Lib/sunau.py

   1 # Stuff to parse Sun and NeXT audio files.
   2 #
   3 # An audio consists of a header followed by the data.  The structure
   4 # of the header is as follows.
   5 #
   6 #       +---------------+
   7 #       | magic word    |
   8 #       +---------------+
   9 #       | header size   |
  10 #       +---------------+
  11 #       | data size     |
  12 #       +---------------+
  13 #       | encoding      |
  14 #       +---------------+
  15 #       | sample rate   |
  16 #       +---------------+
  17 #       | # of channels |
  18 #       +---------------+
  19 #       | info          |
  20 #       |               |
  21 #       +---------------+
  22 #
  23 # The magic word consists of the 4 characters '.snd'.  Apart from the
  24 # info field, all header fields are 4 bytes in size.  They are all
  25 # 32-bit unsigned integers encoded in big-endian byte order.
  26 #
  27 # The header size really gives the start of the data.
  28 # The data size is the physical size of the data.  From the other
  29 # parameter the number of frames can be calculated.
  30 # The encoding gives the way in which audio samples are encoded.
  31 # Possible values are listed below.
  32 # The info field currently consists of an ASCII string giving a
  33 # human-readable description of the audio file.  The info field is
  34 # padded with NUL bytes to the header size.
  35 #
  36 # Usage.
  37 #
  38 # Reading audio files:
  39 #       f = sunau.open(file, 'r')
  40 # where file is either the name of a file or an open file pointer.
  41 # The open file pointer must have methods read(), seek(), and close().
  42 # When the setpos() and rewind() methods are not used, the seek()
  43 # method is not  necessary.
  44 #
  45 # This returns an instance of a class with the following public methods:
  46 #       getnchannels()  -- returns number of audio channels (1 for
  47 #                          mono, 2 for stereo)
  48 #       getsampwidth()  -- returns sample width in bytes
  49 #       getframerate()  -- returns sampling frequency
  50 #       getnframes()    -- returns number of audio frames
  51 #       getcomptype()   -- returns compression type ('NONE' for AIFF files)
  52 #       getcompname()   -- returns human-readable version of
  53 #                          compression type ('not compressed' for AIFF files)
  54 #       getparams()     -- returns a tuple consisting of all of the
  55 #                          above in the above order
  56 #       getmarkers()    -- returns None (for compatibility with the
  57 #                          aifc module)
  58 #       getmark(id)     -- raises an error since the mark does not
  59 #                          exist (for compatibility with the aifc module)
  60 #       readframes(n)   -- returns at most n frames of audio
  61 #       rewind()        -- rewind to the beginning of the audio stream
  62 #       setpos(pos)     -- seek to the specified position
  63 #       tell()          -- return the current position
  64 #       close()         -- close the instance (make it unusable)
  65 # The position returned by tell() and the position given to setpos()
  66 # are compatible and have nothing to do with the actual postion in the
  67 # file.
  68 # The close() method is called automatically when the class instance
  69 # is destroyed.
  70 #
  71 # Writing audio files:
  72 #       f = sunau.open(file, 'w')
  73 # where file is either the name of a file or an open file pointer.
  74 # The open file pointer must have methods write(), tell(), seek(), and
  75 # close().
  76 #
  77 # This returns an instance of a class with the following public methods:
  78 #       setnchannels(n) -- set the number of channels
  79 #       setsampwidth(n) -- set the sample width
  80 #       setframerate(n) -- set the frame rate
  81 #       setnframes(n)   -- set the number of frames
  82 #       setcomptype(type, name)
  83 #                       -- set the compression type and the
  84 #                          human-readable compression type
  85 #       setparams(nchannels, sampwidth, framerate, nframes, comptype, compname)
  86 #                       -- set all parameters at once
  87 #       tell()          -- return current position in output file
  88 #       writeframesraw(data)
  89 #                       -- write audio frames without pathing up the
  90 #                          file header
  91 #       writeframes(data)
  92 #                       -- write audio frames and patch up the file header
  93 #       close()         -- patch up the file header and close the
  94 #                          output file
  95 # You should set the parameters before the first writeframesraw or
  96 # writeframes.  The total number of frames does not need to be set,
  97 # but when it is set to the correct value, the header does not have to
  98 # be patched up.
  99 # It is best to first set all parameters, perhaps possibly the
 100 # compression type, and then write audio frames using writeframesraw.
 101 # When all frames have been written, either call writeframes('') or
 102 # close() to patch up the sizes in the header.
 103 # The close() method is called automatically when the class instance
 104 # is destroyed.
