jpayne@69: """ Python 'utf-16' Codec
jpayne@69: 
jpayne@69: 
jpayne@69: Written by Marc-Andre Lemburg (mal@lemburg.com).
jpayne@69: 
jpayne@69: (c) Copyright CNRI, All Rights Reserved. NO WARRANTY.
jpayne@69: 
jpayne@69: """
jpayne@69: import codecs, sys
jpayne@69: 
jpayne@69: ### Codec APIs
jpayne@69: 
jpayne@69: encode = codecs.utf_16_encode
jpayne@69: 
jpayne@69: def decode(input, errors='strict'):
jpayne@69:     return codecs.utf_16_decode(input, errors, True)
jpayne@69: 
jpayne@69: class IncrementalEncoder(codecs.IncrementalEncoder):
jpayne@69:     def __init__(self, errors='strict'):
jpayne@69:         codecs.IncrementalEncoder.__init__(self, errors)
jpayne@69:         self.encoder = None
jpayne@69: 
jpayne@69:     def encode(self, input, final=False):
jpayne@69:         if self.encoder is None:
jpayne@69:             result = codecs.utf_16_encode(input, self.errors)[0]
jpayne@69:             if sys.byteorder == 'little':
jpayne@69:                 self.encoder = codecs.utf_16_le_encode
jpayne@69:             else:
jpayne@69:                 self.encoder = codecs.utf_16_be_encode
jpayne@69:             return result
jpayne@69:         return self.encoder(input, self.errors)[0]
jpayne@69: 
jpayne@69:     def reset(self):
jpayne@69:         codecs.IncrementalEncoder.reset(self)
jpayne@69:         self.encoder = None
jpayne@69: 
jpayne@69:     def getstate(self):
jpayne@69:         # state info we return to the caller:
jpayne@69:         # 0: stream is in natural order for this platform
jpayne@69:         # 2: endianness hasn't been determined yet
jpayne@69:         # (we're never writing in unnatural order)
jpayne@69:         return (2 if self.encoder is None else 0)
jpayne@69: 
jpayne@69:     def setstate(self, state):
jpayne@69:         if state:
jpayne@69:             self.encoder = None
jpayne@69:         else:
jpayne@69:             if sys.byteorder == 'little':
jpayne@69:                 self.encoder = codecs.utf_16_le_encode
jpayne@69:             else:
jpayne@69:                 self.encoder = codecs.utf_16_be_encode
jpayne@69: 
jpayne@69: class IncrementalDecoder(codecs.BufferedIncrementalDecoder):
jpayne@69:     def __init__(self, errors='strict'):
jpayne@69:         codecs.BufferedIncrementalDecoder.__init__(self, errors)
jpayne@69:         self.decoder = None
jpayne@69: 
jpayne@69:     def _buffer_decode(self, input, errors, final):
jpayne@69:         if self.decoder is None:
jpayne@69:             (output, consumed, byteorder) = \
jpayne@69:                 codecs.utf_16_ex_decode(input, errors, 0, final)
jpayne@69:             if byteorder == -1:
jpayne@69:                 self.decoder = codecs.utf_16_le_decode
jpayne@69:             elif byteorder == 1:
jpayne@69:                 self.decoder = codecs.utf_16_be_decode
jpayne@69:             elif consumed >= 2:
jpayne@69:                 raise UnicodeError("UTF-16 stream does not start with BOM")
jpayne@69:             return (output, consumed)
jpayne@69:         return self.decoder(input, self.errors, final)
jpayne@69: 
jpayne@69:     def reset(self):
jpayne@69:         codecs.BufferedIncrementalDecoder.reset(self)
jpayne@69:         self.decoder = None
jpayne@69: 
jpayne@69:     def getstate(self):
jpayne@69:         # additional state info from the base class must be None here,
jpayne@69:         # as it isn't passed along to the caller
jpayne@69:         state = codecs.BufferedIncrementalDecoder.getstate(self)[0]
