Commit graph

730 commits

Author SHA1 Message Date
Leonardo de Moura
81618e30f3 fix(tests/lean/run): adjust some tests to changes in the standard library 2015-11-08 14:04:56 -08:00
Leonardo de Moura
28a5ca5809 fix(frontends/lean): fixes #830 2015-09-18 07:51:02 -07:00
Leonardo de Moura
68dc39c106 fix(tests/lean/run/rewriter12): broken test, now ^[ is a token 2015-09-16 08:37:43 -07:00
Leonardo de Moura
df3100d2cd fix(library/local_context): bug in abstract_locals procedure 2015-09-12 17:17:13 -07:00
Leonardo de Moura
1fdbd681cc feat(frontends/lean/builtin_exprs): name hypothesis in suffices
closes #817
2015-09-03 16:09:39 -07:00
Leonardo de Moura
c84e886c7b fix(frontends/lean/notation_cmd): fixes #808
This commit and 2b1d2c fixes #808
2015-08-31 18:05:58 -10:00
Leonardo de Moura
08169c5ac2 fix(library/unifier): fixes #809
Daniel is correct when he says the interaction between choice
case-splits, delta case-splits, and coercions can be subtle.

I believe the following condition
https://github.com/leanprover/lean/blob/master/src/frontends/lean/elaborator.cpp#L111
reduces counter-intuitive behavior. Example, the coercion should not
influence the resulting type.
BTW, by removing this condition, many files in the library broke when I
tried to compile from scratch

      make clean-olean
      make

I used the following workaround. Given a delta-delta constraint

           f a =?= f b

If the terms are types, and no case-split will be performed, then
the delta-delta constraint is eagerly solved.
In principle, we don't need the condition that the terms are types.
However, many files break if we remove it. The problem is that many files in the standard
library are abusing the higher-order unification procedure. The
elaboration problems are quite tricky to solve.
I use the extra condition "the terms are types" because usually if they
are, "f" is morally injective, and we don't really want to unfold it.

Note that the following two cases do not work

     check '{1, 2, 3}
     check insert 1 (insert 2 (insert 3 empty))

Well, they work if we the num namespace is open, and they are
interpreted as having type (finset num)
2015-08-31 17:59:30 -10:00
Leonardo de Moura
2b1d2c21ad fix(frontends/lean/util): bug when parsing priorities and numerals are overloaded 2015-08-31 15:08:21 -10:00
Leonardo de Moura
a3c404ac3b feat(library/tactic/apply_tactic): do not report elaboration failure in apply tactic when proof_state.report_failure() is false 2015-08-21 15:45:52 -07:00
Leonardo de Moura
87349dc355 feat(frontends/lean/token_table): add 'proposition' keyword 2015-08-19 08:05:31 -07:00
Leonardo de Moura
3a72cd9621 fix(frontends/lean): rename multiword keyword "suffices to show" to "suffices" 2015-08-18 17:57:53 -07:00
Leonardo de Moura
3ce8c5d6f7 feat(frontends/lean): add "suffices to show A, from B, C" construct 2015-08-18 17:04:38 -07:00
Leonardo de Moura
21c41f50ea fix(frontends/lean/elaborator): fixes #803 2015-08-17 14:56:41 -07:00
Leonardo de Moura
ea04414058 feat(frontends/lean): allow user to overload notation containing foldr/foldl and/or scoped expressions
see new tests for a summary of new features

see issue #800
2015-08-16 18:24:30 -07:00
Leonardo de Moura
eb8f586dba fix(library/normalize): fixes #801 2015-08-16 14:22:02 -07:00
Leonardo de Moura
5a6a4b45c1 fix(library/definitional/equations): fixes #796 2015-08-14 14:39:23 -07:00
Leonardo de Moura
d2eb99bf11 refactor(library/logic): move logic/choice.lean to init/classical.lean
choice axiom is now in the classical namespace.
2015-08-12 18:37:33 -07:00
Leonardo de Moura
b4024982a2 refactor(library/data): move vector as indexed family to examples folder 2015-08-12 15:05:14 -07:00
Leonardo de Moura
66a59d5b51 feat(frontends/lean/util): remove hack that overrides priority namespace
closes #789
2015-08-11 18:01:40 -07:00
Leonardo de Moura
0b8f57841a feat(frontends/lean/decl_cmds): closes #791 2015-08-11 17:53:33 -07:00
Jeremy Avigad
305f72bf4f fix(tests/lean): fix three tests broken by setext renaming 2015-08-09 22:14:25 -04:00
Leonardo de Moura
6c5832a564 feat(frontends/lean/decl_cmds): allow recursive examples
closes #774
2015-08-08 08:26:25 -07:00
Leonardo de Moura
f21647899f feat(frontends/lean/builtin_exprs): rename 'show' hidden name to 'this'
This is useful if 'show' is recursive
2015-08-07 13:29:21 -07:00
Leonardo de Moura
308af87b69 feat(library): add 'noncomputable' keyword for the standard library 2015-07-28 21:56:35 -07:00
Leonardo de Moura
80e3da0526 fix(library/util): fixes #751 2015-07-28 16:30:20 -07:00
Leonardo de Moura
844caf32e4 feat(frontends/lean/bultin_cmds): add 'print [congr]' command for displaying active congruence rules 2015-07-23 18:52:59 -07:00
Leonardo de Moura
933f056fff feat(library/simplifier): add API for extracting simplification rules defined in a given namespace 2015-07-22 18:47:56 -07:00
Leonardo de Moura
092c8d05b9 feat(frontends/lean,library): rename '[rewrite]' to '[simp]' 2015-07-22 09:01:42 -07:00
Leonardo de Moura
23dd47d27f fix(tests/lean): adjust tests to recent changes to the standard library 2015-07-19 21:32:42 -07:00
Leonardo de Moura
5112232d6d fix(tests/lean/run/finset): adjust test to recent changes to the
standard library
2015-07-19 11:53:21 -07:00
Leonardo de Moura
3ab0e07ba9 feat(frontends/lean): add simp tactic frontend stub
This commit also removes the fake_simplifier. It doesn't work anymore
because simp is now a reserved word.
2015-07-14 09:54:53 -04:00
Leonardo de Moura
f8d472c9f1 feat(frontends/lean/parse_rewrite_tactic): change the semantics of rewrite[↑f] when f is recursive
After this commit it behaves like 'unfold f'.
That is, it will unfold f even if it fails to fold recursive
applications. Now, only 'esimp[f]' will not unfold f-applications when
it cannot fold the recursive applications.