 105
 106 # from <multimedia/audio_filehdr.h>
 107 AUDIO_FILE_MAGIC = 0x2e736e64
 108 AUDIO_FILE_ENCODING_MULAW_8 = 1
 109 AUDIO_FILE_ENCODING_LINEAR_8 = 2
 110 AUDIO_FILE_ENCODING_LINEAR_16 = 3
 111 AUDIO_FILE_ENCODING_LINEAR_24 = 4
 112 AUDIO_FILE_ENCODING_LINEAR_32 = 5
 113 AUDIO_FILE_ENCODING_FLOAT = 6
 114 AUDIO_FILE_ENCODING_DOUBLE = 7
 115 AUDIO_FILE_ENCODING_ADPCM_G721 = 23
 116 AUDIO_FILE_ENCODING_ADPCM_G722 = 24
 117 AUDIO_FILE_ENCODING_ADPCM_G723_3 = 25
 118 AUDIO_FILE_ENCODING_ADPCM_G723_5 = 26
 119 AUDIO_FILE_ENCODING_ALAW_8 = 27
 120
 121 # from <multimedia/audio_hdr.h>
 122 AUDIO_UNKNOWN_SIZE = 0xFFFFFFFFL        # ((unsigned)(~0))
 123
 124 _simple_encodings = [AUDIO_FILE_ENCODING_MULAW_8,
 125                      AUDIO_FILE_ENCODING_LINEAR_8,
 126                      AUDIO_FILE_ENCODING_LINEAR_16,
 127                      AUDIO_FILE_ENCODING_LINEAR_24,
 128                      AUDIO_FILE_ENCODING_LINEAR_32,
 129                      AUDIO_FILE_ENCODING_ALAW_8]
 130
 131 def _read_u32(file):
 132         x = 0L
 133         for i in range(4):
 134                 byte = file.read(1)
 135                 if byte == '':
 136                         raise EOFError
 137                 x = x*256 + ord(byte)
 138         return x
 139
 140 def _write_u32(file, x):
 141         data = []
 142         for i in range(4):
 143                 d, m = divmod(x, 256)
 144                 data.insert(0, m)
 145                 x = d
 146         for i in range(4):
 147                 file.write(chr(int(data[i])))
 148
 149 class Au_read:
 150         access _file, _soundpos, _hdr_size, _data_size, _encoding, \
 151                   _sampwidth, _framesize, _framerate, _nchannels, \
 152                   _info: private
 153
 154         def __init__(self, f):
 155                 if type(f) == type(''):
 156                         import __builtin__
 157                         f = __builtin__.open(f, 'r')
 158                 self.initfp(f)
 159
 160         def __del__(self):
 161                 if self._file:
 162                         self.close()
 163
 164         def initfp(self, file):
 165                 self._file = file
 166                 self._soundpos = 0
 167                 magic = int(_read_u32(file))
 168                 if magic != AUDIO_FILE_MAGIC:
 169                         raise Error, 'bad magic number'
 170                 self._hdr_size = int(_read_u32(file))
 171                 if self._hdr_size < 24:
 172                         raise Error, 'header size too small'
 173                 if self._hdr_size > 100:
 174                         raise Error, 'header size rediculously large'
 175                 self._data_size = _read_u32(file)
 176                 if self._data_size != AUDIO_UNKNOWN_SIZE:
 177                         self._data_size = int(self._data_size)
 178                 self._encoding = int(_read_u32(file))
 179                 if self._encoding not in _simple_encodings:
 180                         raise Error, 'encoding not (yet) supported'
 181                 if self._encoding in (AUDIO_FILE_ENCODING_MULAW_8,
 182                           AUDIO_FILE_ENCODING_LINEAR_8,
 183                           AUDIO_FILE_ENCODING_ALAW_8):
 184                         self._sampwidth = 2
 185                         self._framesize = 1
 186                 elif self._encoding == AUDIO_FILE_ENCODING_LINEAR_16:
 187                         self._framesize = self._sampwidth = 2
 188                 elif self._encoding == AUDIO_FILE_ENCODING_LINEAR_24:
 189                         self._framesize = self._sampwidth = 3
 190                 elif self._encoding == AUDIO_FILE_ENCODING_LINEAR_32:
 191                         self._framesize = self._sampwidth = 4
 192                 else:
 193                         raise Error, 'unknown encoding'
 194                 self._framerate = int(_read_u32(file))
 195                 self._nchannels = int(_read_u32(file))
 196                 self._framesize = self._framesize * self._nchannels
 197                 if self._hdr_size > 24:
 198                         self._info = file.read(self._hdr_size - 24)
 199                         for i in range(len(self._info)):
 200                                 if self._info[i] == '\0':
 201                                         self._info = self._info[:i]
 202                                         break
 203                 else:
 204                         self._info = ''
 205
 206         def getfp(self):
 207                 return self._file
 208
 209         def getnchannels(self):
 210                 return self._nchannels
 211
 212         def getsampwidth(self):
 213                 return self._sampwidth
 214
 215         def getframerate(self):
 216                 return self._framerate
 217
 218         def getnframes(self):
 219                 if self._data_size == AUDIO_UNKNOWN_SIZE:
 220                         return AUDIO_UNKNOWN_SIZE
 221                 if self._encoding in _simple_encodings:
 222                         return self._data_size / self._framesize
 223                 return 0                # XXX--must do some arithmetic here
 224
 225         def getcomptype(self):
 226                 if self._encoding == AUDIO_FILE_ENCODING_MULAW_8:
 227                         return 'ULAW'
 228                 elif self._encoding == AUDIO_FILE_ENCODING_ALAW_8:
