Commit Graph

3093 Commits

Author SHA1 Message Date
MithicSpirit
acd9c09a05
Reduce ambiguity in dependent types filter example
`p` was used for both the filter predicate and the length of the
filtered vector. This renames the latter to n'.
2022-10-16 00:44:54 -04:00
Zoe Stafford
bc8b6bb7a7
Merge pull request #2715 from Z-snails/parser-fc
[ parser ] fix FC for type declarations
2022-10-14 21:30:07 +01:00
Zoe Stafford
c2e4773ed7 [ parser ] fix FC for type declarations 2022-10-14 18:56:28 +01:00
G. Allais
51d00638b0
[ cosmetic ] Say when you start compiling the executable (#2714) 2022-10-14 16:21:19 +01:00
G. Allais
30866fa892
[ new ] do not split on dotted patterns (#2706) 2022-10-14 16:21:06 +01:00
Denis Buzdalov
eb4cc0054b [ golden ] Make options to store more flexible tests filter 2022-10-14 15:03:05 +01:00
Steve Dunham
b15be4a70a [ new ] Allow forward declarations of records 2022-10-14 14:58:29 +01:00
CodingCellist
bb602643b1
[ fix #2712 ] Don't consider CyclicMeta recoverable (#2713) 2022-10-14 14:57:55 +01:00
Thomas E. Hansen
6823915dd0 [ tests ] Check REPL help text and module import cmds
The help text check is for making sure we update it when modifying
things, more than anything else.
2022-10-13 11:31:05 +02:00
Thomas E. Hansen
a598fec4f7 [ repl ] Add :import command 2022-10-13 11:31:05 +02:00
Denis Buzdalov
3ca065b85d [ fix #2088 ] Make %runElab expression propagate its rig to check 2022-10-07 10:55:22 +02:00
Denis Buzdalov
2744a3a5a2
[ golden ] Truncate long test names when printing the results (#2553) 2022-10-06 19:18:34 +01:00
Zoe Stafford
c69f439c2d
[ cleanup ] Compiler/* modules (#2694)
* cleanup(Compiler/*)
cleanup some messy/unclear code
Also use primitives directly where possible, instead of idris's wrappers

* Fix tests
2022-10-06 10:17:38 +02:00
Thomas E. Hansen
ff6ffad907 fix formatting 2022-10-06 10:10:16 +02:00
Denis Buzdalov
693ed5543a [ doc, re #2193 ] Document expected behaviour of namespaces in functions 2022-10-06 10:10:16 +02:00
Thomas E. Hansen
bdbc0c72bf Make the linter happy 2022-10-05 14:30:08 +02:00
Thomas E. Hansen
cfe18b3d01 [ papers ] A bit of tidying up and clarification 2022-10-05 14:30:08 +02:00
Thomas E. Hansen
d25db9e353 [ papers ] Add implementation of Dekker's algorithm
This completes the implementation of the examples in the paper
"Applications of Applicative Proof Search" (Liam O'Connor, 2016).
Unfortunately, the final example is an example of something that _can_
be expressed, but _cannot_ be model-checked by the approach in the
paper.

(Side note on `petersonsCorrect`: The paper mentions that it checks in
~3 minutes on a 2013 MacBook Pro. Assuming they mean "type-checks", this
is roughly consistent with our observations of just short of 2 minutes.
I doubt that they evaluated it, since an attempt at doing this on a
reasonably modern server (Intel Xeon Gold 5220R, 502 GB of RAM) was
killed after just over 3 hours, producing the following resource log:

  Command exited with non-zero status 255
  Time: 11320.46s user, 35.12s system, 3h09m46s elapsed, 99%CPU
  Memory: 57644016 Kbytes RAM

)
2022-10-05 14:30:08 +02:00
Thomas E. Hansen
260d9ecb58 [ papers ] Comment out petersonsCorrect for performance reasons
On average across 10 runs on an Intel Core i7-8750H with 15.2GB of
available system memory (16GB installed in system), type-checking
`libs/papers/Search/GCL.idr` WITH `petersonsCorrect` takes:

  * 1 minute  48.7 seconds, consuming 3.92GB of RAM

By contrast, commenting `petersonsCorrect` out results in type-checking
taking on average (same #runs, same config):

  * 0 minutes  1.2 seconds, consuming 0.25GB of RAM

And good luck trying to evaluate the thing!
(This might be a good performance test at some point, but uh, we're not
there yet...)
2022-10-05 14:30:08 +02:00
Thomas E. Hansen
c5b114209e [ papers ] Tidy the Peterson's Prop type
Inspired by `Search/Properties.Pythagoras.formula`, having the CSP as
its own type allows for some better docs+understanding than `Prop ? ?`.
2022-10-05 14:30:08 +02:00
Thomas E. Hansen
d5c9253e7e [ papers ] Fix impl.n s.t. model-checking Peterson's works
Thanks to the debug info supplied by #2673, I was able to spot which
functions were blocking and `public import` the relevant files in
`papers/Search/Properties.idr`.

