]>
matita.cs.unibo.it Git - helm.git/log
Andrea Asperti [Fri, 20 Oct 2006 15:44:48 +0000 (15:44 +0000)]
Demodulate and applyS moved form saturation to auto.
Andrea Asperti [Fri, 20 Oct 2006 15:43:02 +0000 (15:43 +0000)]
Major changes to auto, documented on the helm mailing list.
Andrea Asperti [Fri, 20 Oct 2006 15:39:46 +0000 (15:39 +0000)]
Minor changes.
Andrea Asperti [Fri, 20 Oct 2006 15:32:34 +0000 (15:32 +0000)]
a. uniform mangement for context and library
b. collapsing flexible terms (X args) into a single meta. Metas in
head position give arity problems with discrimination trees.
c. Even terms with a single meta as conlcusion (e.g. elimination
principles) get indexed.
Andrea Asperti [Fri, 20 Oct 2006 14:58:18 +0000 (14:58 +0000)]
The type of universe_of_goals has slightly changed, omitting
to pass a useless proof. A few prerr_endline have been removed.
Andrea Asperti [Fri, 20 Oct 2006 14:50:40 +0000 (14:50 +0000)]
This is only a temporary patch. The typecheker raises a
CicUniv.UniverseInconsistency that should be properly
captured and transformed into a tyepchecking error.
Andrea Asperti [Fri, 20 Oct 2006 14:37:25 +0000 (14:37 +0000)]
The function exists_a_meta has been modified to capture also
flexible arguments instead of metas.
Andrea Asperti [Fri, 20 Oct 2006 14:35:22 +0000 (14:35 +0000)]
New function pack_coercion_metasenv, used in auto after try_candidates
to clean up the metasenv (othewise application could fail).
Enrico Zoli [Fri, 20 Oct 2006 14:28:55 +0000 (14:28 +0000)]
Up to f_algebras.
Enrico Zoli [Fri, 20 Oct 2006 09:59:39 +0000 (09:59 +0000)]
1. developed up to algebras
2. some lemmas commented out because of a new bug in rewrite
Enrico Zoli [Fri, 20 Oct 2006 08:35:00 +0000 (08:35 +0000)]
Serious bug fixed: without a Lazy.force the user obtained an error comparing
two functional values.
Claudio Sacerdoti Coen [Thu, 19 Oct 2006 11:09:06 +0000 (11:09 +0000)]
Disambiguation errors are now compressed in a maybe better way.
Claudio Sacerdoti Coen [Thu, 19 Oct 2006 09:16:07 +0000 (09:16 +0000)]
Bug fixed: when trying to insert coercions after an Uncertain, a RefineFailure
was raised (instead of re-raising an Uncertain).
Claudio Sacerdoti Coen [Thu, 19 Oct 2006 08:50:31 +0000 (08:50 +0000)]
Potential performance improvement + better disambiguation error messages:
the delift functions used to raise Uncertain every time it failed; now it
raises Failure when the local context contains no Metas (that means that
for every future subst it will not contain new Rels).
Note: if we are working up to conversion (I do not think so), we should check
that the local context is Meta-closed, that is a weaker check.
Claudio Sacerdoti Coen [Wed, 18 Oct 2006 17:39:48 +0000 (17:39 +0000)]
Missing optimization implemented: before starting to analyze the disambiguation
domain we try first without interpreting any symbol. This way we can avoid
lot of refinements when the information in the uninterpreted term already
tells us that every instance will not be well typed.
Claudio Sacerdoti Coen [Wed, 18 Oct 2006 15:06:03 +0000 (15:06 +0000)]
- Disambiguation error exception enriched with more information
(the same returned in case of correct disambiguation)
- The disambiguation error interface now shows the complete environment
in which the error occurs. I am not sure if this is better or worst than
showing only the diffs.
