508 lines
19 KiB
Python
508 lines
19 KiB
Python
#!/usr/bin/python
|
|
|
|
#Audio Tools, a module and set of tools for manipulating audio data
|
|
#Copyright (C) 2007-2011 Brian Langenberger
|
|
|
|
#This program is free software; you can redistribute it and/or modify
|
|
#it under the terms of the GNU General Public License as published by
|
|
#the Free Software Foundation; either version 2 of the License, or
|
|
#(at your option) any later version.
|
|
|
|
#This program is distributed in the hope that it will be useful,
|
|
#but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
#MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
#GNU General Public License for more details.
|
|
|
|
#You should have received a copy of the GNU General Public License
|
|
#along with this program; if not, write to the Free Software
|
|
#Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
|
|
|
|
from audiotools import (AudioFile, ChannelMask, PCMReader,
|
|
transfer_framelist_data, WaveAudio,
|
|
AiffAudio, cStringIO, EncodingError,
|
|
UnsupportedBitsPerSample, InvalidFile,
|
|
PCMReaderError,
|
|
WaveContainer, AiffContainer, to_pcm_progress)
|
|
|
|
import audiotools.decoders
|
|
import os.path
|
|
|
|
|
|
class InvalidShorten(InvalidFile):
|
|
pass
|
|
|
|
|
|
class ShortenAudio(WaveContainer, AiffContainer):
|
|
"""A Shorten audio file."""
|
|
|
|
SUFFIX = "shn"
|
|
NAME = SUFFIX
|
|
|
|
def __init__(self, filename):
|
|
"""filename is a plain string."""
|
|
|
|
AudioFile.__init__(self, filename)
|
|
try:
|
|
f = open(filename, 'rb')
|
|
except IOError, msg:
|
|
raise InvalidShorten(str(msg))
|
|
try:
|
|
if (not ShortenAudio.is_type(f)):
|
|
raise InvalidShorten(_(u'Shorten header not detected'))
|
|
finally:
|
|
f.close()
|
|
|
|
#Why not call __populate_metadata__ here and raise InvalidShorten
|
|
#if it errors out?
|
|
#The problem is that __populate_metadata__ needs to walk
|
|
#through the *entire* file in order to calculate total PCM frames
|
|
#and so on.
|
|
#That's an expensive operation to perform at init-time
|
|
#so it's better to postpone it to an on-demand fetch.
|
|
|
|
def __populate_metadata__(self):
|
|
#set up some default values
|
|
self.__bits_per_sample__ = 16
|
|
self.__channels__ = 2
|
|
self.__channel_mask__ = 0x3
|
|
self.__sample_rate__ = 44100
|
|
self.__total_frames__ = 0
|
|
self.__blocks__ = []
|
|
self.__format__ = None
|
|
|
|
#grab a few pieces of technical metadata from the Shorten file itself
|
|
#which requires a dry-run through the decoder
|
|
try:
|
|
decoder = audiotools.decoders.SHNDecoder(self.filename)
|
|
try:
|
|
|
|
self.__bits_per_sample__ = decoder.bits_per_sample
|
|
self.__channels__ = decoder.channels
|
|
(self.__total_frames__,
|
|
self.__blocks__) = decoder.metadata()
|
|
finally:
|
|
decoder.close()
|
|
|
|
try:
|
|
self.__channel_mask__ = ChannelMask.from_channels(
|
|
self.__channels__)
|
|
except ValueError:
|
|
self.__channel_mask__ = 0
|
|
except (ValueError, IOError):
|
|
#if we hit an error in SHNDecoder while reading
|
|
#technical metadata, the default values will have to do
|
|
return
|
|
|
|
#the remainder requires parsing the file's VERBATIM blocks
|
|
#which may contain Wave, AIFF or Sun AU info
|
|
if (self.__blocks__[0] is not None):
|
|
