| 1 | """ codecs -- Python Codec Registry, API and helpers.
|
| 2 |
|
| 3 |
|
| 4 | Written by Marc-Andre Lemburg (mal@lemburg.com).
|
| 5 |
|
| 6 | (c) Copyright CNRI, All Rights Reserved. NO WARRANTY.
|
| 7 |
|
| 8 | """#"
|
| 9 |
|
| 10 | import __builtin__, sys
|
| 11 |
|
| 12 | ### Registry and builtin stateless codec functions
|
| 13 |
|
| 14 | try:
|
| 15 | from _codecs import *
|
| 16 | except ImportError, why:
|
| 17 | raise SystemError('Failed to load the builtin codecs: %s' % why)
|
| 18 |
|
| 19 | __all__ = ["register", "lookup", "open", "EncodedFile", "BOM", "BOM_BE",
|
| 20 | "BOM_LE", "BOM32_BE", "BOM32_LE", "BOM64_BE", "BOM64_LE",
|
| 21 | "BOM_UTF8", "BOM_UTF16", "BOM_UTF16_LE", "BOM_UTF16_BE",
|
| 22 | "BOM_UTF32", "BOM_UTF32_LE", "BOM_UTF32_BE",
|
| 23 | "CodecInfo", "Codec", "IncrementalEncoder", "IncrementalDecoder",
|
| 24 | "StreamReader", "StreamWriter",
|
| 25 | "StreamReaderWriter", "StreamRecoder",
|
| 26 | "getencoder", "getdecoder", "getincrementalencoder",
|
| 27 | "getincrementaldecoder", "getreader", "getwriter",
|
| 28 | "encode", "decode", "iterencode", "iterdecode",
|
| 29 | "strict_errors", "ignore_errors", "replace_errors",
|
| 30 | "xmlcharrefreplace_errors", "backslashreplace_errors",
|
| 31 | "register_error", "lookup_error"]
|
| 32 |
|
| 33 | ### Constants
|
| 34 |
|
| 35 | #
|
| 36 | # Byte Order Mark (BOM = ZERO WIDTH NO-BREAK SPACE = U+FEFF)
|
| 37 | # and its possible byte string values
|
| 38 | # for UTF8/UTF16/UTF32 output and little/big endian machines
|
| 39 | #
|
| 40 |
|
| 41 | # UTF-8
|
| 42 | BOM_UTF8 = '\xef\xbb\xbf'
|
| 43 |
|
| 44 | # UTF-16, little endian
|
| 45 | BOM_LE = BOM_UTF16_LE = '\xff\xfe'
|
| 46 |
|
| 47 | # UTF-16, big endian
|
| 48 | BOM_BE = BOM_UTF16_BE = '\xfe\xff'
|
| 49 |
|
| 50 | # UTF-32, little endian
|
| 51 | BOM_UTF32_LE = '\xff\xfe\x00\x00'
|
| 52 |
|
| 53 | # UTF-32, big endian
|
| 54 | BOM_UTF32_BE = '\x00\x00\xfe\xff'
|
| 55 |
|
| 56 | if sys.byteorder == 'little':
|
| 57 |
|
| 58 | # UTF-16, native endianness
|
| 59 | BOM = BOM_UTF16 = BOM_UTF16_LE
|
| 60 |
|
| 61 | # UTF-32, native endianness
|
| 62 | BOM_UTF32 = BOM_UTF32_LE
|
| 63 |
|
| 64 | else:
|
| 65 |
|
| 66 | # UTF-16, native endianness
|
| 67 | BOM = BOM_UTF16 = BOM_UTF16_BE
|
| 68 |
|
| 69 | # UTF-32, native endianness
|
| 70 | BOM_UTF32 = BOM_UTF32_BE
|
| 71 |
|
| 72 | # Old broken names (don't use in new code)
|
| 73 | BOM32_LE = BOM_UTF16_LE
|
| 74 | BOM32_BE = BOM_UTF16_BE
|
| 75 | BOM64_LE = BOM_UTF32_LE
|
| 76 | BOM64_BE = BOM_UTF32_BE
|
| 77 |
|
| 78 |
|
| 79 | ### Codec base classes (defining the API)
|
| 80 |
|
| 81 | class CodecInfo(tuple):
|
| 82 | """Codec details when looking up the codec registry"""
|
| 83 |
|
| 84 | # Private API to allow Python to blacklist the known non-Unicode
|
| 85 | # codecs in the standard library. A more general mechanism to
|
| 86 | # reliably distinguish test encodings from other codecs will hopefully
|
| 87 | # be defined for Python 3.5
|
| 88 | #
|
| 89 | # See http://bugs.python.org/issue19619
|
| 90 | _is_text_encoding = True # Assume codecs are text encodings by default
|
| 91 |
|
| 92 | def __new__(cls, encode, decode, streamreader=None, streamwriter=None,
|
| 93 | incrementalencoder=None, incrementaldecoder=None, name=None,
|
| 94 | _is_text_encoding=None):
|
| 95 | self = tuple.__new__(cls, (encode, decode, streamreader, streamwriter))
|
| 96 | self.name = name
|
| 97 | self.encode = encode
|
| 98 | self.decode = decode
|
| 99 | self.incrementalencoder = incrementalencoder
|
| 100 | self.incrementaldecoder = incrementaldecoder
|
| 101 | self.streamwriter = streamwriter
|
| 102 | self.streamreader = streamreader
|
| 103 | if _is_text_encoding is not None:
|
| 104 | self._is_text_encoding = _is_text_encoding
|
| 105 | return self
|
| 106 |
|
| 107 | def __repr__(self):
|
| 108 | return "<%s.%s object for encoding %s at 0x%x>" % (self.__class__.__module__, self.__class__.__name__, self.name, id(self))
|
| 109 |
|
| 110 | class Codec:
|
| 111 |
|
| 112 | """ Defines the interface for stateless encoders/decoders.
