Commit Graph

3333 Commits

Author SHA1 Message Date
Justus Matthiesen
5718be2b45 [ cleanup ] ignore generated file 2022-12-12 14:47:33 +00:00
Justus Matthiesen
e673d05a67
[ fix ] jump-to-definition (":name-at" IDE command) (#2811) 2022-12-09 17:02:57 +00:00
Markus Pfeiffer
c3bbdb30a1
Add mustWork in GADT data declaration parser (#2805)
Co-authored-by: Guillaume Allais <guillaume.allais@ens-lyon.org>
2022-12-09 16:04:11 +00:00
Guillaume Allais
1d73e359e0 [ ci ] only run latest job
As discussed on discord a while ago, this kills runs when a new
commit lands on the same branch. It should save some CI ressources.
2022-12-09 14:40:14 +00:00
Guillaume Allais
caeb55b495 [ ci ] fix initialise step
Github is now failing hard on multiline commit messages it seems.
2022-12-09 14:40:14 +00:00
Guillaume Allais
9bcaa104f3 [ new ] logging topics autocompletion in the emacs REPL 2022-12-08 21:06:39 +00:00
locriacyber
cb30d7cda8
Install *.ttm on idris2 --install (#2796)
Co-authored-by: Justus Matthiesen <mail@justusmatthiesen.com>
Co-authored-by: Guillaume Allais <guillaume.allais@ens-lyon.org>
2022-12-08 21:06:17 +00:00
Thomas E. Hansen
7a41606573 [ admin ] Fix PR template 2022-12-08 11:24:37 +00:00
Ruslan
d44ddc7f11
Let qualified do-notation apply to pure and to idiom brackets (#2786) 2022-12-07 22:27:58 +00:00
Ellis Kesterton
9931f35b6b
Improve parser error messages when using reserved identifiers in decls (#2803)
Co-authored-by: Ellis Kesterton <erk4@st-andrews.ac.uk>
2022-12-07 16:09:26 +00:00
G. Allais
32f92746b3
[ perf ] compile Core.CompileExpr.Pretty faster (#2800) 2022-12-07 10:08:31 +00:00
Guillaume Allais
085ea348ae [ perf ] use integers in the buffer state 2022-12-06 21:13:55 +00:00
Guillaume Allais
ebc406c519 [ perf ] cherrypick fast weakenNVar from yaffle 2022-12-06 12:28:16 +00:00
Guillaume Allais
001f48fba8 [ test ] for io_bind inlining 2022-12-06 11:55:28 +00:00
G. Allais
fcbd9e9190
[ fix #2794 ] Do not ignore notinline lets in identity detection (#2795) 2022-12-06 11:41:30 +00:00
Denis Buzdalov
ff6afb0c59 [ minor ] Make NotBothZero parameter of standard gcd to be erased 2022-12-05 15:10:16 +00:00
Denis Buzdalov
8d7b993cca [ cleanup ] Use several library functions instead of their code 2022-12-05 15:10:16 +00:00
Denis Buzdalov
33fe3f44db [ cleanup ] Clean up existing code in the Arrow module 2022-12-05 15:10:16 +00:00
Denis Buzdalov
c8f625cad9 [ typo ] Fix typo in an issue template 2022-12-05 15:10:16 +00:00
Guillaume Allais
99eda4dd2b [ papers ] fill last hole in lambda pi formalisation 2022-12-05 15:02:04 +00:00
Denis Buzdalov
59aadd650f [ ci ] Replace deprecated feature with its functional equivalent 2022-12-01 20:35:22 +00:00
G. Allais
255cce9d9c
[ ci ] fix katla scripts following the TTC version subdirectory (#2787) 2022-12-01 20:20:01 +00:00
G. Allais
85bb822f3b
[ perf ] manually eta-expand unsaturated io_bind calls (#2785) 2022-12-01 15:31:00 +00:00
G. Allais
f96c5ca596
[ papers ] LambdaPi (#2780) 2022-12-01 13:48:48 +00:00
Steve Dunham
f443723f4e [ parser ] fix issue where indentation is not checked in record parameters 2022-11-25 08:22:37 +00:00
Tim Engler
69f680e10a Update introapp.rst
Further explain why the App passed to ``run`` can't throw an exception.
2022-11-21 10:55:52 +00:00
0xd34df00d
c6a8c9e7a7 [ base ] Move the worker outside of the Vect's reverse definition 2022-11-21 10:52:47 +00:00
Robert Wright
7d9d39c045 Use escaped version of system in Idris src 2022-11-21 10:51:43 +00:00
Robert Wright
edacf65182 Bump version of Idris used in CI 2022-11-21 10:51:43 +00:00
Steve Dunham
56a9bc6be4 [ error ] Improve locality of parse errors in implicit arguments 2022-11-21 10:40:40 +00:00
Tim Engler
bff18428b4 Added seqL to Control.App and updated docs to fix #2761
Also updated test real002 to use the actual Control.App from
libs/base/Control/App.idr. Before it was using a different version that
existed within its test directory, tests/idris2/real002/Control/App.idr
2022-11-21 10:39:43 +00:00
Marek L
0e657aeba1 [ ide-mode ] Do not include whole error message in building file
display result message.