 229                         return 'ALAW'
 230                 else:
 231                         return 'NONE'
 232
 233         def getcompname(self):
 234                 if self._encoding == AUDIO_FILE_ENCODING_MULAW_8:
 235                         return 'CCITT G.711 u-law'
 236                 elif self._encoding == AUDIO_FILE_ENCODING_ALAW_8:
 237                         return 'CCITT G.711 A-law'
 238                 else:
 239                         return 'not compressed'
 240
 241         def getparams(self):
 242                 return self.getnchannels(), self.getsampwidth(), \
 243                           self.getframerate(), self.getnframes(), \
 244                           self.getcomptype(), self.getcompname()
 245
 246         def getmarkers(self):
 247                 return None
 248
 249         def getmark(self, id):
 250                 raise Error, 'no marks'
 251
 252         def readframes(self, nframes):
 253                 if self._encoding in _simple_encodings:
 254                         if nframes == AUDIO_UNKNOWN_SIZE:
 255                                 data = self._file.read()
 256                         else:
 257                                 data = self._file.read(nframes * self._framesize * self._nchannels)
 258                         if self._encoding == AUDIO_FILE_ENCODING_MULAW_8:
 259                                 import audioop
 260                                 data = audioop.ulaw2lin(data, self._sampwidth)
 261                         return data
 262                 return None             # XXX--not implemented yet
 263
 264         def rewind(self):
 265                 self._soundpos = 0
 266                 self._file.seek(self._hdr_size)
 267
 268         def tell(self):
 269                 return self._soundpos
 270
 271         def setpos(self, pos):
 272                 if pos < 0 or pos > self.getnframes():
 273                         raise Error, 'position not in range'
 274                 self._file.seek(pos * self._framesize + self._hdr_size)
 275                 self._soundpos = pos
 276
 277         def close(self):
 278                 self._file = None
 279
 280 class Au_write:
 281         access _file, _framerate, _nchannels, _sampwidth, _framesize, \
 282                   _nframes, _nframeswritten, _datawritten, _info, \
 283                   _comptype: private
 284
 285         def __init__(self, f):
 286                 if type(f) == type(''):
 287                         import __builtin__
 288                         f = __builtin__.open(f, 'w')
 289                 self.initfp(f)
 290
 291         def __del__(self):
 292                 if self._file:
 293                         self.close()
 294
 295         def initfp(self, file):
 296                 self._file = file
 297                 self._framerate = 0
 298                 self._nchannels = 0
 299                 self._sampwidth = 0
 300                 self._framesize = 0
 301                 self._nframes = AUDIO_UNKNOWN_SIZE
 302                 self._nframeswritten = 0
 303                 self._datawritten = 0
 304                 self._datalength = 0
 305                 self._info = ''
 306                 self._comptype = 'ULAW' # default is U-law
 307
 308         def setnchannels(self, nchannels):
 309                 if self._nframeswritten:
 310                         raise Error, 'cannot change parameters after starting to write'
 311                 if nchannels not in (1, 2, 4):
 312                         raise Error, 'only 1, 2, or 4 channels supported'
 313                 self._nchannels = nchannels
 314
 315         def getnchannels(self):
 316                 if not self._nchannels:
 317                         raise Error, 'number of channels not set'
 318                 return self._nchannels
 319
 320         def setsampwidth(self, sampwidth):
 321                 if self._nframeswritten:
 322                         raise Error, 'cannot change parameters after starting to write'
 323                 if sampwidth not in (1, 2, 4):
 324                         raise Error, 'bad sample width'
 325                 self._sampwidth = sampwidth
 326
 327         def getsampwidth(self):
 328                 if not self._framerate:
 329                         raise Error, 'sample width not specified'
 330                 return self._sampwidth
 331
 332         def setframerate(self, framerate):
 333                 if self._nframeswritten:
 334                         raise Error, 'cannot change parameters after starting to write'
 335                 self._framerate = framerate
 336
 337         def getframerate(self):
 338                 if not self._framerate:
 339                         raise Error, 'frame rate not set'
 340                 return self._framerate
 341
 342         def setnframes(self, nframes):
 343                 if self._nframeswritten:
 344                         raise Error, 'cannot change parameters after starting to write'
 345                 if nframes < 0:
 346                         raise Error, '# of frames cannot be negative'
 347                 self._nframes = nframes
 348
 349         def getnframes(self):
 350                 return self._nframeswritten
 351
 352         def setcomptype(self, type, name):
 353                 if type in ('NONE', 'ULAW'):
 354                         self._comptype = type
 355                 else:
