Commit graph

1980 commits

Author SHA1 Message Date
Leonardo de Moura
368f17d0b1 feat(library/tactic/simplify): add simp! 2018-01-16 17:29:24 -08:00
Leonardo de Moura
c195d7c2a1 feat(library/tactic/simp_tactic): improve mk_simp_attr
- An new simp attribute may depend on other existing attributes

- Add `[norm]` simp attribute. It is an extension of the default `[simp]` attribute.
  It should be used to add extra rules for normalizing goals.
  These extra rules are used to produce normal forms and/or reduce the
  number of constants used in a goal. Here is an example coming from a
  discussion with @kha. When working with monads, we may want to
  eliminate `<$>` by using the lemma `f <$> x = x >>= pure ∘ f`.
  This lemma is in the `[norm]` simp set, but it is not in `[simp]`
2018-01-16 16:47:30 -08:00
Leonardo de Moura
6635f6c8c1 chore(library/init/coe): document why @[reducible] annotation is needed 2018-01-16 11:31:43 -08:00
Leonardo de Moura
cebde17bec feat(library/tactic/simplify): simp reduces c a_1 ... a_n = c b_1 ... b_n into a_1 = b_1 /\ ... /\ a_n = b_n 2018-01-12 18:18:56 -08:00
Leonardo de Moura
9eb22cd548 feat(library/constructions/injective): automatically generate auxiliary lemma *.inj_eq for constructors
We are going to use these lemmas in the simplifier.
2018-01-12 16:41:12 -08:00
Leonardo de Moura
58fce78282 feat(library/init/meta/interactive): add interactive tactic subst_vars 2018-01-12 14:37:11 -08:00
Leonardo de Moura
5bad6d5372 feat(library/init/meta/tactic): subst supports heterogeneous equalities that are actually homogeneous 2018-01-12 14:32:49 -08:00
Leonardo de Moura
1437d7209a feat(library/tactic/simplify): add support for generalized inductive datatypes 2018-01-12 11:51:49 -08:00
Leonardo de Moura
c5df94ed17 feat(library/tactic): add support for auto params at simp tactic 2018-01-11 16:47:22 -08:00
Leonardo de Moura
c2bd8626d7 fix(library/type_context): put back same_head_symbol heuristic
The timeout at qed_perf_bug test demonstrates why it is useful.
2018-01-09 15:31:33 -08:00
Leonardo de Moura
96792cd9aa chore(tests/lean): fix tests 2018-01-09 15:09:32 -08:00
Leonardo de Moura
43d7bac49b feat(library/init/meta): add support for new unify at rw tactic 2018-01-04 13:05:55 -08:00
Leonardo de Moura
040722c7e7 feat(library/init/meta): add unify config option to apply_cfg
This commit also fixes a problem in the `apply` tactic error messages.
2018-01-04 12:51:59 -08:00
Leonardo de Moura
d0dfb3f9f9 fix(library/tactic/simp_lemmas): closes #1863 2017-12-20 15:12:02 -08:00
Leonardo de Moura
c6ddc51c2b fix(library/app_builder): typo at mk_congr
closes #1893
2017-12-20 13:13:55 -08:00
Leonardo de Moura
a2dc2b6549 fix(util): null_output_channel
On OSX, Lean was often crashing when using trace messages.
I identified a problem in the thread finalization process.
In OSX, the `silent_ios_helper` at `library/trace.cpp` was being
finalized after the `null_streambuf` at `util/null_ostream.cpp`.
There was also a memory corruption problem also related to
`null_streambuf`.

This commit fixes this problem by using the following recipe
for creating null output stream buffers in C++.

