Rename "morituri" module to "whipper".

Fixes https://github.com/JoeLametta/whipper/issues/100
This commit is contained in:
Frederik “Freso” S. Olesen
2017-04-26 16:51:11 +02:00
parent a8af9b79ab
commit ff309e468c
114 changed files with 198 additions and 198 deletions

View File

207
whipper/image/cue.py Normal file
View File

@@ -0,0 +1,207 @@
# -*- Mode: Python; test-case-name: whipper.test.test_image_cue -*-
# vi:si:et:sw=4:sts=4:ts=4
# Morituri - for those about to RIP
# Copyright (C) 2009 Thomas Vander Stichele
# This file is part of whipper.
#
# whipper 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 3 of the License, or
# (at your option) any later version.
#
# whipper 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 whipper. If not, see <http://www.gnu.org/licenses/>.
"""
Reading .cue files
See http://digitalx.org/cuesheetsyntax.php
"""
import re
import codecs
from whipper.common import common
from whipper.image import table
import logging
logger = logging.getLogger(__name__)
_REM_RE = re.compile("^REM\s(\w+)\s(.*)$")
_PERFORMER_RE = re.compile("^PERFORMER\s(.*)$")
_TITLE_RE = re.compile("^TITLE\s(.*)$")
_FILE_RE = re.compile(r"""
^FILE # FILE
\s+"(?P<name>.*)" # 'file name' in quotes
\s+(?P<format>\w+)$ # format (WAVE/MP3/AIFF/...)
""", re.VERBOSE)
_TRACK_RE = re.compile(r"""
^\s+TRACK # TRACK
\s+(?P<track>\d\d) # two-digit track number
\s+(?P<mode>.+)$ # mode (AUDIO, MODEx/2xxx, ...)
""", re.VERBOSE)
_INDEX_RE = re.compile(r"""
^\s+INDEX # INDEX
\s+(\d\d) # two-digit index number
\s+(\d\d) # minutes
:(\d\d) # seconds
:(\d\d)$ # frames
""", re.VERBOSE)
class CueFile(object):
"""
I represent a .cue file as an object.
@type table: L{table.Table}
@ivar table: the index table.
"""
logCategory = 'CueFile'
def __init__(self, path):
"""
@type path: unicode
"""
assert type(path) is unicode, "%r is not unicode" % path
self._path = path
self._rems = {}
self._messages = []
self.leadout = None
self.table = table.Table()
def parse(self):
state = 'HEADER'
currentFile = None
currentTrack = None
counter = 0
logger.info('Parsing .cue file %r', self._path)
handle = codecs.open(self._path, 'r', 'utf-8')
for number, line in enumerate(handle.readlines()):
line = line.rstrip()
m = _REM_RE.search(line)
if m:
tag = m.expand('\\1')
value = m.expand('\\2')
if state != 'HEADER':
self.message(number, 'REM %s outside of header' % tag)
else:
self._rems[tag] = value
continue
# look for FILE lines
m = _FILE_RE.search(line)
if m:
counter += 1
filePath = m.group('name')
fileFormat = m.group('format')
currentFile = File(filePath, fileFormat)
# look for TRACK lines
m = _TRACK_RE.search(line)
if m:
if not currentFile:
self.message(number, 'TRACK without preceding FILE')
continue
state = 'TRACK'
trackNumber = int(m.group('track'))
#trackMode = m.group('mode')
logger.debug('found track %d', trackNumber)
currentTrack = table.Track(trackNumber)
self.table.tracks.append(currentTrack)
continue
# look for INDEX lines
m = _INDEX_RE.search(line)
if m:
if not currentTrack:
self.message(number, 'INDEX without preceding TRACK')
print 'ouch'
continue
indexNumber = int(m.expand('\\1'))
minutes = int(m.expand('\\2'))
seconds = int(m.expand('\\3'))
frames = int(m.expand('\\4'))
frameOffset = frames \
+ seconds * common.FRAMES_PER_SECOND \
+ minutes * common.FRAMES_PER_SECOND * 60
logger.debug('found index %d of track %r in %r:%d',
indexNumber, currentTrack, currentFile.path, frameOffset)
# FIXME: what do we do about File's FORMAT ?
currentTrack.index(indexNumber,
path=currentFile.path, relative=frameOffset,
counter=counter)
continue
def message(self, number, message):
"""
Add a message about a given line in the cue file.
@param number: line number, counting from 0.
"""
self._messages.append((number + 1, message))
def getTrackLength(self, track):
# returns track length in frames, or -1 if can't be determined and
# complete file should be assumed
# FIXME: this assumes a track can only be in one file; is this true ?
i = self.table.tracks.index(track)
if i == len(self.table.tracks) - 1:
# last track, so no length known
return -1
thisIndex = track.indexes[1] # FIXME: could be more
nextIndex = self.table.tracks[i + 1].indexes[1] # FIXME: could be 0
c = thisIndex.counter
if c is not None and c == nextIndex.counter:
# they belong to the same source, so their relative delta is length
return nextIndex.relative - thisIndex.relative
# FIXME: more logic
return -1
def getRealPath(self, path):
"""
Translate the .cue's FILE to an existing path.
@type path: unicode
"""
return common.getRealPath(self._path, path)
class File:
"""
I represent a FILE line in a cue file.
"""
def __init__(self, path, format):
"""
@type path: unicode
"""
assert type(path) is unicode, "%r is not unicode" % path
self.path = path
self.format = format
def __repr__(self):
return '<File %r of format %s>' % (self.path, self.format)

255
whipper/image/image.py Normal file
View File

@@ -0,0 +1,255 @@
# -*- Mode: Python; test-case-name: whipper.test.test_image_image -*-
# vi:si:et:sw=4:sts=4:ts=4
# Morituri - for those about to RIP
# Copyright (C) 2009 Thomas Vander Stichele
# This file is part of whipper.
#
# whipper 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 3 of the License, or
# (at your option) any later version.
#
# whipper 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 whipper. If not, see <http://www.gnu.org/licenses/>.
"""
Wrap on-disk CD images based on the .cue file.