header = cStringIO.StringIO(self.__blocks__[0])
|
|
for format in WaveAudio, AiffAudio:
|
|
header.seek(0, 0)
|
|
if (format.is_type(header)):
|
|
self.__format__ = format
|
|
break
|
|
if (self.__format__ is WaveAudio):
|
|
for (chunk_id, chunk_data) in self.__wave_chunks__():
|
|
if (chunk_id == 'fmt '):
|
|
fmt_chunk = WaveAudio.FMT_CHUNK.parse(chunk_data)
|
|
self.__sample_rate__ = fmt_chunk.sample_rate
|
|
if (fmt_chunk.compression == 0xFFFE):
|
|
self.__channel_mask__ = \
|
|
WaveAudio.fmt_chunk_to_channel_mask(
|
|
fmt_chunk.channel_mask)
|
|
elif (self.__format__ is AiffAudio):
|
|
for (chunk_id, chunk_data) in self.__aiff_chunks__():
|
|
if (chunk_id == 'COMM'):
|
|
comm_chunk = AiffAudio.COMM_CHUNK.parse(chunk_data)
|
|
self.__sample_rate__ = comm_chunk.sample_rate
|
|
|
|
def __wave_chunks__(self):
|
|
total_size = sum([len(block) for block in self.__blocks__
|
|
if block is not None])
|
|
wave_data = cStringIO.StringIO("".join([block for block in
|
|
self.__blocks__
|
|
if block is not None]))
|
|
|
|
wave_data.read(12) # skip the RIFFxxxxWAVE header data
|
|
total_size -= 12
|
|
|
|
#iterate over all the non-data chunks
|
|
while (total_size > 0):
|
|
header = WaveAudio.CHUNK_HEADER.parse_stream(wave_data)
|
|
total_size -= 8
|
|
if (header.chunk_id != 'data'):
|
|
yield (header.chunk_id, wave_data.read(header.chunk_length))
|
|
total_size -= header.chunk_length
|
|
else:
|
|
continue
|
|
|
|
def __aiff_chunks__(self):
|
|
total_size = sum([len(block) for block in self.__blocks__
|
|
if block is not None])
|
|
aiff_data = cStringIO.StringIO("".join([block for block in
|
|
self.__blocks__
|
|
if block is not None]))
|
|
|
|
aiff_data.read(12) # skip the FORMxxxxAIFF header data
|
|
total_size -= 12
|
|
|
|
#iterate over all the chunks
|
|
while (total_size > 0):
|
|
header = AiffAudio.CHUNK_HEADER.parse_stream(aiff_data)
|
|
total_size -= 8
|
|
if (header.chunk_id != 'SSND'):
|
|
yield (header.chunk_id, aiff_data.read(header.chunk_length))
|
|
total_size -= header.chunk_length
|
|
else:
|
|
#This presumes that audiotools encoded
|
|
#the Shorten file from an AIFF source.
|
|
#The reference encoder places the 8 alignment
|
|
#bytes in the PCM stream itself, which is wrong.
|
|
yield (header.chunk_id, aiff_data.read(8))
|
|
total_size -= 8
|
|
|
|
@classmethod
|
|
def is_type(cls, file):
|
|
"""Returns True if the given file object describes this format.
|
|
|
|
Takes a seekable file pointer rewound to the start of the file."""
|
|
|
|
return (file.read(4) == 'ajkg') and (ord(file.read(1)) == 2)
|
|
|
|
def bits_per_sample(self):
|
|
"""Returns an integer number of bits-per-sample this track contains."""
|
|
|
|
if (not hasattr(self, "__bits_per_sample__")):
|
|
self.__populate_metadata__()
|
|
return self.__bits_per_sample__
|
|
|
|
def channels(self):
|
|
"""Returns an integer number of channels this track contains."""
|
|
|
|
if (not hasattr(self, "__channels__")):
|
|
self.__populate_metadata__()
|
|
return self.__channels__
|
|
|
|
def channel_mask(self):
|
|
"""Returns a ChannelMask object of this track's channel layout."""
|
|
|
|
if (not hasattr(self, "__channel_mask__")):
|
|
self.__populate_metadata__()
|
|
return self.__channel_mask__
|
|
|
|
def lossless(self):
|
|
"""Returns True."""
