From 487c0469c6ec81b47ac4209364dc2e5e672e949a Mon Sep 17 00:00:00 2001
From: Cieons <109508894+cieons@users.noreply.github.com>
Date: Fri, 13 Sep 2024 23:10:45 +0800
Subject: [PATCH] fix(route): people charset decoding (#16734) (#16743)

---
 lib/routes/people/index.ts | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

diff --git a/lib/routes/people/index.ts b/lib/routes/people/index.ts
index 48c352385..9543a452a 100644
--- a/lib/routes/people/index.ts
+++ b/lib/routes/people/index.ts
@@ -32,8 +32,12 @@ async function handler(ctx) {
         responseType: 'buffer',
     });
 
-    const encoding = site === 'www' ? 'gbk' : 'utf-8';
-    const decodedResponse = iconv.decode(response, encoding);
+    // not seen Content-Type in response headers
+    // try to parse charset from meta tag
+    let decodedResponse = iconv.decode(response, 'utf-8');
+    const parsedCharset = decodedResponse.match(/<meta.*?charset=["']?([^"'>]+)["']?/i);
+    const encoding = parsedCharset ? parsedCharset[1].toLowerCase() : 'utf-8';
+    decodedResponse = encoding === 'utf-8' ? decodedResponse : iconv.decode(response, encoding);
     const $ = load(decodedResponse);
 
     $('em').remove();