Whenever a document is changed helix maps various positions like the
cursor or diagnostics through the `ChangeSet` applied to the document.
Currently, this mapping handles replacements as follows:
* Move position to the left for `Assoc::Before` (start of selection)
* Move position to the right for `Assoc::After` (end of selection)
However, when text is exactly replaced this can produce weird results
where the cursor is moved when it shouldn't. For example if `foo` is
selected and a separate cursor is placed on each character (`s.<ret>`)
and the text is replaced (for example `rx`) then the cursors are moved
to the side instead of remaining in place.
This change adds a special case to the mapping code of replacements:
If the deleted and inserted text have the same (char) length then
the position is returned as if the replacement doesn't exist.
only keep selections invariant under replacement
Keeping selections unchanged if they are inside an exact replacement
is intuitive. However, for diagnostics this is not desirable as
helix would otherwise fail to remove diagnostics if replacing parts
of the document.
This is a temporary limitation because of the way that command sequences
are executed. Each command is currently executed back-to-back
synchronously, but macros are by design queued up for the compositor.
So macros mixed into a command sequence will behave undesirably: they
will be executed after the rest of the static and/or typable commands
in the sequence.
This is pending a larger refactor of how we handle commands.
<https://redirect.github.com/helix-editor/helix/issues/5555> has
further details and <https://redirect.github.com/helix-editor/helix/issues/4508>
discusses a similar problem faced by the command palette.
The repository reference used here was a fork from the actual
repository, which has now been moved under ionide organization, where it is
in active maintenance and development.
The commit SHA is the currently latest commit from main branch.
The injections.scm is copied as is from the fsharp treesitter repo
[queries](https://github.com/ionide/tree-sitter-fsharp/blame/main/queries).
The locals.scm is copied from the repo and the capture names are to follow
the standard names:
- Replace @local.definition.var @local.definition.function, and @local.definition.parameter with @local.definition
- Remove (#set! "definition.function.scope" "parent")
The highlights.scm is copied as well from the fsharp
treesitter repo, but modified here to match helix highlight scopes based
on my best guesstimates. The changes made:
- Remove @spell scopes
- Split @comment into @comment.line and @comment.block
- Replace @comment.documentation with @comment.block.documentation
- Replace @character.special with @special
- Replace @variable.member with @variable.other.member
- Replace @type.definition with @type
- Replace @function.member with @function.method
- Replace @module with @namespace
- Replace @constant.macro with @function.macro
- Replace @property with @variable.other.member
- Replace @variable.member with @variable.other.member
- Replace @variable.parameter.builtin with @variable.builtin
- Replace @function.call with @function
- Replace @number with @constant.numeric.integer and @constant.numeric.float
- Replace @boolean with @constant.builtin.boolean
- Replace @keyword.conditional with @keyword.control.conditional
- Replace @keyword.return with @keyword.control.return
- Replace @keyword.repeate with @keyword.control.repeat
- Replace @keyword.import with @keyword.control.import
- Replace @keyword.modifier with @keyword.storage.modifier
- Replace @keyword.type with @keyword.storage.type
- Replace @keyword.exception with @keyword.control.exception
- Replace @module.builtin with @namespace
* Use `try_lock` in `diff_handle` for Diff gutter
- Add the method `try_load() -> Option<Diff>` to `DiffHandle`, using `try_lock` to avoid deadlocks.
- Use said method in `gutter.rs diff()`, which will use a blank `GutterFn` instead when met with a locked `Diff`.
* Revert changes
* Replace `Mutex` with `RwLock` in `Diff`
---------
Co-authored-by: Kaniel Kirby <pirate7007@runbox.com>
* Add TypeSpec support
Adds support for TypeSpec <https://typespec.io> in helix.
* Resolve PR comments
* Pull in LICENSE
Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
---------
Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
* refactor(commands): output `stderr` in `:sh` popup
* refactor(commands): switch to `from_utf8_lossy`
This way something is always displayed.
* refactor: no longer log stderr output on failure
The primary executable that comes with Tcl is `tclsh`. Not really sure what `tclish` is, as I initially thought it was a typo. However, there seems to be references to it based on a quick search (e.g. [here](https://wiki.tcl-lang.org/page/Tclish) and [here](https://tclish.sourceforge.net/)), so maybe it's a valid executable that I just haven't been aware of. I was hesitant to replace it and instead opted to just add `tclsh`.
The `take_while` should limit the layers to those that can match the
input range so we don't always scan the entire `injection_layers`. We
can limit `depth == 1` layers to those that start before the search
`end`. Deeper layers overlap with shallower layers though so we need
to allow those layers as well in the `take_while`.
For example
```vue
<script setup lang="ts">
const foo = 'bar'.match(/foo/);
const bar = foo;
</script>
```
L2 and L3 are a typescript layer and the `/foo/` part is a small regex
layer. If you used `A-o` before the regex layer you would select the
entire typescript layer. The search in `layer_id_containing_byte_range`
would not consider the typescript layer since the regex layer comes
earlier in `injection_ranges` and that layer's start is after `end`.
The regex layer has a depth of `2` though so the change in this commit
allows scanning through that layer.
Co-authored-by: Pascal Kuthe <pascalkuthe@pm.me>
* feat: pertty header
* 更新 themes.md
Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
---------
Co-authored-by: Michael Davis <mcarsondavis@gmail.com>
Closes#11351
Also fixed some minor issues related to log
message contents, and removed unnecessary use
of `.as_mut()` as per code review comments on
the PR.