jpayne@69:         # additional state info we pass to the caller:
jpayne@69:         # 0: stream is in natural order for this platform
jpayne@69:         # 1: stream is in unnatural order
jpayne@69:         # 2: endianness hasn't been determined yet
jpayne@69:         if self.decoder is None:
jpayne@69:             return (state, 2)
jpayne@69:         addstate = int((sys.byteorder == "big") !=
jpayne@69:                        (self.decoder is codecs.utf_16_be_decode))
jpayne@69:         return (state, addstate)
jpayne@69: 
jpayne@69:     def setstate(self, state):
jpayne@69:         # state[1] will be ignored by BufferedIncrementalDecoder.setstate()
jpayne@69:         codecs.BufferedIncrementalDecoder.setstate(self, state)
jpayne@69:         state = state[1]
jpayne@69:         if state == 0:
jpayne@69:             self.decoder = (codecs.utf_16_be_decode
jpayne@69:                             if sys.byteorder == "big"
jpayne@69:                             else codecs.utf_16_le_decode)
jpayne@69:         elif state == 1:
jpayne@69:             self.decoder = (codecs.utf_16_le_decode
jpayne@69:                             if sys.byteorder == "big"
jpayne@69:                             else codecs.utf_16_be_decode)
jpayne@69:         else:
jpayne@69:             self.decoder = None
jpayne@69: 
jpayne@69: class StreamWriter(codecs.StreamWriter):
jpayne@69:     def __init__(self, stream, errors='strict'):
jpayne@69:         codecs.StreamWriter.__init__(self, stream, errors)
jpayne@69:         self.encoder = None
jpayne@69: 
jpayne@69:     def reset(self):
jpayne@69:         codecs.StreamWriter.reset(self)
jpayne@69:         self.encoder = None
jpayne@69: 
jpayne@69:     def encode(self, input, errors='strict'):
jpayne@69:         if self.encoder is None:
jpayne@69:             result = codecs.utf_16_encode(input, errors)
jpayne@69:             if sys.byteorder == 'little':
jpayne@69:                 self.encoder = codecs.utf_16_le_encode
jpayne@69:             else:
jpayne@69:                 self.encoder = codecs.utf_16_be_encode
jpayne@69:             return result
jpayne@69:         else:
jpayne@69:             return self.encoder(input, errors)
jpayne@69: 
jpayne@69: class StreamReader(codecs.StreamReader):
jpayne@69: 
jpayne@69:     def reset(self):
jpayne@69:         codecs.StreamReader.reset(self)
jpayne@69:         try:
jpayne@69:             del self.decode
jpayne@69:         except AttributeError:
jpayne@69:             pass
jpayne@69: 
jpayne@69:     def decode(self, input, errors='strict'):
jpayne@69:         (object, consumed, byteorder) = \
jpayne@69:             codecs.utf_16_ex_decode(input, errors, 0, False)
jpayne@69:         if byteorder == -1:
jpayne@69:             self.decode = codecs.utf_16_le_decode
jpayne@69:         elif byteorder == 1:
jpayne@69:             self.decode = codecs.utf_16_be_decode
jpayne@69:         elif consumed>=2:
jpayne@69:             raise UnicodeError("UTF-16 stream does not start with BOM")
jpayne@69:         return (object, consumed)
jpayne@69: 
jpayne@69: ### encodings module API
jpayne@69: 
jpayne@69: def getregentry():
jpayne@69:     return codecs.CodecInfo(
jpayne@69:         name='utf-16',
jpayne@69:         encode=encode,
jpayne@69:         decode=decode,
jpayne@69:         incrementalencoder=IncrementalEncoder,
jpayne@69:         incrementaldecoder=IncrementalDecoder,
jpayne@69:         streamreader=StreamReader,
jpayne@69:         streamwriter=StreamWriter,
jpayne@69:     )