andr3/zls
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
1,390 Commits 1 Branch 0 Tags 3.6 MiB
bf19ed3ea9
Commit Graph

355 Commits

Author SHA1 Message Date
Alex Kladov
73d6264cab
Fix missing nodes in outline (#990) 
* Remove some boilerplate from test

* Fix missing nodes in outline
 2023-02-11 14:20:01 -05:00
Techatrix
8ec82321d4
remove arena allocator from analysis (#981) 
* remove arena allocator from analysis

* fuzzer fix

* deinit `BoundTypeParams`
 2023-02-07 17:44:04 -05:00
bing
384f227cb7
fix doc comment not showing on container field (#947) 2023-02-03 18:58:53 -05:00
Techatrix
fe54fb7cfa
Use new ast helper functions (#939) 
* use new ast helper functions

* use new ast helper functions

* update min build version

* fix merge conflicts
 2023-01-24 18:55:38 -05:00
Techatrix
d2586f79a1
replace ArrayList with MultiArrayList in DocumentScope (#938) 2023-01-24 15:43:48 -05:00
Álan Crístoffer
903f85ab94
Fix lsp weird behaviour on block cursors (#891) (#905) 
* Fix lsp weird behaviour on block cursors (#891)

Adds lookahead option to getPositionContext.
 2023-01-22 15:47:53 -05:00
Aleksey Kladov
ea05916e69 Goto definition works when the cursor is at the start of the identifier. 
Before, the code lexed only a prefix of the line up to cursor position.
Now, we lex the entire line, and do look at the token just after the
cursor.

This subtly changes sematncih of `getPostionContext`: now it is becomes
oblivious of the _exact_ position of the cursor and returns the whole
token at cursor's position.

I believe this is semantically right approach -- _most_ of the callsite
should not worry at all about such details. Something like completion
_might_ want to know more, but it's better to make that call site's
problem.

It might be the case that some existing code relies on the past
behavior. It's hard to tell though -- we don't have a lot of tests for
_features_, and changes to unit-tests don't explain if the changes are
meaningful for user-observable behavior or not.

In general, for LSP-shaped thing, I feel that the bulk of testing should
be focused on end-to-end behaviors....
 2023-01-21 18:31:26 +00:00
Techatrix
8d53a5382d
resolve type of @typeInfo (#915) 2023-01-16 13:46:33 -05:00
Techatrix
3f2700eaa5
improve completion on error and enums (#887) 2023-01-06 13:59:20 -05:00
Techatrix
a574ac68ba
use more explicit error sets (#880) 2023-01-03 10:21:58 -05:00
Techatrix
c718e12d16
Autofix improvements (#879) 
* improve autofix stability and client support

* run zig fmt
 2023-01-02 14:59:01 -05:00
Techatrix
94ec3a0a86
Debugging utilities (#860) 
* add debug printing for Ast and DocumentScope

* add optional failing allocator
 2022-12-30 18:42:53 -05:00
Techatrix
f473088b64
fix crashes found through fuzzing (#866) 2022-12-30 18:42:28 -05:00
Techatrix
61c0981294
Use zig-lsp-codegen (#850) 
* add lsp.zig

* change references from types.zig to lsp.zig

* remove types.zig and requests.zig

* add tres as a submodule

* transition codebase from types.zig to lsp.zig

* update lsp.zig

* completely overhaul message handler

* fix memory errors

* partially transition tests to lsp.zig

* update lsp.zig

* more test fixes

* disable failing tests

* fix message handling bugs

* fix remaining tests

* access correct union in diff.applyTextEdits

* more message handler fixes

* run zig fmt

* update tres submodule

* fix memory access to freed memory

* simplify initialize_msg for testing

* check if publishDiagnostics is supported
 2022-12-27 01:47:57 -05:00
Techatrix
941882371c
fix memory lifetime issues (#851) 2022-12-27 00:52:15 -05:00
Travis Staloch
e09c773005
analysis.zig getPositionContext() - check for null 
closes #754 and #832

this pr just adds `.label => {}` to the switch as suggested by
@nullptrdevs, thereby preventing the null unwrap.  i checked that zls no
longer crashes when positioning the cursor on Server.zig:2287:41 which
is this line:
  `break :blk .{ .WorkspaceEdit = edits };`
                              ^ cursor here previously crashed zls
 2022-12-17 15:37:49 -08:00
Techatrix
68ac6f9f30
remove ast.tokenSlice() (#824) 2022-12-15 17:01:42 -05:00
Techatrix
c39a4eb2ab
simplify & refactor analysis code (#823) 2022-12-15 13:03:09 -05:00
Auguste Rame
0ab34abc0f
Bug stream fixes (#818) 
* Fix glaring inlay hint issue; thanks for the report Nameless