"""
import os
from whipper.common import encode
from whipper.common import common
from whipper.common import checksum
from whipper.image import cue, table
from whipper.extern.task import task
from whipper.program.soxi import AudioLengthTask
import logging
logger = logging.getLogger(__name__)
class Image(object):
"""
@ivar table: The Table of Contents for this image.
@type table: L{table.Table}
"""
logCategory = 'Image'
def __init__(self, path):
"""
@type path: unicode
@param path: .cue path
"""
assert type(path) is unicode, "%r is not unicode" % path
self._path = path
self.cue = cue.CueFile(path)
self.cue.parse()
self._offsets = [] # 0 .. trackCount - 1
self._lengths = [] # 0 .. trackCount - 1
self.table = None
def getRealPath(self, path):
"""
Translate the .cue's FILE to an existing path.
@param path: .cue path
"""
assert type(path) is unicode, "%r is not unicode" % path
return self.cue.getRealPath(path)
def setup(self, runner):
"""
Do initial setup, like figuring out track lengths, and
constructing the Table of Contents.
"""
logger.debug('setup image start')
verify = ImageVerifyTask(self)
logger.debug('verifying image')
runner.run(verify)
logger.debug('verified image')
# calculate offset and length for each track
# CD's have a standard lead-in time of 2 seconds;
# checksums that use it should add it there
if verify.lengths.has_key(0):
offset = verify.lengths[0]
else:
offset = self.cue.table.tracks[0].getIndex(1).relative
tracks = []
for i in range(len(self.cue.table.tracks)):
length = self.cue.getTrackLength(self.cue.table.tracks[i])
if length == -1:
length = verify.lengths[i + 1]
t = table.Track(i + 1, audio=True)
tracks.append(t)
# FIXME: this probably only works for non-compliant .CUE files
# where pregap is put at end of previous file
t.index(1, absolute=offset,
path=self.cue.table.tracks[i].getIndex(1).path,
relative=0)
offset += length
self.table = table.Table(tracks)
self.table.leadout = offset
logger.debug('setup image done')
class AccurateRipChecksumTask(task.MultiSeparateTask):
"""
I calculate the AccurateRip checksums of all tracks.
"""
description = "Checksumming tracks"
# TODO MW: Update this further for V2 code
def __init__(self, image):
task.MultiSeparateTask.__init__(self)
self._image = image
cue = image.cue
self.checksums = []
logger.debug('Checksumming %d tracks' % len(cue.table.tracks))
for trackIndex, track in enumerate(cue.table.tracks):
index = track.indexes[1]
length = cue.getTrackLength(track)
if length < 0:
logger.debug('track %d has unknown length' % (trackIndex + 1, ))
else:
logger.debug('track %d is %d samples long' % (
trackIndex + 1, length))
path = image.getRealPath(index.path)
checksumTask = checksum.FastAccurateRipChecksumTask(path,
trackNumber=trackIndex + 1, trackCount=len(cue.table.tracks),
wave=True, v2=False)
self.addTask(checksumTask)
def stop(self):
self.checksums = [t.checksum for t in self.tasks]
task.MultiSeparateTask.stop(self)
class ImageVerifyTask(task.MultiSeparateTask):
"""
I verify a disk image and get the necessary track lengths.
"""
logCategory = 'ImageVerifyTask'
description = "Checking tracks"
lengths = None
def __init__(self, image):
task.MultiSeparateTask.__init__(self)
self._image = image
cue = image.cue
self._tasks = []
self.lengths = {}
try:
htoa = cue.table.tracks[0].indexes[0]
track = cue.table.tracks[0]
path = image.getRealPath(htoa.path)
assert type(path) is unicode, "%r is not unicode" % path
logger.debug('schedule scan of audio length of %r', path)
taskk = AudioLengthTask(path)
self.addTask(taskk)
self._tasks.append((0, track, taskk))
except (KeyError, IndexError):
logger.debug('no htoa track')
for trackIndex, track in enumerate(cue.table.tracks):
logger.debug('verifying track %d', trackIndex + 1)
index = track.indexes[1]
length = cue.getTrackLength(track)
if length == -1:
path = image.getRealPath(index.path)
assert type(path) is unicode, "%r is not unicode" % path
logger.debug('schedule scan of audio length of %r', path)
taskk = AudioLengthTask(path)
self.addTask(taskk)
self._tasks.append((trackIndex + 1, track, taskk))
else:
logger.debug('track %d has length %d', trackIndex + 1, length)
def stop(self):
for trackIndex, track, taskk in self._tasks:
if taskk.exception:
logger.debug('subtask %r had exception %r, shutting down' % (
taskk, taskk.exception))
self.setException(taskk.exception)
break
if taskk.length is None:
raise ValueError("Track length was not found; look for "
"earlier errors in debug log (set RIP_DEBUG=4)")
# print '%d has length %d' % (trackIndex, taskk.length)
index = track.indexes[1]
assert taskk.length % common.SAMPLES_PER_FRAME == 0
end = taskk.length / common.SAMPLES_PER_FRAME
self.lengths[trackIndex] = end - index.relative
task.MultiSeparateTask.stop(self)
class ImageEncodeTask(task.MultiSeparateTask):
"""
I encode a disk image to a different format.