Claudio Sacerdoti Coen [Wed, 18 Oct 2006 14:13:25 +0000 (14:13 +0000)]
Bug fixed: the diff component of the exception raised when the term cannot
be disambiguated used to lack the last choice in case of the look-ahead
optimization.
Claudio Sacerdoti Coen [Wed, 18 Oct 2006 13:18:37 +0000 (13:18 +0000)]
EXPERIMENTAL: new interface for disambiguation errors.
[ Moreover, a dead dialog window has been removed ]
Claudio Sacerdoti Coen [Wed, 18 Oct 2006 13:17:10 +0000 (13:17 +0000)]
Dead dialog window removed.
Claudio Sacerdoti Coen [Wed, 18 Oct 2006 13:16:06 +0000 (13:16 +0000)]
Dead dialog window removed.
Claudio Sacerdoti Coen [Wed, 18 Oct 2006 12:11:48 +0000 (12:11 +0000)]
Disambiguation errors now carry more information (i.e. the patches to apply to
the current interpretation to re-obtain the refinement errors).
Claudio Sacerdoti Coen [Wed, 18 Oct 2006 12:10:00 +0000 (12:10 +0000)]
Dama is now in the night benchmarks.
Claudio Sacerdoti Coen [Wed, 18 Oct 2006 09:42:40 +0000 (09:42 +0000)]
Too verbose error message (probably activated by Enrico without really wanting
it) removed.
Ferruccio Guidi [Tue, 17 Oct 2006 20:45:12 +0000 (20:45 +0000)]
new objects for the LambdaDelta development (4th conjecture proved)
Ferruccio Guidi [Tue, 17 Oct 2006 10:52:58 +0000 (10:52 +0000)]
more new objects for the LambdaDelta contribution
Enrico Zoli [Mon, 16 Oct 2006 16:49:51 +0000 (16:49 +0000)]
Beginning of the development of integration algebras.
Claudio Sacerdoti Coen [Fri, 13 Oct 2006 16:59:04 +0000 (16:59 +0000)]
Content level representation of LetRec changed.
Claudio Sacerdoti Coen [Fri, 13 Oct 2006 16:57:10 +0000 (16:57 +0000)]
Content level representation of LetRec changed.
Claudio Sacerdoti Coen [Fri, 13 Oct 2006 16:53:54 +0000 (16:53 +0000)]
New content level representations for LetRec, Inductive and CoInductive.
For LetRec the pretty-printer is now in sync with the parser.
Ferruccio Guidi [Thu, 12 Oct 2006 21:14:23 +0000 (21:14 +0000)]
files with newest objects (to be included in the respective developments)
Enrico Tassi [Thu, 12 Oct 2006 16:10:53 +0000 (16:10 +0000)]
carr_of_term now returns Fun if a Prod is encountered
Enrico Tassi [Thu, 12 Oct 2006 16:10:23 +0000 (16:10 +0000)]
fixed defaultauto behaviour. not the cache is preserveed
Enrico Tassi [Thu, 12 Oct 2006 13:34:32 +0000 (13:34 +0000)]
timeout if unspecfied should be set to infinity, not 0, since the timeout inside the flagfs structure is a time in the future, not a decreasing quantity.
every loop of auto the check gettimeofday() > timeout raise Fail
Claudio Sacerdoti Coen [Thu, 12 Oct 2006 10:37:17 +0000 (10:37 +0000)]
The default for paramodulation is now back to false (I set it to true
by mistake).
acciavat [Thu, 12 Oct 2006 09:53:08 +0000 (09:53 +0000)]
auto => auto new.
There is still one auto left that does not work with the new implementation
(because of implicative hypothesis for local lemmas?)
Claudio Sacerdoti Coen [Thu, 12 Oct 2006 09:46:37 +0000 (09:46 +0000)]
Inclusion "improved".
Claudio Sacerdoti Coen [Thu, 12 Oct 2006 09:44:03 +0000 (09:44 +0000)]
CoRN integrated in the night benchmarks.
acciavat [Thu, 12 Oct 2006 09:39:45 +0000 (09:39 +0000)]
Manual porting of CoRN to Matita.
