Class Latin1Converter


  • public class Latin1Converter
    extends java.lang.Object
    Since:
    12.10.2006
    • Field Summary

      Fields 
      Modifier and Type Field Description
      private static int STATE_START  
      private static int STATE_UTF8CHAR  
    • Constructor Summary

      Constructors 
      Modifier Constructor Description
      private Latin1Converter()
      Private constructor
    • Method Summary

      All Methods Static Methods Concrete Methods 
      Modifier and Type Method Description
      static ByteBuffer convert​(ByteBuffer buffer)
      A converter that processes a byte buffer containing a mix of UTF8 and Latin-1/Cp1252 chars.
      private static byte[] convertToUTF8​(byte ch)
      Converts a Cp1252 char (contains all Latin-1 chars above 0x80) into a UTF-8 byte sequence.
      • Methods inherited from class java.lang.Object

        clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
    • Constructor Detail

      • Latin1Converter

        private Latin1Converter()
        Private constructor
    • Method Detail

      • convert

        public static ByteBuffer convert​(ByteBuffer buffer)
        A converter that processes a byte buffer containing a mix of UTF8 and Latin-1/Cp1252 chars. The result is a buffer where those chars have been converted to UTF-8; that means it contains only valid UTF-8 chars.

        Explanation of the processing: First the encoding of the buffer is detected looking at the first four bytes (that works only if the buffer starts with an ASCII-char, like xmls '<'). UTF-16/32 flavours do not require further proccessing.

        In the case, UTF-8 is detected, it assumes wrong UTF8 chars to be a sequence of Latin-1/Cp1252 encoded bytes and converts the chars to their corresponding UTF-8 byte sequence.

        The 0x80..0x9F range is undefined in Latin-1, but is defined in Windows code page 1252. The bytes 0x81, 0x8D, 0x8F, 0x90, and 0x9D are formally undefined by Windows 1252. These are in XML's RestrictedChar set, so we map them to a space.

        The official Latin-1 characters in the range 0xA0..0xFF are converted into the Unicode Latin Supplement range U+00A0 - U+00FF.

        Example: If an Euro-symbol (€) appears in the byte buffer (0xE2, 0x82, 0xAC), it will be left as is. But if only the first two bytes are appearing, followed by an ASCII char a (0xE2 - 0x82 - 0x41), it will be converted to 0xC3, 0xA2 (â) - 0xE2, 0x80, 0x9A (‚) - 0x41 (a).

        Parameters:
        buffer - a byte buffer contain
        Returns:
        Returns a new buffer containing valid UTF-8
      • convertToUTF8

        private static byte[] convertToUTF8​(byte ch)
        Converts a Cp1252 char (contains all Latin-1 chars above 0x80) into a UTF-8 byte sequence. The bytes 0x81, 0x8D, 0x8F, 0x90, and 0x9D are formally undefined by Windows 1252 and therefore replaced by a space (0x20).
        Parameters:
        ch - an Cp1252 / Latin-1 byte
        Returns:
        Returns a byte array containing a UTF-8 byte sequence.