| Age | Commit message (Collapse) | Author |
|---|
|
|
|
This is documented in dev/doc/changes.md.
|
|
|
|
This provides several advantages to people serializing tactic
scripts. Appearance of the involved constructors is common enough as
to bother to submit this PR.
|
|
|
|
|
|
This is a step towards limiting calls to the global environment.
Incidentally unify naming evd -> sigma in Termops.
|
|
This avoids all the side effects associated with the manipulation of an
unresolvable flag. In the new design:
- The evar_map stores a set of evars that are candidates for typeclass
resolution, which can be retrieved and set.
We maintain the invariant that it always contains only undefined
evars.
- At the creation time of an evar (new_evar), we classify it as a
potential candidate of resolution.
- This uses a hook to test if the conclusion ends in a typeclass
application. (hook set in typeclasses.ml)
- This is an approximation if the conclusion is an existential (i.e.
not yet determined). In that case we register the evar as
potentially a typeclass instance, and later phases must consider
that case, dropping the evar if it is not a typeclass.
- One can pass the ~typeclass_candidate:false flag to new_evar to
prevent classification entirely. Typically this is for new goals
which should not ever be considered to be typeclass resolution
candidates.
- One can mark a subset of evars unresolvable later if
needed. Typically for clausenv, and marking future goals as
unresolvable even if they are typeclass goals. For clausenv for
example, after turing metas into evars we first (optionally) try a
typeclass resolution on the newly created evars and only then mark
the remaining newly created evars as subgoals. The intent of the
code looks clearer now.
This should prevent keeping testing if undefined evars are classes
all the time and crawling large sets when no typeclasses are present.
- Typeclass candidate evars stay candidates through
restriction/evar-evar solutions.
- Evd.add uses ~typeclass_candidate:false to avoid recomputing if the new
evar is a candidate. There's a deficiency in the API, in most use
cases of Evd.add we should rather use a:
`Evd.update_evar_info : evar_map -> Evar.t -> (evar_info -> evar_info)
-> evar_map`
Usually it is only about nf_evar'ing the evar_info's contents, which
doesn't change the evar candidate status.
- Typeclass resolution can now handle the set of candidates
functionally: it always starts from the set of candidates (and not the
whole undefined_map) and a filter on it, potentially splitting it in
connected components, does proof search for each component in an
evar_map with an empty set of typeclass evars (allowing clean
reentrancy), then reinstates the potential remaining unsolved
components and filtered out typeclass evars at the end of
resolution.
This means no more marking of resolvability/unresolvability
everywhere, and hopefully a more efficient implementation in general.
- This is on top of the cleanup of evar_info's currently but can
be made independent.
[typeclasses] Fix cases.ml: none of the new_evars should be typeclass candidates
Solve bug in inheritance of flags in evar-evar solutions.
Renaming unresolvable to typeclass_candidate (positive) and fix maybe_typeclass_hook
|
|
|
|
We refactor the `Coqlib` API to locate objects over a namespace
`module.object.property`.
This introduces the vernacular command `Register g as n` to expose the
Coq constant `g` under the name `n` (through the `register_ref`
function). The constant can then be dynamically located using the
`lib_ref` function.
Co-authored-by: Emilio Jesús Gallego Arias <e+git@x80.org>
Co-authored-by: Maxime Dénès <mail@maximedenes.fr>
Co-authored-by: Vincent Laporte <Vincent.Laporte@fondation-inria.fr>
|
|
It was only used in ltac/rewrite which as a plugin is too late to
affect init.
|
|
|
|
reference was defined as Ident or Qualid, but the qualid type already
permits empty paths. So we had effectively two representations for
unqualified names, that were not seen as equal by eq_reference.
We remove the reference type and replace its uses by qualid.
|
|
- move_location to proofs/logic.
- intro_pattern_naming to Namegen.
|
|
We address the easy ones, but they should probably be all removed.
|
|
|
|
We bootstrap the circular evar_map <-> econstr dependency by moving
the internal EConstr.API module to Evd.MiniEConstr. Then we make the
Evd functions use econstr.
|
|
The `reference` type contains some ad-hoc locations in its
constructors, but there is no reason not to handle them with the
standard attribute container provided by `CAst.t`.
An orthogonal topic to this commit is whether the `reference` type
should contain a location or not at all.