|
| 113 |
|
| 114 | The .encode()/.decode() methods may use different error
|
| 115 | handling schemes by providing the errors argument. These
|
| 116 | string values are predefined:
|
| 117 |
|
| 118 | 'strict' - raise a ValueError error (or a subclass)
|
| 119 | 'ignore' - ignore the character and continue with the next
|
| 120 | 'replace' - replace with a suitable replacement character;
|
| 121 | Python will use the official U+FFFD REPLACEMENT
|
| 122 | CHARACTER for the builtin Unicode codecs on
|
| 123 | decoding and '?' on encoding.
|
| 124 | 'xmlcharrefreplace' - Replace with the appropriate XML
|
| 125 | character reference (only for encoding).
|
| 126 | 'backslashreplace' - Replace with backslashed escape sequences
|
| 127 | (only for encoding).
|
| 128 |
|
| 129 | The set of allowed values can be extended via register_error.
|
| 130 |
|
| 131 | """
|
| 132 | def encode(self, input, errors='strict'):
|
| 133 |
|
| 134 | """ Encodes the object input and returns a tuple (output
|
| 135 | object, length consumed).
|
| 136 |
|
| 137 | errors defines the error handling to apply. It defaults to
|
| 138 | 'strict' handling.
|
| 139 |
|
| 140 | The method may not store state in the Codec instance. Use
|
| 141 | StreamWriter for codecs which have to keep state in order to
|
| 142 | make encoding efficient.
|
| 143 |
|
| 144 | The encoder must be able to handle zero length input and
|
| 145 | return an empty object of the output object type in this
|
| 146 | situation.
|
| 147 |
|
| 148 | """
|
| 149 | raise NotImplementedError
|
| 150 |
|
| 151 | def decode(self, input, errors='strict'):
|
| 152 |
|
| 153 | """ Decodes the object input and returns a tuple (output
|
| 154 | object, length consumed).
|
| 155 |
|
| 156 | input must be an object which provides the bf_getreadbuf
|
| 157 | buffer slot. Python strings, buffer objects and memory
|
| 158 | mapped files are examples of objects providing this slot.
|
| 159 |
|
| 160 | errors defines the error handling to apply. It defaults to
|
| 161 | 'strict' handling.
|
| 162 |
|
| 163 | The method may not store state in the Codec instance. Use
|
| 164 | StreamReader for codecs which have to keep state in order to
|
| 165 | make decoding efficient.
|
| 166 |
|
| 167 | The decoder must be able to handle zero length input and
|
| 168 | return an empty object of the output object type in this
|
| 169 | situation.
|
| 170 |
|
| 171 | """
|
| 172 | raise NotImplementedError
|
| 173 |
|
| 174 | class IncrementalEncoder(object):
|
| 175 | """
|
| 176 | An IncrementalEncoder encodes an input in multiple steps. The input can be
|
| 177 | passed piece by piece to the encode() method. The IncrementalEncoder remembers
|
| 178 | the state of the Encoding process between calls to encode().
|
| 179 | """
|
| 180 | def __init__(self, errors='strict'):
|
| 181 | """
|
| 182 | Creates an IncrementalEncoder instance.
|
| 183 |
|
| 184 | The IncrementalEncoder may use different error handling schemes by
|
| 185 | providing the errors keyword argument. See the module docstring
|
| 186 | for a list of possible values.
|
| 187 | """
|
| 188 | self.errors = errors
|
| 189 | self.buffer = ""
|
| 190 |
|
| 191 | def encode(self, input, final=False):
|
| 192 | """
|
| 193 | Encodes input and returns the resulting object.
|
| 194 | """
|
| 195 | raise NotImplementedError
|
| 196 |
|
| 197 | def reset(self):
|
| 198 | """
|
| 199 | Resets the encoder to the initial state.
|
| 200 | """
|
| 201 |
|
| 202 | def getstate(self):
|
| 203 | """
|
| 204 | Return the current state of the encoder.
|
| 205 | """
|
| 206 | return 0
|
| 207 |
|
| 208 | def setstate(self, state):
|
| 209 | """
|
| 210 | Set the current state of the encoder. state must have been
|
| 211 | returned by getstate().
|
| 212 | """
|
| 213 |
|
| 214 | class BufferedIncrementalEncoder(IncrementalEncoder):
|
| 215 | """
|
| 216 | This subclass of IncrementalEncoder can be used as the baseclass for an
|
| 217 | incremental encoder if the encoder must keep some of the output in a
|
| 218 | buffer between calls to encode().
|
| 219 | """
|
| 220 | def __init__(self, errors='strict'):
|
| 221 | IncrementalEncoder.__init__(self, errors)
|
| 222 | self.buffer = "" # unencoded input that is kept between calls to encode()
|
| 223 |
|
| 224 | def _buffer_encode(self, input, errors, final):
|
| 225 | # Overwrite this method in subclasses: It must encode input
|
| 226 | # and return an (output, length consumed) tuple
|
| 227 | raise NotImplementedError
|
| 228 |
|
| 229 | def encode(self, input, final=False):
|
| 230 | # encode input (taking the buffer into account)
|
| 231 | data = self.buffer + input
|
| 232 | (result, consumed) = self._buffer_encode(data, self.errors, final)
|
| 233 | # keep unencoded input until the next call
|
| 234 | self.buffer = data[consumed:]
|
| 235 | return result
|
| 236 |
|
| 237 | def reset(self):
|
| 238 | IncrementalEncoder.reset(self)
|
| 239 | self.buffer = ""
|
| 240 |
|
| 241 | def getstate(self):
|
| 242 | return self.buffer or 0
|
| 243 |
|
| 244 | def setstate(self, state):
|
| 245 | self.buffer = state or ""
|
| 246 |
|
| 247 | class IncrementalDecoder(object):
|
| 248 | """
|
| 249 | An IncrementalDecoder decodes an input in multiple steps. The input can be
|
| 250 | passed piece by piece to the decode() method. The IncrementalDecoder
|
| 251 | remembers the state of the decoding process between calls to decode().
|
| 252 | """
|
| 253 | def __init__(self, errors='strict'):
|
| 254 | """
|
| 255 | Creates an IncrementalDecoder instance.
