diff options
| author | Emilio Jesus Gallego Arias | 2020-03-26 02:26:43 -0400 |
|---|---|---|
| committer | Emilio Jesus Gallego Arias | 2020-03-31 05:16:22 -0400 |
| commit | b755869a7fdb34dcf7dacc9d5fd93c768d71d751 (patch) | |
| tree | 68eb026dc0f9e214afcb44739e2c617d3d5f8be6 /tactics/proof_global.mli | |
| parent | 5828dffb05022ff667f44b1ad9a89f677647e0b4 (diff) | |
[proof] Split delayed and regular proof closing functions, part II
We make the types of the delayed / non-delayed declaration path
different, as the latter is just creating futures that are forced
right away.
TTBOMK the new code should behave identically w.r.t. old one, modulo
the equation `Future.(force (from_val x)) = x`.
There are some questions as what the code is doing, but in this PR
I've opted to keep the code 100% faithful to the original one, unless
I did a mistake.
Diffstat (limited to 'tactics/proof_global.mli')
| -rw-r--r-- | tactics/proof_global.mli | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/tactics/proof_global.mli b/tactics/proof_global.mli index d820fc8b40..44b78ee911 100644 --- a/tactics/proof_global.mli +++ b/tactics/proof_global.mli @@ -67,7 +67,7 @@ val update_global_env : t -> t (* Takes a function to add to the exceptions data relative to the state in which the proof was built *) -val close_proof : opaque:opacity_flag -> keep_body_ucst_separate:bool -> Future.fix_exn -> t -> proof_object +val close_proof : opaque:opacity_flag -> keep_body_ucst_separate:bool -> t -> proof_object (* Intermediate step necessary to delegate the future. * Both access the current proof state. The former is supposed to be |