Merge remote-tracking branch 'origin/master' into debug

1 files changed, 86 insertions, 0 deletions

diff --git a/helix-core/src/graphemes.rs b/helix-core/src/graphemes.rs
index c6398875..aa898684 100644
--- a/helix-core/src/graphemes.rs
+++ b/helix-core/src/graphemes.rs
@@ -120,6 +120,43 @@ pub fn nth_next_grapheme_boundary(slice: RopeSlice, char_idx: usize, n: usize) -
     chunk_char_idx + tmp
 }
 
+#[must_use]
+pub fn nth_next_grapheme_boundary_byte(slice: RopeSlice, mut byte_idx: usize, n: usize) -> usize {
+    // Bounds check
+    debug_assert!(byte_idx <= slice.len_bytes());
+
+    // Get the chunk with our byte index in it.
+    let (mut chunk, mut chunk_byte_idx, mut _chunk_char_idx, _) = slice.chunk_at_byte(byte_idx);
+
+    // Set up the grapheme cursor.
+    let mut gc = GraphemeCursor::new(byte_idx, slice.len_bytes(), true);
+
+    // Find the nth next grapheme cluster boundary.
+    for _ in 0..n {
+        loop {
+            match gc.next_boundary(chunk, chunk_byte_idx) {
+                Ok(None) => return slice.len_bytes(),
+                Ok(Some(n)) => {
+                    byte_idx = n;
+                    break;
+                }
+                Err(GraphemeIncomplete::NextChunk) => {
+                    chunk_byte_idx += chunk.len();
+                    let (a, _, _c, _) = slice.chunk_at_byte(chunk_byte_idx);
+                    chunk = a;
+                    // chunk_char_idx = c;
+                }
+                Err(GraphemeIncomplete::PreContext(n)) => {
+                    let ctx_chunk = slice.chunk_at_byte(n - 1).0;
+                    gc.provide_context(ctx_chunk, n - ctx_chunk.len());
+                }
+                _ => unreachable!(),
+            }
+        }
+    }
+    byte_idx
+}
+
 /// Finds the next grapheme boundary after the given char position.
 #[must_use]
 #[inline(always)]
@@ -127,6 +164,13 @@ pub fn next_grapheme_boundary(slice: RopeSlice, char_idx: usize) -> usize {
     nth_next_grapheme_boundary(slice, char_idx, 1)
 }
 
+/// Finds the next grapheme boundary after the given byte position.
+#[must_use]
+#[inline(always)]
+pub fn next_grapheme_boundary_byte(slice: RopeSlice, byte_idx: usize) -> usize {
+    nth_next_grapheme_boundary_byte(slice, byte_idx, 1)
+}
+
 /// Returns the passed char index if it's already a grapheme boundary,
 /// or the next grapheme boundary char index if not.
 #[must_use]
@@ -151,6 +195,23 @@ pub fn ensure_grapheme_boundary_prev(slice: RopeSlice, char_idx: usize) -> usize
     }
 }
 
+/// Returns the passed byte index if it's already a grapheme boundary,
+/// or the next grapheme boundary byte index if not.
+#[must_use]
+#[inline]
+pub fn ensure_grapheme_boundary_next_byte(slice: RopeSlice, byte_idx: usize) -> usize {
+    if byte_idx == 0 {
+        byte_idx
+    } else {
+        // TODO: optimize so we're not constructing grapheme cursor twice
+        if is_grapheme_boundary_byte(slice, byte_idx) {
+            byte_idx
+        } else {
+            next_grapheme_boundary_byte(slice, byte_idx)
+        }
+    }
+}
+
 /// Returns whether the given char position is a grapheme boundary.
 #[must_use]
 pub fn is_grapheme_boundary(slice: RopeSlice, char_idx: usize) -> bool {
@@ -179,6 +240,31 @@ pub fn is_grapheme_boundary(slice: RopeSlice, char_idx: usize) -> bool {
     }
 }
 
+/// Returns whether the given byte position is a grapheme boundary.
+#[must_use]
+pub fn is_grapheme_boundary_byte(slice: RopeSlice, byte_idx: usize) -> bool {
+    // Bounds check
+    debug_assert!(byte_idx <= slice.len_bytes());
+
+    // Get the chunk with our byte index in it.
+    let (chunk, chunk_byte_idx, _, _) = slice.chunk_at_byte(byte_idx);
+
+    // Set up the grapheme cursor.
+    let mut gc = GraphemeCursor::new(byte_idx, slice.len_bytes(), true);
+
+    // Determine if the given position is a grapheme cluster boundary.
+    loop {
+        match gc.is_boundary(chunk, chunk_byte_idx) {
+            Ok(n) => return n,
+            Err(GraphemeIncomplete::PreContext(n)) => {
+                let (ctx_chunk, ctx_byte_start, _, _) = slice.chunk_at_byte(n - 1);
+                gc.provide_context(ctx_chunk, ctx_byte_start);
+            }
+            Err(_) => unreachable!(),
+        }
+    }
+}
+
 /// An iterator over the graphemes of a `RopeSlice`.
 #[derive(Clone)]
 pub struct RopeGraphemes<'a> {