|
| 256 |
|
| 257 | The IncrementalDecoder may use different error handling schemes by
|
| 258 | providing the errors keyword argument. See the module docstring
|
| 259 | for a list of possible values.
|
| 260 | """
|
| 261 | self.errors = errors
|
| 262 |
|
| 263 | def decode(self, input, final=False):
|
| 264 | """
|
| 265 | Decodes input and returns the resulting object.
|
| 266 | """
|
| 267 | raise NotImplementedError
|
| 268 |
|
| 269 | def reset(self):
|
| 270 | """
|
| 271 | Resets the decoder to the initial state.
|
| 272 | """
|
| 273 |
|
| 274 | def getstate(self):
|
| 275 | """
|
| 276 | Return the current state of the decoder.
|
| 277 |
|
| 278 | This must be a (buffered_input, additional_state_info) tuple.
|
| 279 | buffered_input must be a bytes object containing bytes that
|
| 280 | were passed to decode() that have not yet been converted.
|
| 281 | additional_state_info must be a non-negative integer
|
| 282 | representing the state of the decoder WITHOUT yet having
|
| 283 | processed the contents of buffered_input. In the initial state
|
| 284 | and after reset(), getstate() must return (b"", 0).
|
| 285 | """
|
| 286 | return (b"", 0)
|
| 287 |
|
| 288 | def setstate(self, state):
|
| 289 | """
|
| 290 | Set the current state of the decoder.
|
| 291 |
|
| 292 | state must have been returned by getstate(). The effect of
|
| 293 | setstate((b"", 0)) must be equivalent to reset().
|
| 294 | """
|
| 295 |
|
| 296 | class BufferedIncrementalDecoder(IncrementalDecoder):
|
| 297 | """
|
| 298 | This subclass of IncrementalDecoder can be used as the baseclass for an
|
| 299 | incremental decoder if the decoder must be able to handle incomplete byte
|
| 300 | sequences.
|
| 301 | """
|
| 302 | def __init__(self, errors='strict'):
|
| 303 | IncrementalDecoder.__init__(self, errors)
|
| 304 | self.buffer = "" # undecoded input that is kept between calls to decode()
|
| 305 |
|
| 306 | def _buffer_decode(self, input, errors, final):
|
| 307 | # Overwrite this method in subclasses: It must decode input
|
| 308 | # and return an (output, length consumed) tuple
|
| 309 | raise NotImplementedError
|
| 310 |
|
| 311 | def decode(self, input, final=False):
|
| 312 | # decode input (taking the buffer into account)
|
| 313 | data = self.buffer + input
|
| 314 | (result, consumed) = self._buffer_decode(data, self.errors, final)
|
| 315 | # keep undecoded input until the next call
|
| 316 | self.buffer = data[consumed:]
|
| 317 | return result
|
| 318 |
|
| 319 | def reset(self):
|
| 320 | IncrementalDecoder.reset(self)
|
| 321 | self.buffer = ""
|
| 322 |
|
| 323 | def getstate(self):
|
| 324 | # additional state info is always 0
|
| 325 | return (self.buffer, 0)
|
| 326 |
|
| 327 | def setstate(self, state):
|
| 328 | # ignore additional state info
|
| 329 | self.buffer = state[0]
|
| 330 |
|
| 331 | #
|
| 332 | # The StreamWriter and StreamReader class provide generic working
|
| 333 | # interfaces which can be used to implement new encoding submodules
|
| 334 | # very easily. See encodings/utf_8.py for an example on how this is
|
| 335 | # done.
|
| 336 | #
|
| 337 |
|
| 338 | class StreamWriter(Codec):
|
| 339 |
|
| 340 | def __init__(self, stream, errors='strict'):
|
| 341 |
|
| 342 | """ Creates a StreamWriter instance.
|
| 343 |
|
| 344 | stream must be a file-like object open for writing
|
| 345 | (binary) data.
|
| 346 |
|
| 347 | The StreamWriter may use different error handling
|
| 348 | schemes by providing the errors keyword argument. These
|
| 349 | parameters are predefined:
|
| 350 |
|
| 351 | 'strict' - raise a ValueError (or a subclass)
|
| 352 | 'ignore' - ignore the character and continue with the next
|
| 353 | 'replace'- replace with a suitable replacement character
|
| 354 | 'xmlcharrefreplace' - Replace with the appropriate XML
|
| 355 | character reference.
|
| 356 | 'backslashreplace' - Replace with backslashed escape
|
| 357 | sequences (only for encoding).
|
| 358 |
|
| 359 | The set of allowed parameter values can be extended via
|
| 360 | register_error.
|
| 361 | """
|
| 362 | self.stream = stream
|
| 363 | self.errors = errors
|
| 364 |
|
| 365 | def write(self, object):
|
| 366 |
|
| 367 | """ Writes the object's contents encoded to self.stream.
|
| 368 | """
|
| 369 | data, consumed = self.encode(object, self.errors)
|
| 370 | self.stream.write(data)
|
| 371 |
|
| 372 | def writelines(self, list):
|
| 373 |
|
| 374 | """ Writes the concatenated list of strings to the stream
|
| 375 | using .write().
|
| 376 | """
|
| 377 | self.write(''.join(list))
|
| 378 |
|
| 379 | def reset(self):
|
| 380 |
|
| 381 | """ Flushes and resets the codec buffers used for keeping state.
|
| 382 |
|
| 383 | Calling this method should ensure that the data on the
|
| 384 | output is put into a clean state, that allows appending
|
| 385 | of new fresh data without having to rescan the whole
|
| 386 | stream to recover state.
|
| 387 |
|
| 388 | """
|
| 389 | pass
|
| 390 |
|
| 391 | def seek(self, offset, whence=0):
|
| 392 | self.stream.seek(offset, whence)
|
| 393 | if whence == 0 and offset == 0:
|
| 394 | self.reset()
|
| 395 |
|
| 396 | def __getattr__(self, name,
|
| 397 | getattr=getattr):
|
| 398 |
|
| 399 | """ Inherit all other methods from the underlying stream.
