| Commit message (Collapse) | Author | Age |
|
|
|
| |
of syntax in test file ltac.v.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
capturing bound names unexpectingly).
We moved renaming to after finding bindings, i.e. in pretyping
"fun x y => x + y" in an ltac environment where y is ltac-bound to x,
we type the body in environment "x y |-" but build "fun y y => Rel 2 + Rel 1"
(which later on is displayed into "fun y y0 => y + y0").
We renounced to renaming in "match" patterns, which was anyway not
working well, as e.g. in
let x := ipattern:y in (forall z, match z with S x => x | x => x end = 0).
which was producing
forall z : nat, match z with 0 => z | S y => y end = 0
because the names in default branches are treated specifically.
It would be easy to support renaming in match again, by putting the
ltac env in the Cases.pattern_matching_problem state and to rename the
names in "typs" (from build_branch), just before returning them at the
end of the function (and similarly in abstract_predicate for the names
occurring in the return predicate).
However, we rename binders in fix which were not interpreted.
There are some difficulties with evar because we want them defined in
the interpreted env (see comment to ltac_interp_name_env).
fix ltac name
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Some functions from pretyping/typing.ml and their derivatives were potential
source of evarmap leaks, as they dropped their resulting evarmap. This commit
clarifies the situation by renaming them according to a unsafe_* scheme. Their
sound variant is likewise renamed to their old name. The following renamings
were made.
- Typing.type_of -> unsafe_type_of
- Typing.e_type_of -> type_of
- A new e_type_of function that matches the e_ prefix policy
- Tacmach.pf_type_of -> pf_unsafe_type_of
- A new safe pf_type_of function.
All uses of unsafe_* functions should be eventually eliminated.
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
| |
environment.
A closure is supposedly closed: all the relevant Ltac variables should be then. The last field [ltac_genargs], if I'm not mistaken, is there to represent the Ltac variables which are bound but not to something which makes sense in a term. They should be irrelevant at this point, since the uconstr has already been interpreted and these checks are supposed to have happened. (though I'm not entirely sure they do, it can be an interesting exercise to try and make [uconstr] behave weirdly)
I'm not quite sure why it caused #3679, though. But it still seems to be solved.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Instead of modifying exceptions to wear additional information, we instead use
a dedicated type now. All exception-using functions were modified to support
this new type, in particular Future's fix_exn-s and the tactic monad.
To solve the problem of enriching exceptions at raise time and recover this
data in the try-with handler, we use a global datastructure recording the
given piece of data imperatively that we retrieve in the try-with handler.
We ensure that such instrumented try-with destroy the data so that there
may not be confusion with another exception. To further harden the correction
of this structure, we also check for pointer equality with the last raised
exception.
The global data structure is not thread-safe for now, which is incorrect as
the STM uses threads and enriched exceptions. Yet, we splitted the patch in
two parts, so that we do not introduce dependencies to the Thread library
immediatly. This will allow to revert only the second patch if ever we
switch to OCaml-coded lightweight threads.
|
|
|
|
| |
pattern-matching predicate.
|
|
|
|
|
|
|
|
| |
pattern-matching.
In the syntax [let (x1,…,xn) := b in e] the name [x1…xn] were not interpreted with respect to the Ltac context. Hence typeable term would raise type-errors.
The same problem also existed in pattern-matching.
|
|
|
|
|
|
| |
Pretyping of [if] cases was missing a convertibility check with the ambient
type constraint, and just dropped it. This was making the pretyper construct
ill-typed terms.
|
| |
|
|
|
|
|
| |
Instead of keeping checking that evars are pending, we cache the pending
evars in a proper set computed once and for all.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The main change is that selection of subterm is made similar whether
the given term is fully applied or not.
