| Commit message (Collapse) | Author | Age |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
==========
This big patch is commited here with a HUGE experimental tag on it. It
is probably not a finished job. The aim of committing it now, as
agreed with Hugo, is to get some feedback from potential users to
identify more clearly the directions the implementation could take. So
please feel free to mail me any remarks, bug reports or advices at
<puech@cs.unibo.it>.
Here are the changes induced by it :
For the user
============
* Search tools have been reimplemented to be faster and more
general. Affected are [SearchPattern], [SearchRewrite] and [Search]
(not [SearchAbout] yet). Changes are:
- All of them accept general constructions, and previous syntactical
limitations are abolished. In particular, one can for example
[SearchPattern (nat -> Prop)], which will find [isSucc], but also
[le], [gt] etc.
- Patterns are typed. This means that you cannot search mistyped
expressions anymore. I'm not sure if it's a good or a bad thing
though (especially regarding coercions)...
* New tool to automatically infer (some) Record/Typeclasses instances.
Usage : [Record/Class *Infer* X := ...] flags a record/class as
subject to instance search. There is also an option to
activate/deactivate the search [Set/Unset Autoinstance]. It works
by finding combinations of definitions (actually all kinds of
objects) which forms a record instance, possibly parameterized. It
is activated at two moments:
- A complete search is done when defining a new record, to find all
possible instances that could have been formed with past
definitions. Example:
Require Import List.
Record Infer Monoid A (op:A->A->A) e :=
{ assoc : forall x y z, op x (op y z) = op (op x y) z;
idl : forall x, x = op x e ;
idr : forall x, x = op e x }.
new instance Monoid_autoinstance_1 : (Monoid nat plus 0)
[...]
- At each new declaration (Definition, Axiom, Inductive), a search
is made to find instances involving the new object. Example:
Parameter app_nil_beg : forall A (l:list A), l = nil ++ l.
new instance Build_Monoid_autoinstance_12 :
(forall H : Type, Monoid (list H) app nil) :=
(fun H : Type =>
Build_Monoid (list H) app nil ass_app (app_nil_beg H)
(app_nil_end H))
For the developper
==================
* New yet-to-be-named datastructure in [lib/dnet.ml]. Should do
efficient one-to-many or many-to-one non-linear first-order
filtering, faster than traditional methods like discrimination nets
(so yes, the name of the file should probably be changed).
* Comes with its application to Coq's terms
[pretyping/term_dnet.ml]. Terms are represented so that you can
search for patterns under products as fast as you would do not under
products, and facilities are provided to express other kind of
searches (head of application, under equality, whatever you need
that can be expressed as a pattern)
* A global repository of all objects defined and imported is
maintained [toplevel/libtypes.ml], with all search facilities
described before.
* A certain kind of proof search in [toplevel/autoinstance.ml]. For
the moment it is specialized on finding instances, but it should be
generalizable and reusable (more on this in a few months :-).
The bad news
============
* Compile time should increase by 0 to 15% (depending on the size of
the Requires done). This could be optimized greatly by not
performing substitutions on modules which are not functors I
think. There may also be some inefficiency sources left in my code
though...
* Vo's also gain a little bit of weight (20%). That's inevitable if I
wanted to store the big datastructure of objects, but could also be
optimized some more.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11794 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
| |
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11792 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
|
|
|
|
|
|
|
|
| |
- Really unify with types of metas when they contain metas
_or_ evars (why not always?) (fixes bug #2027).
- Better handling of evars in rewrite lemmas when using setoid_rewrite
through rewrite (reported by Ralf Hinze).
- Use retyping with metas when possible (check?) and improve an
obscure error message in retyping.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11776 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
|
|
|
|
| |
(detection of Miller's pattern)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11748 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
|
|
|
|
| |
by user) and #2017 (unification pattern test too crude leading to
regression wrt to 8.1).
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11743 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
|
|
|
|
|
|
|
|
| |
- matching_subterm was activating partial_app to true in matches_core even
when no partial_app was expected,
- "match goal" (hence "extended_matches") was called with partial_app
in 8.2 (currently "matches" but not in trunk; what to do with
(legacy) "matches" remains unclear.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11733 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
|
|
|
|
|
|
| |
splay_prod_n, lam_it -> it_mkLambda, splay_lambda -> splay_lam). Added
shortcuts for "fst (decompose_prod t)" and co.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11727 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- Backtrack on precise unfolding of "iff" in "tauto": it has effects on
the naming of hypotheses (especially when doing "case H" with H of
type "{x|P<->Q}" since not unfolding will eventually introduce a name
"i" while unfolding will eventually introduce a name "a" (deep sigh).