|
|
|
|
return True
|
|
|
|
def total_frames(self):
|
|
"""Returns the total PCM frames of the track as an integer."""
|
|
|
|
if (not hasattr(self, "__total_frames__")):
|
|
self.__populate_metadata__()
|
|
return self.__total_frames__
|
|
|
|
def sample_rate(self):
|
|
"""Returns the rate of the track's audio as an integer number of Hz."""
|
|
|
|
if (not hasattr(self, "__sample_rate__")):
|
|
self.__populate_metadata__()
|
|
return self.__sample_rate__
|
|
|
|
def to_pcm(self):
|
|
"""Returns a PCMReader object containing the track's PCM data."""
|
|
|
|
try:
|
|
sample_rate = self.sample_rate()
|
|
channels = self.channels()
|
|
channel_mask = int(self.channel_mask())
|
|
bits_per_sample = self.bits_per_sample()
|
|
|
|
decoder = audiotools.decoders.SHNDecoder(self.filename)
|
|
decoder.sample_rate = sample_rate
|
|
decoder.channel_mask = channel_mask
|
|
return decoder
|
|
except (IOError, ValueError), msg:
|
|
#these may not be accurate if the Shorten file is broken
|
|
#but if it is broken, there'll be no way to
|
|
#cross-check the results anyway
|
|
return PCMReaderError(error_message=str(msg),
|
|
sample_rate=44100,
|
|
channels=2,
|
|
channel_mask=0x3,
|
|
bits_per_sample=16)
|
|
|
|
@classmethod
|
|
def from_pcm(cls, filename, pcmreader, compression=None,
|
|
block_size=256):
|
|
"""Encodes a new file from PCM data.
|
|
|
|
Takes a filename string, PCMReader object
|
|
and optional compression level string.
|
|
Encodes a new audio file from pcmreader's data
|
|
at the given filename with the specified compression level
|
|
and returns a new ShortenAudio object."""
|
|
|
|
if (pcmreader.bits_per_sample not in (8, 16)):
|
|
raise UnsupportedBitsPerSample(filename, pcmreader.bits_per_sample)
|
|
|
|
import tempfile
|
|
|
|
f = tempfile.NamedTemporaryFile(suffix=".wav")
|
|
try:
|
|
w = WaveAudio.from_pcm(f.name, pcmreader)
|
|
return cls.from_wave(filename, f.name, compression, block_size)
|
|
finally:
|
|
if (os.path.isfile(f.name)):
|
|
f.close()
|
|
else:
|
|
f.close_called = True
|
|
|
|
def to_wave(self, wave_filename, progress=None):
|
|
"""Writes the contents of this file to the given .wav filename string.
|
|
|
|
Raises EncodingError if some error occurs during decoding."""
|
|
|
|
if (not hasattr(self, "__format__")):
|
|
try:
|
|
self.__populate_metadata__()
|
|
except IOError, msg:
|
|
raise EncodingError(str(msg))
|
|
|
|
if (self.__format__ is WaveAudio):
|
|
try:
|
|
f = open(wave_filename, 'wb')
|
|
except IOError, msg:
|
|
raise EncodingError(str(msg))
|
|
for block in self.__blocks__:
|
|
if (block is not None):
|
|
f.write(block)
|
|
else:
|
|
try:
|
|
total_frames = self.total_frames()
|
|
current_frames = 0
|
|
decoder = audiotools.decoders.SHNDecoder(self.filename)
|
|
frame = decoder.read(4096)
|
|
while (len(frame) > 0):
|
|
f.write(frame.to_bytes(False, True))
|
|
current_frames += frame.frames
|
|
if (progress is not None):
|
|
progress(current_frames, total_frames)
|
|
frame = decoder.read(4096)
|
|
except IOError, msg:
|
|
raise EncodingError(str(msg))
|
|
else:
|
|
WaveAudio.from_pcm(wave_filename, to_pcm_progress(self, progress))
|
|
|
|
def to_aiff(self, aiff_filename, progress=None):
|
|
"""Writes the contents of this file to the given .aiff filename string.