The current version shows a bug in coercions to function classes.
Claudio Sacerdoti Coen [Thu, 12 Oct 2006 08:06:48 +0000 (08:06 +0000)]
Bug fixed: the conversion was done with the wront argument (a sort of typo).
Claudio Sacerdoti Coen [Wed, 11 Oct 2006 14:58:23 +0000 (14:58 +0000)]
Some hocus-pocus to avoid a common race condition (Gtk/??? code not
reentrant). I bet the "avoided" problem will surface again...
Claudio Sacerdoti Coen [Wed, 11 Oct 2006 14:26:59 +0000 (14:26 +0000)]
Unlocking the interface was not performed as the last action of the callback.
Claudio Sacerdoti Coen [Wed, 11 Oct 2006 12:25:27 +0000 (12:25 +0000)]
The RELATIONAL contrib must be compiled before the others since the others
depend on it.
Claudio Sacerdoti Coen [Wed, 11 Oct 2006 09:49:27 +0000 (09:49 +0000)]
My previous commit changed the regular timeout of paramodulation from
infinity to 30s. Old behaviour restored.
Ferruccio Guidi [Tue, 10 Oct 2006 19:51:39 +0000 (19:51 +0000)]
makefiles fixups
Claudio Sacerdoti Coen [Tue, 10 Oct 2006 14:06:44 +0000 (14:06 +0000)]
fguidi removed from RT in makefiles
Claudio Sacerdoti Coen [Tue, 10 Oct 2006 14:03:40 +0000 (14:03 +0000)]
Implemented:
"by ... we proved ... that is equivalent to ... done"
and
"we need to prove ... that is equivalent to ..."
Claudio Sacerdoti Coen [Tue, 10 Oct 2006 13:12:44 +0000 (13:12 +0000)]
auto => auto new
Claudio Sacerdoti Coen [Tue, 10 Oct 2006 13:10:08 +0000 (13:10 +0000)]
auto => auto new
added depth=4 to some file in order to make it pass
Claudio Sacerdoti Coen [Tue, 10 Oct 2006 13:09:35 +0000 (13:09 +0000)]
I do not understand at all why Enrico removed the contribs from the bench!
Claudio Sacerdoti Coen [Tue, 10 Oct 2006 08:00:32 +0000 (08:00 +0000)]
Sorry, bug introduced by me yesterday now fixed.
auto should be the new one even when the paramodulation flag is given.
Claudio Sacerdoti Coen [Mon, 9 Oct 2006 18:26:05 +0000 (18:26 +0000)]
Bugs fixed in merging of composite coercions. In particular the imperative
status was not handled properly and letins were handled inefficiently.
Claudio Sacerdoti Coen [Mon, 9 Oct 2006 18:23:49 +0000 (18:23 +0000)]
One auto modified in an apply since auto is no longer supposed to apply
elimination principles automatically.
Claudio Sacerdoti Coen [Mon, 9 Oct 2006 18:18:24 +0000 (18:18 +0000)]
added to applyS in nat/gcd.ma a timeout large enough for compilation in bytecode
mode.
Claudio Sacerdoti Coen [Mon, 9 Oct 2006 18:17:50 +0000 (18:17 +0000)]
1. applyS now uses its ~params
2. autoTactic.ml* is going to be deprecated.
auto.ml* has now the expected interface (that of a tactic)
and it is almost ready to replace autoTactic.ml
Claudio Sacerdoti Coen [Mon, 9 Oct 2006 17:48:45 +0000 (17:48 +0000)]
applyS now receives the same parameters that auto receives.
The parameters are not used yet.
Andrea Asperti [Mon, 9 Oct 2006 16:33:24 +0000 (16:33 +0000)]
Theorems from the library and from the context are indexed also in "unfolded"
version.