"""
description = "Encoding tracks"
def __init__(self, image, outdir):
task.MultiSeparateTask.__init__(self)
self._image = image
cue = image.cue
self._tasks = []
self.lengths = {}
def add(index):
path = image.getRealPath(index.path)
assert type(path) is unicode, "%r is not unicode" % path
logger.debug('schedule encode of %r', path)
root, ext = os.path.splitext(os.path.basename(path))
outpath = os.path.join(outdir, root + '.' + 'flac')
logger.debug('schedule encode to %r', outpath)
taskk = encode.FlacEncodeTask(path, os.path.join(outdir,
root + '.' + 'flac'))
self.addTask(taskk)
try:
htoa = cue.table.tracks[0].indexes[0]
logger.debug('encoding htoa track')
add(htoa)
except (KeyError, IndexError):
logger.debug('no htoa track')
pass
for trackIndex, track in enumerate(cue.table.tracks):
logger.debug('encoding track %d', trackIndex + 1)
index = track.indexes[1]
add(index)

871
whipper/image/table.py Normal file
View File

@@ -0,0 +1,871 @@
# -*- Mode: Python; test-case-name: whipper.test.test_image_table -*-
# vi:si:et:sw=4:sts=4:ts=4
# Morituri - for those about to RIP
# Copyright (C) 2009 Thomas Vander Stichele
# This file is part of whipper.
#
# whipper 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 3 of the License, or
# (at your option) any later version.
#
# whipper 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 whipper. If not, see <http://www.gnu.org/licenses/>.
"""
Wrap Table of Contents.
"""
import copy
import urllib
import urlparse
import whipper
from whipper.common import common
import logging
logger = logging.getLogger(__name__)
# FIXME: taken from libcdio, but no reference found for these
CDTEXT_FIELDS = [
'ARRANGER',
'COMPOSER',
'DISCID',
'GENRE',
'MESSAGE',
'ISRC',
'PERFORMER',
'SIZE_INFO',
'SONGWRITER',
'TITLE',
'TOC_INFO',
'TOC_INFO2',
'UPC_EAN',
]
class Track:
"""
I represent a track entry in an Table.
@ivar number: track number (1-based)
@type number: int
@ivar audio: whether the track is audio
@type audio: bool
@type indexes: dict of number -> L{Index}
@ivar isrc: ISRC code (12 alphanumeric characters)
@type isrc: str
@ivar cdtext: dictionary of CD Text information; see L{CDTEXT_KEYS}.
@type cdtext: str -> unicode
@ivar pre_emphasis: whether track is pre-emphasised
@type pre_emphasis: bool
"""
number = None
audio = None
indexes = None
isrc = None
cdtext = None
session = None
pre_emphasis = None
def __repr__(self):
return '<Track %02d>' % self.number
def __init__(self, number, audio=True, session=None):
self.number = number
self.audio = audio
self.indexes = {}
self.cdtext = {}
def index(self, number, absolute=None, path=None, relative=None,
counter=None):
"""
@type path: unicode or None
"""
if path is not None:
assert type(path) is unicode, "%r is not unicode" % path
i = Index(number, absolute, path, relative, counter)
self.indexes[number] = i
def getIndex(self, number):
return self.indexes[number]
def getFirstIndex(self):
"""
Get the first chronological index for this track.
Typically this is INDEX 01; but it could be INDEX 00 if there's
a pre-gap.
"""
indexes = self.indexes.keys()
indexes.sort()
return self.indexes[indexes[0]]
def getLastIndex(self):
indexes = self.indexes.keys()
indexes.sort()
return self.indexes[indexes[-1]]
def getPregap(self):
"""
Returns the length of the pregap for this track.
The pregap is 0 if there is no index 0, and the difference between
index 1 and index 0 if there is.
"""
if 0 not in self.indexes:
return 0
return self.indexes[1].absolute - self.indexes[0].absolute
class Index:
"""
@ivar counter: counter for the index source; distinguishes between
the matching FILE lines in .cue files for example
@type path: unicode or None
"""
number = None
absolute = None
path = None
relative = None
counter = None
def __init__(self, number, absolute=None, path=None, relative=None,
counter=None):
if path is not None:
assert type(path) is unicode, "%r is not unicode" % path
self.number = number
self.absolute = absolute
self.path = path
self.relative = relative
self.counter = counter
def __repr__(self):
return '<Index %02d absolute %r path %r relative %r counter %r>' % (
self.number, self.absolute, self.path, self.relative, self.counter)
class Table(object):
"""
I represent a table of indexes on a CD.
@ivar tracks: tracks on this CD
@type tracks: list of L{Track}
@ivar catalog: catalog number
@type catalog: str
@type cdtext: dict of str -> str
"""
tracks = None # list of Track
leadout = None # offset where the leadout starts
catalog = None # catalog number; FIXME: is this UPC ?
cdtext = None
mbdiscid = None
classVersion = 4
def __init__(self, tracks=None):
if not tracks:
tracks = []
self.tracks = tracks
self.cdtext = {}
# done this way because just having a class-defined instance var
# gets overridden when unpickling
self.instanceVersion = self.classVersion
self.unpickled()
def unpickled(self):
self.logName = "Table 0x%08x v%d" % (id(self), self.instanceVersion)
logger.debug('set logName')
def getTrackStart(self, number):
"""
@param number: the track number, 1-based
@type number: int
@returns: the start of the given track number's index 1, in CD frames
@rtype: int
"""
track = self.tracks[number - 1]
return track.getIndex(1).absolute
def getTrackEnd(self, number):
"""
@param number: the track number, 1-based
@type number: int
@returns: the end of the given track number (ie index 1 of next track)
@rtype: int
"""
# default to end of disc
end = self.leadout - 1
# if not last track, calculate it from the next track
if number < len(self.tracks):
end = self.tracks[number].getIndex(1).absolute - 1
# if on a session border, subtract the session leadin
thisTrack = self.tracks[number - 1]
nextTrack = self.tracks[number]
if nextTrack.session > thisTrack.session:
gap = self._getSessionGap(nextTrack.session)
end -= gap
return end
def getTrackLength(self, number):
"""
@param number: the track number, 1-based
@type number: int
@returns: the length of the given track number, in CD frames
@rtype: int
"""
return self.getTrackEnd(number) - self.getTrackStart(number) + 1
def getAudioTracks(self):
"""
@returns: the number of audio tracks on the CD
@rtype: int
"""
return len([t for t in self.tracks if t.audio])
def hasDataTracks(self):
"""
@returns: whether this disc contains data tracks
"""
return len([t for t in self.tracks if not t.audio]) > 0
def _cddbSum(self, i):
ret = 0
while i > 0:
ret += (i % 10)
i /= 10
return ret
def getCDDBValues(self):
"""
Get all CDDB values needed to calculate disc id and lookup URL.