Why:
Makes the error message being displayed twice in Emacs.
This change aligns IDEMode/REPL.idr with Idris/REPL.idr which
works as expected.

Relates to:
https://github.com/idris-hackers/idris-mode/pull/563
2022-11-21 10:34:23 +00:00
Tim Engler
811dcd864f Similar to pull request #2136, commit 8ffad8878f
except now packageDirs keeps getting added to. So rather than just
saving and replacing extraDirs, I'm replacing the whole dirs directory
within defs.options
2022-11-21 10:33:56 +00:00
Kamiλ Shakirov
dbb3853484 Do not build docs for papers 2022-11-15 16:38:13 +01:00
Kamil Shakirov
73733f03ff Cleanup after rebase 2022-11-15 16:38:13 +01:00
Kamil Shakirov
8a35a9dc74 Fix IDRIS2_BOOT_PATH 2022-11-15 16:38:13 +01:00
Kamil Shakirov
8ebec8ebae Make linter happy 2022-11-15 16:38:13 +01:00
Kamil Shakirov
7c6ab4e1b6 Do not use IDRIS2_BOOT_PATH in bootstrap-stage2.sh
It is not needed as bootstrapped intermediate version knows where to find import files.
`IDRIS2_BOOT_PATH` wasn't pointing to the correct directories anyways.
2022-11-15 16:38:13 +01:00
Kamil Shakirov
502aa6f4fc Do not install papers library 2022-11-15 16:38:13 +01:00
Kamil Shakirov
dccb80d0bf [ fix ] Fix Makefile targets for linear and papers libraries 2022-11-15 16:38:13 +01:00
Hattori, Hiroki
5c9f8e36a1
[ RefC ] Add 16 and 32 bit access to base/Data.Buffer . (#2609)
* Fix symbom mangling

* Revert "Fix symbom mangling"

This reverts commit 6481e80155.

* Fix typo

* [RefC] Add missed prims of setBuffer* .

* [ fix ] formatting

* [ re #2609 ] Use 'UInt' instead of 'Word'

More descriptive/to the point / Less assumed knowledge.

There are no *LE suffixes for UInt8, since endianness is to do with
multiple bytes and UInt8 is a single one.

Co-authored-by: Guillaume Allais <guillaume.allais@ens-lyon.org>
Co-authored-by: Thomas E. Hansen <teh6@st-andrews.ac.uk>
2022-11-15 12:15:06 +01:00
CodingCellist
037b48eb29
[ re #2649 ] Update CHANGELOG (#2765) 2022-11-15 10:58:00 +01:00
CodingCellist
70ef197cf6
[ base ] Deprecate setByte in favour of setBits8 (#2764)
* [ base ] Deprecate setByte in favour of setBits8
           Deprecate getByte; fix Core.Binary.Prims

Along with `setByte`, the `getByte` function should similarly be
deprecated since it also assumes the value will have the given size,
rather than guaranteeing it in the type.

CI highlighted some required changes in `Core.Binary.Prims` thanks to
`-Werror`. The fix was to add some `cast` calls where the old `getByte`
and `setByte` used to be.

The RefC buffer test will need updating once we remove the functions
completely. Added a note for future peeps.
2022-11-15 10:42:07 +01:00
Denis Buzdalov
f5f7e2eb80 [ error ] Make failing IAlternative w/ FirstSuccess print all errors 2022-11-15 10:28:59 +01:00
Tim Engler
774529fd70 Altered docs for `Exception` to reflect the code changes in Control.App
Exception is no longer an interface but just a synonym of ``HasErr``. Updated docs to reflect this.
2022-11-13 09:28:43 +00:00
Zoe Stafford
d0e761d63b
Merge pull request #2772 from dunhamsteve/issue2769
[ error ] Improve error messages on record and interface constructors (#2769)
2022-11-13 09:17:40 +00:00
Steve Dunham
85c93c8018 [ error ] Add FC for errors on interface constructors (#2769) 2022-11-12 22:17:42 -08:00
Steve Dunham
ea733181c5 [ error ] Add an FC to record constructors for better error messages (#2769) 2022-11-12 21:59:16 -08:00
Zoe Stafford
d1e1d0636c
Merge pull request #2767 from madman-bob/erase-fin-fromInteger-upper-bound
Erase Fin fromInteger upper bound
2022-11-12 09:45:51 +00:00
Robert Wright
b2c42280bb Erase Fin fromInteger upper bound 2022-11-11 15:34:58 +00:00