As a result the GCL file now type-checks, albeit extremely slowly!
I stopped an evaluation of `petersonsCorrect` at the REPL after 15
minutes and ~14GB of RAM consumed.
2022-10-05 14:30:08 +02:00
Thomas E. Hansen
bad8631c77 [ fix ] Delayed values are concrete
Co-authored-by: Edwin Brady <ecb10@st-andrews.ac.uk>
2022-10-05 14:30:08 +02:00
Thomas E. Hansen
36554eb56f [ papers ] Make ops total, implement termination mc
* Currying the `ops` function makes the totality checker spot that it
  _is_ actually total.
* Instance arguments are heavily abused in the paper, along with
  implicit `open` magic, but Idris allows no such ~~luxury~~
  obfuscation, so we have to pass things explicitly.
* `decSo` is not `public export`ed, so we have to define `IsTT` by
  pattern-matching (which is fine).

Currently, it gets stuck on checking `petersonsCorrect` for some,
currently unknown, reason. (And the log output is loooooong O.O)

Once again, this would not have been possible without gallais insigths.
Many thanks!

Co-authored-by: Guillaume Allais <guillaume.allais@ens-lyon.org>
2022-10-05 14:30:08 +02:00
Thomas E. Hansen
8c76118f2f [ papers ] Finish Search.CTL
* Switch to `Inf` to actually use codata/corecursion.
* Add `%hint`s to mark the interface implementations as such, despite
  use of a record for `DepthInv` (this is necessary for other stuff).
* Pass in `Oh` to `reaches10.evidence` in order for things to work.

With huge thanks to gallais for helping me put the final things in
place!

Co-authored-by: Guillaume Allais <guillaume.allais@ens-lyon.org>
2022-10-05 14:30:08 +02:00
Thomas E. Hansen
0d72964f21 [ papers ] Missed some public exports 2022-10-05 14:30:08 +02:00
Thomas E. Hansen
eb4749d381 [ papers ] Figure out types, start termination check
Termination checking needs figuring out. There is some funky stuff going
on with the half-deciders and their constructors. Other than that, I
**think** it's nearly done. God knows how much RAM it'll take though...
2022-10-05 14:30:08 +02:00
Thomas E. Hansen
6421973282 [ papers ] Add mutex and starv-free properties 2022-10-05 14:30:08 +02:00
Thomas E. Hansen
5742afe3a5 [ papers ] Set the stage for mc-ing Peterson's alg 2022-10-05 14:30:08 +02:00
Thomas E. Hansen
8ca48e0033 [ papers ] Implement reaches10 proof
Seems to be very slow though...
And Idris is unable to find the depth-inv instance for `r10Proof`.
Could be that auto-search is not as strong as Agda's? Or more likely,
I've set things up slightly wrong...
2022-10-05 14:30:08 +02:00
Thomas E. Hansen
5f48fbe857 [ papers ] Add derived GCL control structures 2022-10-05 14:30:08 +02:00
Thomas E. Hansen
92ed8607e3 [ papers ] Start implementing GCL
I am upset about the amount of computation done in that view...
2022-10-05 14:30:08 +02:00
Thomas E. Hansen
b3a8d344e4 [ papers ] Implement proof-search for EF, AF, EG, AG
This also caught an implementation error in the Global formula
definitions:
AG f = A[ f U (f AND' Completed) ]
and **not**
AG f = A[ (f U f) AND' Completed ]
(both of which are valid parsings of the original
 AG f = A[ f u f AND' Completed]
 )
2022-10-05 14:30:08 +02:00
Thomas E. Hansen
f96b25a3d7 [ papers ] Keep AF, EF, AG, EG naming consistent 2022-10-05 14:30:08 +02:00
Thomas E. Hansen
81df879c77 [ papers ] Define Finally operators 2022-10-05 14:30:08 +02:00
Thomas E. Hansen
58b6ccdaf3 [ papers ] Change to LazyList for Computation Tree
This makes more sense in terms of `EU` being efficient and only
evaluating as much as it needs to. However, I'm not sure `model.follow`
is implemented correctly (Agda delays the call to `model.followAll`,
which I'm unsure if we can do (and if so, how to do it) in Idris)...
2022-10-05 14:30:08 +02:00
Thomas E. Hansen
68f8e69ea2 [ papers ] Impl.t proof-search for AU and EU
This reveals an unfortunate problem/misunderstanding: For `ExistsUntil`
to make sense, in terms of evaluation speedups, the list needs to be
lazy. Which is _not_ what `Lazy (List a)` does /!\
I need to switch to LazyList...
2022-10-05 14:30:08 +02:00
Thomas E. Hansen
40780c8f85 [ papers ] Figure out isCompleted by trial and error
And here's a good case against allowing custom unicode syntax:
〈$〉 is `<$>`, i.e. the infix notation for `map`. That's fine; If you
happen to know it!
ESPECIALLY, if your paper defines 〈_〉 as custom notation for a guarded
expression! Then there is **no way** to tell that the expression 〈$〉
is not a guarded expression over `$`, but is instead the alias for
`map`!! You just have to magically know this beforehand!

