fix: force yt-dlp pipe output to UTF-8

Author: Aimeast

YoutubeDLSharp/YoutubeDLProcess.cs

@@ -57,7 +57,18 @@ public YoutubeDLProcess(string executablePath = "yt-dlp.exe")
         }
 
         internal string ConvertToArgs(string[] urls, OptionSet options)
-            => options.ToString() + " -- " + (urls != null ? String.Join(" ", urls.Select(s => $"\"{s}\"")) : String.Empty);
+        {
+            var args = options.ToString();
+            // By default, force yt-dlp to write its standard output and error output in UTF-8,
+            // which is the encoding the process output is always read with below. Without this,
+            // yt-dlp (a Python program) uses the system locale's ANSI code page (e.g. GBK on
+            // Chinese Windows) for redirected pipes, and decoding those bytes as UTF-8 turns
+            // every non-ASCII character into a U+FFFD replacement character. An explicit
+            // OptionSet.Encoding is respected and left untouched.
+            if (options.Encoding == null)
+                args += " --encoding utf-8";
+            return args + " -- " + (urls != null ? String.Join(" ", urls.Select(s => $"\"{s}\"")) : String.Empty);
+        }
 
         internal void RedirectToError(DataReceivedEventArgs e)
             => ErrorReceived?.Invoke(this, e);