From e02467ef36257859ed29a35f672bafe24df97c66 Mon Sep 17 00:00:00 2001
From: Nicolas DUBIEN <github@dubien.org>
Date: Sat, 11 May 2024 15:27:01 +0200
Subject: [PATCH] fix: support inspect on strings with unicode (#79)

* Support inspect on strings with unicode

* Update src/helpers.ts

* fix lint
---
 src/helpers.ts  | 10 +++++++++-
 test/strings.js | 19 +++++++++++++++++++
 2 files changed, 28 insertions(+), 1 deletion(-)

diff --git a/src/helpers.ts b/src/helpers.ts
index 39a494f..1b41d29 100644
--- a/src/helpers.ts
+++ b/src/helpers.ts
@@ -93,6 +93,10 @@ export function normaliseOptions(
   return options
 }
 
+function isHighSurrogate(char: string): boolean {
+  return char >= '\ud800' && char <= '\udbff'
+}
+
 export function truncate(string: string | number, length: number, tail: typeof truncator = truncator) {
   string = String(string)
   const tailLength = tail.length
@@ -101,7 +105,11 @@ export function truncate(string: string | number, length: number, tail: typeof t
     return tail
   }
   if (stringLength > length && stringLength > tailLength) {
-    return `${string.slice(0, length - tailLength)}${tail}`
+    let end = length - tailLength
+    if (end > 0 && isHighSurrogate(string[end - 1])) {
+      end = end - 1
+    }
+    return `${string.slice(0, end)}${tail}`
   }
   return string
 }
diff --git a/test/strings.js b/test/strings.js
index aae1140..afe1fe7 100644
--- a/test/strings.js
+++ b/test/strings.js
@@ -59,6 +59,25 @@ describe('strings', () => {
       expect(inspect('foobarbaz', { truncate: 3 })).to.equal("'…'")
     })
 
+    it('truncates strings involving surrogate pairs longer than truncate (7)', () => {
+      // not '🐱🐱\ud83d…' (length 7) but '🐱🐱…' (length 6)
+      expect(inspect('🐱🐱🐱', { truncate: 7 })).to.equal("'🐱🐱…'")
+    })
+
+    it('truncates strings involving surrogate pairs longer than truncate (6)', () => {
+      expect(inspect('🐱🐱🐱', { truncate: 6 })).to.equal("'🐱…'")
+    })
+
+    it('truncates strings involving surrogate pairs longer than truncate (5)', () => {
+      // not '🐱\ud83d…' (length 5) but '🐱…' (length 4)
+      expect(inspect('🐱🐱🐱', { truncate: 5 })).to.equal("'🐱…'")
+    })
+
+    it('truncates strings involving graphemes than truncate (5)', () => {
+      // partial support: valid string for unicode
+      expect(inspect('👨‍👩‍👧‍👧', { truncate: 5 })).to.equal("'👨…'")
+    })
+
     it('disregards truncate when it cannot truncate further (2)', () => {
       expect(inspect('foobarbaz', { truncate: 2 })).to.equal("'…'")
     })