This includes:
- CDDB disc id
- number of audio tracks
- offset of index 1 of each track
- length of disc in seconds (including data track)
@rtype: list of int
"""
result = []
result.append(self.getAudioTracks())
# cddb disc id takes into account data tracks
# last byte is the number of tracks on the CD
n = 0
# CD's have a standard lead-in time of 2 seconds
# which gets added for CDDB disc id's
delta = 2 * common.FRAMES_PER_SECOND
#if self.getTrackStart(1) > 0:
# delta = 0
debug = [str(len(self.tracks))]
for track in self.tracks:
offset = self.getTrackStart(track.number) + delta
result.append(offset)
debug.append(str(offset))
seconds = offset / common.FRAMES_PER_SECOND
n += self._cddbSum(seconds)
# the 'real' leadout, not offset by 150 frames
# print 'THOMAS: disc leadout', self.leadout
last = self.tracks[-1]
leadout = self.getTrackEnd(last.number) + 1
logger.debug('leadout LBA: %d', leadout)
# FIXME: we can't replace these calculations with the getFrameLength
# call because the start and leadout in the algorithm get rounded
# before making the difference
startSeconds = self.getTrackStart(1) / common.FRAMES_PER_SECOND
leadoutSeconds = leadout / common.FRAMES_PER_SECOND
t = leadoutSeconds - startSeconds
# durationFrames = self.getFrameLength(data=True)
# duration = durationFrames / common.FRAMES_PER_SECOND
# assert t == duration, "%r != %r" % (t, duration)
debug.append(str(leadoutSeconds + 2)) # 2 is the 150 frame cddb offset
result.append(leadoutSeconds)
value = (n % 0xff) << 24 | t << 8 | len(self.tracks)
result.insert(0, value)
# compare this debug line to cd-discid output
logger.debug('cddb values: %r', result)
logger.debug('cddb disc id debug: %s',
" ".join(["%08x" % value, ] + debug))
return result
def getCDDBDiscId(self):
"""
Calculate the CDDB disc ID.
@rtype: str
@returns: the 8-character hexadecimal disc ID
"""
values = self.getCDDBValues()
return "%08x" % values[0]
def getMusicBrainzDiscId(self):
"""
Calculate the MusicBrainz disc ID.
@rtype: str
@returns: the 28-character base64-encoded disc ID
"""
if self.mbdiscid:
logger.debug('getMusicBrainzDiscId: returning cached %r'
% self.mbdiscid)
return self.mbdiscid
values = self._getMusicBrainzValues()
# MusicBrainz disc id does not take into account data tracks
# P2.3
try:
import hashlib
sha1 = hashlib.sha1
except ImportError:
from sha import sha as sha1
import base64
sha = sha1()
# number of first track
sha.update("%02X" % values[0])
# number of last track
sha.update("%02X" % values[1])
sha.update("%08X" % values[2])
# offsets of tracks
for i in range(1, 100):
try:
offset = values[2 + i]
except IndexError:
#print 'track', i - 1, '0 offset'
offset = 0
sha.update("%08X" % offset)
digest = sha.digest()
assert len(digest) == 20, \
"digest should be 20 chars, not %d" % len(digest)
# The RFC822 spec uses +, /, and = characters, all of which are special
# HTTP/URL characters. To avoid the problems with dealing with that, I
# (Rob) used ., _, and -
# base64 altchars specify replacements for + and /
result = base64.b64encode(digest, '._')
# now replace =
result = "-".join(result.split("="))
assert len(result) == 28, \
"Result should be 28 characters, not %d" % len(result)
logger.debug('getMusicBrainzDiscId: returning %r' % result)
self.mbdiscid = result
return result
def getMusicBrainzSubmitURL(self):
host = 'musicbrainz.org'
discid = self.getMusicBrainzDiscId()
values = self._getMusicBrainzValues()
query = urllib.urlencode({
'id': discid,
'toc': ' '.join([str(v) for v in values]),
'tracks': self.getAudioTracks(),
})
return urlparse.urlunparse((
'https', host, '/cdtoc/attach', '', query, ''))
def getFrameLength(self, data=False):
"""
Get the length in frames (excluding HTOA)
@param data: whether to include the data tracks in the length
"""
# the 'real' leadout, not offset by 150 frames
if data:
last = self.tracks[-1]
else:
last = self.tracks[self.getAudioTracks() - 1]
leadout = self.getTrackEnd(last.number) + 1
logger.debug('leadout LBA: %d', leadout)
durationFrames = leadout - self.getTrackStart(1)
return durationFrames
def duration(self):
"""
Get the duration in ms for all audio tracks (excluding HTOA).
"""
return int(self.getFrameLength() * 1000.0 / common.FRAMES_PER_SECOND)
def _getMusicBrainzValues(self):
"""
Get all MusicBrainz values needed to calculate disc id and submit URL.
This includes:
- track number of first track
- number of audio tracks
- leadout of disc
- offset of index 1 of each track
@rtype: list of int
"""
# MusicBrainz disc id does not take into account data tracks
result = []
# number of first track
result.append(1)
# number of last audio track
result.append(self.getAudioTracks())
leadout = self.leadout
# if the disc is multi-session, last track is the data track,
# and we should subtract 11250 + 150 from the last track's offset
# for the leadout
if self.hasDataTracks():
assert not self.tracks[-1].audio
leadout = self.tracks[-1].getIndex(1).absolute - 11250 - 150
# treat leadout offset as track 0 offset
result.append(150 + leadout)
# offsets of tracks
for i in range(1, 100):
try:
track = self.tracks[i - 1]
if not track.audio:
continue
offset = track.getIndex(1).absolute + 150
result.append(offset)
except IndexError:
pass
logger.debug('Musicbrainz values: %r', result)
return result
def getAccurateRipIds(self):
"""
Calculate the two AccurateRip ID's.