|
|
|
|
Raises EncodingError if some error occurs during decoding."""
|
|
|
|
if (not hasattr(self, "__format__")):
|
|
try:
|
|
self.__populate_metadata__()
|
|
except IOError, msg:
|
|
raise EncodingError(str(msg))
|
|
|
|
if (self.__format__ is AiffAudio):
|
|
try:
|
|
f = open(aiff_filename, 'wb')
|
|
except IOError, msg:
|
|
raise EncodingError(str(msg))
|
|
for block in self.__blocks__:
|
|
if (block is not None):
|
|
f.write(block)
|
|
else:
|
|
try:
|
|
total_frames = self.total_frames()
|
|
current_frames = 0
|
|
decoder = audiotools.decoders.SHNDecoder(self.filename)
|
|
frame = decoder.read(4096)
|
|
while (len(frame) > 0):
|
|
f.write(frame.to_bytes(True, True))
|
|
current_frames += frame.frames
|
|
if (progress is not None):
|
|
progress(current_frames, total_frames)
|
|
frame = decoder.read(4096)
|
|
except IOError, msg:
|
|
raise EncodingError(str(msg))
|
|
else:
|
|
AiffAudio.from_pcm(aiff_filename, to_pcm_progress(self, progress))
|
|
|
|
@classmethod
|
|
def from_wave(cls, filename, wave_filename, compression=None,
|
|
block_size=256, progress=None):
|
|
"""Encodes a new AudioFile from an existing .wav file.
|
|
|
|
Takes a filename string, wave_filename string
|
|
of an existing WaveAudio file
|
|
and an optional compression level string.
|
|
Encodes a new audio file from the wave's data
|
|
at the given filename with the specified compression level
|
|
and returns a new ShortenAudio object."""
|
|
|
|
wave = WaveAudio(wave_filename)
|
|
|
|
if (wave.bits_per_sample() not in (8, 16)):
|
|
raise UnsupportedBitsPerSample(filename, wave.bits_per_sample())
|
|
|
|
(head, tail) = wave.pcm_split()
|
|
if (len(tail) > 0):
|
|
blocks = [head, None, tail]
|
|
else:
|
|
blocks = [head, None]
|
|
|
|
import audiotools.encoders
|
|
|
|
try:
|
|
audiotools.encoders.encode_shn(
|
|
filename=filename,
|
|
pcmreader=to_pcm_progress(wave, progress),
|
|
block_size=block_size,
|
|
file_type={8: 2,
|
|
16: 5}[wave.bits_per_sample()],
|
|
verbatim_chunks=blocks)
|
|
|
|
return cls(filename)
|
|
except IOError, err:
|
|
cls.__unlink__(filename)
|
|
raise EncodingError(str(err))
|
|
except Exception, err:
|
|
cls.__unlink__(filename)
|
|
raise err
|
|
|
|
@classmethod
|
|
def from_aiff(cls, filename, aiff_filename, compression=None,
|
|
block_size=256, progress=None):
|
|
"""Encodes a new AudioFile from an existing .aiff file.
|
|
|
|
Takes a filename string, aiff_filename string
|
|
of an existing WaveAudio file
|
|
and an optional compression level string.
|
|
Encodes a new audio file from the aiff's data
|
|
at the given filename with the specified compression level
|
|
and returns a new ShortenAudio object."""
|
|
|
|
aiff = AiffAudio(aiff_filename)
|
|
|
|
if (aiff.bits_per_sample() not in (8, 16)):
|
|
raise UnsupportedBitsPerSample(filename, aiff.bits_per_sample())
|
|
|
|
(head, tail) = aiff.pcm_split()
|
|
if (len(tail) > 0):
|
|
blocks = [head, None, tail]
|
|
else:
|
|
blocks = [head, None]
|
|
|
|
import audiotools.encoders
|
|
|
|
try:
|
|
audiotools.encoders.encode_shn(
|
|
filename=filename,
|
|
pcmreader=to_pcm_progress(aiff, progress),
|
|
block_size=block_size,
|
|
file_type={8: 1, # 8-bit AIFF seems to be signed
|
|
16: 3}[aiff.bits_per_sample()],
|
|
verbatim_chunks=blocks)
|
|
|
|
return cls(filename)
|
|
except IOError, err:
|
|
cls.__unlink__(filename)
|
|
raise EncodingError(str(err))
|
|
except Exception, err:
|
|
cls.__unlink__(filename)
|
|
raise err
|
|
|
|
def convert(self, target_path, target_class, compression=None,
|
|
progress=None):
|
|
"""Encodes a new AudioFile from existing AudioFile.
