| Age | Commit message (Collapse) | Author |
|---|
|
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@15371 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
Same for check_leq instead of check_geq
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@15081 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
- use evar candidates instead of postponed conversion problems when
it is known (according to the projection heuristic used) that an
evar has only a fixed number of possible instances (as e.g. in
equation ?n[x,x] = x, with x a variable);
- this allows to be more robust in solving remaining problems: if
several instanciations exist, and one is not compatible with a previous
instantiation made among several choices for another evar,
backtracking is now possible;
- this allows in particular to fix regression #2670 (two postponed
conversion problems solved in an inconsistent way);
- but this requires more code.
At the same time, a refactoring of the code has been made so as to
hopefully clarify the elementary pieces of the algorithm. For
instance, there are now generic functions for both applying a filter
and giving candidates. The filter is systematically generalized so as
to have the ccl of the evar well-typed even in situations where we
could try on the contrary to restrict the evars occurring in the
ccl.
Anyway, when the representation of instances will be optimized using
identity substitutions, it will no longer be useful to use the filter
to shorten the size of the instances. Then, the filters will have,
like candidates, the only role of restricting the search space for
potential solutions to the unification problems.
Also, solve_refl can now be used to restrict equations ?x[t1..tn]=?x[u1..un]
up to conversion instead of up to unification. This (potententially)
looses constraints but this avoids looping at the time of considering
remaining problems and applying heuristics to them.
Also added printing of evar candidates in debugging evar printers
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@15061 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
Util only depends on Ocaml stdlib and Utf8 tables.
Generic pretty printing and loc functions are in Pp.
Generic errors are in Errors.
+ Training white-spaces, useless open, prlist copies random erasure.
Too many "open Errors" on the contrary.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@15020 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
instantiating them in the unification algorithm used for tactics.
This allows to discard ill-typed uses of first-order unification which
otherwise would have been fatal (this incidentally allows to partially
restore some compatibility with 8.3 that was broken after eta was
added in unification).
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@14812 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
known in advance to be instantiable by only a finite number of terms.
When an evar with candidates remain unsolved after unification, the
first candidate is taken as a heuristic.
This is used in particular to reduce the number of pending conversion
problems when trying to infer the return clause of a pattern-matching
problem. As an example, this repairs test 2615.v.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@14797 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
Here Foo is Univ.constraints, Univ.universes, Evd.evar_map, Evd.Metamap
Ok, all these structures are currently ocaml's maps or similar,
with a unique empty value, and (=) can be used on them in this
particular case. But using Foo.is_empty is safer : it will work
even if the underlying representation changes. Example : for
spotting non-legitimate use of (=) we might embed a type into
a record with a functional field.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@14614 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@14595 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
global ref)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@14578 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@14546 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
(replace_vars was anyway optimized)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@14536 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@14535 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
evar_define so that it can recursively deal with evar/evar problems.
Also, check_evar_instance now called after each instantiation.
Also did a bit of file reformatting.
The commit apparently induces a loss of some 0,4% on the compilation
of the standard library. Maybe, introducing a heuristic to decide when
to call check_evar_instance (which I guess is responsible for the
overhead) might be a good thing to look at?
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@14534 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@14380 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
particular, new printer for evar_map which displays undefined evars +
defined evars that were instantiated by these undefined evars and
recursively, up to some arbitrary level n chosen to be in practice
n=2 (thanks to Arnaud).
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@14231 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
flag to forbid rewriting tactics to instantiate an evar of the goal
while looking for subterms (this is not clear that we always want that
for rewrite but we certainly want it for autorewrite; see comments
by Charguéraud on coqdev Oct 2010).
In a few cases in the theories, a pre-existing evar of an hyp used for
rewriting is instantiated by the rewriting step. Let's accept this at
the current time.
We have to make progress towards documenting and stabilizing the
strategy for matching/unifying subterms in rewrite/induction/set...
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@14190 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@14121 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
"apply" unification.
