Commit graph

1616 commits

Author SHA1 Message Date
Leonardo de Moura
7772c16033 refactor(kernel): add unfold_opaque flag to normalizer, modify how type checker uses the opaque flag, remove hidden_defs, and mark most builtin definitions as opaque
After this commit, in the type checker, when checking convertability, we first compute a normal form without expanding opaque terms.
If the terms are convertible, then we are done, and saved a lot of time by not expanding unnecessary definitions.
If they are not, instead of throwing an error, we try again expanding the opaque terms.
This seems to be the best of both worlds.
The opaque flag is a hint for the type checker, but it would never prevent us from type checking  a valid term.

Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-20 12:47:47 -08:00
Leonardo de Moura
026f666526 feat(kernel/type_checker): use expression before normalization in the unification constraints generated by the typechecker
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-20 11:22:13 -08:00
Leonardo de Moura
daef2b7b24 feat(util/sexpr/options): add is_eqp predicate for options
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-20 10:53:53 -08:00
Leonardo de Moura
4838c055b8 feat(kernel/environment): add set_opaque method
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-20 10:45:44 -08:00
Leonardo de Moura
812c1a2960 feat(library/elaborator): only expand definitions that are not marked as hidden
The elaborator produces better proof terms. This is particularly important when we have to prove the remaining holes using tactics.
For example, in one of the tests, the elaborator was producing the sub-expression

 (λ x : N, if ((λ x::1 : N, if (P a x x::1) ⊥ ⊤) == (λ x : N, ⊤)) ⊥ ⊤)

After, this commit it produces

 (λ x : N, ¬ ∀ x::1 : N, ¬ P a x x::1)

The expressions above are definitionally equal, but the second is easier to work with.

Question: do we really need hidden definitions?
Perhaps, we can use only the opaque flag.

Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-20 02:16:49 -08:00
Leonardo de Moura
cb48fbf3c4 fix(library/elaborator): missing case
The elaborator was failing in the following scenario:
- Failing constraint of the form
     ctx |- ?m1 =:= ?m2
where
     ?m2 is assigned to ?m1,
     and ?m1 is unassigned.

has_metavar(?m2, ?m1) returns true, and a cycle is incorrectly reported.

Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-20 01:39:43 -08:00
Leonardo de Moura
96ea8b81c8 feat(frontends/lean/parser): change show-expression binder name
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-20 01:30:18 -08:00
Leonardo de Moura
c730dd7872 feat(frontends/lean/parser): propagate position information to expressions created by macro implemented in Lua
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-20 01:07:37 -08:00
Leonardo de Moura
3eb4de6760 fix(frontends/lean/parser): fix deadlock in macro parser
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-19 21:40:00 -08:00
Leonardo de Moura
c77464703f feat(frontends/lean): macro definition using Lua
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-19 19:08:10 -08:00
Leonardo de Moura
b08c606696 fix(library/io_state): bug in the io_state Lua bindings
This commit also includes a new test that exposes the problem.
The options in the io_state object were being lost.

Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-19 16:39:42 -08:00
Leonardo de Moura
2648f41eaa test(tests/lean): add new test script that checks if Lean can parse the output produced by its pretty printer
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-19 16:16:56 -08:00
Leonardo de Moura
d9e692f506 feat(frontends/lean): improve coercion manangement
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-19 15:45:46 -08:00
Leonardo de Moura
f43db96e1f fix(frontends/lean/pp): pretty printer for Type
Add parenthesis around Type when it has a universe.

Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-19 15:24:45 -08:00
Leonardo de Moura
ae01d3818d fix(frontends/lean/parser): parse_type method
The parser had a nasty ambiguity. For example,
    f Type 1
had two possible interpretations
    (f (Type) (1))
or
    (f (Type 1))

To fix this issue, whenever we want to specify a particular universe, we have to precede 'Type' with a parenthesis.
Examples:
    (Type 1)
    (Type U)
    (Type M + 1)

Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-19 15:24:34 -08:00
Leonardo de Moura
46627289b8 fix(kernel/expr): avoid '_' as a binder name, we use '_' as a placeholder in the Lean frontend
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-19 14:47:53 -08:00
Leonardo de Moura
6cc83dbe2a fix(kernel/kernel_exception): incorrect pp method
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-19 14:46:22 -08:00
Leonardo de Moura
d3d24696f4 feat(frontends/lean): hide builtin object in the 'Show Environment' command
The user can still display builtin objects by using