This commit also closes #692. It is part of a series of commits that
addresses this issue.

closes #692
2015-07-12 13:20:21 -04:00
Leonardo de Moura
4c0a656ecc fix(library/tactic/unfold_rec): support indexed families + brec_on at unfold_rec
see issue #692
2015-07-12 12:45:05 -04:00
Leonardo de Moura
8e8e08cfe7 feat(library/tactic): use occurrence object in unfold tactic family 2015-07-11 18:53:45 -04:00
Leonardo de Moura
554a42b407 fix(library/tactic/unfold_rec): add annother brec pattern that should be checked in the unfold recursive definition tactic 2015-07-10 22:16:23 -04:00
Leonardo de Moura
a9515ac7a4 feat(library/tactic/rewrite_tactic): try to fold nested recursive applications after unfolding a recursive function
See issue #692.
The implementation still has some rough spots.
It is not clear what the right semantic is.
Moreover, the folds in e_closure could not be eliminated automatically.
2015-07-08 21:19:18 -04:00
Leonardo de Moura
4b1b3e277f feat(frontends/lean): rename '[unfold-c]' to '[unfold]' and '[unfold-f]' to '[unfold-full]'
see issue #693
2015-07-07 16:37:06 -07:00
Leonardo de Moura
b0c56273e2 fix(frontends/lean/elaborator): fixes #724 2015-07-06 15:19:19 -07:00
Leonardo de Moura
7e0844a9e6 fix(tests): to reflect recent changes in the standard library 2015-07-06 15:05:01 -07:00
Leonardo de Moura
aa338f6002 test(tests/lean/run): add test showing new coercion module addresses issue #668 2015-07-01 16:41:19 -07:00
Leonardo de Moura
cabe30ba71 feat(frontends/lean): allow user to assign priorities to notation declarations 2015-06-30 17:10:27 -07:00
Leonardo de Moura
0fc2efe88e fix(library/tactic/rewrite_tactic): fixes #702 2015-06-28 20:37:17 -07:00
Leonardo de Moura
95720b1670 fix(frontends/lean/elaborator): fixes #687 2015-06-28 19:58:57 -07:00
Leonardo de Moura
ecfc01b2d0 test(tests/lean/run): add test for <d notation
see issue #695

remark: we have to fix the tutorial the ASCII notation for fold is `<d`
instead of `<D`
2015-06-28 13:10:15 -07:00
Leonardo de Moura
d1eaa7bcda feat(frontends/lean/parse_rewrite_tactic): accept trailing comman in rewrite tactic
see issue #695
2015-06-28 11:45:30 -07:00
Leonardo de Moura
2aa64034df fix(tests/lean): adjust tests to reflect changes in the elaboration process 2015-06-26 17:18:30 -07:00
Floris van Doorn
9ef1ae0848 fix(tests): update tests to reflect the change of notation from \~ to ~ 2015-06-25 22:55:05 -04:00
Leonardo de Moura
1b414d36e7 refactor(library/init): define prod as an inductive datatype
Motivation: prod is used internally in the definitional package.
If we define prod as a structure, then Lean will tag pr1 and pr2 as
projections. This creates problems when we add special support for
projections in the elaborator. The heuristics avoid some case-splits
that are currently performed, and without them some files break.
2015-06-25 17:59:06 -07:00
Floris van Doorn
fa1979c128 feat(datatypes): let the type of unit be the lowest non-Prop universe
The definitional package (brec_on and cases_on) now use poly_unit instead of unit

closes #698
2015-06-25 17:33:46 -07:00
Leonardo de Moura
a4c0699e81 feat(library/tactic/constructor_tactic): restore 'constructor' tactic old semantics, add 'fconstructor' tactic
See issue #676

Add new test demonstrating why it is useful to have the old semantics
for 'constructor'
2015-06-17 23:48:54 -07:00