- The selection of subterm now works as follows depending on whether
the "at" is given, of whether the subterm is fully applied or not,
and whether there are incompatible subterms matching the pattern. In
particular, we have:
"at" given
| subterm fully applied
| | incompatible subterms
| | |
Y Y - it works like in 8.4
Y N - this was broken in 8.4 ("at" was ineffective and it was finding
all subterms syntactically equal to the first one which matches)
N Y Y it now finds all subterms like the first one which matches
while in 8.4 it used to fail (I hope it is not a too risky in-draft
for a semantics we would regret...) (e.g. "destruct (S _)" on
goal "S x = S y + S x" now selects the two occurrences of "S x"
while it was failing before)
N Y N it works like in 8.4
N N - it works like in 8.4, selecting all subterms like the
first one which matches
- Note that the "historical" semantics, when looking for a subterm, to
select all subterms that syntactically match the first subterm to
match the pattern (looking from left to right) is now internally called
"like first".
- Selection of subterms can now find the type by pattern-matching (useful e.g.
for "induction (nat_rect _ _ _ _)")
- A version of Unification.w_unify w/o any conversion is used for
finding the subterm: it could be easily replaced by an other
matching algorithm.
In particular, "destruct H" now works on a goal such as "H:True -> x<=y |- P y".
Secondary change is in the interpretation of terms with existential
variables:
- When several arguments are given, interpretation is delayed at the
time of execution
- Because we aim at eventually accepting "edestruct c" with unresolved
holes in c, we need the sigma obtained from c to be an extension of
the sigma of the tactics, while before, we just type-checked c
independently of the sigma of the tactic
- Finishing the resolution of evars (using type classes, candidates,
pending conversion problems) is made slightly cleaner: it now takes
three states: a term is evaluated in state sigma, leading to state
sigma' >= sigma, with evars finally solved in state sigma'' >=
sigma'; we solve evars in the diff of sigma' and sigma and report
the solution in sigma''
- We however renounce to give now a success semantics to "edestruct c"
when "c" has unresolved holes, waiting instead for a decision on
what to do in the case of a similar eapply (see mail to coqdev).
An auxiliary change is that an "in" clause can be attached to each component
of a "destruct t, u, v", etc.
Incidentally, make_abstraction does not do evar resolution itself any longer.
|
|
|
|
| |
to maintain compatibility (HoTT bug #557).
|
|
|
|
|
| |
it became possible to have binding names themselves bound to ltac
variables (2fcc458af16b).
|
|
|
|
|
|
|
|
| |
will name the goal id; writing ?[?id] will use the first
fresh name available based with prefix id.
Tactics intro, rename, change, ... from logic.ml now preserve goal
name; cut preserves goal name on its main premise.
|
|
|
|
|
|
|
|
| |
so as to reproduce correctly the reduction behavior of existing
projections, i.e. delta + iota. Make [projection] an abstract datatype
in Names.ml, most of the patch is about using that abstraction.
Fix unification.ml which tried canonical projections too early in
presence of primitive projections.
|
|
|
|
|
| |
and pretyping which was not contracting the eta-expanded forms in presence
of coercions.
|
|
|
|
|
|
|
|
| |
contortions in internalization/externalization. It uses a fully typed
version of detyping, requiring the environment, to move from
primitive projection applications to regular applications of
the eta-expanded version. The kernel is unchanged, and only
constrMatching needs compatibility code now.
|
|
|
|
|
| |
substitution at the right place.
I used to change [id] to its interpretation before calling [pretype_id]. But it's incorrect: we need to use the Ltac interpretation only when looking up the rel context (where it has been interpreted previously). It would not be to use the interpreted identifier to look up the named context or the Ltac context.
|
|
|
| |
The ident closure was not propagated when pretying a [uconstr] coming from a [uconstr] closure. This bug had never been reported, as far as I'm aware.
|
| |
|
|
|
|
| |
in instances.
|
| |
|
| |
|
|
|
|
|
| |
instances still to do). Using heuristics to name after the quantifier
name it comes. Also added a "sigma" to almost all printing functions.
|
| |
|
|
|
|
| |
context for goal contexts.
|
|
|
| |
Binder names are interpreted as the Ltac specified one if available.