 356                         raise Error, 'unknown compression type'
 357
 358         def getcomptype(self):
 359                 return self._comptype
 360
 361         def getcompname(self):
 362                 if self._comptype == 'ULAW':
 363                         return 'CCITT G.711 u-law'
 364                 elif self._comptype == 'ALAW':
 365                         return 'CCITT G.711 A-law'
 366                 else:
 367                         return 'not compressed'
 368
 369         def setparams(self, (nchannels, sampwidth, framerate, nframes, comptype, compname)):
 370                 self.setnchannels(nchannels)
 371                 self.setsampwidth(sampwidth)
 372                 self.setframerate(framerate)
 373                 self.setnframes(nframes)
 374                 self.setcomptype(comptype, compname)
 375
 376         def getparams(self):
 377                 return self.getnchannels(), self.getsampwidth(), \
 378                           self.getframerate(), self.getnframes(), \
 379                           self.getcomptype(), self.getcompname()
 380
 381         def tell(self):
 382                 return self._nframeswritten
 383
 384         def writeframesraw(self, data):
 385                 self._ensure_header_written()
 386                 nframes = len(data) / self._framesize
 387                 if self._comptype == 'ULAW':
 388                         import audioop
 389                         data = audioop.lin2ulaw(data, self._sampwidth)
 390                 self._file.write(data)
 391                 self._nframeswritten = self._nframeswritten + nframes
 392                 self._datawritten = self._datawritten + len(data)
 393
 394         def writeframes(self, data):
 395                 self.writeframesraw(data)
 396                 if self._nframeswritten != self._nframes or \
 397                           self._datalength != self._datawritten:
 398                         self._patchheader()
 399
 400         def close(self):
 401                 self._ensure_header_written()
 402                 if self._nframeswritten != self._nframes or \
 403                           self._datalength != self._datawritten:
 404                         self._patchheader()
 405                 self._file.flush()
 406                 self._file = None
 407
 408         #
 409         # private methods
 410         #
 411         access *: private
 412
 413         def _ensure_header_written(self):
 414                 if not self._nframeswritten:
 415                         if not self._nchannels:
 416                                 raise Error, '# of channels not specified'
 417                         if not self._sampwidth:
 418                                 raise Error, 'sample width not specified'
 419                         if not self._framerate:
 420                                 raise Error, 'frame rate not specified'
 421                         self._write_header()
 422
 423         def _write_header(self):
 424                 if self._comptype == 'NONE':
 425                         if self._sampwidth == 1:
 426                                 encoding = AUDIO_FILE_ENCODING_LINEAR_8
 427                                 self._framesize = 1
 428                         elif self._sampwidth == 2:
 429                                 encoding = AUDIO_FILE_ENCODING_LINEAR_16
 430                                 self._framesize = 2
 431                         elif self._sampwidth == 4:
 432                                 encoding = AUDIO_FILE_ENCODING_LINEAR_32
 433                                 self._framesize = 4
 434                         else:
 435                                 raise Error, 'internal error'
 436                 elif self._comptype == 'ULAW':
 437                         encoding = AUDIO_FILE_ENCODING_MULAW_8
 438                         self._framesize = 1
 439                 else:
 440                         raise Error, 'internal error'
 441                 self._framesize = self._framesize * self._nchannels
 442                 _write_u32(self._file, AUDIO_FILE_MAGIC)
 443                 header_size = 25 + len(self._info)
 444                 header_size = (header_size + 7) & ~7
 445                 _write_u32(self._file, header_size)
 446                 if self._nframes == AUDIO_UNKNOWN_SIZE:
 447                         length = AUDIO_UNKNOWN_SIZE
 448                 else:
 449                         length = self._nframes * self._framesize
 450                 _write_u32(self._file, length)
 451                 self._datalength = length
 452                 _write_u32(self._file, encoding)
 453                 _write_u32(self._file, self._framerate)
 454                 _write_u32(self._file, self._nchannels)
 455                 self._file.write(self._info)
 456                 self._file.write('\0'*(header_size - len(self._info) - 24))
 457
 458         def _patchheader(self):
 459                 self._file.seek(8)
 460                 _write_u32(self._file, self._datawritten)
 461                 self._datalength = self._datawritten
 462                 self._file.seek(0, 2)
 463
 464 def open(f, mode):
 465         if mode == 'r':
 466                 return Au_read(f)
 467         elif mode == 'w':
 468                 return Au_write(f)
 469         else:
 470                 raise Error, "mode must be 'r' or 'w'"
 471
 472 openfp = open