Assuming w_unify_0 is not eventually abandoned, it remains to merge
unify_with_eta into unify_0 (what unify_with_eta does and that unify_0
does not do is to select of two instances of the same meta the one
with less lambda's; it is unclear whether this is useful heuristic).
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@14091 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
- Better printing of unif constraints in evar_map's
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@13925 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
evar, and simultaneously make type inference with universes work better.
This only exports more functions from kernel/univ, to be able to work
with a set of universe variables during type inference. Universe
constraints are gradually added during type checking, adding information
necessary e.g. to lower the level of unknown Type variables to Prop or
Set. There does not seem to be a disastrous performance hit on the
stdlib, but might have one on some contribs (hence the "Tentative").
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@13905 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
- Normalize evars in typeclasses eauto also before [intro].
- Disallow use of nf_evars variants that drop unif. constraints.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@13904 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
in the 8.3 patch.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@13903 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
- In clenv, this used to forget the existing constraints, there are none
in the other evar_map.
- Other uses (evar_refiner, class_tactics) always reset with an enriched
version of the same evar_map.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@13901 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
conversion.
- Fix trans_fconv* to use evars correctly.
- Normalize the goal with respect to evars before rewriting in
[rewrite], allowing to see instanciations from other subgoals.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@13844 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
- A Evd.defined_evars to keep only this part of the evar_map
- One Evd.fold less in Typeclasses.mark_unresolvables
- We check that only undefined evar_map could be set unresolvable
- A duplicated function in himsg.ml
TODO: some calls to Evd.fold(_undefined) would be faster if written
as Map.map or Map.mapi.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@13716 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
In particular, the unused lib/tlm.ml and lib/gset.ml are removed
In addition, to simplify code, Libobject.record_object returning only the
('a->obj) function, which is enough almost all the time.
Use Libobject.record_object_full if you really need also the (obj->'a).
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@13460 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@13323 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@13179 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
- made the example work (a call to whd_meta was missing)
- replaced the internal error messages of w_unify_to_subterm_list into
user-understandable messages
- incidentally fixed the meaning of whd_meta (which now takes an evd)
and meta_name (which now does what it means and do not treat differently
the instantiated metas)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@13122 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
evar_map into a map for defined evars and a map for undefined evars.
Even before Spiwack's new proof engine, some Evd.fold were very
costly, e.g. in check_evars or progress_evar_map. With the new proof
engine, undefined evars traversals are apparently even more common (at
least, it improves significantly the complexity of some calls to omega
in JordanCurveTheorem - a new factor 5-7 after the factor 5-6 obtained
by removal of evar_merge in clenv_fchain in commit 13007, arriving to
figures comparable to the 8.3 ones).
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@13011 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
- Many of them were broken, some of them after Pierre B's rework
of mli for ocamldoc, but not only (many bad annotation, many files
with no svn property about Id, etc)
- Useless for those of us that work with git-svn (and a fortiori
in a forthcoming git-only setting)
- Even in svn, they seem to be of little interest
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@12972 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
This is a fairly large commit (around 140 files and 7000 lines of code
impacted), it will cause some troubles for sure (I've listed the know
regressions below, there is bound to be more).
At this state of developpement it brings few features to the user, as
the old tactics were
ported with no change. Changes are on the side of the developer mostly.
Here comes a list of the major changes. I will stay brief, but the code
is hopefully well documented so that it is reasonably easy to infer the
details from it.
Feature developer-side:
* Primitives for a "real" refine tactic (generating a goal for each
evar).
* Abstract type of tactics, goals and proofs
* Tactics can act on several goals (formally all the focused goals). An
interesting consequence of this is that the tactical (. ; [ . | ... ])
can be separated in two
tacticals (. ; .) and ( [ . | ... ] ) (although there is a conflict for
this particular syntax). We can also imagine a tactic to reorder the
goals.
* Possibility for a tactic to pass a value to following tactics (a
typical example is
an intro function which tells the following tactics which name it
introduced).