|
|
|
|
|
|
|
|
|
|
|
| |
Now kernel/indtypes builds the corresponding terms (has to be trusted)
while translate_constant just binds a constant name to the
already entered projection body, avoiding the dubious "check"
of user given terms. "case" Pattern-matching on primitive records is
now disallowed, and the default scheme is implemented using
projections and eta (all elimination tactics now use projections
as well). Elaborate "let (x, y) := p in t" using let bindings
for the projections of p too.
|
| |
|
|
|
|
|
| |
This avoids relying on detyping. As Matthieu Sozeau pointed out to me, [understand∘detyping] has no reason to be the identity. This may create surprising behaviour some times (when a detyped term loses its relations to the current context, in particular in terms of universes), and downright incompatibilities in the case of refine.
As a bonus this should be a faster implementation of [uconstr] with a leaner memory profile.
|
|
|
|
|
|
|
| |
variables.
Simplifies instantiation of constants/inductives, requiring less allocation and Map.find's.
Abstraction by variables is handled mostly inside the kernel but could be moved outside.
|
|
|
|
|
|
|
|
|
|
|
|
| |
- realargs: refers either to the indices of an inductive, or to the proper args
of a constructor
- params: refers to parameters (which are common to inductive and constructors)
- allargs = params + realargs
- realdecls: refers to the defining context of indices or proper args
of a constructor (it includes letins)
- paramdecls: refers to the defining context of params (it includes letins)
- alldecls = paramdecls + realdecls
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
a global reference that the current (goal) env contains all the
section variables that the global reference expects to be present.
Note that the test for inclusion might be costly: everytime a
conversion happens in a section variable copied in a goal, this
conversion has to be redone when referring to a constant dependent on
this section variable.
It is unclear to me whether we should not instead give global names to
section variables so that they exist even if they are not listed in
the context of the current goal.
Here are two examples which are still problematic:
Section A.
Let B := True : Type.
Definition C := eq_refl : B = True.
Theorem D : Type.
clearbody B.
set (x := C).
unfold C in x.
(* inconsistent context *)
or
Section A.
Let B : Type.
exact True.
Qed.
Definition C := eq_refl : B = True. (* Note that this violated the Qed. *)
Theorem D : Type.
set (x := C).
unfold C in x.
(* inconsistent context *)
|
|
|
|
|
|
|
| |
the computed direction argument. In case pbty is conv, no refreshing is done
as the evar body must be convertible with the given term, however refreshing
of template application evar arguments can still happen.
(Re)-Closing bug #2966.
|
|
|
|
| |
for building polymorphic instances of template polymorphic inductives.
|
|
|
|
| |
universe instance.
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
for helping fixing this).
Now the issue is handled solely through refreshing of the terms assigned
to evars during unification.
If ?X = list ?Y, then Y's type is refreshed so that it doesn't mention
a template universe and in turn, ?X won't. Same goes when typechecking
(nil ?X, nil ?Y), the pair constructor levels will be set higher than fresh
universes for the lists carriers. This also handles user-defined functions
on template polymorphic inductives, which was fragile before.
Pretyping and Evd are now uncluttered from template-specific code.
|
| |
|
| |
|
|
|
|
| |
applications earlier.
|
|
|
|
|
|
|
| |
conclusion, and results of
unifying the lemma with subterms. Using Retyping.get_type_of instead results in 3x
speedup in Ncring_polynom.
|
|
|
|
| |
different places
|
|
|
|
|
|
|
| |
- Add a tentative syntax for specifying universes: Type{"i"} and foo@{Type{"i"},Type{"j"}}.
These are always rigid.
- Use level-to-level substitutions where the more general level-to-universe substitutions
were previously used.
|
|
|
|
|
|
|
| |
eagerly solve l <= k constraints as k := l when k is a fresh variable coming
from a template type. This has the effect of fixing the variable at the first
instantiation of the parameters of template polymorphic inductive and avoiding
to generate useless <= constraints that need to be minimized afterwards.
|
| |
|
|
|
|
|
| |
- Fix in canonical structure inferce, we have to check that the heads
are convertible and keep universe information around.
|