| Age | Commit message (Collapse) | Author |
|---|
|
We perform some cleanup and remove dependency of `proofs/` on
`interp/`, which seems logical.
In fact, `interp` + `parsing` are quite self-contained, so if there is
interest we could also make tactics to depend directly on proofs.
|
|
Ack-by: SkySkimmer
Reviewed-by: ejgallego
Reviewed-by: ppedrot
|
|
|
|
We change the API; after some thinking the tradeoff is clear in favor
of the more radical functional option from the start.
We also guarante the existence of a proof is by typing now,
so exceptions `NoCurrentProof` and `NoSuchGoal` are gone.
TODO: Review what's going on with focusing now.
|
|
|
|
Kernel should be mostly correct, higher levels do random stuff at
times.
|
|
Ack-by: gares
Ack-by: herbelin
Ack-by: mattam82
Ack-by: ppedrot
|
|
I think the usage looks cleaner this way.
|
|
In order to do so we place the polymorphic status and name in the
read-only part of the monad.
Note the added comments, as well as the fact that almost no part of
tactics depends on `proofs` nor `interp`, thus they should be placed
just after pretyping.
Gaëtan Gilbert noted that ideally, abstract should not depend on the
polymorphic status, should we be able to defer closing of the
constant, however this will require significant effort.
Also, we may deprecate nameless abstract, thus rending both of the
changes this PR need unnecessary.
|
|
|
|
|
|
Named evar_abstract_arguments, this field indicates if the evar
arguments corresponding to certain hypothesis can be immitated during
inversion or not. If the argument comes from an abstraction (the evar
was of arrow type), then imitation is disallowed as it gives unnatural
solutions, and lambda abstraction is preferred.
|
|
We remove all calls to `Flags.is_program_mode` except one (to compute
the default value of the attribute). Everything else is passed
explicitely, and we remove the special logic in the interpretation loop
to set/unset the flag.
This is especially important since the value of the flag has an impact on
proof modes, so on the separation of parsing and execution phases.
|
|
|
|
DAG nodes hold now a system state and a parsing state.
The latter is always passed to the parser.
This paves the way to decoupling the effect of commands on the parsing
state and the system state, and hence never force to interpret, say,
Notation.
Handling proof modes is now done explicitly in the STM, not by interpreting
VernacStartLemma.
Similarly Notation execution could be split in two phases in order to obtain a
parsing state without fully executing it (that requires executing all
commands before it).
Co-authored-by: Maxime Dénès <maxime.denes@inria.fr>
Co-authored-by: Emilio Jesus Gallego Arias <e+git@x80.org>
|
|
This error was more or less a debug tool (checking that no tactic breaks
the invariant). But some users may want to support other models, see
https://github.com/Mtac2/Mtac2/pull/139 for an example discussion.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
workers
|
|
- deprecate the old 5-tuple accessor in favor of a view record,
- move `name` and `kind` proof data from `Proof_global` to `Proof`,
this will prove useful in subsequent functionalizations of the
interface, in particular this is what abstract, which lives in the
monads, needs in order no to access global state.
- Note that `Proof.t` and `Proof_global.t` are redundant anyways.
|
|
This should improve correctness and will be needed for the PRs that
remove global access to the proof state.
|
|
|
|
|
|
These modules do actually belong there.
We have to slightly reorganize printers, removing a couple of
duplicated ones in the way.
|
|
This is a pre-requisite to use automated formatting tools such as
`ocamlformat`, also, there were quite a few places where the comments
had basically no effect, thus it was confusing for the developer.
p.s: Reading some comments was a lot of fun :)
|
|
|
|
As per https://github.com/coq/coq/pull/8965#issuecomment-441440779
|
|
write_function
|
|
|
|
|
|
|
|
We remove the `Proof_types` file which was a trivial stub, we also
cleanup a few layers of aliases.
This is not a lot but every little step helps.
|
|
|
|
|
|
|
|
This way we only have 2 `start_proof` entries, in `Lemmas` and
`Proof_global`; which they should be unified / brought closer in the
future.
|
|
|
|
This is documented in dev/doc/changes.md.
|
|
|
|
This is inspired and an alternative to #8981. We consolidate the "open
proof" exception, allowing clients to explicitly capture it and
removing some ugly duplicated code in the way.
The `Solve Obligation tac` semantics are then tweaked as to removed
the wide-scope "catch-all" and indeed will now relay errors in `tac`
as it will only absorb tactics that don't error but fail to close the
goal such as `auto`. For the rest of the cases, we introduce a
warning, and may move to a full error in later releases.
We also remove an unnecessary `tclCOMPLETE` call to code that will
actually call `close_proof`. In this case, it is better to delegate
error management to the core function.
Some error messages have changed [as we consolidate two error paths]
so this PR may require adjustment in that area.
|
|
This is barely used and not very useful, clients should use the
close_proof API directly.
|
|
|
|
Allow for new goals that don't map to old goals
Include background_goals in all_goals return value
Fix incorrect change to raw diffs in shorten_diff_span
Fixes #8922
|
|
This reverts commit 8d8200d4bff3ffc44efc51ad44dccee9eb14ec6a.
Fix #7936
# Conflicts:
# proofs/clenvtac.ml
|
|
This emphasizes that it works only on inductive types.
Also, the name is_template_polymorphic will be reused for a more
general version.
|
|
Clenv.make_evar_clause.
|
