Merge 4d64d670dd into 7ebf650029

Cargo.lock

@@ -1345,12 +1345,11 @@ dependencies = [
  "slotmap",
  "smallvec",
  "smartstring",
- "textwrap",
  "toml",
  "tree-sitter",
  "unicode-general-category",
  "unicode-segmentation",
- "unicode-width 0.1.12",
+ "unicode-width",
  "url",
 ]
 
@@ -2427,12 +2426,6 @@ dependencies = [
  "version_check",
 ]
 
-[[package]]
-name = "smawk"
-version = "0.3.2"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "b7c388c1b5e93756d0c740965c41e8822f866621d41acbdf6336a6a168f8840c"
-
 [[package]]
 name = "socket2"
 version = "0.5.7"
@@ -2505,17 +2498,6 @@ dependencies = [
  "home",
 ]
 
-[[package]]
-name = "textwrap"
-version = "0.16.2"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "c13547615a44dc9c452a8a534638acdf07120d4b6847c8178705da06306a3057"
-dependencies = [
- "smawk",
- "unicode-linebreak",
- "unicode-width 0.2.0",
-]
-
 [[package]]
 name = "thiserror"
 version = "1.0.69"
@@ -2701,12 +2683,6 @@ version = "1.0.12"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "3354b9ac3fae1ff6755cb6db53683adb661634f67557942dea4facebec0fee4b"
 
-[[package]]
-name = "unicode-linebreak"
-version = "0.1.5"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "3b09c83c3c29d37506a3e260c08c03743a6bb66a9cd432c6934ab501a190571f"
-
 [[package]]
 name = "unicode-normalization"
 version = "0.1.23"
@@ -2728,12 +2704,6 @@ version = "0.1.12"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "68f5e5f3158ecfd4b8ff6fe086db7c8467a2dfdac97fe420f2b7c4aa97af66d6"
 
-[[package]]
-name = "unicode-width"
-version = "0.2.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "1fc81956842c57dac11422a97c3b8195a1ff727f06e85c84ed2e8aa277c9a0fd"
-
 [[package]]
 name = "url"
 version = "2.5.4"

helix-core/Cargo.toml

@@ -53,8 +53,6 @@ encoding_rs = "0.8"
 
 chrono = { version = "0.4", default-features = false, features = ["alloc", "std"] }
 
-textwrap = "0.16.2"
-
 nucleo.workspace = true
 parking_lot.workspace = true
 globset = "0.4.16"

helix-core/src/doc_formatter.rs

@@ -24,7 +24,7 @@ use helix_stdx::rope::{RopeGraphemes, RopeSliceExt};
 use crate::graphemes::{Grapheme, GraphemeStr};
 use crate::syntax::Highlight;
 use crate::text_annotations::TextAnnotations;
-use crate::{Position, RopeSlice};
+use crate::{movement, Change, LineEnding, Position, RopeSlice, Tendril};
 