@returns: the two 8-character hexadecimal disc ID's
@rtype: tuple of (str, str)
"""
# AccurateRip does not take into account data tracks,
# but does count the data track to determine the leadout offset
discId1 = 0
discId2 = 0
for track in self.tracks:
if not track.audio:
continue
offset = self.getTrackStart(track.number)
discId1 += offset
discId2 += (offset or 1) * track.number
# also add end values, where leadout offset is one past the end
# of the last track
last = self.tracks[-1]
offset = self.getTrackEnd(last.number) + 1
discId1 += offset
discId2 += offset * (self.getAudioTracks() + 1)
discId1 &= 0xffffffff
discId2 &= 0xffffffff
return ("%08x" % discId1, "%08x" % discId2)
def getAccurateRipURL(self):
"""
Return the full AccurateRip URL.
@returns: the AccurateRip URL
@rtype: str
"""
discId1, discId2 = self.getAccurateRipIds()
return "http://www.accuraterip.com/accuraterip/" \
"%s/%s/%s/dBAR-%.3d-%s-%s-%s.bin" % (
discId1[-1], discId1[-2], discId1[-3],
self.getAudioTracks(), discId1, discId2, self.getCDDBDiscId())
def cue(self, cuePath='', program='whipper'):
"""
@param cuePath: path to the cue file to be written. If empty,
will treat paths as if in current directory.
Dump our internal representation to a .cue file content.
@rtype: C{unicode}
"""
logger.debug('generating .cue for cuePath %r', cuePath)
lines = []
def writeFile(path):
targetPath = common.getRelativePath(path, cuePath)
line = 'FILE "%s" WAVE' % targetPath
lines.append(line)
logger.debug('writeFile: %r' % line)
# header
main = ['PERFORMER', 'TITLE']
for key in CDTEXT_FIELDS:
if key not in main and key in self.cdtext:
lines.append(" %s %s" % (key, self.cdtext[key]))
assert self.hasTOC(), "Table does not represent a full CD TOC"
lines.append('REM DISCID %s' % self.getCDDBDiscId().upper())
lines.append('REM COMMENT "%s %s"' % (program, whipper.__version__))
if self.catalog:
lines.append("CATALOG %s" % self.catalog)
for key in main:
if key in self.cdtext:
lines.append('%s "%s"' % (key, self.cdtext[key]))
# FIXME:
# - the first FILE statement goes before the first TRACK, even if
# there is a non-file-using PREGAP
# - the following FILE statements come after the last INDEX that
# use that FILE; so before a next TRACK, PREGAP silence, ...
# add the first FILE line; EAC always puts the first FILE
# statement before TRACK 01 and any possible PRE-GAP
firstTrack = self.tracks[0]
index = firstTrack.getFirstIndex()
indexOne = firstTrack.getIndex(1)
counter = index.counter
track = firstTrack
while not index.path:
t, i = self.getNextTrackIndex(track.number, index.number)
track = self.tracks[t - 1]
index = track.getIndex(i)
counter = index.counter
if index.path:
logger.debug('counter %d, writeFile' % counter)
writeFile(index.path)
for i, track in enumerate(self.tracks):
logger.debug('track i %r, track %r' % (i, track))
# FIXME: skip data tracks for now
if not track.audio:
continue
indexes = track.indexes.keys()
indexes.sort()
wroteTrack = False
for number in indexes:
index = track.indexes[number]
logger.debug('index %r, %r' % (number, index))
# any time the source counter changes to a higher value,
# write a FILE statement
# it has to be higher, because we can run into the HTOA
# at counter 0 here
if index.counter > counter:
if index.path:
logger.debug('counter %d, writeFile' % counter)
writeFile(index.path)
logger.debug('setting counter to index.counter %r' %
index.counter)
counter = index.counter
# any time we hit the first index, write a TRACK statement
if not wroteTrack:
wroteTrack = True
line = " TRACK %02d %s" % (i + 1, 'AUDIO')
lines.append(line)
logger.debug('%r' % line)
for key in CDTEXT_FIELDS:
if key in track.cdtext:
lines.append(' %s "%s"' % (
key, track.cdtext[key]))
if track.isrc is not None:
lines.append(" ISRC %s" % track.isrc)
if track.pre_emphasis is not None:
lines.append(" FLAGS PRE")
# handle TRACK 01 INDEX 00 specially
if 0 in indexes:
index00 = track.indexes[0]
if i == 0:
# if we have a silent pre-gap, output it
if not index00.path:
length = indexOne.absolute - index00.absolute
lines.append(" PREGAP %s" %
common.framesToMSF(length))
continue
# handle any other INDEX 00 after its TRACK
lines.append(" INDEX %02d %s" % (0,
common.framesToMSF(index00.relative)))
if number > 0:
# index 00 is output after TRACK up above
lines.append(" INDEX %02d %s" % (number,
common.framesToMSF(index.relative)))
lines.append("")
return "\n".join(lines)
### methods that modify the table
def clearFiles(self):
"""
Clear all file backings.
Resets indexes paths and relative offsets.
"""
# FIXME: do a loop over track indexes better, with a pythonic
# construct that allows you to do for t, i in ...
t = self.tracks[0].number
index = self.tracks[0].getFirstIndex()
i = index.number
logger.debug('clearing path')
while True:
track = self.tracks[t - 1]
index = track.getIndex(i)
logger.debug('Clearing path on track %d, index %d', t, i)
index.path = None
index.relative = None
try:
t, i = self.getNextTrackIndex(t, i)
except IndexError:
break
def setFile(self, track, index, path, length, counter=None):
"""
Sets the given file as the source from the given index on.