* backtracking primitives for tactics (it is now possible to implement a
tactical '+'
with (a+b);c equivalent to (a;c+b;c) (itself equivalent to
(a;c||b;c)). This is a valuable
tool to implement tactics like "auto" without nowing of the
implementation of tactics.
* A notion of proof modes, which allows to dynamically change the parser
for tactics. It is controlled at user level with the keywords Set
Default Proof Mode (this is the proof mode which is loaded at the start
of each proof) and Proof Mode (switches the proof mode of the current
proof) to control them.
* A new primitive Evd.fold_undefined which operates like an Evd.fold,
except it only goes through the evars whose body is Evar_empty. This is
a common operation throughout the code,
some of the fold-and-test-if-empty occurences have been replaced by
fold_undefined. For now,
it is only implemented as a fold-and-test, but we expect to have some
optimisations coming some day, as there can be a lot of evars in an
evar_map with this new implementation (I've observed a couple of
thousands), whereas there are rarely more than a dozen undefined ones.
Folding being a linear operation, this might result in a significant
speed-up.
* The declarative mode has been moved into the plugins. This is made
possible by the proof mode feature. I tried to document it so that it
can serve as a tutorial for a tactic mode plugin.
Features user-side:
* Unfocus does not go back to the root of the proof if several Focus-s
have been performed.
It only goes back to the point where it was last focused.
* experimental (non-documented) support of keywords
BeginSubproof/EndSubproof:
BeginSubproof focuses on first goal, one can unfocus only with
EndSubproof, and only
if the proof is completed for that goal.
* experimental (non-documented) support for bullets ('+', '-' and '*')
they act as hierarchical BeginSubproof/EndSubproof:
First time one uses '+' (for instance) it focuses on first goal, when
the subproof is
completed, one can use '+' again which unfocuses and focuses on next
first goal.
Meanwhile, one cas use '*' (for instance) to focus more deeply.
Known regressions:
* The xml plugin had some functions related to proof trees. As the
structure of proof changed significantly, they do not work anymore.
* I do not know how to implement info or show script in this new engine.
Actually I don't even know what they were suppose to actually mean in
earlier versions either. I wager they would require some calm thinking
before going back to work.
* Declarative mode not entirely working (in particular proofs by
induction need to be restored).
* A bug in the inversion tactic (observed in some contributions)
* A bug in Program (observed in some contributions)
* Minor change in the 'old' type of tactics causing some contributions
to fail.
* Compilation time takes about 10-15% longer for unknown reasons (I
suspect it might be linked to the fact that I don't perform any
reduction at QED-s, and also to some linear operations on evar_map-s
(see Evd.fold_undefined above)).
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@12961 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
(Forgot the file in last commit).
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@12868 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
- to type patterns w/o losing the information of what subterm is a hole
would need to remember where holes were in "understand", but "understand"
needs sometimes to instantiate evars to ensure the type of an evar
is not its original type but the type of its instance (what can
e.g. lower a universe level); we would need here to update evars
type at the same time we define them but this would need in turn to
check the convertibility of the actual and expected type since otherwise
type-checking constraints may disappear;
- typing pattern is apparently expensive in time; is it worth to do it
for the benefit of pattern-matching compilation and coercion insertion?
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@12607 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@12605 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
no longer considered a progress (this prepares generally having tactics
with arguments that contains holes that are added to the goal sigma).
Incidentally, made that "clear" now restricts evars only if the
restriction is really needed.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@12602 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@12502 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@12337 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
major changes in [w_unify] and the conversion functions used by it to
handle the sort constraints correctly.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@12159 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
conversion when sort variables are involved and always call it
with an empty sort constraint set to avoid [whd_sort_variable] reducing
a universe variable to an algebraic universe.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@12156 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
[set_eq_sort_variable] for cases where two universes should be equal,
fix [evars_reset_evd] to keep sort constraints and use [whd_sort_var]
directly in [whd_evar].
