修正 fetchScriptBody 的 Accept 设定，避免触发 Error 406

Author: cyfung1031

src/pages/install/App.tsx

@@ -67,7 +67,8 @@ const fetchScriptBody = async (url: string, { onProgress }: { [key: string]: any
   const response = await fetch(url, {
     headers: {
       "Cache-Control": "no-cache",
-      Accept: "text/javascript,application/javascript,text/plain,application/octet-stream,application/force-download",
+      /* 不指定 application/octet-stream 和 application/force-download 避免触发伺服器端 Error 406 */
+      Accept: "text/javascript, application/javascript, */*", // prefer JavaScript, but anything is acceptable
       // 参考：加权 Accept-Encoding 值说明
       // https://developer.mozilla.org/en-US/docs/Web/HTTP/Reference/Headers/Accept-Encoding#weighted_accept-encoding_values
       "Accept-Encoding": "br;q=1.0, gzip;q=0.8, *;q=0.1",
@@ -83,8 +84,21 @@ const fetchScriptBody = async (url: string, { onProgress }: { [key: string]: any
   if (!response.body || !response.headers) {
     throw new Error("No response body or headers");
   }
-  if (response.headers.get("content-type")?.includes("text/html")) {
-    throw new Error("Response is text/html, not a valid UserScript");
+  const contentType = response.headers.get("content-type");
+
+  if (contentType) {
+    // 不接受非 JavaScript文本 的回应
+    const contentTypeLower = contentType.toLowerCase();
+    const m = /^\s*([\w-]+)[^\w-]+([\w-]+)/.exec(contentTypeLower);
+    if (m) {
+      const contentTypeOK =
+        (m[2] === "javascript" && (m[1] === "text" || m[1] === "application")) ||
+        (m[1] === "application" && (m[2] === "octet-stream" || m[2] === "force-download"));
+      if (!contentTypeOK) {
+        throw new Error(`Response is ${contentType}, not a valid UserScript`);
+        // e.g. Response is text/html, not a valid UserScript
+      }
+    }
   }
 
   const reader = response.body.getReader();
@@ -113,7 +127,6 @@ const fetchScriptBody = async (url: string, { onProgress }: { [key: string]: any
   }
 
   // 检测编码：优先使用 Content-Type，回退到 chardet（仅检测前16KB）
-  const contentType = response.headers.get("content-type");
   const encode = detectEncoding(chunksAll, contentType);
 
   // 使用检测到的 charset 解码