|
| 400 | """
|
| 401 | return getattr(self.stream, name)
|
| 402 |
|
| 403 | def __enter__(self):
|
| 404 | return self
|
| 405 |
|
| 406 | def __exit__(self, type, value, tb):
|
| 407 | self.stream.close()
|
| 408 |
|
| 409 | ###
|
| 410 |
|
| 411 | class StreamReader(Codec):
|
| 412 |
|
| 413 | def __init__(self, stream, errors='strict'):
|
| 414 |
|
| 415 | """ Creates a StreamReader instance.
|
| 416 |
|
| 417 | stream must be a file-like object open for reading
|
| 418 | (binary) data.
|
| 419 |
|
| 420 | The StreamReader may use different error handling
|
| 421 | schemes by providing the errors keyword argument. These
|
| 422 | parameters are predefined:
|
| 423 |
|
| 424 | 'strict' - raise a ValueError (or a subclass)
|
| 425 | 'ignore' - ignore the character and continue with the next
|
| 426 | 'replace'- replace with a suitable replacement character;
|
| 427 |
|
| 428 | The set of allowed parameter values can be extended via
|
| 429 | register_error.
|
| 430 | """
|
| 431 | self.stream = stream
|
| 432 | self.errors = errors
|
| 433 | self.bytebuffer = ""
|
| 434 | # For str->str decoding this will stay a str
|
| 435 | # For str->unicode decoding the first read will promote it to unicode
|
| 436 | self.charbuffer = ""
|
| 437 | self.linebuffer = None
|
| 438 |
|
| 439 | def decode(self, input, errors='strict'):
|
| 440 | raise NotImplementedError
|
| 441 |
|
| 442 | def read(self, size=-1, chars=-1, firstline=False):
|
| 443 |
|
| 444 | """ Decodes data from the stream self.stream and returns the
|
| 445 | resulting object.
|
| 446 |
|
| 447 | chars indicates the number of characters to read from the
|
| 448 | stream. read() will never return more than chars
|
| 449 | characters, but it might return less, if there are not enough
|
| 450 | characters available.
|
| 451 |
|
| 452 | size indicates the approximate maximum number of bytes to
|
| 453 | read from the stream for decoding purposes. The decoder
|
| 454 | can modify this setting as appropriate. The default value
|
| 455 | -1 indicates to read and decode as much as possible. size
|
| 456 | is intended to prevent having to decode huge files in one
|
| 457 | step.
|
| 458 |
|
| 459 | If firstline is true, and a UnicodeDecodeError happens
|
| 460 | after the first line terminator in the input only the first line
|
| 461 | will be returned, the rest of the input will be kept until the
|
| 462 | next call to read().
|
| 463 |
|
| 464 | The method should use a greedy read strategy meaning that
|
| 465 | it should read as much data as is allowed within the
|
| 466 | definition of the encoding and the given size, e.g. if
|
| 467 | optional encoding endings or state markers are available
|
| 468 | on the stream, these should be read too.
|
| 469 | """
|
| 470 | # If we have lines cached, first merge them back into characters
|
| 471 | if self.linebuffer:
|
| 472 | self.charbuffer = "".join(self.linebuffer)
|
| 473 | self.linebuffer = None
|
| 474 |
|
| 475 | # read until we get the required number of characters (if available)
|
| 476 | while True:
|
| 477 | # can the request be satisfied from the character buffer?
|
| 478 | if chars >= 0:
|
| 479 | if len(self.charbuffer) >= chars:
|
| 480 | break
|
| 481 | elif size >= 0:
|
| 482 | if len(self.charbuffer) >= size:
|
| 483 | break
|
| 484 | # we need more data
|
| 485 | if size < 0:
|
| 486 | newdata = self.stream.read()
|
| 487 | else:
|
| 488 | newdata = self.stream.read(size)
|
| 489 | # decode bytes (those remaining from the last call included)
|
| 490 | data = self.bytebuffer + newdata
|
| 491 | try:
|
| 492 | newchars, decodedbytes = self.decode(data, self.errors)
|
| 493 | except UnicodeDecodeError, exc:
|
| 494 | if firstline:
|
| 495 | newchars, decodedbytes = self.decode(data[:exc.start], self.errors)
|
| 496 | lines = newchars.splitlines(True)
|
| 497 | if len(lines)<=1:
|
| 498 | raise
|
| 499 | else:
|
| 500 | raise
|
| 501 | # keep undecoded bytes until the next call
|
| 502 | self.bytebuffer = data[decodedbytes:]
|
| 503 | # put new characters in the character buffer
|
| 504 | self.charbuffer += newchars
|
| 505 | # there was no data available
|
| 506 | if not newdata:
|
| 507 | break
|
| 508 | if chars < 0:
|
| 509 | # Return everything we've got
|
| 510 | result = self.charbuffer
|
| 511 | self.charbuffer = ""
|
| 512 | else:
|
| 513 | # Return the first chars characters
|
| 514 | result = self.charbuffer[:chars]
|
| 515 | self.charbuffer = self.charbuffer[chars:]
|
| 516 | return result
|
| 517 |
|
| 518 | def readline(self, size=None, keepends=True):
|
| 519 |
|
| 520 | """ Read one line from the input stream and return the
|
| 521 | decoded data.
|
| 522 |
|
| 523 | size, if given, is passed as size argument to the
|
| 524 | read() method.
|
| 525 |
|
| 526 | """
|
| 527 | # If we have lines cached from an earlier read, return
|
| 528 | # them unconditionally
|
| 529 | if self.linebuffer:
|
| 530 | line = self.linebuffer[0]
|
| 531 | del self.linebuffer[0]
|
| 532 | if len(self.linebuffer) == 1:
|
| 533 | # revert to charbuffer mode; we might need more data
|
| 534 | # next time
|
| 535 | self.charbuffer = self.linebuffer[0]
|
| 536 | self.linebuffer = None
|
| 537 | if not keepends:
|
| 538 | line = line.splitlines(False)[0]
|
| 539 | return line
|
| 540 |
|
| 541 | readsize = size or 72
|
| 542 | line = ""
|
| 543 | # If size is given, we call read() only once
|
| 544 | while True:
|
| 545 | data = self.read(readsize, firstline=True)
|
| 546 | if data:
|
| 547 | # If we're at a "\r" read one extra character (which might
|
| 548 | # be a "\n") to get a proper line ending. If the stream is
|
| 549 | # temporarily exhausted we return the wrong line ending.