We also need an explicit `Lazy` annotation for Idris to be happy with
the implicit `ms` in the `IsCompleted` constructor.
2022-10-05 14:30:08 +02:00
Thomas E. Hansen
67218e3eac [ papers ] Move LTE' outside parameters block 2022-10-05 14:30:08 +02:00
Thomas E. Hansen
3e0d5acfa4 [ papers ] Map depth-invariant prfs for AU and EU
The proofs of depth-invariance for Always Until and Exists Until require
mapping the proofs over the Formulae's internal `All` and `Any`
respectively. Idris provides some functions for this, but they erase the
list and so don't quite work. Instead we need to implement our own,
which don't erase the list.
2022-10-05 14:30:08 +02:00
Thomas E. Hansen
40e1a53ab2 [ papers ] Cleaned things up a bit; utterly confused
Don't you love when papers introduce syntax and functions which you've
never seen before and don't seem to match the types of the existing
stuff?

P.S. YEET! (aka. that's probably enough for today ^^)
2022-10-05 14:30:08 +02:00
Thomas E. Hansen
db30dd8d4a [ papers ] Fix things by adding a parameter block
The original Agda code declares the module with L and Sigma (Lbls and
Sts) with type Set. This is apparently close to a parameter block, which
solves the unification error I was having with `now`! Huge thanks to
gallais for showing me that!
2022-10-05 14:30:08 +02:00
Thomas E. Hansen
0c72f83fe8 [ papers ] Start implem.g the model-checking part of Liam's paper
I should have put this under version-control WAAAAAY sooner than this!
Oh well, better late than never...

There are some fun problems to solve in terms of type-mismatch and
erasure, but that's for another day.
2022-10-05 14:30:08 +02:00
Thomas E. Hansen
a90fe03ff7 Add DOI to Liam's search paper stuff 2022-10-05 14:30:08 +02:00
CodingCellist
23eea52c50
[ fix ] also usleep in the producer of channels006 to guarantee correct output (#2702)
Co-authored-by: Guillaume Allais <guillaume.allais@ens-lyon.org>
2022-10-05 12:03:20 +01:00
G. Allais
1f3809c49a
[ re #2675 ] Do not build libs/{contribs,papers} during bootstrap (#2677)
Co-authored-by: Ben Hormann <benhormann@users.noreply.github.com>
2022-10-04 13:37:45 +01:00
Zoe Stafford
138452273c
Merge pull request #2701 from stefan-hoeck/snocfold
[ performance ] efficient foldr and foldMap for SnocList
2022-10-04 12:52:39 +01:00
stefan-hoeck
0e956249cc [ performance ] efficient foldr and foldMap for SnocList 2022-10-04 12:19:32 +02:00
Sam Phillips
bacc418aa0 Remove codegen option from channel tests 2022-10-04 10:24:27 +02:00
Sam Phillips
54a426356a Move channel tests to allschemes 2022-10-04 10:24:27 +02:00
Sam Phillips
358c107c53 Update CHANGELOG.md and CONTRIBUTORS 2022-10-04 10:24:27 +02:00