Leonardo de Moura
e3f32a6108
chore(library/init/lean/compiler): cleanup
2019-05-02 15:02:35 -07:00
Leonardo de Moura
e90224b435
fix(library/init/lean/compiler/simpcase): bugs at simpCase
2019-05-02 15:02:21 -07:00
Leonardo de Moura
76a49ec256
chore(library/compiler): add ir::test
2019-05-02 14:40:04 -07:00
Leonardo de Moura
3628b39cb6
feat(library/init/lean/compiler): add simpcase transformation
2019-05-02 12:40:37 -07:00
Leonardo de Moura
db30a23055
fix(library/init/lean/compiler): bug fixes
2019-05-02 12:22:00 -07:00
Leonardo de Moura
81d11db5d2
chore(runtime/object): rename runtime primitives
2019-05-02 10:55:29 -07:00
Leonardo de Moura
4c9a488446
chore(library/init/data/array): naming convention
...
@kha Trying again :)
I started using the prefix `f` for the `Fin` version (e.g., `fswap` and
`fswapAt`). So, it seemed natural to use the prefix `f` for the `Fin`
versions of `get` and `set`.
BTW, is it too crazy to use `a[i]` (without spaces) as notation for
`a.get i`? The constraint is to make sure `f a [i]` is not ambiguous.
That is, `f a[i]` is `f (a.get i)` and `f a [i]` is `f a (i::nil)`.
This is doable with the new parser.
Then, we could have `a[i]` as notation for `a.fget i` if `i` is a `Fin`,
and `a.get i` if `i` is `Nat`.
Similarly, `a[i := v]` would be notation for `a.fset i v` if `i` is
`Fin`, and `a.set i` if `i` is `Nat`.
It would also be awesome to have
```
let a[i] := v in
e
```
as notation for
```
let a := a[i := v] in
```
2019-05-02 10:23:53 -07:00
Leonardo de Moura
970941db2c
feat(library/init/data/array/basic): add Array.filter
2019-05-02 09:51:08 -07:00
Leonardo de Moura
46c3a14f83
chore(library/init/lean/compiler/pushproj): move comment
2019-05-02 09:48:17 -07:00
Leonardo de Moura
719eb67114
chore(library/init/lean/compiler): VarIdxSet => IndexSet
2019-05-02 08:23:59 -07:00
Leonardo de Moura
3a612bfd8b
feat(library/init/lean/compiler): add elimDead transformation
2019-05-02 08:06:45 -07:00
Leonardo de Moura
5a83a2d7bb
feat(library/init/data/array/basic): add swapAt
2019-05-02 07:46:11 -07:00
Leonardo de Moura
94fe3c18d0
fix(library/init/data/rbtree/basic): add HasInsert instance
...
It is useful for defining finite trees
2019-05-01 21:19:07 -07:00
Leonardo de Moura
e52e787ad5
fix(library/init/lean/compiler/pushproj): bug and cleanup
2019-05-01 21:01:03 -07:00
Leonardo de Moura
2991b966e5
featg(library/init/lean/compiler): add pushproj
2019-05-01 17:38:44 -07:00
Leonardo de Moura
1707628b6b
feat(library/init/data/array/basic): Array.reverse
2019-05-01 17:38:44 -07:00
Leonardo de Moura
af1d521706
feat(library/init/data/array/basic): add Array.swap
2019-05-01 17:38:44 -07:00
Leonardo de Moura
cfec797e69
feat(library/init/lean/compiler/ir): helper functions
2019-05-01 17:38:44 -07:00
Leonardo de Moura
45d09d3044
fix(library/compiler/ir): bug at LLNF -> IR
2019-05-01 17:38:44 -07:00
Leonardo de Moura
ed5e461130
feat(library/init/lean/compiler/ir): add maxVar
2019-05-01 17:38:44 -07:00
Leonardo de Moura
2614b95a8b
refactor(library/init/lean/compiler/ir): use Nat instead of Name for local vars
2019-05-01 17:38:44 -07:00
Leonardo de Moura
fa0b4bff40
chore(tests/playground/parser/syntax): fix experiment
2019-04-30 18:06:03 -07:00
Leonardo de Moura
0c9fa13763
feat(library/init/lean/compiler): convert LLNF into Lean IR
2019-04-30 17:55:43 -07:00
Leonardo de Moura
d433811c64
test(tests/playground): add "checkpoint" variant for rbmap benchmark
...
@kha I created a variant of the `rbmap` example where we create a tree
but also save "checkpoints". The idea is to simulate the idiom frequently
used in a backtracking search where we "save" the "context" before each
case split in a "trail stack".
The benchmark has two parameters: the number of nodes to be inserted, and a "frequency" (how often we create a "checkpoint").