    Show Environment all

Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-19 14:00:58 -08:00
Leonardo de Moura
ad3f771b1d feat(frontends/lean): hide 'explicit' version of objects with implicit arguments
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-19 13:12:39 -08:00
Leonardo de Moura
bff5a6bfb2 fix(frontends/lean/pp): make sure pp and parser are using the same precedences
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-19 12:46:14 -08:00
Leonardo de Moura
02bd166793 chore(library/basic_thms): fix typo in comment
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-19 12:22:05 -08:00
Leonardo de Moura
9b14744405 chore(examples/lean): use the same notation in the example set.lean
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-18 21:24:04 -08:00
Leonardo de Moura
dd72269b13 feat(frontends/lean): rename command Set to SetOption
It is not nice to have Set as a reserved keyword. See example examples/lean/set.lean

Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-18 21:18:48 -08:00
Leonardo de Moura
d7886c4f5f doc(examples/lean): new example
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-18 21:03:16 -08:00
Leonardo de Moura
42cb5d1d3c fix(tests/lean): typo
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-18 20:11:49 -08:00
Leonardo de Moura
8cfe5cf9ed fix(frontends/lean/pp): pretty printer was ignoring notation decls in the local scope
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-18 18:00:37 -08:00
Leonardo de Moura
79fa6e4940 feat(frontends/lean): Scopes in the default Lean frontend
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-18 17:40:21 -08:00
Leonardo de Moura
97b872a05c refactor(frontends/lean): remove frontend class, it is not needed anymore
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-18 14:37:55 -08:00
Leonardo de Moura
2aaa9a5273 feat(frontends/lean/parser): change function application precedence
Now, we can write

  Pi (x y : A), R x y -> R y x

instead of

  Pi (x y : A), (R x y) -> (R y x)

Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-18 12:44:15 -08:00
Leonardo de Moura
47c7bb1bde refactor(*): uses aliases for unordered_map and unordered_set
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-18 12:30:45 -08:00
Leonardo de Moura
1e4fa76a47 feat(util/name_map): add template alias
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-18 11:34:40 -08:00
Leonardo de Moura
7b2fea3fab fix(kernel/normalizer): compilation problem with clang++
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-18 08:46:36 -08:00
Leonardo de Moura
418623b874 feat(kernel/replace_fn): add template replace that captures commonly used pattern
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-17 18:31:59 -08:00
Leonardo de Moura
23e518001a feat(kernel/normalizer): avoid unnecessary creation of closures for n-ary functions
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-17 18:10:13 -08:00
Leonardo de Moura
10f28c7bec feat(kernel/replace_fn): non-recursive replace_fn
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-17 16:35:39 -08:00
Leonardo de Moura
af4a6c9364 fix(kernel/normalizer): cache problems
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-17 14:52:14 -08:00
Leonardo de Moura
33789fad4c fix(kernel/builtin): make sure the if-then-else semantic attachment is not a simplifier
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-17 14:34:40 -08:00
Leonardo de Moura
c53233ea26 fix(kernel/normalizer): avoid svalue hack, use 'semantic attachments' for implementing closures, include context in the closure
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-17 14:33:42 -08:00
Leonardo de Moura
836357c65c fix(kernel/normalizer): bug in Let normalization
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-17 12:35:25 -08:00
Leonardo de Moura
84bfe2a222 fix(library/elaborator): bug in process_meta_app
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-17 10:56:20 -08:00
Leonardo de Moura
b887a2b900 fix(examples/lean/tactic_in_lua): try to make sure the example works when using LuaJIT
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-17 08:58:06 -08:00
Leonardo de Moura
09b51a0fb7 fix(library/elaborator): missing condition
The elaborator was missing solutions because of the missing condition at is_simple_ho_match.

This commit also adds a new test that exposes the problem.

Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-16 17:13:36 -08:00
Leonardo de Moura
f09fd0fc04 feat(kernel/printer): include de Bruijn index in the debug printer
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-16 16:56:46 -08:00
Leonardo de Moura
91f4ced83b feat(library/elaborator): do not create trivial constraints of the form 'ctx |- t =:= t'
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-16 16:41:20 -08:00
Leonardo de Moura
5b29fd8d93 test(tests/lean): add ExistsElim test
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-16 15:43:09 -08:00
Leonardo de Moura
7792561b20 fix(library/type_inferer): another incorrect use of scoped_map
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-16 15:17:19 -08:00
Leonardo de Moura
af42078205 fix(kernel): incorrect use of scoped_map
This commit also adds a new test that exposes the problem.
The scoped_map should not be used for caching values in the normalizer and type_checker. When we extend the context, the meaning of all variables is modified (we are essentially performing a lift). So, the values stored in the cache are not correct in the new context.

Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-16 15:11:39 -08:00
Leonardo de Moura
2fee2def72 feat(library/basic_thms): simplify DoubleNegElim
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-16 13:19:19 -08:00
Leonardo de Moura
de53e92de8 feat(library/basic_thms): add ExistsElim theorem
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-16 12:43:34 -08:00
Leonardo de Moura
8f5c2b7d9f feat(library/basic_thms): add Refute theorem
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
2013-12-16 12:03:31 -08:00