|
| 550 | if data.endswith("\r"):
|
| 551 | data += self.read(size=1, chars=1)
|
| 552 |
|
| 553 | line += data
|
| 554 | lines = line.splitlines(True)
|
| 555 | if lines:
|
| 556 | if len(lines) > 1:
|
| 557 | # More than one line result; the first line is a full line
|
| 558 | # to return
|
| 559 | line = lines[0]
|
| 560 | del lines[0]
|
| 561 | if len(lines) > 1:
|
| 562 | # cache the remaining lines
|
| 563 | lines[-1] += self.charbuffer
|
| 564 | self.linebuffer = lines
|
| 565 | self.charbuffer = None
|
| 566 | else:
|
| 567 | # only one remaining line, put it back into charbuffer
|
| 568 | self.charbuffer = lines[0] + self.charbuffer
|
| 569 | if not keepends:
|
| 570 | line = line.splitlines(False)[0]
|
| 571 | break
|
| 572 | line0withend = lines[0]
|
| 573 | line0withoutend = lines[0].splitlines(False)[0]
|
| 574 | if line0withend != line0withoutend: # We really have a line end
|
| 575 | # Put the rest back together and keep it until the next call
|
| 576 | self.charbuffer = "".join(lines[1:]) + self.charbuffer
|
| 577 | if keepends:
|
| 578 | line = line0withend
|
| 579 | else:
|
| 580 | line = line0withoutend
|
| 581 | break
|
| 582 | # we didn't get anything or this was our only try
|
| 583 | if not data or size is not None:
|
| 584 | if line and not keepends:
|
| 585 | line = line.splitlines(False)[0]
|
| 586 | break
|
| 587 | if readsize<8000:
|
| 588 | readsize *= 2
|
| 589 | return line
|
| 590 |
|
| 591 | def readlines(self, sizehint=None, keepends=True):
|
| 592 |
|
| 593 | """ Read all lines available on the input stream
|
| 594 | and return them as list of lines.
|
| 595 |
|
| 596 | Line breaks are implemented using the codec's decoder
|
| 597 | method and are included in the list entries.
|
| 598 |
|
| 599 | sizehint, if given, is ignored since there is no efficient
|
| 600 | way to finding the true end-of-line.
|
| 601 |
|
| 602 | """
|
| 603 | data = self.read()
|
| 604 | return data.splitlines(keepends)
|
| 605 |
|
| 606 | def reset(self):
|
| 607 |
|
| 608 | """ Resets the codec buffers used for keeping state.
|
| 609 |
|
| 610 | Note that no stream repositioning should take place.
|
| 611 | This method is primarily intended to be able to recover
|
| 612 | from decoding errors.
|
| 613 |
|
| 614 | """
|
| 615 | self.bytebuffer = ""
|
| 616 | self.charbuffer = u""
|
| 617 | self.linebuffer = None
|
| 618 |
|
| 619 | def seek(self, offset, whence=0):
|
| 620 | """ Set the input stream's current position.
|
| 621 |
|
| 622 | Resets the codec buffers used for keeping state.
|
| 623 | """
|
| 624 | self.stream.seek(offset, whence)
|
| 625 | self.reset()
|
| 626 |
|
| 627 | def next(self):
|
| 628 |
|
| 629 | """ Return the next decoded line from the input stream."""
|
| 630 | line = self.readline()
|
| 631 | if line:
|
| 632 | return line
|
| 633 | raise StopIteration
|
| 634 |
|
| 635 | def __iter__(self):
|
| 636 | return self
|
| 637 |
|
| 638 | def __getattr__(self, name,
|
| 639 | getattr=getattr):
|
| 640 |
|
| 641 | """ Inherit all other methods from the underlying stream.
|
| 642 | """
|
| 643 | return getattr(self.stream, name)
|
| 644 |
|
| 645 | def __enter__(self):
|
| 646 | return self
|
| 647 |
|
| 648 | def __exit__(self, type, value, tb):
|
| 649 | self.stream.close()
|
| 650 |
|
| 651 | ###
|
| 652 |
|
| 653 | class StreamReaderWriter:
|
| 654 |
|
| 655 | """ StreamReaderWriter instances allow wrapping streams which
|
| 656 | work in both read and write modes.
|
| 657 |
|
| 658 | The design is such that one can use the factory functions
|
| 659 | returned by the codec.lookup() function to construct the
|
| 660 | instance.
|
| 661 |
|
| 662 | """
|
| 663 | # Optional attributes set by the file wrappers below
|
| 664 | encoding = 'unknown'
|
| 665 |
|
| 666 | def __init__(self, stream, Reader, Writer, errors='strict'):
|
| 667 |
|
| 668 | """ Creates a StreamReaderWriter instance.
|
| 669 |
|
| 670 | stream must be a Stream-like object.
|
| 671 |
|
| 672 | Reader, Writer must be factory functions or classes
|
| 673 | providing the StreamReader, StreamWriter interface resp.
|
| 674 |
|
| 675 | Error handling is done in the same way as defined for the
|
| 676 | StreamWriter/Readers.
|
| 677 |
|
| 678 | """
|
| 679 | self.stream = stream
|
| 680 | self.reader = Reader(stream, errors)
|
| 681 | self.writer = Writer(stream, errors)
|
| 682 | self.errors = errors
|
| 683 |
|
| 684 | def read(self, size=-1):
|
| 685 |
|
| 686 | return self.reader.read(size)
|
| 687 |
|
| 688 | def readline(self, size=None):
|
| 689 |
|
| 690 | return self.reader.readline(size)
|
| 691 |
|
| 692 | def readlines(self, sizehint=None):
|
| 693 |
|
| 694 | return self.reader.readlines(sizehint)
|
| 695 |
|
| 696 | def next(self):
|
| 697 |
|
| 698 | """ Return the next decoded line from the input stream."""