The command `rbmap_checkpoint.lean.out n n` behaves like the original
`rbmap` benchmark, and `rbmap_checkpoint.lean.out n 1` creates a
checkpoint after each insertion. The frequency provides a simple way to
control the amount of sharing. We can provide performance numbers for
different frequencies, and show the impact on the `reset/reuse` optimization.
BTW, the performance numbers are much better than I expected. For
example,
`./rbmap_checkpoint.lean.out 1000000 10` is only 30% slower than
`./rbmap_checkpoint.lean.out 1000000 1000000`
although 100k checkpoints were created.
Another good news is that we are faster than Haskell even for
`./rbmap_checkpoint.lean.out 1000000 1`
2019-04-30 10:41:23 -07:00
Leonardo de Moura
952eb0f515
feat(library/compiler): C++ API for Lean IR
2019-04-29 18:23:19 -07:00
Leonardo de Moura
8befa9bea5
chore(stage0): update
...
Make sure Lean IR primitives are available at Stage0.
2019-04-29 17:16:27 -07:00
Leonardo de Moura
2db7cd158b
feat(library/init/lean/compiler/ir): export primitives to C++
...
Remark: we don't need to expose all of them since the C++ code only
generates code in the Lambda_pure fragment.
2019-04-29 17:10:42 -07:00
Leonardo de Moura
a32fcf33c7
feat(util/array_ref): simple wrapper for creating Lean array objects in C++
2019-04-29 17:10:35 -07:00
Leonardo de Moura
e0821be132
chore(tests/playground/ir): fix test
2019-04-29 10:50:10 -07:00
Leonardo de Moura
c09df2d8c3
feat(library/init/lean/compiler/ir): use Array instead of List
...
The idea is to use the same approach we have used at `tests/playground/parser/syntax.lean`
2019-04-29 10:48:33 -07:00
Leonardo de Moura
711fab451b
refactor(library/init/data/array): Array.foldl argument order
2019-04-29 10:48:33 -07:00
Leonardo de Moura
63442ebde7
feat(library/init/data/array/basic): add iterate₂ and foldl₂
2019-04-29 10:48:33 -07:00
Leonardo de Moura
5298841def
test(tests/playground/ir): small test
2019-04-29 10:47:51 -07:00
Leonardo de Moura
df84868ab4
feat(library/init/data/array/basic): array helper functions
2019-04-28 10:10:24 -07:00
Leonardo de Moura
dd5fa4626f
chore(runtime/object): style
2019-04-28 09:18:54 -07:00
Leonardo de Moura
807c9bff6d
refactor(library/init): HasToFormat ==> HasFormat, toFormat ==> format
2019-04-28 09:16:47 -07:00
Leonardo de Moura
79e2abe33f
feat(CMakeLists): put configuration options relevant to leanc at config.h
2019-04-27 21:04:41 -07:00
Leonardo de Moura
13d2398fb3
feat(library/init/lean/compiler/ir): formatter
2019-04-27 17:55:27 -07:00
Leonardo de Moura
03f42fda34
refactor(library/init/data/rbmap/basic): use [inline] instead of auxiliary "corpse"
2019-04-27 08:10:26 -07:00
Leonardo de Moura
aefe49d575
test(tests/playground): add new rbmap experiment
2019-04-27 08:02:32 -07:00
Leonardo de Moura
35317139fd
chore(library/compiler/util): style
2019-04-26 16:34:22 -07:00
Leonardo de Moura
e1a84d2f2c
fix(library/compiler/struct_cases_on): performance problem exposed by badupdate1.lean
2019-04-26 16:30:19 -07:00
Leonardo de Moura
240ca3fc68
test(tests/playground/badupdate1): add test for exposing performance bug at struct_cases_on_fn
...
```
./run.sh badupdate1.lean 4000
8000
test1 2.12s
8000
test2 1.11ms
```
2019-04-26 15:04:39 -07:00
Leonardo de Moura
3c52183e3c
fix(library/compiler/struct_cases_on): bug
2019-04-26 15:04:16 -07:00
Leonardo de Moura
c6d0083456
fear(library/init/data/array/basic): add Array.modify
2019-04-26 14:41:17 -07:00
Leonardo de Moura
1a1105b533
test(tests/playground/lazylist): perf tests
2019-04-26 13:12:39 -07:00
Leonardo de Moura
a5128484a1
test(tests/playground/lazylist): perf tests
2019-04-26 12:05:16 -07:00
Leonardo de Moura
07a68375f5
chore(library/init/lean/compiler/ir): style
2019-04-26 09:15:37 -07:00
Leonardo de Moura
5c7849a869
feat(runtime): eager heap initialization
2019-04-25 18:10:36 -07:00
Leonardo de Moura
9a39eb254a
chore(tests/playground): fix tests
2019-04-25 16:48:06 -07:00