Will loop over all indexes that fall within the given length,
to adjust the path.
Assumes all indexes have an absolute offset and will raise if not.
@type track: C{int}
@type index: C{int}
"""
logger.debug('setFile: track %d, index %d, path %r, '
'length %r, counter %r', track, index, path, length, counter)
t = self.tracks[track - 1]
i = t.indexes[index]
start = i.absolute
assert start is not None, "index %r is missing absolute offset" % i
end = start + length - 1 # last sector that should come from this file
# FIXME: check border conditions here, esp. wrt. toc's off-by-one bug
while i.absolute <= end:
i.path = path
i.relative = i.absolute - start
i.counter = counter
logger.debug('Setting path %r, relative %r on '
'track %d, index %d, counter %r',
path, i.relative, track, index, counter)
try:
track, index = self.getNextTrackIndex(track, index)
t = self.tracks[track - 1]
i = t.indexes[index]
except IndexError:
break
def absolutize(self):
"""
Calculate absolute offsets on indexes as much as possible.
Only possible for as long as tracks draw from the same file.
"""
t = self.tracks[0].number
index = self.tracks[0].getFirstIndex()
i = index.number
# the first cut is the deepest
counter = index.counter
#for t in self.tracks: print t, t.indexes
logger.debug('absolutizing')
while True:
track = self.tracks[t - 1]
index = track.getIndex(i)
assert track.number == t
assert index.number == i
if index.counter is None:
logger.debug('Track %d, index %d has no counter', t, i)
break
if index.counter != counter:
logger.debug('Track %d, index %d has a different counter', t, i)
break
logger.debug('Setting absolute offset %d on track %d, index %d',
index.relative, t, i)
if index.absolute is not None:
if index.absolute != index.relative:
msg = 'Track %d, index %d had absolute %d,' \
' overriding with %d' % (
t, i, index.absolute, index.relative)
raise ValueError(msg)
index.absolute = index.relative
try:
t, i = self.getNextTrackIndex(t, i)
except IndexError:
break
def merge(self, other, session=2):
"""
Merges the given table at the end.
The other table is assumed to be from an additional session,
@type other: L{Table}
"""
gap = self._getSessionGap(session)
trackCount = len(self.tracks)
sourceCounter = self.tracks[-1].getLastIndex().counter
for track in other.tracks:
t = copy.deepcopy(track)
t.number = track.number + trackCount
t.session = session
for i in t.indexes.values():
if i.absolute is not None:
i.absolute += self.leadout + gap
logger.debug('Fixing track %02d, index %02d, absolute %d' % (
t.number, i.number, i.absolute))
if i.counter is not None:
i.counter += sourceCounter
logger.debug('Fixing track %02d, index %02d, counter %d' % (
t.number, i.number, i.counter))
self.tracks.append(t)
self.leadout += other.leadout + gap # FIXME
logger.debug('Fixing leadout, now %d', self.leadout)
def _getSessionGap(self, session):
# From cdrecord multi-session info:
# For the first additional session this is 11250 sectors
# lead-out/lead-in overhead + 150 sectors for the pre-gap of the first
# track after the lead-in = 11400 sectos.
# For all further session this is 6750 sectors lead-out/lead-in
# overhead + 150 sectors for the pre-gap of the first track after the
# lead-in = 6900 sectors.
gap = 11400
if session > 2:
gap = 6900
return gap
### lookups
def getNextTrackIndex(self, track, index):
"""
Return the next track and index.
@param track: track number, 1-based
@raises IndexError: on last index
@rtype: tuple of (int, int)
"""
t = self.tracks[track - 1]
indexes = t.indexes.keys()
position = indexes.index(index)
if position + 1 < len(indexes):
return track, indexes[position + 1]
track += 1
if track > len(self.tracks):
raise IndexError("No index beyond track %d, index %d" % (
track - 1, index))
t = self.tracks[track - 1]
indexes = t.indexes.keys()
return track, indexes[0]
# various tests for types of Table
def hasTOC(self):
"""
Check if the Table has a complete TOC.
a TOC is a list of all tracks and their Index 01, with absolute
offsets, as well as the leadout.
"""
if not self.leadout:
logger.debug('no leadout, no TOC')
return False
for t in self.tracks:
if 1 not in t.indexes.keys():
logger.debug('no index 1, no TOC')
return False
if t.indexes[1].absolute is None:
logger.debug('no absolute index 1, no TOC')
return False
return True
def canCue(self):
"""
Check if this table can be used to generate a .cue file
"""
if not self.hasTOC():
logger.debug('No TOC, cannot cue')
return False
for t in self.tracks:
for i in t.indexes.values():
if i.relative is None:
logger.debug('Track %02d, Index %02d does not have relative',
t.number, i.number)
return False
return True

445
whipper/image/toc.py Normal file
View File

@@ -0,0 +1,445 @@
# -*- Mode: Python; test-case-name: whipper.test.test_image_toc -*-
# vi:si:et:sw=4:sts=4:ts=4
# Morituri - for those about to RIP
# Copyright (C) 2009 Thomas Vander Stichele
# This file is part of whipper.
#
# whipper 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 3 of the License, or
# (at your option) any later version.
#
# whipper 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 whipper. If not, see <http://www.gnu.org/licenses/>.
"""
Reading .toc files
The .toc file format is described in the man page of cdrdao
"""
import re
import codecs
from whipper.common import common
from whipper.image import table
import logging
logger = logging.getLogger(__name__)
# shared
_CDTEXT_CANDIDATE_RE = re.compile(r'(?P<key>\w+) "(?P<value>.+)"')
# header
_CATALOG_RE = re.compile(r'^CATALOG "(?P<catalog>\d+)"$')
# pre emphasis
_PRE_EMPHASIS_RE = re.compile(r'^PRE_EMPHASIS$')
# records
_TRACK_RE = re.compile(r"""
^TRACK # TRACK
\s(?P<mode>.+)$ # mode (AUDIO, MODE2_FORM_MIX, MODEx/2xxx, ...)
