| Age | Commit message (Collapse) | Author |
|---|
|
|
|
|
|
|
|
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.
|
|
This is more efficient in general, because Termops.dependent doesn't take
advantage of the knowledge of its pattern argument.
|
|
We forbid calling `EConstr.to_constr` on terms that are not evar-free,
as to progress towards enforcing the invariant that `Constr.t` is
evar-free. [c.f. #6308]
Due to compatibility constraints we provide an optional parameter to
`to_constr`, `abort` which can be used to overcome this restriction
until we fix all parts of the code.
Now, grepping for `~abort:false` should return the questionable
parts of the system.
Not a lot of places had to be fixed, some comments:
- problems with the interface due to `Evd/Constr` [`Evd.define` being
the prime example] do seem real!
- inductives also look bad with regards to `Constr/EConstr`.
- code in plugins needs work.
A notable user of this "feature" is `Obligations/Program` that seem to
like to generate kernel-level entries with free evars, then to scan
them and workaround this problem by generating constants.
|
|
Previously [fun x : Ind@{i} => x : Ind@{j}] with Ind some cumulative
inductive would try to generate a constraint [i = j] and use
cumulativity only if this resulted in an inconsistency. This is
confusingly different from the behaviour with [Type] and means
cumulativity can only be used to lift between universes related by
strict inequalities. (This isn't a kernel restriction so there might
be some workaround to send the kernel the right constraints, but
not in a nice way.)
See modified test for more details of what is now possible.
Technical notes:
When universe constraints were inferred by comparing the shape of
terms without reduction, cumulativity was not used and so too-strict
equality constraints were generated. Then in order to use cumulativity
we had to make this comparison fail to fall back to full conversion.
When unifiying 2 instances of a cumulative inductive type, if there
are any Irrelevant universes we try to unify them if they are
flexible.
|
|
|
|
We mirror the structure of EConstr and move the destructors from `Term`
to `Constr`.
This is a step towards having a single module for `Constr`.
|
|
We do up to `Term` which is the main bulk of the changes.
|
|
|
|
We refine the criterion for selecting a projection. Before PR#924 it
was alphabetic (i.e. morally "random" up to alpha-conversion). After
PR#924 it was chronological.
We refine a bit more by giving priority to simple projections when
they exist over projections which include an evar instantiation (and
which may actually be ill-typed).
|
|
This is a first step towards some of the solutions proposed in #6008.
|
|
alphabet
|
|
|
|
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).
|
|
This surprising bug was caused by an Id.Set which was ordering
solutions to variable-projection problems in ascii order.
We fix it by re-considering the variables involved in the solutions
using the declaration order.
Note that in practice, this implies preferring a dependent solution
over a non-dependent one.
|
|
The use of template polymorphism in constants was quite limited, as it
only applied to definitions that were exactly inductive types without any
parameter whatsoever. Furthermore, it seems that following the introduction
of polymorphic definitions, the code path enforced regular polymorphism as
soon as the type of a definition was given, which was in practice almost
always.
Removing this feature had no observable effect neither on the test-suite,
nor on any development that we monitor on Travis. I believe it is safe to
assume it was nowadays useless.
|
|
|
|
|
|
|
|
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.
|
|
Use an explicit label ~algebraic for make_flexible_variable as well.
|
|
As per https://github.com/coq/coq/pull/716#issuecomment-305140839
Partially using
```bash
git grep --name-only 'anomaly\s*\(~label:"[^"]*"\s*\)\?\(Pp.\)\?(\(\(Pp.\)\?str\)\?\s*".*[^\.!]")' | xargs sed s'/\(anomaly\s*\(~label:"[^"]*"\s*\)\?\(Pp.\)\?(\(\(Pp.\)\?str\)\?\s*".*\s*[^\.! ]\)\s*")/\1.")/g' -i
```
and
```bash
git grep --name-only ' !"' | xargs sed s'/ !"/!"/g' -i
```
The rest were manually edited by looking at the results of
```bash
git grep anomaly | grep '\.ml' | grep -v 'anomaly\s*\(~label:"[^"]*"\s*\)\?\(Pp\.\)\?(\(\(Pp.\)\?str\)\?\s*".*\(\.\|!\)")' | grep 'anomaly\($\|[^_]\)' | less
```
|
|
|
|
This is the continuation of #244, we now deprecate `CErrors.error`,
the single entry point in Coq is `user_err`.
The rationale is to allow for easier grepping, and to ease a future
cleanup of error messages. In particular, we would like to
systematically classify all error messages raised by Coq and be sure
they are properly documented.
We restore the two functions removed in #244 to improve compatibility,
but mark them deprecated.
|
|
|
|
|
|
For now we only normalize sorts, and we leave instances for the next
commit.
|
|
|
|
|
|
|
|
This removes quite a few unsafe casts. Unluckily, I had to reintroduce
the old non-module based names for these data structures, because I could
not reproduce easily the same hierarchy in EConstr.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Universes are kept in normal form w.r.t. equality but not the <=
relation, so the previous check worked almost always but was actually
too strict! In cases like (max(Set,u) = u) when u is declared >= Set it
was failing to find an equality. Applying the KISS principle:
u = v <-> u <= v /\ v <= u.
Fix invariant breakage that triggered the discovery of the check_eq bug as well. No algebraic universes should appear in a term position (on the left of
a colon in a typing judgment), this was not the case when an algebraic
universe instantiated an evar that appeared in the term. We force their
universe variable status to change in refresh_universes to avoid this.
Fix ind sort inference: Use syntactic universe equality for inductive
sort inference instead of check_leq (which now correctly takes
constraints into account) and simplify code
|