 /// TODO make Highlight a u32 to reduce the size of this enum to a single word.
 #[derive(Debug, Clone, Copy)]
@@ -478,3 +478,91 @@ impl<'t> Iterator for DocumentFormatter<'t> {
         Some(grapheme)
     }
 }
+
+pub struct ReflowOpts<'a> {
+    pub width: usize,
+    pub line_ending: LineEnding,
+    pub comment_tokens: &'a [String],
+}
+
+impl ReflowOpts<'_> {
+    fn find_indent<'a>(&self, line: usize, doc: RopeSlice<'a>) -> RopeSlice<'a> {
+        let line_start = doc.line_to_char(line);
+        let mut indent_end = movement::skip_while(doc, line_start, |ch| matches!(ch, ' ' | '\t'))
+            .unwrap_or(line_start);
+        let slice = doc.slice(indent_end..);
+        if let Some(token) = self
+            .comment_tokens
+            .iter()
+            .filter(|token| slice.starts_with(token))
+            .max_by_key(|x| x.len())
+        {
+            indent_end += token.chars().count();
+        }
+        let indent_end = movement::skip_while(doc, indent_end, |ch| matches!(ch, ' ' | '\t'))
+            .unwrap_or(indent_end);
+        return doc.slice(line_start..indent_end);
+    }
+}
+
+/// reflow wraps long lines in text to be less than opts.width.
+pub fn reflow(text: RopeSlice, char_pos: usize, opts: &ReflowOpts) -> Vec<Change> {
+    // A constant so that reflow behaves consistently across
+    // different configurations.
+    const TAB_WIDTH: u16 = 8;
+
+    let line_idx = text.char_to_line(char_pos.min(text.len_chars()));
+    let mut char_pos = text.line_to_char(line_idx);
+
+    let mut col = 0;
+    let mut word_width = 0;
+    let mut last_word_boundary = None;
+    let mut changes = Vec::new();
+    for grapheme in text.slice(char_pos..).graphemes() {
+        let grapheme_chars = grapheme.len_chars();
+        let mut grapheme = Grapheme::new(GraphemeStr::from(Cow::from(grapheme)), col, TAB_WIDTH);
+        if col + grapheme.width() > opts.width && !grapheme.is_whitespace() {
+            if let Some(n) = last_word_boundary {
+                let indent = opts.find_indent(text.char_to_line(n - 1), text);
+                let mut whitespace_start = n;
+                let mut whitespace_end = n;
+                while whitespace_start > 0 && text.char(whitespace_start - 1) == ' ' {
+                    whitespace_start -= 1;
+                }
+                while whitespace_end < text.chars().len() && text.char(whitespace_end) == ' ' {
+                    whitespace_end += 1;
+                }
+                changes.push((
+                    whitespace_start,
+                    whitespace_end,
+                    Some(Tendril::from(format!(
+                        "{}{}",
+                        opts.line_ending.as_str(),
+                        &indent
+                    ))),
+                ));
+
+                col = 0;
+                for g in indent.graphemes() {
+                    let g = Grapheme::new(GraphemeStr::from(Cow::from(g)), col, TAB_WIDTH);
+                    col += g.width();
+                }
+                col += word_width;
+                last_word_boundary = None;
+                grapheme.change_position(col, TAB_WIDTH);
+            }
+        }
+        col += grapheme.width();
+        word_width += grapheme.width();
+        if grapheme == Grapheme::Newline {
+            col = 0;
+            word_width = 0;
+            last_word_boundary = None;
+        } else if grapheme.is_whitespace() {
+            last_word_boundary = Some(char_pos);
+            word_width = 0;
+        }
+        char_pos += grapheme_chars;
+    }
+    changes
+}

helix-core/src/lib.rs

@@ -32,7 +32,6 @@ pub mod text_annotations;
 pub mod textobject;
 mod transaction;
 pub mod uri;
-pub mod wrap;
 
 pub mod unicode {
     pub use unicode_general_category as category;

helix-core/src/wrap.rs

@@ -1,11 +0,0 @@
-use smartstring::{LazyCompact, SmartString};
-use textwrap::{Options, WordSplitter::NoHyphenation};
-
-/// Given a slice of text, return the text re-wrapped to fit it
-/// within the given width.
-pub fn reflow_hard_wrap(text: &str, text_width: usize) -> SmartString<LazyCompact> {
-    let options = Options::new(text_width)
-        .word_splitter(NoHyphenation)
-        .word_separator(textwrap::WordSeparator::AsciiSpace);
-    textwrap::refill(text, options).into()
-}

helix-term/src/commands/typed.rs

@@ -7,6 +7,7 @@ use crate::job::Job;
 use super::*;
 
 use helix_core::command_line::{Args, Flag, Signature, Token, TokenKind};
+use helix_core::doc_formatter::ReflowOpts;
 use helix_core::fuzzy::fuzzy_match;
 use helix_core::indent::MAX_INDENT;
 use helix_core::line_ending;
@@ -2159,14 +2160,24 @@ fn reflow(cx: &mut compositor::Context, args: Args, event: PromptEvent) -> anyho
         .unwrap_or_else(|| doc.text_width());
 
     let rope = doc.text();
+    let opts = ReflowOpts {
+        width: text_width,
+        line_ending: doc.line_ending,
+        comment_tokens: doc
+            .language_config()
+            .and_then(|config| config.comment_tokens.as_deref())
+            .unwrap_or(&[]),
+    };
 
