Commit graph

26741 commits

Author SHA1 Message Date
Simon Hudon
92c8773137 feat: file IO using handles 2020-01-12 08:02:48 -08:00
Leonardo de Moura
354678a5e4 chore: missing files 2020-01-11 19:17:21 -08:00
Leonardo de Moura
b508f1854d chore: update stage0 2020-01-11 15:35:11 -08:00
Leonardo de Moura
bfe8c6d060 feat: prevent adversarial users from using hugeFuel in actual code 2020-01-11 15:34:39 -08:00
Leonardo de Moura
c358d688c6 chore: fix tests 2020-01-11 15:20:37 -08:00
Leonardo de Moura
6d77aa20aa feat: catch deep recursion at MetaM, TermElabM and CommandElabM
cc @Kha
2020-01-11 15:03:58 -08:00
Leonardo de Moura
f83678f19b chore: fix test 2020-01-11 13:55:31 -08:00
Leonardo de Moura
6762f8ea9a test: old frontend exports at new frontend
We can now "see" exports made in the old frontend in the new one.

cc @dselsam @kha
2020-01-11 13:50:18 -08:00
Leonardo de Moura
3e233b6f3d chore: update stage0 2020-01-11 13:46:10 -08:00
Leonardo de Moura
5eebda7e34 chore: add workaround for allowing new frontend to see old frontend exports 2020-01-11 13:44:22 -08:00
Leonardo de Moura
246e0a5532 fix: propagate type before synthesizeSyntheticMVars 2020-01-11 12:01:12 -08:00
Leonardo de Moura
c729973742 fix: restore state when catching postpone
Reason: avoid error messages to be generated multiple times; avoid
internal postponed mvars to leak; etc
2020-01-11 11:57:39 -08:00
Leonardo de Moura
2735b487f3 chore: update stage0 2020-01-11 09:05:05 -08:00
Leonardo de Moura
82a36fbfe2 feat: declare_syntax_cat without importing Init.Lean
cc @Kha
2020-01-11 09:02:50 -08:00
Leonardo de Moura
95ed5bd468 chore: fix test 2020-01-10 21:26:09 -08:00
Leonardo de Moura
e817257922 feat: elaborate declare_syntax_cat
TODO: `registerParserCategory` uses `registerAttribute` which relies
on the environment having a declaration of type `AttributeImpl`.
This is bad since forces users to import `Init.Lean`.

@Kha The key problem is that we cannot serialize `AttributeImpl`.
I will try to address this issue tomorrow. I am considering different
workarounds.
2020-01-10 21:10:02 -08:00
Leonardo de Moura
142991ed4d chore: update stage0 2020-01-10 20:38:23 -08:00
Leonardo de Moura
80e00a87c6 feat: add declare_syntax_cat command parser 2020-01-10 20:38:08 -08:00
Leonardo de Moura
799914daf2 feat: registerParserCategory 2020-01-10 20:32:16 -08:00
Leonardo de Moura
3c8d8c7434 feat: add attributeExtension 2020-01-10 19:51:53 -08:00
Leonardo de Moura
48600dbbfc refactor: registerAttribute ==> registerBuiltinAttribute 2020-01-10 17:08:12 -08:00
Leonardo de Moura
6fc158f550 chore: update stage0 2020-01-10 15:22:16 -08:00
Leonardo de Moura
f783115d21 refactor: SyntaxNode => Syntax at TermElab and CommandElab
@Kha It is too annoying to write `.val` all over the place. Moreover,
we now have `match_syntax`.
2020-01-10 15:17:58 -08:00
Leonardo de Moura
f73ff914eb feat: extensible elaboration functions
@kha `termParserAttr.lean` has a small example
2020-01-10 15:05:14 -08:00
Leonardo de Moura
f60c8c6249 test: add kabstract tests 2020-01-10 13:30:50 -08:00
Leonardo de Moura
63cd3a8c41 fix: nontermination at isDefEqOffset 2020-01-10 13:25:20 -08:00
Leonardo de Moura
06b6beb9f7 feat: add kabstract 2020-01-10 13:10:03 -08:00
Leonardo de Moura
3077d13b56 feat: add HeadIndex 2020-01-10 11:58:22 -08:00
Leonardo de Moura
947438afbe feat: add Occurrences 2020-01-10 11:06:36 -08:00
Leonardo de Moura
a76b104f61 chore: move Message to Lean 2020-01-10 10:58:50 -08:00
Leonardo de Moura
f8c7d6ca6b feat: elaborate optParam 2020-01-10 09:32:01 -08:00
Leonardo de Moura
d7b23cf297 fix: typo 2020-01-10 08:36:48 -08:00
Daniel Selsam
27158eb3e2 fix: DiscrTree typos 2020-01-10 08:07:41 -08:00
Leonardo de Moura
c5111de0f0 feat: add option set_option synthInstance.maxSteps <num>
cc @dselsam @kha
2020-01-10 08:04:25 -08:00
Leonardo de Moura
733ea89000 feat: improve application type mismatch error message 2020-01-09 16:41:27 -08:00
Leonardo de Moura
693d7dcf62 test: add class projection test 2020-01-09 16:25:12 -08:00
Leonardo de Moura
a4c2b0303e chore: update stage0 2020-01-09 16:12:42 -08:00
Leonardo de Moura
e2ad834a2c fix: weird bug that only occurs in debug mode 2020-01-09 16:11:33 -08:00
Leonardo de Moura
c45218be1a chore: update-stage0 2020-01-09 15:43:46 -08:00
Leonardo de Moura
56d6961529 fix: old constant names 2020-01-09 15:40:55 -08:00
Leonardo de Moura
de24a3b205 chore: update stage0 2020-01-09 15:40:18 -08:00
Leonardo de Moura
bba9cdd8ff feat: improve support for nat literals 2020-01-09 15:36:46 -08:00
Leonardo de Moura
654b464747 chore: improve old pretty printer on numeric literals 2020-01-09 13:48:15 -08:00
Leonardo de Moura
3d1ee3f408 fix: add isNewAnswer predicate
reactivate `typeclass_coerce.lean` test
2020-01-09 13:37:21 -08:00
Leonardo de Moura
125faad936 fix: missing do
This is nasty. The following
```
if entry.answers.contains answer then pure () -- if answer was already found, then do nothing
else
  let newEntry := { answers := entry.answers.push answer, .. entry };
  modify $ fun s => { tableEntries := s.tableEntries.insert key newEntry, .. s };
  entry.waiters.forM (wakeUp answer)
```
was being parsed as
```
if entry.answers.contains answer then pure () -- if answer was already found, then do nothing
else
  (let newEntry := { answers := entry.answers.push answer, .. entry };
   modify $ fun s => { tableEntries := s.tableEntries.insert key newEntry, .. s });
entry.waiters.forM (wakeUp answer)
```
2020-01-09 12:44:41 -08:00
Leonardo de Moura
c49e43f69f chore: reactivate typeclass test 2020-01-09 11:47:48 -08:00
Leonardo de Moura
bc393a750d chore: reactivate typeclass test
I had to split the test because `new_frontend` command consumes the
rest of the file.
2020-01-09 11:46:10 -08:00
Leonardo de Moura
178e411031 chore: reactivate typeclass_append 2020-01-09 11:40:34 -08:00
Leonardo de Moura
f3d4005fc6 fix: outParam support 2020-01-09 11:39:12 -08:00
Leonardo de Moura
78c50125dd fix: bug preprocessArgs 2020-01-09 11:13:25 -08:00