|
| 699 | return self.reader.next()
|
| 700 |
|
| 701 | def __iter__(self):
|
| 702 | return self
|
| 703 |
|
| 704 | def write(self, data):
|
| 705 |
|
| 706 | return self.writer.write(data)
|
| 707 |
|
| 708 | def writelines(self, list):
|
| 709 |
|
| 710 | return self.writer.writelines(list)
|
| 711 |
|
| 712 | def reset(self):
|
| 713 |
|
| 714 | self.reader.reset()
|
| 715 | self.writer.reset()
|
| 716 |
|
| 717 | def seek(self, offset, whence=0):
|
| 718 | self.stream.seek(offset, whence)
|
| 719 | self.reader.reset()
|
| 720 | if whence == 0 and offset == 0:
|
| 721 | self.writer.reset()
|
| 722 |
|
| 723 | def __getattr__(self, name,
|
| 724 | getattr=getattr):
|
| 725 |
|
| 726 | """ Inherit all other methods from the underlying stream.
|
| 727 | """
|
| 728 | return getattr(self.stream, name)
|
| 729 |
|
| 730 | # these are needed to make "with codecs.open(...)" work properly
|
| 731 |
|
| 732 | def __enter__(self):
|
| 733 | return self
|
| 734 |
|
| 735 | def __exit__(self, type, value, tb):
|
| 736 | self.stream.close()
|
| 737 |
|
| 738 | ###
|
| 739 |
|
| 740 | class StreamRecoder:
|
| 741 |
|
| 742 | """ StreamRecoder instances provide a frontend - backend
|
| 743 | view of encoding data.
|
| 744 |
|
| 745 | They use the complete set of APIs returned by the
|
| 746 | codecs.lookup() function to implement their task.
|
| 747 |
|
| 748 | Data written to the stream is first decoded into an
|
| 749 | intermediate format (which is dependent on the given codec
|
| 750 | combination) and then written to the stream using an instance
|
| 751 | of the provided Writer class.
|
| 752 |
|
| 753 | In the other direction, data is read from the stream using a
|
| 754 | Reader instance and then return encoded data to the caller.
|
| 755 |
|
| 756 | """
|
| 757 | # Optional attributes set by the file wrappers below
|
| 758 | data_encoding = 'unknown'
|
| 759 | file_encoding = 'unknown'
|
| 760 |
|
| 761 | def __init__(self, stream, encode, decode, Reader, Writer,
|
| 762 | errors='strict'):
|
| 763 |
|
| 764 | """ Creates a StreamRecoder instance which implements a two-way
|
| 765 | conversion: encode and decode work on the frontend (the
|
| 766 | input to .read() and output of .write()) while
|
| 767 | Reader and Writer work on the backend (reading and
|
| 768 | writing to the stream).
|
| 769 |
|
| 770 | You can use these objects to do transparent direct
|
| 771 | recodings from e.g. latin-1 to utf-8 and back.
|
| 772 |
|
| 773 | stream must be a file-like object.
|
| 774 |
|
| 775 | encode, decode must adhere to the Codec interface, Reader,
|
| 776 | Writer must be factory functions or classes providing the
|
| 777 | StreamReader, StreamWriter interface resp.
|
| 778 |
|
| 779 | encode and decode are needed for the frontend translation,
|
| 780 | Reader and Writer for the backend translation. Unicode is
|
| 781 | used as intermediate encoding.
|
| 782 |
|
| 783 | Error handling is done in the same way as defined for the
|
| 784 | StreamWriter/Readers.
|
| 785 |
|
| 786 | """
|
| 787 | self.stream = stream
|
| 788 | self.encode = encode
|
| 789 | self.decode = decode
|
| 790 | self.reader = Reader(stream, errors)
|
| 791 | self.writer = Writer(stream, errors)
|
| 792 | self.errors = errors
|
| 793 |
|
| 794 | def read(self, size=-1):
|
| 795 |
|
| 796 | data = self.reader.read(size)
|
| 797 | data, bytesencoded = self.encode(data, self.errors)
|
| 798 | return data
|
| 799 |
|
| 800 | def readline(self, size=None):
|
| 801 |
|
| 802 | if size is None:
|
| 803 | data = self.reader.readline()
|
| 804 | else:
|
| 805 | data = self.reader.readline(size)
|
| 806 | data, bytesencoded = self.encode(data, self.errors)
|
| 807 | return data
|
| 808 |
|
| 809 | def readlines(self, sizehint=None):
|
| 810 |
|
| 811 | data = self.reader.read()
|
| 812 | data, bytesencoded = self.encode(data, self.errors)
|
| 813 | return data.splitlines(1)
|
| 814 |
|
| 815 | def next(self):
|
| 816 |
|
| 817 | """ Return the next decoded line from the input stream."""
|
| 818 | data = self.reader.next()
|
| 819 | data, bytesencoded = self.encode(data, self.errors)
|
| 820 | return data
|
| 821 |
|
| 822 | def __iter__(self):
|
| 823 | return self
|
| 824 |
|
| 825 | def write(self, data):
|
| 826 |
|
| 827 | data, bytesdecoded = self.decode(data, self.errors)
|
| 828 | return self.writer.write(data)
|
| 829 |
|
| 830 | def writelines(self, list):
|
| 831 |
|
| 832 | data = ''.join(list)
|
| 833 | data, bytesdecoded = self.decode(data, self.errors)
|
| 834 | return self.writer.write(data)
|
| 835 |
|
| 836 | def reset(self):
|
| 837 |
|
| 838 | self.reader.reset()
|
| 839 | self.writer.reset()
|
| 840 |
|
| 841 | def __getattr__(self, name,
|
| 842 | getattr=getattr):
|
| 843 |
|
| 844 | """ Inherit all other methods from the underlying stream.
