audiotools 3.1.1-1.1 / usr / bin / tracksplit

#!/usr/bin/python3

# Audio Tools, a module and set of tools for manipulating audio data
# Copyright (C) 2007-2015  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

import sys
import os
import os.path
import audiotools
import audiotools.cue
import audiotools.ui
import audiotools.text as _
import termios
from fractions import Fraction


def merge_metadatas(metadatas):
    if len(metadatas) == 0:
        return audiotools.MetaData()
    elif len(metadatas) == 1:
        return metadatas[0]
    else:
        merged = metadatas[0]
        for to_merge in metadatas[1:]:
            merged = merged.intersection(to_merge)
        return merged


def split(progress, source_audiofile, destination_filename,
          destination_class, compression, metadata,
          pcm_frames_offset, total_pcm_frames):
    try:
        pcmreader = source_audiofile.to_pcm()

        # if PCMReader has seek()
        # use it to reduce the amount of frames to skip
        if hasattr(pcmreader, "seek") and callable(pcmreader.seek):
            pcm_frames_offset -= pcmreader.seek(pcm_frames_offset)

        destination_audiofile = destination_class.from_pcm(
            str(destination_filename),
            audiotools.PCMReaderProgress(
                audiotools.PCMReaderWindow(pcmreader,
                                           pcm_frames_offset,
                                           total_pcm_frames),
                total_pcm_frames,
                progress),
            compression,
            total_pcm_frames)

        if metadata is not None:
            destination_audiofile.set_metadata(metadata)
    except KeyboardInterrupt:
        # remove partially split file, if any
        try:
            os.unlink(str(destination_filename))
        except OSError:
            pass

    return str(destination_filename)


def split_raw(progress, source_filename,
              sample_rate, channels, channel_mask, bits_per_sample,
              destination_filename, destination_class, compression,
              metadata, pcm_frames_offset, total_pcm_frames):
    f = open(source_filename, "rb")
    try:
        # skip initial offset
        f.seek(pcm_frames_offset * channels * (bits_per_sample // 8))

        destination_audiofile = destination_class.from_pcm(
            str(destination_filename),
            audiotools.PCMReaderProgress(
                audiotools.PCMReaderHead(
                    audiotools.PCMFileReader(file=f,
                                             sample_rate=sample_rate,
                                             channels=channels,
                                             channel_mask=channel_mask,
                                             bits_per_sample=bits_per_sample),
                    total_pcm_frames),
                total_pcm_frames,
                progress),
            compression,
            total_pcm_frames)

        if metadata is not None:
            destination_audiofile.set_metadata(metadata)

        return str(destination_filename)
    except KeyboardInterrupt:
        # remove partially split file, if any
        try:
            os.unlink(str(destination_filename))
        except OSError:
            pass
    finally:
        f.close()


if (__name__ == '__main__'):
    import argparse

    parser = argparse.ArgumentParser(description=_.DESCRIPTION_TRACKSPLIT)

    parser.add_argument("--version",
                        action="version",
                        version="Python Audio Tools %s" % (audiotools.VERSION))

    parser.add_argument("-I", "--interactive",
                        action="store_true",
                        default=False,
                        dest="interactive",
                        help=_.OPT_INTERACTIVE_OPTIONS)

    parser.add_argument("--cue",
                        dest="cuesheet",
                        metavar="FILENAME",
                        help=_.OPT_CUESHEET_TRACKSPLIT)

    parser.add_argument("-V", "--verbose",
                        dest="verbosity",
                        choices=audiotools.VERBOSITY_LEVELS,
                        default=audiotools.DEFAULT_VERBOSITY,
                        help=_.OPT_VERBOSE)

    conversion = parser.add_argument_group(_.OPT_CAT_ENCODING)

    conversion.add_argument(
        "-t", "--type",
        dest="type",
        choices=sorted(list(t.NAME for t in audiotools.AVAILABLE_TYPES
                            if t.supports_from_pcm()) + ["help"]),
        help=_.OPT_TYPE)

    conversion.add_argument("-q", "--quality",
                            dest="quality",
                            help=_.OPT_QUALITY)

    conversion.add_argument("-d", "--dir",
                            dest="dir",
                            default=".",
                            help=_.OPT_DIR)

    conversion.add_argument("--format",
                            default=None,
                            dest="format",
                            help=_.OPT_FORMAT)

    conversion.add_argument("-j", "--joint",
                            type=int,
                            default=audiotools.MAX_JOBS,
                            dest="max_processes",
                            help=_.OPT_JOINT)

    lookup = parser.add_argument_group(_.OPT_CAT_CD_LOOKUP)

    lookup.add_argument("--musicbrainz-server",
                        dest="musicbrainz_server",
                        default=audiotools.MUSICBRAINZ_SERVER,
                        metavar="HOSTNAME")

