michael/lean2
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions
3041 commits 3 branches 0 tags 28 MiB
Commit graph

514 commits

Author SHA1 Message Date
Leonardo de Moura
aff766430d fix(frontends/lean/pp): universe pretty printer 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-09 19:17:28 -07:00
Leonardo de Moura
43eba857cb feat(frontends/lean): add let-expr pretty printer, reduce default indentation to 2 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-09 18:47:10 -07:00
Leonardo de Moura
2ef7b9be2f feat(frontends/lean): add basic pretty printer 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-09 01:12:36 -07:00
Leonardo de Moura
2ad1a93657 test(tests/lean/run): add append tactic test 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-08 18:21:22 -07:00
Leonardo de Moura
a1601e7a5f feat(library/tactic/apply_tactic): add option for 'refreshing' universe metavariables in the 'apply' tactic 
The new test ../../tests/lean/run/tactic27.lean demonstrates why we need this feature. The tactic 'apply @refl' is actually 'apply @refl.{?l}'. It is used inside of a repeat tactical. Each iteration of the 'repeat' may need to use a different value for ?l. Before this commit, there was not way to say we want a fresh ?l each iteration.

Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-08 16:55:23 -07:00
Leonardo de Moura
4505016154 feat(frontends/lean): allow tactic_hints to be applied when class-instance mechanism fails 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-08 15:08:13 -07:00
Leonardo de Moura
5e836092cc feat(frontends/lean): allow user to suppress proofs in theorems, and let them be inferred automatically using tactic_hints 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-08 15:08:13 -07:00
Leonardo de Moura
a3be63af73 feat(frontends/lean): add tactic_hint command 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-08 15:08:13 -07:00
Leonardo de Moura
da4c1922e3 feat(frontends/lean): add '_root_' prefix for referencing names in the root namespace 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-07 19:15:46 -07:00
Leonardo de Moura
b43fb7448c feat(frontends/lean): search for identifiers in the stack of namespaces; reject non-atomic names as local names 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-07 19:00:06 -07:00
Leonardo de Moura
e6d4c01b88 feat(frontends/lean): check whether namespace exists or not in the 'using' command, add to_valid_namespace_name helper function 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-07 18:17:10 -07:00
Leonardo de Moura
bd1873f6b1 feat(frontends/lean): add coercion modifier 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-07 17:48:20 -07:00
Leonardo de Moura
08465f049a feat(frontends/lean): remove [class] annotation and 'class' command, they are redundant, we only need [instance] and 'instance' 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-07 16:09:30 -07:00
Leonardo de Moura
112353861c feat(frontends/lean): rename command 'reset_proof_qed' to 'set_proof_qed' 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-07 15:48:16 -07:00
Leonardo de Moura
c16951aba6 fix(library/aliases): aliasing behavior 
The new test '../../tests/lean/run/alias3.lean' demonstrates the issue being fixed.

Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-07 15:40:55 -07:00
Leonardo de Moura
b9d08ff28c feat(frontends/lean/builtin_cmds): allow many namespaces in the same 'using' command 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-07 14:53:06 -07:00
Leonardo de Moura
d30f387e72 feat(library/standard): add namespace 'pair' 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-07 14:48:19 -07:00
Leonardo de Moura
10b0dfeb37 feat(frontends/lean/class): allow many instances to provided with a single 'instance' command 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-07 14:41:14 -07:00
Leonardo de Moura
6ceecf6a15 feat(library/aliases): remove duplicates from aliasing tables 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-07 12:59:57 -07:00
Leonardo de Moura
b956ce68d2 feat(frontends/lean/elaborator): keep postponing delayed coercions until the type can be inferred 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-07 12:50:43 -07:00
Leonardo de Moura
831de22bcd fix(frontends/lean): bugs in notation management 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-07 09:31:42 -07:00
Leonardo de Moura
e8bd267a00 fet(frontends/lean): allow coercions to sort-class in the types of variable and definitions 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-06 21:54:16 -07:00
Leonardo de Moura
48b28ad75c fix(library/unifier): missing test in flex_rigid 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-06 21:36:23 -07:00
Leonardo de Moura
dcf7cf00ff fix(*): bugs in the type checker, inductive datatypes, and unifier 
The bugs were indentified when performing the tiny change in the file
tests/lean/run/group.lean
 2014-07-06 18:44:56 -07:00
Leonardo de Moura
9a13bef4f3 fix(frontends/lean): fix (and simplify) parameter universe inference 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-06 16:56:54 -07:00
Leonardo de Moura
8e5ac54ae1 test(tests/lean/run): group experiments 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-05 22:31:57 -07:00
Leonardo de Moura
55894f01e3 feat(frontends/lean): add 'opaque_hint' command 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-05 18:58:20 -07:00
Leonardo de Moura
e68a3e5251 test(tests/lean/run): add another class-instance example 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-05 17:41:08 -07:00
Leonardo de Moura
59755289e4 feat(library/unifier): case split on constraints of the form (f ...) =?= (f ...), where f can be unfolded, and there are metavariables in the arguments 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-05 15:52:40 -07:00
Leonardo de Moura
a52c9f4e2b feat(library/unifier): add option 'unifier.unfold_opaque', remove option 'unifier.use_exceptions' (the user should not be able to change this) 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-05 09:43:16 -07:00
Leonardo de Moura
efabd2280c feat(library/unifier): allow unifier to unfold opaque definitions of the current module 
It is not clear whether this is a good idea or not. In some cases, it seems to do more harm than good.

Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-05 00:46:03 -07:00
Leonardo de Moura
99fb6431a6 fix(frontends/lean/elaborator): support for local instances 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-04 18:38:08 -07:00
Leonardo de Moura
8ab0b5bee3 feat(frontends/lean/elaborator): use local declarations as class instances 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-04 18:18:25 -07:00
Leonardo de Moura
00e1a7db23 feat(frontends/lean/elaborator): add class instance elaboration 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-04 15:45:50 -07:00
Leonardo de Moura
b94ce412ae fix(library/unifier): non-termination 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-04 10:32:01 -07:00
Leonardo de Moura
a97f82be1a fix(library/standard): orelse notation, avoid conflict with inductive datatype declaration 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-04 10:10:05 -07:00
Leonardo de Moura
e0501104e2 feat(library/tactic): add 'fixpoint' tactic 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-04 01:30:28 -07:00
Leonardo de Moura
7fb2b0f6d8 feat(kernel): add method 'may_reduce_later' to normalizer_extension, and improve unifier 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-03 22:31:05 -07:00
Leonardo de Moura
855ffcba34 feat(library/standard): add pairs 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-03 20:43:16 -07:00
Leonardo de Moura
110b622b83 feat(library/unifier): add support for unification constraints of the form "(elim ... (?m ...)) =?= t", where elim is an eliminator 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-03 20:41:51 -07:00
Leonardo de Moura
b5f63e78ca feat(frontends/lean/notation_cmd): reuse existing precedence to increase compatibility with existing notation 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-03 17:23:29 -07:00
Leonardo de Moura
fa1857e6a9 fix(frontends/lean/notation_cmd): fix default, add 'prev' action 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-03 16:44:44 -07:00
Leonardo de Moura
abbd054b51 feat(library/tactic): add eassumption tactic, and remove redundant 'subgoals' from apply tactic 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-03 13:04:46 -07:00
Leonardo de Moura
d63ccbcf94 fix(library/unifier): missing case 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-03 10:51:59 -07:00
Leonardo de Moura
0ff145e59b feat(library/tactic): add apply tactic 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-03 09:20:01 -07:00
Leonardo de Moura
c1538bfc40 test(tests/lean/run): add 'apply subst' test 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-03 09:07:56 -07:00
Leonardo de Moura
a7d660f875 feat(frontends/lean): add command for customizing the behavior of proof-qed blocks: we can automatically register tactics to be automatically applied before each component 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-02 20:45:10 -07:00
Leonardo de Moura
5527955ba8 feat(frontends/lean): add 'proof-qed' notation 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-02 19:30:48 -07:00
Leonardo de Moura
60c637fb9d feat(library/tactic): add 'exact' tactic 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-02 18:37:32 -07:00
Leonardo de Moura
37b5b7c4c2 feat(library/tactic): rename 'exact' to 'assumption', 'exact' is a different tactic 
Signed-off-by: Leonardo de Moura <leonardo@microsoft.com>
 2014-07-02 18:10:42 -07:00