Ayaz Hafiz
ff4b5f58ab
Avoid over-eager disjoint variable merging during lambda set compaction
...
During the unspecialized lambda set compaction procedure, we might end
up trying to merge too many disjoint variables during unspecialized
lambda unification. Avoid doing so, by checking if we're in the
compaction procedure.
2022-07-29 14:18:47 -04:00
Ayaz Hafiz
bda52b0d39
Implement unspecialized lambda set unification as list merging
2022-07-29 14:18:47 -04:00
Ayaz Hafiz
fe864a445d
Add a test for big encoders enabled by disjoint specialization variables
2022-07-29 14:18:47 -04:00
Ayaz Hafiz
7926499900
Implement disjoint type variable handling in the lambda set specialization algorithm
...
This completes the last known hole I am aware of in the current
lambda set specialization algorithm.
Closes #3421
2022-07-29 14:18:46 -04:00
Ayaz Hafiz
1976e435a0
Turn a couple more solve tests back on
2022-07-29 08:43:19 -04:00
Ayaz Hafiz
41eb3ad9a4
Separate ability members and their specializations in reference checking
2022-07-29 08:43:19 -04:00
Ayaz Hafiz
7a4c57d3dc
Clippy
2022-07-29 08:43:19 -04:00
Ayaz Hafiz
4049eb6780
Remove unused OpportunisticResolve field
2022-07-29 08:43:19 -04:00
Ayaz Hafiz
31220d5598
Fix spelling
2022-07-29 08:43:19 -04:00
Ayaz Hafiz
f145f29b1b
Make sure records don't de-generalize function types, fixing ability let-generalization
...
Closes #3641
2022-07-29 08:43:19 -04:00
Ayaz Hafiz
ce8c8f7264
Update solve test
...
Use named recursive calls for now in ability members
2022-07-29 08:43:18 -04:00
Ayaz Hafiz
f2d4bf20ba
Collect awaited lambda set specializations to be solved when a specialization is known
...
Despite our best efforts, sometimes we still can't specialize lambda
sets on the fly, if a specialization lambda set's specialization type
isn't yet well-known! This commit adds an `AwaitingSpecializations`
data structure to keep track of the lambda sets blocked for
specialization behind a specialization's full resolution in the module.
After the specialization is resolved, its blocked lambda sets can be
eagerly compacted.
2022-07-29 08:43:18 -04:00
Ayaz Hafiz
350d9cd59b
Simplify opportunistic resolution
...
We don't need to unify the type state anymore because resolved able
variables are determined only by one type variable.
2022-07-29 08:43:18 -04:00
Ayaz Hafiz
403b2b14b7
Add remove_dependent_unspecialized_lambda_sets to UlsOfVar
2022-07-29 08:43:18 -04:00
Ayaz Hafiz
5a42acc11c
Debug specialization keys
2022-07-29 08:43:18 -04:00
Ayaz Hafiz
240a48bc1c
De-duplicate unspecialized lambda sets by root var
2022-07-29 08:43:17 -04:00
Ayaz Hafiz
76fe397aa1
Consolidate exposed types and derived module in a derived environment
2022-07-29 08:43:17 -04:00
Ayaz Hafiz
0ec92c12f7
Move lambda set specialization to its own module in solve
2022-07-29 08:43:16 -04:00
Ayaz Hafiz
c0e976f544
Wrap unification subs in Env
2022-07-28 08:57:33 -04:00
Ayaz Hafiz
0989b2cb82
Move solve problems to their own crate
2022-07-28 08:57:32 -04:00
Ayaz Hafiz
ae4916ceb7
Remove unused variables
2022-07-28 08:57:14 -04:00
Ayaz Hafiz
753d12e912
Eagerly resolve obligations during solving
...
Replaces the previously-used `DeferredObligations` structure used for
accumulating and then acting over ability obligations during module
solving in favor of just the `ObligationCache`. The `ObligationCache`
stays alive for the entirety of a module solving and provides a
convenient mechanism for answering obligation queries with a backed
cache.
2022-07-28 08:57:13 -04:00
Folkert de Vries
01eb161ffe
Merge pull request #3637 from rtfeldman/can-abilities4
...
