fix: typo start position corrected for multiple code point unicode

resolves #22

crates/typos-lsp/Cargo.toml

@@ -19,6 +19,7 @@ serde = { version = "1.0", features = ["derive"] }
 ignore = "0.4.20"
 matchit = "0.7.1"
 shellexpand = "3.1.0"
+unicode-segmentation = "1.10.1"
 
 [dev-dependencies]
 test-log = { version = "0.2.11", features = ["trace"] }

crates/typos-lsp/src/lsp.rs

@@ -458,8 +458,13 @@ impl AccumulatePosition {
             .map(|s| s + 1)
             .unwrap_or(0);
 
+        let before_typo = String::from_utf8_lossy(&buffer[line_start..byte_offset]);
+        let line_pos =
+            unicode_segmentation::UnicodeSegmentation::graphemes(before_typo.as_ref(), true)
+                .count();
+
         self.line_num = line_num;
-        self.line_pos = byte_offset - line_start;
+        self.line_pos = line_pos;
         self.last_offset = byte_offset;
 
         (self.line_num, self.line_pos)

crates/typos-lsp/tests/integration_test.rs

@@ -336,8 +336,6 @@ async fn test_custom_config_file_e2e() {
         workspace_folder_uri,
     );
 
-    println!("{}", initialize);
-
     let did_open_diag_txt = format!(
         r#"{{
             "jsonrpc": "2.0",
@@ -382,6 +380,60 @@ async fn test_custom_config_file_e2e() {
     );
 }
 
+#[test_log::test(tokio::test)]
+async fn test_unicode_diagnostics() {
+    let initialize = r#"{
+        "jsonrpc": "2.0",
+        "method": "initialize",
+        "params": {
+          "initializationOptions": {
+            "diagnosticSeverity": "Warning"
+          },
+          "capabilities": {
+            "textDocument": { "publishDiagnostics": { "dataSupport": true } }
+          }
+        },
+        "id": 1
+      }
+    "#;
+
+    let did_open = r#"{
+        "jsonrpc": "2.0",
+        "method": "textDocument/didOpen",
+        "params": {
+          "textDocument": {
+            "uri": "file:///diagnostics.txt",
+            "languageId": "plaintext",
+            "version": 1,
+            "text": "¿Qué hace él?"
+          }
+        }
+      }
+    "#;
+
+    let (mut req_client, mut resp_client) = start_server();
+    let mut buf = vec![0; 1024];
+
+    req_client
+        .write_all(req(initialize).as_bytes())
+        .await
+        .unwrap();
+    let _ = resp_client.read(&mut buf).await.unwrap();
+
+    tracing::debug!("{}", did_open);
+    req_client
+        .write_all(req(did_open).as_bytes())
+        .await
+        .unwrap();
+    let n = resp_client.read(&mut buf).await.unwrap();
+
+    // start position should count graphemes with multiple code points as one visible character
+    similar_asserts::assert_eq!(
+        body(&buf[..n]).unwrap(),
+        r#"{"jsonrpc":"2.0","method":"textDocument/publishDiagnostics","params":{"diagnostics":[{"data":{"corrections":["have"]},"message":"`hace` should be `have`","range":{"end":{"character":9,"line":0},"start":{"character":5,"line":0}},"severity":2,"source":"typos"}],"uri":"file:///diagnostics.txt","version":1}}"#,
+    );
+}
+
 fn start_server() -> (tokio::io::DuplexStream, tokio::io::DuplexStream) {
     let (req_client, req_server) = tokio::io::duplex(1024);
     let (resp_server, resp_client) = tokio::io::duplex(1024);