Moura
8e84a8c9ec feat: Float from big numbers 2020-04-06 14:05:17 -07:00
Leonardo de Moura
746615d81d chore: remove as alternative for => 2020-04-06 13:45:22 -07:00
Leonardo de Moura
582fe3ac8a test: add test that exposed unboxing issue 2020-04-03 18:24:48 -07:00
Leonardo de Moura
51f1893a8c test: add basic float tests 2020-04-03 17:39:47 -07:00
Leonardo de Moura
22fc56f5c8 chore: fix test 2020-03-31 10:10:57 -07:00
Leonardo de Moura
af8bc28be0 fix: typo at IR.Expr.alphaEqv 2020-03-31 10:04:59 -07:00
Daniel Selsam
722ea2b694 failing test for suspicious behavior 2020-03-31 10:03:22 -07:00
Sebastian Ullrich
b6fc9428f1 fix: support Windows newlines and '\r' escape 2020-03-27 13:21:21 -07:00
Leonardo de Moura
0595e8a78a feat: add tryPureCoe
@Kha It seems to work reasonably well. It didn't break any test.
2020-03-27 11:20:43 -07:00
Sebastian Ullrich
fcdbdeecaf feat: delaborator: use implicit lambdas where possible
/cc @leodemoura it's not bullet-proof (unless `pp.explicit` is set), but let's
see if it is good enough in practice
2020-03-27 14:55:46 +01:00
Sebastian Ullrich
db93638a96 feat: delab forallE 2020-03-27 11:59:58 +01:00
Leonardo de Moura
02acdee9ce feat: disable implicit lambda insertion for fun containing {} or []
cc @Kha
2020-03-26 14:32:12 -07:00
Leonardo de Moura
b4b5f2736f feat: elaborate [] and {} binder annotations in lambda abstractions
We can simulate Lean3 behavior.
2020-03-26 14:20:39 -07:00
Leonardo de Moura
07ffa535fe refactor: decide whether to insert implicit lambdas AFTER macro expansion
cc @Kha
2020-03-26 13:53:15 -07:00
Leonardo de Moura
2b3bdbe0cf test: implicit lambdas + notation 2020-03-26 12:31:04 -07:00
Leonardo de Moura
e682c1c88f fix: make sure @ disable new syntax and macros too
cc @Kha
2020-03-26 12:14:55 -07:00
Leonardo de Moura
6625656940 refactor: remove support for fun {a : A} => ...
The semantics was weird. It seems Agda is also having problems with
it. Here is an example that demonstrates how weird the semantics is:

```lean
check (fun {β α} (a : α) (b : β) => (b, a) : {α : Type} → {β : Type} → (a : α) → (b : β) → β × α)

-- Same example using `def`
def f : {α : Type} → {β : Type} → α → β → β × α :=
fun {β : Type} {α : Type} (a : α) (b : β) => (b, a)
```

Both commands were being accepted before this commit. Note that it
flips `β` and `α`.

Here is an example that did not work before this commit and would
confuse users.

```lean
check
  let id := fun {α} (a : α) => a;
  id [id 1]
```

users would have to write

```lean
check
  let id {α} (a : α) := a;
  id [id 1]
```

@Kha The Delaborator.lean test broke and I "fixed" by removing the
`{}` from it, and copying `produced` over `expected`. Please make sure
it still makes sense.
2020-03-26 10:40:16 -07:00
Leonardo de Moura
d79cfa38e3 test: generalizeTelescope tests 2020-03-25 16:00:55 -07:00
Sebastian Ullrich
6ffe310c48 feat: delab coercions 2020-03-24 17:27:10 +01:00
Sebastian Ullrich
3c71c4f0ea chore: fix benchmark 2020-03-24 17:27:02 +01:00
Sebastian Ullrich
fdc03c9d74 feat: delab projections 2020-03-24 17:27:02 +01:00
Leonardo de Moura
2ca347fcd6 chore: fix test 2020-03-23 15:47:51 -07:00
Leonardo de Moura
dc80b76106 feat: add IO.FS.lines 2020-03-23 14:53:27 -07:00
Leonardo de Moura
41949ee801 refactor: String.toNat ==> String.toNat? and String.toNat! 2020-03-23 14:29:48 -07:00
Leonardo de Moura
94f016a3e8 fix: readLine and test 2020-03-23 14:06:17 -07:00
Leonardo de Moura
db914052ce fix: IO.getLine 2020-03-23 12:26:09 -07:00