From 6bf2c9a1c7ebc5b1cc0e32bb83bd7affc89becae Mon Sep 17 00:00:00 2001
From: Alexandros Naskos <alex_naskos@hotmail.com>
Date: Fri, 3 Jul 2020 12:39:58 +0300
Subject: [PATCH] Added encoding offset aware documentRange

---
 src/main.zig    |  2 +-
 src/offsets.zig | 46 ++++++++++++++++++++++++++++++++++++++++++++++
 src/types.zig   | 21 ---------------------
 3 files changed, 47 insertions(+), 22 deletions(-)

diff --git a/src/main.zig b/src/main.zig
index 71cc132..3dd9c3b 100644
--- a/src/main.zig
+++ b/src/main.zig
@@ -1245,7 +1245,7 @@ fn formattingHandler(arena: *std.heap.ArenaAllocator, id: types.RequestId, req:
                     .result = .{
                         .TextEdits = &[1]types.TextEdit{
                             .{
-                                .range = handle.document.range(),
+                                .range = try offsets.documentRange(handle.document, offset_encoding),
                                 .newText = stdout_bytes,
                             },
                         },
diff --git a/src/offsets.zig b/src/offsets.zig
index 9a689b8..8f5ef6a 100644
--- a/src/offsets.zig
+++ b/src/offsets.zig
@@ -113,3 +113,49 @@ pub fn tokenLength(tree: *std.zig.ast.Tree, token: std.zig.ast.TokenIndex, encod
     }
     return utf16_len;
 }
+
+pub fn documentRange(doc: types.TextDocument, encoding: Encoding) !types.Range {
+    var line_idx: i64 = 0;
+    var curr_line: []const u8 = doc.text;
+
+    var split_iterator = std.mem.split(doc.text, "\n");
+    while (split_iterator.next()) |line| : (line_idx += 1) {
+        curr_line = line;
+    }
+
+    if (encoding == .utf8) {
+        return types.Range{
+            .start = .{
+                .line = 0,
+                .character = 0,
+            },
+            .end = .{
+                .line = line_idx,
+                .character = @intCast(i64, curr_line.len),
+            },
+        };
+    } else {
+        var utf16_len: usize = 0;
+        var line_utf8_idx: usize = 0;
+        while (line_utf8_idx < curr_line.len) {
+            const n = try std.unicode.utf8ByteSequenceLength(curr_line[line_utf8_idx]);
+            const codepoint = try std.unicode.utf8Decode(curr_line[line_utf8_idx .. line_utf8_idx + n]);
+            if (codepoint < 0x10000) {
+                utf16_len += 1;
+            } else {
+                utf16_len += 2;
+            }
+            line_utf8_idx += n;
+        }
+        return types.Range{
+            .start = .{
+                .line = 0,
+                .character = 0,
+            },
+            .end = .{
+                .line = line_idx,
+                .character = @intCast(i64, utf16_len),
+            },
+        };
+    }
+}
diff --git a/src/types.zig b/src/types.zig
index 3280fa3..f39ff1e 100644
--- a/src/types.zig
+++ b/src/types.zig
@@ -138,27 +138,6 @@ pub const TextDocument = struct {
     text: String,
     // This holds the memory that we have actually allocated.
     mem: []u8,
-
-    pub fn range(self: TextDocument) Range {
-        var line_idx: i64 = 0;
-        var curr_line: []const u8 = self.text;
-
-        var split_iterator = std.mem.split(self.text, "\n");
-        while (split_iterator.next()) |line| : (line_idx += 1) {
-            curr_line = line;
-        }
-
-        return .{
-            .start = .{
-                .line = 0,
-                .character = 0,
-            },
-            .end = .{
-                .line = line_idx,
-                .character = @intCast(i64, curr_line.len),
-            },
-        };
-    }
 };
 
 pub const WorkspaceEdit = struct {