It seems that many places would become a bit clearer by splitting
`reference` into non-located `reference` and `lreference`, however
some other places become messier so we maintain the current status-quo
for now.
|
|
UState normalize -> minimize, Evd nf_constraints -> minimize_universes
|
|
|
|
|
|
|
|
This commit was motivated by true spurious conversions arising in my
`to_constr` debug branch.
The changes here need careful review as the tradeoffs are subtle and
still a lot of clean up remains to be done in `vernac/*`.
We have opted for penalize [minimally] the few users coming from true
`Constr`-land, but I am sure we can tweak code in a much better way.
In particular, it is not clear if internalization should take an
`evar_map` even in the cases where it is not triggered, see the
changes under `plugins` for a good example.
Also, the new return type of `Pretyping.understand` should undergo
careful review.
We don't touch `Impargs` as it is not clear how to proceed, however,
the current type of `compute_implicits_gen` looks very suspicious as
it is called often with free evars.
Some TODOs are:
- impargs was calling whd_all, the Econstr equivalent can be either
+ Reductionops.whd_all [which does refolding and no sharing]
+ Reductionops.clos_whd_flags with all as a flag.
|
|
|
|
We follow the suggestions in #402 and turn uses of `Loc.located` in
`vernac` into `CAst.t`. The impact should be low as this change mostly
affects top-level vernaculars.
With this change, we are even closer to automatically map a text
document to its AST in a programmatic way.
|
|
|
|
|
|
Some code in typeclasses was even breaking the invariant that
use_polymorphic_flag should not be called twice, but that code was
morally dead it seems, so we remove it.
|
|
|
|
This is useful as it allows to reflect program_mode behavior as an attribute.
|
|
|
|
The exception needs to carry aroud a pair of `env, sigma` so printing
is correct. This gets rid of a few global calls, and it is IMO the
right thing to do.
While we are at it, we incorporate some fixes to a couple of
additional printing functions missing the `env, sigma` pair.
|
|
New module introduced in OCaml 4.05 I think, can create problems when
linking with the OCaml toplevel for `Drop`.
|
|
This reduces conversions between ContextSet/UContext and encodes
whether we are polymorphic by which constructor we use rather than
using some boolean.
|
|
Also use constant_universes_entry instead of a bool flag to indicate
polymorphism in ParameterEntry.
There are a few places where we convert back to ContextSet because
check_univ_decl returns a UContext, this could be improved.
|
|
We can enforce properties through check_univ_decl, or get an arbitrary
ordered context with UState.context / Evd.to_universe_context (the
later being a new wrapper of the former).
|
|
|
|
To that extent we introduce a new prototype vernacular extension macro
`VERNAC COMMAND FUNCTIONAL EXTEND` that will take a function with the
proper parameters and attributes.
This of course needs more refinement, in particular we should move
`vernac_command` to its own file and make `Vernacentries` consistent
wrt it.
|
|
|
|
We do up to `Term` which is the main bulk of the changes.
|
|
The manual has long stated that these forms are deprecated. We add a
warning for them, as indeed `Add Morphism` is an "proof evil" [*]
command, and we may want to remove it in the future.
We've also fixed the stdlib not to emit the warning.
[*] https://ncatlab.org/nlab/show/principle+of+equivalence
|
|
The old algorithm was relying on list membership, which is O(n). This was
nefarious for terms with many binders. We use instead sets in O(log n).
|
|
We dont care about the order of the binder map ([map] in the code) so
no need to do tricky things with it.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Reminder of (some of) the reasons for removal:
- Despite the claim in sigma.mli, it does *not* prevent evar
leaks, something like:
fun env evd ->
let (evd',ev) = new_evar env evd in
(evd,ev)
will typecheck even with Sigma-like type annotations (with a proof of
reflexivity)
- The API stayed embryonic. Even typing functions were not ported to
Sigma.
- Some unsafe combinators (Unsafe.tclEVARS) were replaced with slightly
less unsafe ones (e.g. s_enter), but those ones were not marked unsafe
at all (despite still being so).
- There was no good story for higher order functions manipulating evar
maps. Without higher order, one can most of the time get away with
reusing the same name for the updated evar map.
- Most of the code doing complex things with evar maps was using unsafe
casts to sigma. This code should be fixed, but this is an orthogonal
issue.
Of course, this was showing a nice and elegant use of GADTs, but the
cost/benefit ratio in practice did not seem good.
|