""", re.VERBOSE)
_ISRC_RE = re.compile(r'^ISRC "(?P<isrc>\w+)"$')
# a HTOA is marked in the cdrdao's TOC as SILENCE
_SILENCE_RE = re.compile(r"""
^SILENCE # SILENCE
\s(?P<length>.*)$ # pre-gap length
""", re.VERBOSE)
# ZERO is used as pre-gap source when switching mode
_ZERO_RE = re.compile(r"""
^ZERO # ZERO
\s(?P<mode>.+) # mode (AUDIO, MODEx/2xxx, ...)
\s(?P<length>.*)$ # zero length
""", re.VERBOSE)
_FILE_RE = re.compile(r"""
^FILE # FILE
\s+"(?P<name>.*)" # 'file name' in quotes
\s+(?P<start>.+) # start offset
\s(?P<length>.+)$ # length in frames of section
""", re.VERBOSE)
_DATAFILE_RE = re.compile(r"""
^DATAFILE # DATA FILE
\s+"(?P<name>.*)" # 'file name' in quotes
\s+(?P<length>\S+) # start offset
\s*.* # possible // comment
""", re.VERBOSE)
# FIXME: start can be 0
_START_RE = re.compile(r"""
^START # START
\s(?P<length>.*)$ # pre-gap length
""", re.VERBOSE)
_INDEX_RE = re.compile(r"""
^INDEX # INDEX
\s(?P<offset>.+)$ # start offset
""", re.VERBOSE)
class Sources:
"""
I represent the list of sources used in the .toc file.
Each SILENCE and each FILE is a source.
If the filename for FILE doesn't change, the counter is not increased.
"""
def __init__(self):
self._sources = []
def append(self, counter, offset, source):
"""
@param counter: the source counter; updates for each different
data source (silence or different file path)
@type counter: int
@param offset: the absolute disc offset where this source starts
"""
logger.debug('Appending source, counter %d, abs offset %d, source %r' % (
counter, offset, source))
self._sources.append((counter, offset, source))
def get(self, offset):
"""
Retrieve the source used at the given offset.
"""
for i, (c, o, s) in enumerate(self._sources):
if offset < o:
return self._sources[i - 1]
return self._sources[-1]
def getCounterStart(self, counter):
"""
Retrieve the absolute offset of the first source for this counter
"""
for i, (c, o, s) in enumerate(self._sources):
if c == counter:
return self._sources[i][1]
return self._sources[-1][1]
class TocFile(object):
def __init__(self, path):
"""
@type path: unicode
"""
assert type(path) is unicode, "%r is not unicode" % path
self._path = path
self._messages = []
self.table = table.Table()
self.logName = '<TocFile %08x>' % id(self)
self._sources = Sources()
def _index(self, currentTrack, i, absoluteOffset, trackOffset):
absolute = absoluteOffset + trackOffset
# this may be in a new source, so calculate relative
c, o, s = self._sources.get(absolute)
logger.debug('at abs offset %d, we are in source %r' % (
absolute, s))
counterStart = self._sources.getCounterStart(c)
relative = absolute - counterStart
currentTrack.index(i, path=s.path,
absolute=absolute,
relative=relative,
counter=c)
logger.debug(
'[track %02d index %02d] trackOffset %r, added %r',
currentTrack.number, i, trackOffset,
currentTrack.getIndex(i))
def parse(self):
# these two objects start as None then get set as real objects,
# so no need to complain about them here
__pychecker__ = 'no-objattrs'
currentFile = None
currentTrack = None
state = 'HEADER'
counter = 0 # counts sources for audio data; SILENCE/ZERO/FILE
trackNumber = 0
indexNumber = 0
absoluteOffset = 0 # running absolute offset of where each track starts
relativeOffset = 0 # running relative offset, relative to counter src
currentLength = 0 # accrued during TRACK record parsing;
# length of current track as parsed so far;
# reset on each TRACK statement
totalLength = 0 # accrued during TRACK record parsing, total disc
pregapLength = 0 # length of the pre-gap, current track in for loop
# the first track's INDEX 1 can only be gotten from the .toc
# file once the first pregap is calculated; so we add INDEX 1
# at the end of each parsed TRACK record
handle = codecs.open(self._path, "r", "utf-8")
for number, line in enumerate(handle.readlines()):
line = line.rstrip()
# look for CDTEXT stuff in either header or tracks
m = _CDTEXT_CANDIDATE_RE.search(line)
if m:
key = m.group('key')
value = m.group('value')
