get rid of bounded array usage for wcwidth prev grapheme (#338)

Author: kommander

packages/core/src/edit-buffer.test.ts

@@ -1652,4 +1652,21 @@ describe("EditBuffer Clear Method", () => {
       expect(() => buffer.clear()).toThrow("EditBuffer is destroyed")
     })
   })
+
+  describe("Regression Tests", () => {
+    it("should handle moving left in a long line (potential BoundedArray overflow)", () => {
+      // Create a string longer than 256 chars (the size of BoundedArray in getPrevGraphemeStartWCWidth)
+      const longText = "a".repeat(500)
+      buffer.setText(longText)
+
+      // Move cursor to the end (or near the end)
+      buffer.setCursorToLineCol(0, 500)
+
+      // Move left should not crash
+      buffer.moveCursorLeft()
+
+      const cursor = buffer.getCursorPosition()
+      expect(cursor.col).toBe(499)
+    })
+  })
 })

packages/core/src/zig/edit-buffer.zig

@@ -409,7 +409,9 @@ pub const EditBuffer = struct {
     }
 
     pub fn moveLeft(self: *EditBuffer) void {
-        if (self.cursors.items.len == 0) return;
+        if (self.cursors.items.len == 0) {
+            return;
+        }
         const cursor = &self.cursors.items[0];
 
         if (cursor.col > 0) {

packages/core/src/zig/text-buffer-iterators.zig

@@ -289,13 +289,16 @@ pub fn getPrevGraphemeWidth(rope: *UnifiedRope, mem_registry: *const MemRegistry
                     const pc = prev_chunk.?;
                     const bytes = pc.chunk.getBytes(mem_registry);
                     const prev = utf8.getPrevGraphemeStart(bytes, bytes.len, tab_width, width_method);
-                    if (prev) |res| return res.width;
+                    if (prev) |res| {
+                        return res.width;
+                    }
                     return 0;
                 }
 
                 const bytes = chunk.getBytes(mem_registry);
                 const is_ascii = (chunk.flags & TextChunk.Flags.ASCII_ONLY) != 0;
                 const local_col: u32 = clamped_col - cols_before;
+
                 const here = utf8.findPosByWidth(bytes, local_col, tab_width, is_ascii, false, width_method);
 
                 const grapheme_start_col = here.columns_used;
@@ -325,7 +328,9 @@ pub fn getPrevGraphemeWidth(rope: *UnifiedRope, mem_registry: *const MemRegistry
                 }
 
                 const prev = utf8.getPrevGraphemeStart(bytes, @intCast(here.byte_offset), tab_width, width_method);
-                if (prev) |res| return res.width;
+                if (prev) |res| {
+                    return res.width;
+                }
                 return 0;
             }
 

packages/core/src/zig/utf8.zig

@@ -1499,10 +1499,9 @@ fn getPrevGraphemeStartWCWidth(text: []const u8, byte_offset: usize, tab_width:
     if (byte_offset == 0 or text.len == 0) return null;
     if (byte_offset > text.len) return null;
 
-    // Build a list of all codepoint positions
-    var codepoint_positions = std.BoundedArray(struct { pos: usize, width: u32 }, 256).init(0) catch unreachable;
-
     var pos: usize = 0;
+    var last_result: ?PrevGraphemeResult = null;
+
     while (pos < byte_offset) {
         const b = text[pos];
         const curr_cp: u21 = if (b < 0x80) b else blk: {
@@ -1513,23 +1512,16 @@ fn getPrevGraphemeStartWCWidth(text: []const u8, byte_offset: usize, tab_width:
         const cp_len: usize = if (b < 0x80) 1 else decodeUtf8Unchecked(text, pos).len;
         const cp_width = charWidth(b, curr_cp, tab_width);
 
-        codepoint_positions.appendAssumeCapacity(.{ .pos = pos, .width = cp_width });
-        pos += cp_len;
-    }
-
-    // Find the last non-zero-width codepoint before byte_offset
-    var i: isize = @as(isize, @intCast(codepoint_positions.len)) - 1;
-    while (i >= 0) : (i -= 1) {
-        const idx = @as(usize, @intCast(i));
-        if (codepoint_positions.get(idx).width > 0) {
-            return .{
-                .start_offset = codepoint_positions.get(idx).pos,
-                .width = codepoint_positions.get(idx).width,
+        if (cp_width > 0) {
+            last_result = .{
+                .start_offset = pos,
+                .width = cp_width,
             };
         }
+        pos += cp_len;
     }
 
-    return null;
+    return last_result;
 }
 
 /// Get previous grapheme start using Unicode grapheme cluster segmentation