|
| 845 | """
|
| 846 | return getattr(self.stream, name)
|
| 847 |
|
| 848 | def __enter__(self):
|
| 849 | return self
|
| 850 |
|
| 851 | def __exit__(self, type, value, tb):
|
| 852 | self.stream.close()
|
| 853 |
|
| 854 | ### Shortcuts
|
| 855 |
|
| 856 | def open(filename, mode='rb', encoding=None, errors='strict', buffering=1):
|
| 857 |
|
| 858 | """ Open an encoded file using the given mode and return
|
| 859 | a wrapped version providing transparent encoding/decoding.
|
| 860 |
|
| 861 | Note: The wrapped version will only accept the object format
|
| 862 | defined by the codecs, i.e. Unicode objects for most builtin
|
| 863 | codecs. Output is also codec dependent and will usually be
|
| 864 | Unicode as well.
|
| 865 |
|
| 866 | Files are always opened in binary mode, even if no binary mode
|
| 867 | was specified. This is done to avoid data loss due to encodings
|
| 868 | using 8-bit values. The default file mode is 'rb' meaning to
|
| 869 | open the file in binary read mode.
|
| 870 |
|
| 871 | encoding specifies the encoding which is to be used for the
|
| 872 | file.
|
| 873 |
|
| 874 | errors may be given to define the error handling. It defaults
|
| 875 | to 'strict' which causes ValueErrors to be raised in case an
|
| 876 | encoding error occurs.
|
| 877 |
|
| 878 | buffering has the same meaning as for the builtin open() API.
|
| 879 | It defaults to line buffered.
|
| 880 |
|
| 881 | The returned wrapped file object provides an extra attribute
|
| 882 | .encoding which allows querying the used encoding. This
|
| 883 | attribute is only available if an encoding was specified as
|
| 884 | parameter.
|
| 885 |
|
| 886 | """
|
| 887 | if encoding is not None:
|
| 888 | if 'U' in mode:
|
| 889 | # No automatic conversion of '\n' is done on reading and writing
|
| 890 | mode = mode.strip().replace('U', '')
|
| 891 | if mode[:1] not in set('rwa'):
|
| 892 | mode = 'r' + mode
|
| 893 | if 'b' not in mode:
|
| 894 | # Force opening of the file in binary mode
|
| 895 | mode = mode + 'b'
|
| 896 | file = __builtin__.open(filename, mode, buffering)
|
| 897 | if encoding is None:
|
| 898 | return file
|
| 899 | info = lookup(encoding)
|
| 900 | srw = StreamReaderWriter(file, info.streamreader, info.streamwriter, errors)
|
| 901 | # Add attributes to simplify introspection
|
| 902 | srw.encoding = encoding
|
| 903 | return srw
|
| 904 |
|
| 905 | def EncodedFile(file, data_encoding, file_encoding=None, errors='strict'):
|
| 906 |
|
| 907 | """ Return a wrapped version of file which provides transparent
|
| 908 | encoding translation.
|
| 909 |
|
| 910 | Strings written to the wrapped file are interpreted according
|
| 911 | to the given data_encoding and then written to the original
|
| 912 | file as string using file_encoding. The intermediate encoding
|
| 913 | will usually be Unicode but depends on the specified codecs.
|
| 914 |
|
| 915 | Strings are read from the file using file_encoding and then
|
| 916 | passed back to the caller as string using data_encoding.
|
| 917 |
|
| 918 | If file_encoding is not given, it defaults to data_encoding.
|
| 919 |
|
| 920 | errors may be given to define the error handling. It defaults
|
| 921 | to 'strict' which causes ValueErrors to be raised in case an
|
| 922 | encoding error occurs.
|
| 923 |
|
| 924 | The returned wrapped file object provides two extra attributes
|
| 925 | .data_encoding and .file_encoding which reflect the given
|
| 926 | parameters of the same name. The attributes can be used for
|
| 927 | introspection by Python programs.
|
| 928 |
|
| 929 | """
|
| 930 | if file_encoding is None:
|
| 931 | file_encoding = data_encoding
|
| 932 | data_info = lookup(data_encoding)
|
| 933 | file_info = lookup(file_encoding)
|
| 934 | sr = StreamRecoder(file, data_info.encode, data_info.decode,
|
| 935 | file_info.streamreader, file_info.streamwriter, errors)
|
| 936 | # Add attributes to simplify introspection
|
| 937 | sr.data_encoding = data_encoding
|
| 938 | sr.file_encoding = file_encoding
|
| 939 | return sr
|
| 940 |
|
| 941 | ### Helpers for codec lookup
|
| 942 |
|
| 943 | def getencoder(encoding):
|
| 944 |
|
| 945 | """ Lookup up the codec for the given encoding and return
|
| 946 | its encoder function.
|
| 947 |
|
| 948 | Raises a LookupError in case the encoding cannot be found.
|
| 949 |
|
| 950 | """
|
| 951 | return lookup(encoding).encode
|
| 952 |
|
| 953 | def getdecoder(encoding):
|
| 954 |
|
| 955 | """ Lookup up the codec for the given encoding and return
|
| 956 | its decoder function.
|
| 957 |
|
| 958 | Raises a LookupError in case the encoding cannot be found.
|
| 959 |
|
| 960 | """
|
| 961 | return lookup(encoding).decode
|
| 962 |
|
| 963 | def getincrementalencoder(encoding):
|
| 964 |
|
| 965 | """ Lookup up the codec for the given encoding and return
|
| 966 | its IncrementalEncoder class or factory function.
|
| 967 |
|
| 968 | Raises a LookupError in case the encoding cannot be found
|
| 969 | or the codecs doesn't provide an incremental encoder.
|
| 970 |
|
| 971 | """
|
| 972 | encoder = lookup(encoding).incrementalencoder
|
| 973 | if encoder is None:
|
| 974 | raise LookupError(encoding)
|
| 975 | return encoder
|
| 976 |
|
| 977 | def getincrementaldecoder(encoding):
|
| 978 |
|
| 979 | """ Lookup up the codec for the given encoding and return
|
| 980 | its IncrementalDecoder class or factory function.