# usually, value is encoded with octal escapes and in latin-1
# FIXME: other encodings are possible, does cdrdao handle
# them ?
value = value.decode('string-escape').decode('latin-1')
if key in table.CDTEXT_FIELDS:
# FIXME: consider ISRC separate for now, but this
# is a limitation of our parser approach
if state == 'HEADER':
self.table.cdtext[key] = value
logger.debug('Found disc CD-Text %s: %r', key, value)
elif state == 'TRACK':
if key != 'ISRC' or not currentTrack \
or currentTrack.isrc is not None:
logger.debug('Found track CD-Text %s: %r',
key, value)
currentTrack.cdtext[key] = value
# look for header elements
m = _CATALOG_RE.search(line)
if m:
self.table.catalog = m.group('catalog')
logger.debug("Found catalog number %s", self.table.catalog)
# look for TRACK lines
m = _TRACK_RE.search(line)
if m:
state = 'TRACK'
# set index 1 of previous track if there was one, using
# pregapLength if applicable
if currentTrack:
self._index(currentTrack, 1, absoluteOffset, pregapLength)
# create a new track to be filled by later lines
trackNumber += 1
trackMode = m.group('mode')
audio = trackMode == 'AUDIO'
currentTrack = table.Track(trackNumber, audio=audio)
self.table.tracks.append(currentTrack)
# update running totals
absoluteOffset += currentLength
relativeOffset += currentLength
totalLength += currentLength
# FIXME: track mode
logger.debug('found track %d, mode %s, at absoluteOffset %d',
trackNumber, trackMode, absoluteOffset)
# reset counters relative to a track
currentLength = 0
indexNumber = 1
pregapLength = 0
continue
# look for PRE_EMPHASIS lines
m = _PRE_EMPHASIS_RE.search(line)
if m:
currentTrack.pre_emphasis = True
logger.debug('Track has PRE_EMPHASIS')
# look for ISRC lines
m = _ISRC_RE.search(line)
if m:
isrc = m.group('isrc')
currentTrack.isrc = isrc
logger.debug('Found ISRC code %s', isrc)
# look for SILENCE lines
m = _SILENCE_RE.search(line)
if m:
length = m.group('length')
logger.debug('SILENCE of %r', length)
self._sources.append(counter, absoluteOffset, None)
if currentFile is not None:
logger.debug('SILENCE after FILE, increasing counter')
counter += 1
relativeOffset = 0
currentFile = None
currentLength += common.msfToFrames(length)
# look for ZERO lines
m = _ZERO_RE.search(line)
if m:
if currentFile is not None:
logger.debug('ZERO after FILE, increasing counter')
counter += 1
relativeOffset = 0
currentFile = None
length = m.group('length')
currentLength += common.msfToFrames(length)
# look for FILE lines
m = _FILE_RE.search(line)
if m:
filePath = m.group('name')
start = m.group('start')
length = m.group('length')
logger.debug('FILE %s, start %r, length %r',
filePath, common.msfToFrames(start),
common.msfToFrames(length))
if not currentFile or filePath != currentFile.path:
counter += 1
relativeOffset = 0
logger.debug('track %d, switched to new FILE, '
'increased counter to %d',
trackNumber, counter)
currentFile = File(filePath, common.msfToFrames(start),
common.msfToFrames(length))
self._sources.append(counter, absoluteOffset + currentLength,
currentFile)
#absoluteOffset += common.msfToFrames(start)
currentLength += common.msfToFrames(length)
# look for DATAFILE lines
m = _DATAFILE_RE.search(line)
if m:
filePath = m.group('name')
length = m.group('length')
# print 'THOMAS', length
logger.debug('FILE %s, length %r',
filePath, common.msfToFrames(length))
if not currentFile or filePath != currentFile.path:
counter += 1
relativeOffset = 0
logger.debug('track %d, switched to new FILE, '
'increased counter to %d',
trackNumber, counter)
# FIXME: assume that a MODE2_FORM_MIX track always starts at 0
currentFile = File(filePath, 0, common.msfToFrames(length))
self._sources.append(counter, absoluteOffset + currentLength,
currentFile)
#absoluteOffset += common.msfToFrames(start)
currentLength += common.msfToFrames(length)
# look for START lines
m = _START_RE.search(line)
if m:
if not currentTrack:
self.message(number, 'START without preceding TRACK')
print 'ouch'
continue
length = common.msfToFrames(m.group('length'))
c, o, s = self._sources.get(absoluteOffset)
logger.debug('at abs offset %d, we are in source %r' % (
absoluteOffset, s))
counterStart = self._sources.getCounterStart(c)
relativeOffset = absoluteOffset - counterStart
currentTrack.index(0, path=s and s.path or None,
absolute=absoluteOffset,
relative=relativeOffset, counter=c)
logger.debug('[track %02d index 00] added %r',
currentTrack.number, currentTrack.getIndex(0))
# store the pregapLength to add it when we index 1 for this
# track on the next iteration
pregapLength = length
# look for INDEX lines
m = _INDEX_RE.search(line)
if m:
if not currentTrack:
self.message(number, 'INDEX without preceding TRACK')
print 'ouch'
continue
indexNumber += 1
offset = common.msfToFrames(m.group('offset'))
self._index(currentTrack, indexNumber, absoluteOffset, offset)
# handle index 1 of final track, if any
if currentTrack:
self._index(currentTrack, 1, absoluteOffset, pregapLength)
# totalLength was added up to the penultimate track
self.table.leadout = totalLength + currentLength
logger.debug('parse: leadout: %r', self.table.leadout)
def message(self, number, message):
"""
Add a message about a given line in the cue file.
@param number: line number, counting from 0.
"""
self._messages.append((number + 1, message))
def getTrackLength(self, track):
"""
Returns the length of the given track, from its INDEX 01 to the next
track's INDEX 01
"""
# returns track length in frames, or -1 if can't be determined and
# complete file should be assumed
# FIXME: this assumes a track can only be in one file; is this true ?
i = self.table.tracks.index(track)
if i == len(self.table.tracks) - 1:
# last track, so no length known
return -1
thisIndex = track.indexes[1] # FIXME: could be more
nextIndex = self.table.tracks[i + 1].indexes[1] # FIXME: could be 0
c = thisIndex.counter
if c is not None and c == nextIndex.counter:
# they belong to the same source, so their relative delta is length
return nextIndex.relative - thisIndex.relative
# FIXME: more logic
return -1
def getRealPath(self, path):
"""
Translate the .toc's FILE to an existing path.
@type path: unicode
"""
return common.getRealPath(self._path, path)
class File:
"""
I represent a FILE line in a .toc file.
"""
def __init__(self, path, start, length):
"""
@type path: C{unicode}
@type start: C{int}
@param start: starting point for the track in this file, in frames
@param length: length for the track in this file, in frames
"""
assert type(path) is unicode, "%r is not unicode" % path
self.path = path
self.start = start
self.length = length
def __repr__(self):
return '<File %r>' % (self.path, )