Merge #7657

7657: utf8 r=matklad a=matklad

- Prepare for utf-8 offsets
- reduce code duplication in tests
- Make utf8 default, implement utf16 in terms of it
- Make it easy to add additional context for offset conversion
- Implement utf8 offsets

closes #7453

Co-authored-by: Aleksey Kladov <[email protected]>

2 files changed, 41 insertions, 21 deletions

diff --git a/crates/ide_db/src/line_index.rs b/crates/ide_db/src/line_index.rs
index 41226305e..8e9d8cca2 100644
--- a/crates/ide_db/src/line_index.rs
+++ b/crates/ide_db/src/line_index.rs
@@ -15,11 +15,19 @@ pub struct LineIndex {
 }
 
 #[derive(Clone, Copy, Debug, PartialEq, Eq, Hash)]
-pub struct LineCol {
+pub struct LineColUtf16 {
     /// Zero-based
     pub line: u32,
     /// Zero-based
-    pub col_utf16: u32,
+    pub col: u32,
+}
+
+#[derive(Clone, Copy, Debug, PartialEq, Eq, Hash)]
+pub struct LineCol {
+    /// Zero-based
+    pub line: u32,
+    /// Zero-based utf8 offset
+    pub col: u32,
 }
 
 #[derive(Clone, Debug, Hash, PartialEq, Eq)]
@@ -92,14 +100,21 @@ impl LineIndex {
         let line = partition_point(&self.newlines, |&it| it <= offset) - 1;
         let line_start_offset = self.newlines[line];
         let col = offset - line_start_offset;
-
-        LineCol { line: line as u32, col_utf16: self.utf8_to_utf16_col(line as u32, col) as u32 }
+        LineCol { line: line as u32, col: col.into() }
     }
 
     pub fn offset(&self, line_col: LineCol) -> TextSize {
-        //FIXME: return Result
-        let col = self.utf16_to_utf8_col(line_col.line, line_col.col_utf16);
-        self.newlines[line_col.line as usize] + col
+        self.newlines[line_col.line as usize] + TextSize::from(line_col.col)
+    }
+
+    pub fn to_utf16(&self, line_col: LineCol) -> LineColUtf16 {
+        let col = self.utf8_to_utf16_col(line_col.line, line_col.col.into());
+        LineColUtf16 { line: line_col.line, col: col as u32 }
+    }
+
+    pub fn to_utf8(&self, line_col: LineColUtf16) -> LineCol {
+        let col = self.utf16_to_utf8_col(line_col.line, line_col.col);
+        LineCol { line: line_col.line, col: col.into() }
     }
 
     pub fn lines(&self, range: TextRange) -> impl Iterator<Item = TextRange> + '_ {
diff --git a/crates/ide_db/src/line_index/tests.rs b/crates/ide_db/src/line_index/tests.rs
index 05f7484e8..09f3bca62 100644
--- a/crates/ide_db/src/line_index/tests.rs
+++ b/crates/ide_db/src/line_index/tests.rs
@@ -3,24 +3,29 @@ use super::*;
 #[test]
 fn test_line_index() {
     let text = "hello\nworld";
+    let table = [
+        (00, 0, 0),
+        (01, 0, 1),
+        (05, 0, 5),
+        (06, 1, 0),
+        (07, 1, 1),
+        (08, 1, 2),
+        (10, 1, 4),
+        (11, 1, 5),
+        (12, 1, 6),
+    ];
+
     let index = LineIndex::new(text);
-    assert_eq!(index.line_col(0.into()), LineCol { line: 0, col_utf16: 0 });
-    assert_eq!(index.line_col(1.into()), LineCol { line: 0, col_utf16: 1 });
-    assert_eq!(index.line_col(5.into()), LineCol { line: 0, col_utf16: 5 });
-    assert_eq!(index.line_col(6.into()), LineCol { line: 1, col_utf16: 0 });
-    assert_eq!(index.line_col(7.into()), LineCol { line: 1, col_utf16: 1 });
-    assert_eq!(index.line_col(8.into()), LineCol { line: 1, col_utf16: 2 });
-    assert_eq!(index.line_col(10.into()), LineCol { line: 1, col_utf16: 4 });
-    assert_eq!(index.line_col(11.into()), LineCol { line: 1, col_utf16: 5 });
-    assert_eq!(index.line_col(12.into()), LineCol { line: 1, col_utf16: 6 });
+    for &(offset, line, col) in &table {
+        assert_eq!(index.line_col(offset.into()), LineCol { line, col });
+    }
 
     let text = "\nhello\nworld";
+    let table = [(0, 0, 0), (1, 1, 0), (2, 1, 1), (6, 1, 5), (7, 2, 0)];
     let index = LineIndex::new(text);
-    assert_eq!(index.line_col(0.into()), LineCol { line: 0, col_utf16: 0 });
-    assert_eq!(index.line_col(1.into()), LineCol { line: 1, col_utf16: 0 });
-    assert_eq!(index.line_col(2.into()), LineCol { line: 1, col_utf16: 1 });
-    assert_eq!(index.line_col(6.into()), LineCol { line: 1, col_utf16: 5 });
-    assert_eq!(index.line_col(7.into()), LineCol { line: 2, col_utf16: 0 });
+    for &(offset, line, col) in &table {
+        assert_eq!(index.line_col(offset.into()), LineCol { line, col });
+    }
 }
 
 #[test]