https://stackoverflow.com/questions/11826554/standard-no-op-output-stream
2017-12-20 11:50:41 -08:00
Leonardo de Moura
91ff183b3e chore(library): remove out notation for out_param 2017-12-15 15:47:58 -08:00
Leonardo de Moura
f0352d31a1 feat(library/type_context, library): inout ==> out modifier in type class declarations
@kha: I decided to implement this change before I start the
type_context modifications. The change did not affect the corelib and
test suite much. The only annoying problem is that `out` cannot be
used to name locals anymore.
2017-12-15 14:46:47 -08:00
Leonardo de Moura
0492e49a3f fix(library/type_context): fixes #1888 2017-12-15 08:49:45 -08:00
Leonardo de Moura
6c44dd1b7f feat(frontends/lean): add hide command
cc: @kha
2017-12-13 11:53:21 -08:00
Leonardo de Moura
056a7db7b3 test(tests/lean/run): heap interface experiments
They expose limitations in the elaborator.
2017-12-12 18:19:30 -08:00
Leonardo de Moura
533ddc0279 fix(library/init/meta/interactive): remove buggy generalizing param from with_cases 2017-12-11 16:27:04 -08:00
Leonardo de Moura
5217ae735d feat(library/init/meta/interactive): do not make tag longer when constructor/apply create a single subgoal 2017-12-11 16:27:03 -08:00
Leonardo de Moura
f0231f17bc feat(library/init/meta): propagate tags in constructor-like tactics 2017-12-11 16:27:03 -08:00
Leonardo de Moura
e23db3970a feat(library/init/meta/tactic): apply tactic return parameter name associated with new metavars 2017-12-10 12:11:58 -08:00
Leonardo de Moura
8577fe6984 fix(library/init/meta/interactive): induction ... generalizing ... bug 2017-12-10 08:57:25 -08:00
Leonardo de Moura
24d5a1592d fix(library/init): add simp lemmas for auto_param and opt_param 2017-12-09 09:59:00 -08:00
Leonardo de Moura
ef784ce7b8 fix(library/tactic/simp_lemmas): auto_params when adding simp lemmas 2017-12-09 09:47:39 -08:00
Leonardo de Moura
9fff5ff710 chore(tests/lean/run): fix tests 2017-12-06 16:04:24 -08:00
Leonardo de Moura
a056e87350 fix(library/init/meta/injection_tactic): add support for ginductive datatypes 2017-12-06 09:39:20 -08:00
Leonardo de Moura
9dd382f649 chore(tests/lean): fix tests 2017-12-05 15:36:58 -08:00
Leonardo de Moura
458958b9fc feat(kernel/inductive): use ih to name induction hypothesis (instead of ih_1) when there is only one 2017-12-05 13:50:24 -08:00
Leonardo de Moura
6d96741010 feat(library): provide names for constructor arguments
Motivation: `cases` and `induction` tactics use these names when the
user does not provide them.
2017-12-04 16:25:16 -08:00
Leonardo de Moura
c943576e5a feat(library/init): add funext tactic 2017-12-04 14:54:39 -08:00
Leonardo de Moura
53c9737c70 feat(library/init): new repeat tactic 2017-12-04 12:55:53 -08:00
Leonardo de Moura
b7322e28c1 feat(library): do not using simp lemmas for sorting arguments of AC operators by default 2017-12-03 15:03:58 -08:00
Sebastian Ullrich
0ca9eb16c1 fix(library/type_context): preprocess_class: always solve universe mvars in inout 2017-11-29 17:21:02 -08:00
Leonardo de Moura
960832045f fix(library/type_context): failure condition 2017-11-29 14:35:58 -08:00
Sebastian Ullrich
18cf63e37f fix(frontends/lean/elaborator): avoid assertion error on delayed abstraction in structure notation 2017-11-24 21:27:55 +01:00
Leonardo de Moura
47994fe14e chore(library): remove id_locked 2017-11-22 16:29:04 -08:00
Leonardo de Moura
64f575a2d5 perf(library/equations_compiler): performance problem for definitions that produce many equational lemmas
The new test and comment at src/library/equations_compiler/util.cpp
explains the issue.
2017-11-22 16:16:11 -08:00
Leonardo de Moura
dd9d8e9552 chore(library/equations_compiler): improve comments 2017-11-22 14:55:40 -08:00
Sebastian Ullrich
aa8791a9ee fix(frontends/lean/structure_cmd): support dependent parents in new structure cmd 2017-11-17 16:31:30 -08:00
Leonardo de Moura
17172cbbdd fix(library/tactic/simplify): imp_congr was not preserving binder name 2017-11-17 12:45:04 -08:00
Leonardo de Moura
7f80bf60d1 feat(library/init/data): start rbtree module 2017-11-15 16:17:39 -08:00
Leonardo de Moura
49db6793f0 feat(library/init/data/ordering): cleanup ordering module, and add default cmp implementation 2017-11-13 21:55:41 -08:00
Leonardo de Moura
308bc8de4a chore(tests/lean): fix test suite 2017-11-10 16:59:09 -08:00
Leonardo de Moura
f2ef24696d perf(library/compiler/cse): make sure we eliminate common sub expressions in match-cases associated with 0-ary constructors
The new test exposes the problem. Before this commit, the common
subexpressions at

```
def tst : tree → nat
| (tree.leaf v) := v
| (tree.node v l r) :=
  match f v with
  | tt := tst l + tst l - tst l   -- <<< HERE
  | ff := tst r
  end
```

were not converted into a let-exprs.
2017-11-09 13:32:15 -08:00
Leonardo de Moura
fabf7f6380 perf(library/equations_compiler, library/compiler): expand auxiliary _match_idx definitions when generating byte code
We use the auxiliary procedure pull_nested_rec_fn to pull recursive
application in nested match expressions. This is needed because the
nested match expression is compiled before we process the recursive
procedure that contains it. This transformation may produce
performance problems if the recursive application does not depend on
the data being matched. Here is an example from the new test:

```
def tst : tree → nat
| (tree.leaf v) := v
| (tree.node v l r) :=
  match f v with
  | tt := tst l
  | ff := tst r
  end
```

pull_nested_rec_fn will convert it into

```
def tst : tree → nat
| (tree.leaf v)     := v
| (tree.node v l r) := tst._match_1 (f v) (tst l) (tst r)
```

Since our interpreter uses eager evaluation, both `(tst l)` and `(tst r)`
are executed. This commit fixes this issue by expanding `tst._match_1`
during code generation.
2017-11-09 11:14:57 -08:00
Leonardo de Moura
b40f3af801 feat(library/init/meta): move rb_tree and rb_map to native namespace
cc @kha
2017-11-07 13:41:45 -08:00