Syntactic abilities: Part 4 - simplify obligation solving procedure
2022-07-28 11:32:11 +02:00
Folkert de Vries
66365fba6e
Merge pull request #3649 from rtfeldman/3640-str-split-edge-case
...
Fix `Str.split` edge case when string and delimiter are equal
2022-07-28 11:30:40 +02:00
Folkert de Vries
21fe0ca559
Merge pull request #3639 from rtfeldman/expect-lists
...
Tests for expect output
2022-07-28 08:53:58 +02:00
raleng
c7997c7c38
Fix formatting
2022-07-28 01:57:01 +02:00
raleng
17fc1a297c
Fix Str.split when string equals delimiter
...
countSegments and strSplitHelp check whether the length of the string is
strictly greater-than the length of the delimiter, skipping most of the
logic when this is not the case.
Changing the check to a greater-than-or-equal allows for the case when
the string and the delimiter are equal, giving the expected result of
["", ""].
2022-07-28 01:36:02 +02:00
raleng
71e83746cc
add countSegments test case
...
The test ensures countSegments returns a count of two when the string
and the delimiter are equal. The expected result of Str.split in that
case is ["", ""].
2022-07-28 01:35:08 +02:00
raleng
261ff3224d
Add strSplitHelp test cases
...
Add a test for the case when the string starts with the delimiter, and
when the string and the delimiter are equal.
2022-07-28 01:34:20 +02:00
Folkert
8e5b0daa94
figured out how to serialize nested structures
2022-07-27 23:52:11 +02:00
Folkert
fc01a0ee44
ignore if set_shared_buffer is not defined
2022-07-27 23:50:06 +02:00
Folkert
e62a0ee39a
make string clone return where next extra bytes should be written
2022-07-27 22:48:41 +02:00
Folkert
631c642950
store the start of the value for each variable in a table
2022-07-27 21:32:49 +02:00
Folkert
9bdf1f733f
Wasm32Sized for isize
2022-07-27 20:10:47 +02:00
Folkert
da74585786
switch to load/store instead of phi because of weird bug
2022-07-27 19:27:55 +02:00
Folkert
f7452b8889
pass the shared buffer to zig directly
2022-07-27 17:16:12 +02:00
Folkert
a60a49bce6
dev backend cannot compare floats
2022-07-27 14:46:46 +02:00
Folkert
b132b1fa89
fix non-recursive tag unions in expects
2022-07-27 14:44:43 +02:00
Folkert
b24c9e33d7
make test_gen a package
2022-07-27 13:49:23 +02:00
Folkert
498004c554
clean up expect code when not running tests
2022-07-27 12:26:03 +02:00
Folkert
eefb23e433
DCE expect helper for non-expect code gen
2022-07-27 12:07:51 +02:00
Richard Feldman
fa5bd442f8
Merge pull request #3635 from rtfeldman/joshuawarner32/multiline
...
Strip indents and the first/last newline from multiline strings
2022-07-26 22:09:58 -04:00
Ayaz Hafiz
bb14b649a2
Don't re-report obligations that we know were seen elsewhere
2022-07-26 20:46:35 -04:00
Ayaz Hafiz
6b9c1cb690
Check whether opaque implements ability using store
2022-07-26 20:46:35 -04:00
Ayaz Hafiz
8659ddc684
Remove dominated derive errors
2022-07-26 20:46:35 -04:00
Ayaz Hafiz
3fbf4e228b
Remove dominated derives from obligation checking
2022-07-26 20:46:34 -04:00
Ayaz Hafiz
ad3eb959e9
Don't mark dominated derives anymore
2022-07-26 20:46:34 -04:00
Ayaz Hafiz
04ee74542d
Rename IncompleteAbility
2022-07-26 20:46:34 -04:00
Ayaz Hafiz
c0d26226dd
Remove deferred obligation for opaques during solving
2022-07-26 20:46:34 -04:00
Ayaz Hafiz
c1dbffdad9
Make sure wrong specializations become error types
2022-07-26 20:46:34 -04:00
