From 1ad0779a007ed4a30d9e2443ab2e10d2ba041a1f Mon Sep 17 00:00:00 2001
From: Ariel Davis <ariel.z.davis@icloud.com>
Date: Fri, 5 May 2023 17:35:20 -0700
Subject: [PATCH] Make WideEncoding non-exhaustive

---
 crates/ide-db/src/tests/line_index.rs            |  9 ++-------
 crates/rust-analyzer/src/caps.rs                 | 11 ++++++-----
 crates/rust-analyzer/src/diagnostics/to_proto.rs | 12 +++++-------
 lib/line-index/src/lib.rs                        | 11 +++++++++++
 4 files changed, 24 insertions(+), 19 deletions(-)

diff --git a/crates/ide-db/src/tests/line_index.rs b/crates/ide-db/src/tests/line_index.rs
index c12936071d3..c41b0de5633 100644
--- a/crates/ide-db/src/tests/line_index.rs
+++ b/crates/ide-db/src/tests/line_index.rs
@@ -28,16 +28,11 @@ fn test_every_chars() {
         let got_lin_col = line_index.line_col(got_offset);
         assert_eq!(got_lin_col, lin_col);
 
-        for enc in [WideEncoding::Utf16, WideEncoding::Utf32] {
+        for (enc, col) in [(WideEncoding::Utf16, col_utf16), (WideEncoding::Utf32, col_utf32)] {
             let wide_lin_col = line_index.to_wide(enc, lin_col);
             let got_lin_col = line_index.to_utf8(enc, wide_lin_col);
             assert_eq!(got_lin_col, lin_col);
-
-            let want_col = match enc {
-                WideEncoding::Utf16 => col_utf16,
-                WideEncoding::Utf32 => col_utf32,
-            };
-            assert_eq!(wide_lin_col.col, want_col)
+            assert_eq!(wide_lin_col.col, col)
         }
 
         if c == '\n' {
diff --git a/crates/rust-analyzer/src/caps.rs b/crates/rust-analyzer/src/caps.rs
index 3628670ac98..ab06b96814a 100644
--- a/crates/rust-analyzer/src/caps.rs
+++ b/crates/rust-analyzer/src/caps.rs
@@ -23,13 +23,14 @@ use crate::semantic_tokens;
 
 pub fn server_capabilities(config: &Config) -> ServerCapabilities {
     ServerCapabilities {
-        position_encoding: Some(match negotiated_encoding(config.caps()) {
-            PositionEncoding::Utf8 => PositionEncodingKind::UTF8,
+        position_encoding: match negotiated_encoding(config.caps()) {
+            PositionEncoding::Utf8 => Some(PositionEncodingKind::UTF8),
             PositionEncoding::Wide(wide) => match wide {
-                WideEncoding::Utf16 => PositionEncodingKind::UTF16,
-                WideEncoding::Utf32 => PositionEncodingKind::UTF32,
+                WideEncoding::Utf16 => Some(PositionEncodingKind::UTF16),
+                WideEncoding::Utf32 => Some(PositionEncodingKind::UTF32),
+                _ => None,
             },
-        }),
+        },
         text_document_sync: Some(TextDocumentSyncCapability::Options(TextDocumentSyncOptions {
             open_close: Some(true),
             change: Some(TextDocumentSyncKind::INCREMENTAL),
diff --git a/crates/rust-analyzer/src/diagnostics/to_proto.rs b/crates/rust-analyzer/src/diagnostics/to_proto.rs
index 415fa4e02f2..c8b2c4edb83 100644
--- a/crates/rust-analyzer/src/diagnostics/to_proto.rs
+++ b/crates/rust-analyzer/src/diagnostics/to_proto.rs
@@ -3,7 +3,6 @@
 use std::collections::HashMap;
 
 use flycheck::{Applicability, DiagnosticLevel, DiagnosticSpan};
-use ide_db::line_index::WideEncoding;
 use itertools::Itertools;
 use stdx::format_to;
 use vfs::{AbsPath, AbsPathBuf};
@@ -94,17 +93,16 @@ fn position(
             };
         }
         let mut char_offset = 0;
-        let len_func = match position_encoding {
-            PositionEncoding::Utf8 => char::len_utf8,
-            PositionEncoding::Wide(WideEncoding::Utf16) => char::len_utf16,
-            PositionEncoding::Wide(WideEncoding::Utf32) => |_| 1,
-        };
         for c in line.text.chars() {
             char_offset += 1;
             if char_offset > column_offset {
                 break;
             }
-            true_column_offset += len_func(c) - 1;
+            let len = match position_encoding {
+                PositionEncoding::Utf8 => c.len_utf8(),
+                PositionEncoding::Wide(w) => w.measure(&c.to_string()),
+            };
+            true_column_offset += len - 1;
         }
     }
 
diff --git a/lib/line-index/src/lib.rs b/lib/line-index/src/lib.rs
index 61f907f0604..40815bdcf19 100644
--- a/lib/line-index/src/lib.rs
+++ b/lib/line-index/src/lib.rs
@@ -20,6 +20,7 @@ pub struct LineCol {
 
 /// A kind of wide character encoding.
 #[derive(Clone, Copy, Debug, PartialEq, Eq, Hash)]
+#[non_exhaustive]
 pub enum WideEncoding {
     /// UTF-16.
     Utf16,
@@ -27,6 +28,16 @@ pub enum WideEncoding {
     Utf32,
 }
 
+impl WideEncoding {
+    /// Returns the number of units it takes to encode `text` in this encoding.
+    pub fn measure(&self, text: &str) -> usize {
+        match self {
+            WideEncoding::Utf16 => text.encode_utf16().count(),
+            WideEncoding::Utf32 => text.chars().count(),
+        }
+    }
+}
+
 /// Line/Column information in legacy encodings.
 //
 // Deliberately not a generic type and different from `LineCol`.