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@12149 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
- Solve meta type equations in the order they appeared during
unification: it's sensible because we do an [hnf_constr] on these
types, introducing a bit of delta even when it's not allowed by the
flags, and some code relies on it. A definite solution would involve an
nf variant of hnf_constr or allowing delta-reduction of closed terms
when unifying types.
- Do a bit of betaiota reduction on types in [check_types] while we
haven't got a sort-variable aware [is_trans_fconv] test.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@12140 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
universes as usual, we add the new universes to the sort constraints and
do unification modulo those ([constr_unify_with_sorts]): this allows to
instanciate Type i with Prop for example and keep track of it. The sort
constraints are thrown away at the end of unification for the moment,
but we can detect inconsistencies during unification.
Make unification more symmetric as well w.r.t. substitution of defined
metas.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@12137 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
This commit concerns about the first half of the useless code
mentionned by Oug for coqtop (without plugins). For the moment,
Oug is used in a mode where any elements mentionned in a .mli
is considered to be precious. This already allows to detect and
remove about 600 lines, and more is still to come.
Among the interesting points, the type Entries.specification_entry
and its constructors SPExxx were never used. Large parts of cases.ml
(and hence subtac_cases.ml) were also useless.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@12069 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
it does not cause a time penalty.
- Removing of get_type_of_with_meta made possible by the
evar_defs/evar_map merge.
- Adding unfolding of Meta in reductionops (this assumes that reduction does
not move Metas across binders...)
- Renaming newly created fold_map_rel_context into map_rel_context_in_env.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@12061 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
- The measure can now refer to all the formal arguments
- The recursive calls can make all the arguments vary as well
- Generalized to any relation and measure (new syntax {measure m on R})
This relies on an automatic curryfication transformation, the real
fixpoint combinator is working on a sigma type of the arguments.
Reduces to the previous impl in case only one argument is involved.
The patch also introduces a new flag on implicit arguments that says if
the argument has to be infered (default) or can be turned into a
subgoal/obligation. Comes with a test-suite file.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@12030 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
L'expérience prouve que ce n'est pas franchement concluant.
On peut se risquer à une explication :
- nf_evar, version mémoïsée n'est pas tail recursive
- On retarde la substitution des hypothèses de l'evar en échange de
faire moins de substitutions d'evars. Intuitivement c'est intéressant
seulement si il y a plus de substitutions d'evar dupliquées que
d'hypothèses dupliquées. Ce qui ne doit pas être le cas (ne serait-ce
que parce que dupliquer une evar duplique aussi ses variables libres).
This reverts commit 066a564021788e995eb166ad6ed6e55611d6f593.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11958 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
=20[whd=5Fevar]=20:=20les
=20evar=5Fdefs=20gardent=20un=20cache=20des=20appels=20pr=C3=A9c=C3=A9dents.=20Le=20d=C3=A9faut=20de=20la
=20m=C3=A9thodologie=20est=20que=20=C3=A7a=20int=C3=A9ragit=20assez=20mal=20avec=20la=20substitution=20des
=20hypoth=C3=A8ses=20de=20l'evar=20(qui=20n'est=20pas=20mise=20en=20cache).=20En=20particulier=20les
=20deux=20fonctions=20ne=20sont=20plus=20r=C3=A9cursives=20terminales.=20De=20plus=20un=20appel=20=C3=A0
=20l'une=20des=20deux=20fera=20n=C3=A9cessairement=20un=20parcours=20du=20terme=20pour=20appliquer
=20la=20substitution.?=
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 8bit
D'un point de vue de l'effet observer, ça a un effet assez léger sur le
trunk, je suis curieux de voir les effets sur les contribs.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11950 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
d'entiers à ensemble d'entier. A première vue l'optimisation a un effet
négligeable.
Mais vu l'utilisation qui est fait des des last_mods (des mem à
répétition). C'est plus raisonnable ainsi.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11942 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
L'optimisation semble significative.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11941 85f007b7-540e-0410-9357-904b9bb8a0f7
|