|
| 981 |
|
| 982 | Raises a LookupError in case the encoding cannot be found
|
| 983 | or the codecs doesn't provide an incremental decoder.
|
| 984 |
|
| 985 | """
|
| 986 | decoder = lookup(encoding).incrementaldecoder
|
| 987 | if decoder is None:
|
| 988 | raise LookupError(encoding)
|
| 989 | return decoder
|
| 990 |
|
| 991 | def getreader(encoding):
|
| 992 |
|
| 993 | """ Lookup up the codec for the given encoding and return
|
| 994 | its StreamReader class or factory function.
|
| 995 |
|
| 996 | Raises a LookupError in case the encoding cannot be found.
|
| 997 |
|
| 998 | """
|
| 999 | return lookup(encoding).streamreader
|
| 1000 |
|
| 1001 | def getwriter(encoding):
|
| 1002 |
|
| 1003 | """ Lookup up the codec for the given encoding and return
|
| 1004 | its StreamWriter class or factory function.
|
| 1005 |
|
| 1006 | Raises a LookupError in case the encoding cannot be found.
|
| 1007 |
|
| 1008 | """
|
| 1009 | return lookup(encoding).streamwriter
|
| 1010 |
|
| 1011 | def iterencode(iterator, encoding, errors='strict', **kwargs):
|
| 1012 | """
|
| 1013 | Encoding iterator.
|
| 1014 |
|
| 1015 | Encodes the input strings from the iterator using an IncrementalEncoder.
|
| 1016 |
|
| 1017 | errors and kwargs are passed through to the IncrementalEncoder
|
| 1018 | constructor.
|
| 1019 | """
|
| 1020 | encoder = getincrementalencoder(encoding)(errors, **kwargs)
|
| 1021 | for input in iterator:
|
| 1022 | output = encoder.encode(input)
|
| 1023 | if output:
|
| 1024 | yield output
|
| 1025 | output = encoder.encode("", True)
|
| 1026 | if output:
|
| 1027 | yield output
|
| 1028 |
|
| 1029 | def iterdecode(iterator, encoding, errors='strict', **kwargs):
|
| 1030 | """
|
| 1031 | Decoding iterator.
|
| 1032 |
|
| 1033 | Decodes the input strings from the iterator using an IncrementalDecoder.
|
| 1034 |
|
| 1035 | errors and kwargs are passed through to the IncrementalDecoder
|
| 1036 | constructor.
|
| 1037 | """
|
| 1038 | decoder = getincrementaldecoder(encoding)(errors, **kwargs)
|
| 1039 | for input in iterator:
|
| 1040 | output = decoder.decode(input)
|
| 1041 | if output:
|
| 1042 | yield output
|
| 1043 | output = decoder.decode("", True)
|
| 1044 | if output:
|
| 1045 | yield output
|
| 1046 |
|
| 1047 | ### Helpers for charmap-based codecs
|
| 1048 |
|
| 1049 | def make_identity_dict(rng):
|
| 1050 |
|
| 1051 | """ make_identity_dict(rng) -> dict
|
| 1052 |
|
| 1053 | Return a dictionary where elements of the rng sequence are
|
| 1054 | mapped to themselves.
|
| 1055 |
|
| 1056 | """
|
| 1057 | res = {}
|
| 1058 | for i in rng:
|
| 1059 | res[i]=i
|
| 1060 | return res
|
| 1061 |
|
| 1062 | def make_encoding_map(decoding_map):
|
| 1063 |
|
| 1064 | """ Creates an encoding map from a decoding map.
|
| 1065 |
|
| 1066 | If a target mapping in the decoding map occurs multiple
|
| 1067 | times, then that target is mapped to None (undefined mapping),
|
| 1068 | causing an exception when encountered by the charmap codec
|
| 1069 | during translation.
|
| 1070 |
|
| 1071 | One example where this happens is cp875.py which decodes
|
| 1072 | multiple character to \\u001a.
|
| 1073 |
|
| 1074 | """
|
| 1075 | m = {}
|
| 1076 | for k,v in decoding_map.items():
|
| 1077 | if not v in m:
|
| 1078 | m[v] = k
|
| 1079 | else:
|
| 1080 | m[v] = None
|
| 1081 | return m
|
| 1082 |
|
| 1083 | ### error handlers
|
| 1084 |
|
| 1085 | try:
|
| 1086 | strict_errors = lookup_error("strict")
|
| 1087 | ignore_errors = lookup_error("ignore")
|
| 1088 | replace_errors = lookup_error("replace")
|
| 1089 | xmlcharrefreplace_errors = lookup_error("xmlcharrefreplace")
|
| 1090 | backslashreplace_errors = lookup_error("backslashreplace")
|
| 1091 | except LookupError:
|
| 1092 | # In --disable-unicode builds, these error handler are missing
|
| 1093 | strict_errors = None
|
| 1094 | ignore_errors = None
|
| 1095 | replace_errors = None
|
| 1096 | xmlcharrefreplace_errors = None
|
| 1097 | backslashreplace_errors = None
|
| 1098 |
|
| 1099 | # Tell modulefinder that using codecs probably needs the encodings
|
| 1100 | # package
|
| 1101 | _false = 0
|
| 1102 | if _false:
|
| 1103 | import encodings
|
| 1104 |
|
| 1105 | ### Tests
|
| 1106 |
|
| 1107 | if __name__ == '__main__':
|
| 1108 |
|
| 1109 | # Make stdout translate Latin-1 output into UTF-8 output
|
| 1110 | sys.stdout = EncodedFile(sys.stdout, 'latin-1', 'utf-8')
|
| 1111 |
|
| 1112 | # Have stdin translate Latin-1 input into UTF-8 input
|
| 1113 | sys.stdin = EncodedFile(sys.stdin, 'utf-8', 'latin-1')
|