From fc7ed1bfe4110a45c03c5b1b64697f92c6ef7fe6 Mon Sep 17 00:00:00 2001
From: JT <547158+jntrnr@users.noreply.github.com>
Date: Tue, 21 Dec 2021 11:49:02 +1100
Subject: [PATCH] switch substring to bytes (#538)

* switch substring to bytes

* Add a test
---
 crates/nu-command/src/strings/str_/substring.rs | 16 +++++++++++-----
 src/tests.rs                                    |  8 ++++++++
 2 files changed, 19 insertions(+), 5 deletions(-)

diff --git a/crates/nu-command/src/strings/str_/substring.rs b/crates/nu-command/src/strings/str_/substring.rs
index 9e06840ab7..ec5df7ef43 100644
--- a/crates/nu-command/src/strings/str_/substring.rs
+++ b/crates/nu-command/src/strings/str_/substring.rs
@@ -157,12 +157,18 @@ fn action(input: &Value, options: &Substring, head: Span) -> Value {
                     Ordering::Less => Value::String {
                         val: {
                             if end == isize::max_value() {
-                                s.chars().skip(start as usize).collect::<String>()
+                                String::from_utf8_lossy(
+                                    &s.bytes().skip(start as usize).collect::<Vec<_>>(),
+                                )
+                                .to_string()
                             } else {
-                                s.chars()
-                                    .skip(start as usize)
-                                    .take((end - start) as usize)
-                                    .collect::<String>()
+                                String::from_utf8_lossy(
+                                    &s.bytes()
+                                        .skip(start as usize)
+                                        .take((end - start) as usize)
+                                        .collect::<Vec<_>>(),
+                                )
+                                .to_string()
                             }
                         },
                         span: head,
diff --git a/src/tests.rs b/src/tests.rs
index a87fed402f..fc5bf33838 100644
--- a/src/tests.rs
+++ b/src/tests.rs
@@ -1316,3 +1316,11 @@ fn flatten_should_flatten_inner_table() -> TestResult {
         "123",
     )
 }
+
+#[test]
+fn cjk_in_substrings() -> TestResult {
+    run_test(
+        r#"let s = '[Rust 程序设计语言](title-page.md)'; let start = ($s | str index-of '('); let end = ($s | str index-of ')'); echo ($s | str substring $"($start + 1),($end)")"#,
+        "title-page.md",
+    )
+}