| Walter Dörwald | 41980ca | 2007-08-16 21:55:45 +0000 | [diff] [blame] | 1 | """ |
| 2 | Python 'utf-32' Codec |
| 3 | """ |
| 4 | import codecs, sys |
| 5 | |
| 6 | ### Codec APIs |
| 7 | |
| 8 | encode = codecs.utf_32_encode |
| 9 | |
| 10 | def decode(input, errors='strict'): |
| 11 | return codecs.utf_32_decode(input, errors, True) |
| 12 | |
| 13 | class IncrementalEncoder(codecs.IncrementalEncoder): |
| 14 | def __init__(self, errors='strict'): |
| 15 | codecs.IncrementalEncoder.__init__(self, errors) |
| 16 | self.encoder = None |
| 17 | |
| 18 | def encode(self, input, final=False): |
| 19 | if self.encoder is None: |
| 20 | result = codecs.utf_32_encode(input, self.errors)[0] |
| 21 | if sys.byteorder == 'little': |
| 22 | self.encoder = codecs.utf_32_le_encode |
| 23 | else: |
| 24 | self.encoder = codecs.utf_32_be_encode |
| 25 | return result |
| 26 | return self.encoder(input, self.errors)[0] |
| 27 | |
| 28 | def reset(self): |
| 29 | codecs.IncrementalEncoder.reset(self) |
| 30 | self.encoder = None |
| 31 | |
| 32 | def getstate(self): |
| 33 | # state info we return to the caller: |
| 34 | # 0: stream is in natural order for this platform |
| 35 | # 2: endianness hasn't been determined yet |
| 36 | # (we're never writing in unnatural order) |
| 37 | return (2 if self.encoder is None else 0) |
| 38 | |
| 39 | def setstate(self, state): |
| 40 | if state: |
| 41 | self.encoder = None |
| 42 | else: |
| 43 | if sys.byteorder == 'little': |
| 44 | self.encoder = codecs.utf_32_le_encode |
| 45 | else: |
| 46 | self.encoder = codecs.utf_32_be_encode |
| 47 | |
| 48 | class IncrementalDecoder(codecs.BufferedIncrementalDecoder): |
| 49 | def __init__(self, errors='strict'): |
| 50 | codecs.BufferedIncrementalDecoder.__init__(self, errors) |
| 51 | self.decoder = None |
| 52 | |
| 53 | def _buffer_decode(self, input, errors, final): |
| 54 | if self.decoder is None: |
| 55 | (output, consumed, byteorder) = \ |
| 56 | codecs.utf_32_ex_decode(input, errors, 0, final) |
| 57 | if byteorder == -1: |
| 58 | self.decoder = codecs.utf_32_le_decode |
| 59 | elif byteorder == 1: |
| 60 | self.decoder = codecs.utf_32_be_decode |
| 61 | elif consumed >= 4: |
| 62 | raise UnicodeError("UTF-32 stream does not start with BOM") |
| 63 | return (output, consumed) |
| 64 | return self.decoder(input, self.errors, final) |
| 65 | |
| 66 | def reset(self): |
| 67 | codecs.BufferedIncrementalDecoder.reset(self) |
| 68 | self.decoder = None |
| 69 | |
| 70 | def getstate(self): |
| 71 | # additonal state info from the base class must be None here, |
| 72 | # as it isn't passed along to the caller |
| 73 | state = codecs.BufferedIncrementalDecoder.getstate(self)[0] |
| 74 | # additional state info we pass to the caller: |
| 75 | # 0: stream is in natural order for this platform |
| 76 | # 1: stream is in unnatural order |
| 77 | # 2: endianness hasn't been determined yet |
| 78 | if self.decoder is None: |
| 79 | return (state, 2) |
| 80 | addstate = int((sys.byteorder == "big") != |
| 81 | (self.decoder is codecs.utf_32_be_decode)) |
| 82 | return (state, addstate) |
| 83 | |
| 84 | def setstate(self, state): |
| 85 | # state[1] will be ignored by BufferedIncrementalDecoder.setstate() |
| 86 | codecs.BufferedIncrementalDecoder.setstate(self, state) |
| 87 | state = state[1] |
| 88 | if state == 0: |
| 89 | self.decoder = (codecs.utf_32_be_decode |
| 90 | if sys.byteorder == "big" |
| 91 | else codecs.utf_32_le_decode) |
| 92 | elif state == 1: |
| 93 | self.decoder = (codecs.utf_32_le_decode |
| 94 | if sys.byteorder == "big" |
| 95 | else codecs.utf_32_be_decode) |
| 96 | else: |
| 97 | self.decoder = None |
| 98 | |
| 99 | class StreamWriter(codecs.StreamWriter): |
| 100 | def __init__(self, stream, errors='strict'): |
| 101 | self.bom_written = False |
| 102 | codecs.StreamWriter.__init__(self, stream, errors) |
| 103 | |
| 104 | def encode(self, input, errors='strict'): |
| 105 | self.bom_written = True |
| 106 | result = codecs.utf_32_encode(input, errors) |
| 107 | if sys.byteorder == 'little': |
| 108 | self.encode = codecs.utf_32_le_encode |
| 109 | else: |
| 110 | self.encode = codecs.utf_32_be_encode |
| 111 | return result |
| 112 | |
| 113 | class StreamReader(codecs.StreamReader): |
| 114 | |
| 115 | def reset(self): |
| 116 | codecs.StreamReader.reset(self) |
| 117 | try: |
| 118 | del self.decode |
| 119 | except AttributeError: |
| 120 | pass |
| 121 | |
| 122 | def decode(self, input, errors='strict'): |
| 123 | (object, consumed, byteorder) = \ |
| 124 | codecs.utf_32_ex_decode(input, errors, 0, False) |
| 125 | if byteorder == -1: |
| 126 | self.decode = codecs.utf_32_le_decode |
| 127 | elif byteorder == 1: |
| Walter Dörwald | 19e6238 | 2007-08-17 16:23:21 +0000 | [diff] [blame] | 128 | self.decode = codecs.utf_32_be_decode |
| Walter Dörwald | 41980ca | 2007-08-16 21:55:45 +0000 | [diff] [blame] | 129 | elif consumed>=4: |
| Collin Winter | 4902e69 | 2007-08-30 18:18:27 +0000 | [diff] [blame] | 130 | raise UnicodeError("UTF-32 stream does not start with BOM") |
| Walter Dörwald | 41980ca | 2007-08-16 21:55:45 +0000 | [diff] [blame] | 131 | return (object, consumed) |
| 132 | |
| 133 | ### encodings module API |
| 134 | |
| 135 | def getregentry(): |
| 136 | return codecs.CodecInfo( |
| 137 | name='utf-32', |
| 138 | encode=encode, |
| 139 | decode=decode, |
| 140 | incrementalencoder=IncrementalEncoder, |
| 141 | incrementaldecoder=IncrementalDecoder, |
| 142 | streamreader=StreamReader, |
| 143 | streamwriter=StreamWriter, |
| 144 | ) |