Converting a string into Windows-1252 preserving as much information as possible

Sample.java

private static final Charset WINDOWS_1252_CHARSET = Charset.forName("Windows-1252");  


// Can below be done more effectively (/robust/secure) ?

@Test
public void test() {
  
          // Given
        String inputString = "ĢoogleŰberButÅ";

        // When
        String result = convertToWindows1252(inputString);

        // Then
        assertEquals("GoogleUberButÅ", result); 
        // Or put in another way:
        //  I want to retain Å's accent but not accents for Chars that are outside Windows-1252.
        //  ..and I will rather remove accents (to obtain valid Chars) than throwing them completely away.
        //  ..ie. oogleberButÅ wouldn't be optimal
  
        // The receiver of "result" must get the data in proper Windows-1252 
}

private static String convertToWindows1252(String inputString) {
        CharsetEncoder encoder = WINDOWS_1252_CHARSET.newEncoder();
        return inputString.chars()
                .mapToObj(i -> (char) i)
                .map(c -> convertToWindows1252(encoder, c))
                .filter(Objects::nonNull)
                .collect(StringBuilder::new, StringBuilder::appendCodePoint, StringBuilder::append)
                .toString();
    }

    private static Character convertToWindows1252(CharsetEncoder encoder, char character) {
        if (encoder.canEncode(character)) {
            return character;
        }
        // Check if we can reduce 'character' to an accent-stripped valid Windows-1252 character...
        // It will be split into 2 characters if we can... 
        String decomposed = Normalizer.normalize(String.valueOf(character), Normalizer.Form.NFD);
        return decomposed.length() == 2 ? decomposed.charAt(0) : null;
    }

Hi,
how can I remove "�ÿþ" from my String "�ÿþHello, world�ÿþ"?