Fix #6917: Correct Content-Length for io.StringIO with multi-byte UTF-8

[gqwtb]

Fixes #6917

This PR resolves the incorrect Content-Length calculation when using io.StringIO as the request body. Key changes:

1. Added special handling for io.StringIO in the super_len function
2. Calculates the correct byte length by encoding content as UTF-8
3. Added test cases to validate the fix

Test results:

• Content-Length now accurately reflects byte count for io.StringIO with multi-byte characters
• All existing tests remain passing