-    let selection = doc.selection(view.id);
-    let transaction = Transaction::change_by_selection(rope, selection, |range| {
-        let fragment = range.fragment(rope.slice(..));
-        let reflowed_text = helix_core::wrap::reflow_hard_wrap(&fragment, text_width);
-
-        (range.from(), range.to(), Some(reflowed_text))
-    });
+    let mut changes = Vec::new();
+    for selection in doc.selection(view.id) {
+        changes.append(&mut helix_core::doc_formatter::reflow(
+            rope.slice(..selection.to()),
+            selection.from(),
+            &opts,
+        ));
+    }
+    let transaction = Transaction::change(rope, changes.into_iter());
 
     doc.apply(&transaction, view.id);
     doc.append_changes_to_history(view);

helix-term/tests/test/commands.rs

@@ -820,3 +820,112 @@ async fn macro_play_within_macro_record() -> anyhow::Result<()> {
 
     Ok(())
 }
+
+#[tokio::test(flavor = "multi_thread")]
+async fn test_reflow() -> anyhow::Result<()> {
+    test((
+        "#[|This is a long line bla bla bla]#",
+        ":reflow 5<ret>",
+        "#[|This
+is a
+long
+line
+bla
+bla
+bla]#",
+    ))
+    .await?;
+
+    test((
+        "// #[|This is a really long comment that we want to break onto multiple lines.]#",
+        ":lang rust<ret>:reflow 13<ret>",
+        "// #[|This is a
+// really
+// long
+// comment
+// that we
+// want to
+// break onto
+// multiple
+// lines.]#",
+    ))
+    .await?;
+
+    test((
+        "#[\t// Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod
+\t// tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim
+\t// veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea
+\t// commodo consequat. Duis aute irure dolor in reprehenderit in voluptate
+\t// velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint
+\t// occaecat cupidatat non proident, sunt in culpa qui officia deserunt
+\t// mollit anim id est laborum.
+|]#",
+        ":lang go<ret>:reflow 50<ret>",
+        "#[\t// Lorem ipsum dolor sit amet, consectetur
+\t// adipiscing elit, sed do eiusmod
+\t// tempor incididunt ut labore et dolore
+\t// magna aliqua. Ut enim ad minim
+\t// veniam, quis nostrud exercitation
+\t// ullamco laboris nisi ut aliquip ex ea
+\t// commodo consequat. Duis aute irure
+\t// dolor in reprehenderit in voluptate
+\t// velit esse cillum dolore eu fugiat
+\t// nulla pariatur. Excepteur sint
+\t// occaecat cupidatat non proident, sunt
+\t// in culpa qui officia deserunt
+\t// mollit anim id est laborum.
+|]#",
+    ))
+    .await?;
+
+    test((
+        "    // #[|This document has multiple lines that each need wrapping
+
+    /// currently we wrap each line completely separately in order to preserve existing newlines.]#",
+        ":lang rust<ret>:reflow 40<ret>",
+        "    // #[|This document has multiple lines
+    // that each need wrapping
+
+    /// currently we wrap each line
+    /// completely separately in order
+    /// to preserve existing newlines.]#"
+    ))
+    .await?;
+
+    test((
+        "#[|Very-long-words-should-not-be-broken-by-hard-wrap]#",
+        ":reflow 2<ret>",
+        "#[|Very-long-words-should-not-be-broken-by-hard-wrap]#",
+    ))
+    .await?;
+
+    test((
+        "#[|Spaces             are           removed           when          wrapping]#",
+        ":reflow 2<ret>",
+        "#[|Spaces
+are
+removed
+when
+wrapping]#",
+    ))
+    .await?;
+
+    test((
+        "Test wrapping only part of the text
+#[|wrapping should only modify the lines that are currently selected
+]#",
+        ":reflow 11<ret>",
+        "Test wrapping only part of the text
+#[|wrapping
+should only
+modify the
+lines that
+are
+currently
+selected
+]#",
+    ))
+    .await?;
+
+    Ok(())
+}