- Miscellaneous (error when a plugin is missing, doc hnf, standardization
of names manipulating type constr_pattern, ...).
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11725 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
inductive types was not taken into account).
- Virtually extended tauto to
- support arbitrary-length disjunctions and conjunctions,
- support arbitrary complex forms of disjunctions and
conjunctions when in the contravariant of an implicative hypothesis,
- stick with the purely propositional fragment and not apply reflexivity.
This is virtual in the sense that it is not activated since it breaks
compatibility with the existing tauto.
- Modified the notion of conjunction and unit type used in hipattern in a
way that is closer to the intuitive meaning (forbid dependencies
between parameters in conjunction; forbid indices in unit types).
- Investigated how far "iff" could be turned into a direct inductive
definition; modified tauto.ml4 so that it works with the current and
the alternative definition.
- Fixed a bug in the error message from lookup_eliminator.
- Other minor changes.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11721 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
automation.
- Permitted to use evars in the intermediate steps of "apply in" (as expected
in the test file apply.v).
- Back on the systematic use of eq_rect (r11697) for implementing
rewriting (some proofs, especially lemma DistOoVv in
Lyon/RulerCompassGeometry/C14_Angle_Droit.v and tactic compute_vcg in
Sophia-Antipolis/Semantics/example2.v are explicitly refering
to the name of the lemmas used for rewriting).
- Fixed at the same time a bug in get_sort_of (predicativity of Set was not
taken into account).
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11717 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
|
|
|
|
|
| |
unused methods object. Matthieu please review this change (after
monday), I might have introduced a bug in rebuild_instance.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11659 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
| |
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11655 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
| |
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11653 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
|
|
|
|
|
|
|
| |
- Fixed a virtual bug of unification (ever occurs if w_unify called with
a non-empty context of rel's, which is a priori uncommon).
- Fixed Notation.out test.
- Add better coqide error message in case editor is called on an unnamed file.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11650 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
| |
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11644 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
<<
Axiom A : nat -> bool.
Definition foo :=
match A 0 with
| true => true
| k => k
end.
Print foo.
>>
[A 0] is duplicated in the [k => k] branch!
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11640 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
| |
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11636 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
|
|
| |
closed w.r.t. dependencies
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11634 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
| |
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11614 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
| |
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11612 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
|
|
| |
11585)
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11590 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
|
|
|
|
|
|
|
|
| |
Evarutil.check_and_clear_in_constr in V8.2 revision 11309 and trunk
revision 11300).
- Improved various error messages related to inversion, evars and case
analysis (including the removal of the obsolete dependent/non dependent
distinction).
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11561 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
| |
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11560 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
help the type _checking_ anymore (they don't become typing constraints)
but they permit to coerce a subterm in a type. In particular, when
using a VM cast we avoid unneeded, unexpected conversions using the
default machine (oops!). Also remove the corresponding comment in
pretyping and fix the wrong use of casts in toplevel/command: accept
the trouble of using evars. This has the somewhat adverse effect that
when typing casted object we now have no typing constraints (see
e.g. examples in Cases.v)!
Probably, this will be backtracked partially tomorrow as many contribs
can rely on it and the change could make some unifications fail (in
particular with deep coercions). Let's try anyway!
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11558 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
|
|
|
|
|
|
| |
declaring additional conversion problems when unifying the type of an
evar instance and the evar's declared type. Fix the corresponding
conversion heuristic which failed due to (misplaced?) assertions when
faced with general conversion problems.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11549 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
|
|
|
|
|
|
|
|
| |
instances and the corresponding evar's type if it contains existentials
to avoid dangling evars. No noticeable performance impact (at least on the
stdlib). Subsumes (and fixes) the (broken) fix in unification.ml that was
previously patched by M. Puech.
Improve error messages related to existential variables and type
classes.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11543 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
|
|
|
|
|
| |
for record fields (using "someproj : sometype where not := constr" syntax). Only one
notation allowed currently and no redeclaration after the record
declaration either (will be done for typeclasses).
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11542 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- make the modifiers "value of" and "type of" for "set" working (it was not!),
- clear unselected hypotheses in the "in" clause of "induction/destruct" when
the destructed term is a variable (experimental),
- support for generalization of hypotheses in the induction hypotheses using
the "in" clause of "induction" (e.g. "induction n in m, H" will
generalize over m -- would it be better to have an explicit
"over"/"generalizing" clause ?).
Added clause "as" to "apply in".
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11509 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
|
|
|
|
|
| |
s'est avéré ralentir la compilation des user-contribs au final, sans
compter aussi le bug 1980 apparemment introduit par ce commit).