Claudio Sacerdoti Coen [Mon, 9 Oct 2006 16:28:22 +0000 (16:28 +0000)]
auto => auto new and other minor changes to make it compile.
Claudio Sacerdoti Coen [Mon, 9 Oct 2006 15:33:47 +0000 (15:33 +0000)]
auto => auto new everywhere + minor updates to make more tests pass
Claudio Sacerdoti Coen [Mon, 9 Oct 2006 14:04:31 +0000 (14:04 +0000)]
Comments updated with new reflections.
Andrea Asperti [Mon, 9 Oct 2006 11:00:31 +0000 (11:00 +0000)]
The two coercions sym_eq e eq_f gives BIG TROUBLES when checking composition during
qed. In particular, checking composition calls unification, that calls reductions,
that is problematic when there are a lot of lein in the context (as is it the case
with proofs generated by TPTP: see eg GRP486-1.p.ma )
Claudio Sacerdoti Coen [Mon, 9 Oct 2006 09:23:09 +0000 (09:23 +0000)]
More work to handle -debug properly.
Andrea Asperti [Mon, 9 Oct 2006 06:53:04 +0000 (06:53 +0000)]
Factorized "find_equalities" in demodulation_tac.
Enrico Tassi [Fri, 6 Oct 2006 16:49:16 +0000 (16:49 +0000)]
added support for short name targets
Enrico Tassi [Fri, 6 Oct 2006 16:41:53 +0000 (16:41 +0000)]
resumed ol auto
Enrico Tassi [Fri, 6 Oct 2006 15:56:52 +0000 (15:56 +0000)]
fixed all (that now uses long paths)
Enrico Tassi [Fri, 6 Oct 2006 15:30:37 +0000 (15:30 +0000)]
now the makefile for developments requires the depend file
that now has correct dependencies.
Claudio Sacerdoti Coen [Fri, 6 Oct 2006 07:20:13 +0000 (07:20 +0000)]
1. some "try ... with _ " removed
2. type inference of LetIn terms is now closer than the kernel
3. LetIn code fixed to add the type to the Defs in the environment
4. some typos fixed
Enrico Tassi [Tue, 3 Oct 2006 16:40:35 +0000 (16:40 +0000)]
reduced timeout to 100s
Enrico Tassi [Tue, 3 Oct 2006 16:08:31 +0000 (16:08 +0000)]
commented out are_convertible in is_identity
fixed metadataquery so that only are calculated without the DB (but are calculated)
removed is_meta_convertible left right in is_weak_identity
added extra dependencies between gcd < relevant_equations < ord to make
paramodulation happy. In the past this used to be the order used by make.
added an hack in applys.ma, since it used to work due to convertibility.
maiorino [Tue, 3 Oct 2006 16:05:50 +0000 (16:05 +0000)]
Some declarative tactics did not allow the "done" option in place
of "(id)". Fixed.
Claudio Sacerdoti Coen [Tue, 3 Oct 2006 15:59:34 +0000 (15:59 +0000)]
Inline command implemented.
Claudio Sacerdoti Coen [Tue, 3 Oct 2006 15:58:50 +0000 (15:58 +0000)]
Syntax of paramodulation parameters changed.
Claudio Sacerdoti Coen [Tue, 3 Oct 2006 14:18:34 +0000 (14:18 +0000)]
Fixed handling of exceptions by the worker threads: the exception is not
notified to the user and the thread exits gracefully.
Claudio Sacerdoti Coen [Tue, 3 Oct 2006 13:57:30 +0000 (13:57 +0000)]
Removed /home/tassi from the makefile!
Enrico Tassi [Tue, 3 Oct 2006 13:00:43 +0000 (13:00 +0000)]
Query fixed to handle the cases where 0 tests are failures or successfull.
Andrea Asperti [Tue, 3 Oct 2006 12:56:38 +0000 (12:56 +0000)]
Changed auto from implicit to option and renamed a few functions.