|
|
|
|
Take a filename string, target class and optional compression string.
|
|
Encodes a new AudioFile in the target class and returns
|
|
the resulting object.
|
|
Metadata is not copied during conversion, but embedded
|
|
RIFF chunks are (if any).
|
|
May raise EncodingError if some problem occurs during encoding."""
|
|
|
|
#Note that a Shorten file cannot contain
|
|
#both RIFF chunks and AIFF chunks at the same time.
|
|
|
|
import tempfile
|
|
|
|
if (target_class == WaveAudio):
|
|
self.to_wave(target_path, progress=progress)
|
|
return WaveAudio(target_path)
|
|
elif (target_class == AiffAudio):
|
|
self.to_aiff(target_path, progress=progress)
|
|
return AiffAudio(target_path)
|
|
elif (self.has_foreign_riff_chunks() and
|
|
hasattr(target_class, "from_wave")):
|
|
temp_wave = tempfile.NamedTemporaryFile(suffix=".wav")
|
|
try:
|
|
#we'll only log the second leg of conversion,
|
|
#since that's likely to be the slower portion
|
|
self.to_wave(temp_wave.name)
|
|
return target_class.from_wave(target_path,
|
|
temp_wave.name,
|
|
compression,
|
|
progress=progress)
|
|
finally:
|
|
temp_wave.close()
|
|
elif (self.has_foreign_aiff_chunks() and
|
|
hasattr(target_class, "from_aiff")):
|
|
temp_aiff = tempfile.NamedTemporaryFile(suffix=".aiff")
|
|
try:
|
|
self.to_aiff(temp_aiff.name)
|
|
return target_class.from_aiff(target_path,
|
|
temp_aiff.name,
|
|
compression,
|
|
progress=progress)
|
|
finally:
|
|
temp_aiff.close()
|
|
else:
|
|
return target_class.from_pcm(target_path,
|
|
to_pcm_progress(self, progress),
|
|
compression)
|
|
|
|
def has_foreign_riff_chunks(self):
|
|
"""Returns True if the audio file contains non-audio RIFF chunks.
|
|
|
|
During transcoding, if the source audio file has foreign RIFF chunks
|
|
and the target audio format supports foreign RIFF chunks,
|
|
conversion should be routed through .wav conversion
|
|
to avoid losing those chunks."""
|
|
|
|
if (not hasattr(self, "__format__")):
|
|
self.__populate_metadata__()
|
|
|
|
if (self.__format__ is WaveAudio):
|
|
for (chunk_id, chunk_data) in self.__wave_chunks__():
|
|
if (chunk_id != 'fmt '):
|
|
return True
|
|
else:
|
|
return False
|
|
else:
|
|
return False
|
|
|
|
def has_foreign_aiff_chunks(self):
|
|
"""Returns True if the audio file contains non-audio AIFF chunks.
|
|
|
|
During transcoding, if the source audio file has foreign AIFF chunks
|
|
and the target audio format supports foreign AIFF chunks,
|
|
conversion should be routed through .aiff conversion
|
|
to avoid losing those chunks."""
|
|
|
|
if (not hasattr(self, "__format__")):
|
|
self.__populate_metadata__()
|
|
|
|
if (self.__format__ is AiffAudio):
|
|
for (chunk_id, chunk_data) in self.__aiff_chunks__():
|
|
if ((chunk_id != 'COMM') and (chunk_id != 'SSND')):
|
|
return True
|
|
else:
|
|
return False
|
|
else:
|
|
return False
|