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11505 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
|
|
|
|
| |
avoid trying to resolve classes early in open constr arguments for Ltac,
the tactics themselves should do whatever's appropriate with the constraints.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11503 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
It solves feature request 1852, makes me and Arnaud happy and
will permit to factor some more code in typeclasses.
- Records are introduced using the syntax "{| x := t; y := foo |}" and
"with" clauses are currently parsed but not yet supported in the
elaboration. You are invited to suggest other syntaxes :)
- Missing fields are turned into holes, extra fields cause an error
message. The current implementation finds the type of the record
at pretyping time, from the typing constraint alone (and just expects
an inductive with one constructor). It is then impossible to use
scope information to parse the bodies: that may be wrong. The other
solution I see is using the fields to detect the type earlier, before
internalisation of the bodies, but then we get in name clash hell.
- In funind/contrib/interface I mostly put [assert false] everywhere to
avoid warnings.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11496 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
dans clos_norm_flags (cf échec dans CoRN.Transc.InvTrigonom.Tan_ilim). Ceci
dit :
- cela ne me parait pas moral que clos_norm_flags s'occupe de
normaliser les evars,
- mais comme "evar" n'est pas un flag supporté par closure.ml, on ne
peut pas le donner à la demande comme argument de clos_norm_flags
(question: pourrait-on faire supporter la réduction Evar par closure.ml ??).
Plus généralement, il y a un problème avec la propagation des
instantiations des evars à travers les buts (cf lemme eapply_evar dans
test-suite/success/apply.v).
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11470 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
|
|
|
|
|
|
| |
abusivement sur les clauses.
Nettoyage au passage de metamap qui était utilisé à la fois pour les
substitutions de meta et pour les contextes de typage de meta.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11467 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
|
|
|
|
|
|
| |
* Fixed typo in unify_0 regarding conv_pb
* First attempt to fix a problem related to rels in w_merge. Seems to
be unsuccessful at this point
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11443 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- backtrack on kernel modifications: the monomorphic instance of an inductive
type is constrained to live in an universe higher (or equal) than all the
instances
- improved support for polymorphic inductive types at the refiner level:
introduced type_of_inductive_knowing_conclusion that computes the
instance to match the current conclusion universe.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11435 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
|
|
|
|
|
|
|
|
| |
unfold first. The patch changes the usual order of unifications so some
may differ (only one example in the stdlib breaks because of more
unification happening).
Actually this change was trigerred because of an incompleteness which is
not resolved here. At least this way unfolding is consistent between
w_unify and the kernel.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11427 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
| |
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11425 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
| |
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11423 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
|
|
|
|
|
| |
evars at the end of unification as later evars can refer to
previous ones. This removes the assumption that evars are already
ordered in eterm's code.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11419 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
|
|
|
| |
cases where coercion could not occur as well.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11414 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
| |
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11405 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
| |
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11403 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
|
|
|
| |
processed in the larger context instead of the bodies.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11400 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
|
|
|
| |
overriding the default tactic when adding a definition.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11373 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
|
|
|
|
|
|
| |
evar, do unification between the evar type and the type of the instance
to properly propagate information.
Typical example: in context ?A : Type, ?R : relation ?A. When we instantiate ?R
using a goal like x = y by @eq t, we need to instantiate A to t as well.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11357 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
|
|
|
|
| |
backtracking on coercion classes when a coercion path fails).
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11344 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
|
|
|
| |
instance multiple times at each section closing (still a hack).
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11338 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- evarconv: mauvaise idée d'utiliser la conversion sur la tête d'un
terme applicatif au moment de tester f u1 .. un = g v1 .. vn au
premier ordre : on revient sur l'algo tel qu'il était avant le
commit 11187.
- Bug #1887 (format récursif cassé à cause de la vérification des idents).
- Nouveau choix de formattage du message "Tactic Failure".
- Nettoyage vocabulaire "match context" -> "match goal" au passage.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11305 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
|
|
|
|
|
| |
11126 et qui faisait qu'on ne backtrackait en fait plus sur les
sous-termes d'un terme qui lui filtrait.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11304 85f007b7-540e-0410-9357-904b9bb8a0f7
|
|
|
|
|
|
|
|
|
|
|
| |
canonique que si elle contribue vraiment à une équation canonique,
c'est-à-dire si son argument principal est une evar; sinon on
répercute le comportement historique qui est de préférer le dépliage
du côté droit d'une équation constante/constante.
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@11303 85f007b7-540e-0410-9357-904b9bb8a0f7
|