* Fix label references; closes #728
 2022-12-13 22:07:36 -05:00
Alexandros Naskos
3cac23f8a1
Fix references to now removed, previously deprecated std.ascii declarations 2022-12-10 23:21:55 +02:00
Alexandros Naskos
56a65f42bf
Added tuple type support, fixed compilation with zig master branch (#786) 
* Added tuple type support, fixed compilation with zig master branch

* Removed unneeded comment

* Bumped up minimum zig version required to build zls
 2022-11-26 12:22:16 -05:00
Auguste Rame
87aa4c09e1
Quick comptime interpreter fix, place it behind an option (#761) 
* Quick fix

* Add config option
 2022-11-16 19:28:01 -05:00
Auguste Rame
355d56376f
Merge pull request #724 from zigtools/comptime-bebe-steps 
Implement comptime interpreter
 2022-11-16 18:13:35 -05:00
Techatrix
d75fd3a880
detect comment position context (#756) 2022-11-16 17:34:36 -05:00
Auguste Rame
599c134593
Imports, cross-boundary resolution; can import std but dies on missing builtins / lang features 2022-11-08 14:37:13 -05:00
Auguste Rame
da00751726
Field access, function calls based on function value 2022-11-08 14:37:13 -05:00
Auguste Rame
779c3c0710
Hacky mess but it works (only if your function is the first root decl tho :P) 2022-11-08 14:37:13 -05:00
Auguste Rame
d2640a44ac
comptime interpreter moment 2022-11-08 14:37:01 -05:00
Tom Cheng
af14067911 Fix crash when getting signature of optional 
e.g. typing 'foo.?.bar(' crashes zls because it doesn't recognize ? as a
possible token as part of a function expression, and tries to call
getFieldAccessType with '.bar' instead.

The actual fix is the one line in src/signature_help.zig

getFieldAccessType was reworked to be more resilient to this type of
thing - the `undefined` value of `current_type.type.data` was being used
since it hit the `.period` branch first. This caused the crash.
 2022-11-02 14:55:22 +00:00
Techatrix
2158a201ad add separate function for loading handles with getOrLoadHandle 2022-10-17 20:43:11 +02:00
Techatrix
25ce5dd8aa undo DocumentStore constness changes 2022-10-16 17:17:40 +02:00
Techatrix
d6ef330984 Merge branch 'master' into document-store-refactor 2022-10-10 19:07:15 +02:00
Jonas Enlund
3833566790
Add missing .switch_case_inline, .switch_case_inline_one cases (#699) 
* Add missing .switch_case_inline, .switch_case_inline_one cases

* Include inline semantic token

Co-authored-by: Auguste Rame <19855629+SuperAuguste@users.noreply.github.com>
 2022-10-07 20:00:23 -04:00
Techatrix
5d748767e9 mark DocumentStore & Handle const 2022-10-05 13:40:11 +02:00
Techatrix
dab7eb81cc reimplement document store 2022-10-05 13:23:38 +02:00
Techatrix
a9a1fad13e
remove types.TextDocument (#693) 
* remove types.TextDocument

* resolve compile errors because of previous merge
 2022-09-30 20:45:45 -04:00
Techatrix
6ff19e8b5f analysis: store function node in parameter declaration 2022-09-24 21:25:32 +02:00
Techatrix
e28549fa7d
Completely overhaul offsets.zig (#643) 
* completely overhaul offsets.zig
 2022-09-16 00:33:49 +00:00
max
5aff17afb0
sync with zig master version llvm15 (#652) 2022-09-14 14:56:17 -04:00
Lee Cannon
bd6f60e451
update for zig master (#651) 2022-09-13 16:54:28 -04:00
nullptrdevs
6269eef776
fix: "f..o.o;" crashes the server (#629) 
this is a fix for #381 and a hack for #409
(related, get triggered by "f..o.o;")
 2022-09-07 13:14:11 -04:00
nullptrdevs
54be6d92c6 fix: fix server crash related to parsing incomplete function as a function's parameter 
fixes zigtools#567
 2022-08-25 13:17:38 -07:00
Lee Cannon
a18ec394f1
Merge pull request #602 from Techatrix/unmanage-code 
unmanage all the code
 2022-08-23 21:21:08 +01:00
Techatrix
dc3199cffa unmanage all the code 2022-08-23 12:44:26 +02:00
nullptrdevs
015332b69e
Update analysis.zig 2022-08-22 10:54:29 -07:00
nullptrdevs
2c5ae262f7 fix: fix memory leaks related to updating config variables 
adds an edge case in analysis.makeScopeInternal to prevent
leaking memory when adding duplicate container fields w/ name "other"
 2022-08-21 17:11:50 -07:00
Techatrix
c6892eaafc
Merge branch 'zigtools:master' into cimport-support 2022-08-21 08:53:06 +00:00
Techatrix
8457487b2e better handling for enum field signatures 2022-08-20 18:28:41 +02:00
Techatrix
d57f8e3a64 add basic cImport support 2022-08-19 00:00:46 +02:00
Techatrix
29690f4e0b Refactor Ast helper functions 2022-08-18 00:52:21 +02:00