Claudio Sacerdoti Coen [Tue, 3 Oct 2006 12:15:00 +0000 (12:15 +0000)]
Regular expression fixed to allow '-' into test names.
Claudio Sacerdoti Coen [Tue, 3 Oct 2006 12:02:55 +0000 (12:02 +0000)]
sacerdot is now the user for the daily bench.
Claudio Sacerdoti Coen [Tue, 3 Oct 2006 11:59:30 +0000 (11:59 +0000)]
Crontab updated
Claudio Sacerdoti Coen [Tue, 3 Oct 2006 11:49:19 +0000 (11:49 +0000)]
No more benches without gc.
Claudio Sacerdoti Coen [Tue, 3 Oct 2006 11:33:14 +0000 (11:33 +0000)]
sequent_viewer.xml & co are now generated in /tmp if necessary!
Ferruccio Guidi [Tue, 3 Oct 2006 11:18:21 +0000 (11:18 +0000)]
updated to use destruct instead of disciminate/injection
Claudio Sacerdoti Coen [Tue, 3 Oct 2006 08:31:25 +0000 (08:31 +0000)]
Bug fixed: the script Menu was not locked as expected.
Claudio Sacerdoti Coen [Tue, 3 Oct 2006 07:52:16 +0000 (07:52 +0000)]
The Abort button is now working properly. Cool.
Claudio Sacerdoti Coen [Mon, 2 Oct 2006 17:55:36 +0000 (17:55 +0000)]
Xavier code for killing a thread commented out for a while since it interacts badly with (labl?)Gtk.
Claudio Sacerdoti Coen [Mon, 2 Oct 2006 17:44:01 +0000 (17:44 +0000)]
This commit implements the Abort button for the GUI using a clever trick by Xavier Leroy.
Enrico Tassi [Mon, 2 Oct 2006 16:30:28 +0000 (16:30 +0000)]
restored is_identity instead of is_weak_identity
Claudio Sacerdoti Coen [Mon, 2 Oct 2006 16:22:13 +0000 (16:22 +0000)]
**** EXPERIMENTAL ****
The GUI of Matita is now thread-based.
In principle Ctr-C works unreliably.
In practice it works reasonably well.
Enrico Tassi [Mon, 2 Oct 2006 15:33:24 +0000 (15:33 +0000)]
added a subtle List.rev that makes the order of the equalities taken from the library as it used to be. In this way the applys test works, but is not clear why the other order does not produce a solution in decent time
Enrico Tassi [Mon, 2 Oct 2006 15:30:44 +0000 (15:30 +0000)]
...
Enrico Tassi [Mon, 2 Oct 2006 15:19:25 +0000 (15:19 +0000)]
auto always uses the context (even if paramodulation) to try to close goals left open by paramod
Enrico Tassi [Mon, 2 Oct 2006 15:18:35 +0000 (15:18 +0000)]
returns the right list of goals
Enrico Tassi [Mon, 2 Oct 2006 15:17:01 +0000 (15:17 +0000)]
50 steps on goal are fine for irrat2
Enrico Tassi [Mon, 2 Oct 2006 15:16:02 +0000 (15:16 +0000)]
removed only made with the DB
Enrico Tassi [Mon, 2 Oct 2006 15:15:31 +0000 (15:15 +0000)]
removed a pointless call to auto
Enrico Tassi [Mon, 2 Oct 2006 15:15:03 +0000 (15:15 +0000)]
...
Enrico Tassi [Mon, 2 Oct 2006 15:14:46 +0000 (15:14 +0000)]
added tests for paramod
Enrico Tassi [Mon, 2 Oct 2006 13:56:53 +0000 (13:56 +0000)]
...
Enrico Tassi [Mon, 2 Oct 2006 11:42:43 +0000 (11:42 +0000)]
added missing *)
Enrico Tassi [Mon, 2 Oct 2006 11:29:49 +0000 (11:29 +0000)]
restored old (r6662) behaviour