    lookup.add_argument("--musicbrainz-port",
                        type=int,
                        dest="musicbrainz_port",
                        default=audiotools.MUSICBRAINZ_PORT,
                        metavar="PORT")

    lookup.add_argument("--no-musicbrainz",
                        action="store_false",
                        dest="use_musicbrainz",
                        default=audiotools.MUSICBRAINZ_SERVICE,
                        help=_.OPT_NO_MUSICBRAINZ)

    lookup.add_argument("--freedb-server",
                        dest="freedb_server",
                        default=audiotools.FREEDB_SERVER,
                        metavar="HOSTNAME")

    lookup.add_argument("--freedb-port",
                        type=int,
                        dest="freedb_port",
                        default=audiotools.FREEDB_PORT,
                        metavar="PORT")

    lookup.add_argument("--no-freedb",
                        action="store_false",
                        dest="use_freedb",
                        default=audiotools.FREEDB_SERVICE,
                        help=_.OPT_NO_FREEDB)

    lookup.add_argument("-D", "--default",
                        dest="use_default",
                        action="store_true",
                        default=False,
                        help=_.OPT_DEFAULT)

    metadata = parser.add_argument_group(_.OPT_CAT_METADATA)

    metadata.add_argument("--album-number",
                          dest="album_number",
                          type=int,
                          default=0,
                          help=_.OPT_ALBUM_NUMBER)

    metadata.add_argument("--album-total",
                          dest="album_total",
                          type=int,
                          default=0,
                          help=_.OPT_ALBUM_TOTAL)

    metadata.add_argument("--replay-gain",
                          action="store_true",
                          default=None,
                          dest="add_replay_gain",
                          help=_.OPT_REPLAY_GAIN)

    metadata.add_argument("--no-replay-gain",
                          action="store_false",
                          default=None,
                          dest="add_replay_gain",
                          help=_.OPT_NO_REPLAY_GAIN)

    parser.add_argument("filename",
                        metavar="FILENAME",
                        help=_.OPT_INPUT_FILENAME)

    options = parser.parse_args()

    msg = audiotools.Messenger(options.verbosity == "quiet")

    # ensure interactive mode is available, if selected
    if options.interactive and (not audiotools.ui.AVAILABLE):
        audiotools.ui.not_available_message(msg)
        sys.exit(1)

    # get the AudioFile class we are converted to
    if options.type == 'help':
        audiotools.ui.show_available_formats(msg)
        sys.exit(0)
    if options.type is not None:
        AudioType = audiotools.TYPE_MAP[options.type]
    else:
        AudioType = audiotools.TYPE_MAP[audiotools.DEFAULT_TYPE]

    # ensure the selected compression is compatible with that class
    if options.quality == 'help':
        audiotools.ui.show_available_qualities(msg, AudioType)
        sys.exit(0)
    elif options.quality is None:
        options.quality = audiotools.__default_quality__(AudioType.NAME)
    elif options.quality not in AudioType.COMPRESSION_MODES:
        msg.error(_.ERR_UNSUPPORTED_COMPRESSION_MODE %
                  {"quality": options.quality,
                   "type": AudioType.NAME})
        sys.exit(1)

    try:
        filename = audiotools.Filename(options.filename)
        audiofile = audiotools.open(str(filename))
        input_filename = audiotools.Filename(audiofile.filename)
        input_filenames = {input_filename}
    except audiotools.UnsupportedFile:
        msg.error(_.ERR_UNSUPPORTED_FILE % (filename,))
        sys.exit(1)
    except audiotools.InvalidFile as err:
        msg.error(str(err))
        sys.exit(1)
    except IOError:
        msg.error(_.ERR_OPEN_IOERROR % (filename,))
        sys.exit(1)

    base_directory = options.dir
    encoded_filenames = []

    if options.cuesheet is not None:
        # grab the cuesheet we're using to split tracks
        # (this overrides an embedded cuesheet)
        try:
            cuesheet = audiotools.read_sheet(options.cuesheet)
            input_filenames.add(audiotools.Filename(options.cuesheet))
        except audiotools.SheetException as err:
            msg.error(err)
            sys.exit(1)
    else:
        cuesheet = audiofile.get_cuesheet()
        if cuesheet is None:
            msg.error(_.ERR_TRACKSPLIT_NO_CUESHEET)
            sys.exit(1)

    if not cuesheet.image_formatted():
        msg.error(_.ERR_CUE_INVALID_FORMAT)
        sys.exit(1)

    cuesheet_offsets = [cuesheet.track_offset(t) for t in
                        cuesheet.track_numbers()]

    cuesheet_lengths = [cuesheet.track_length(t) for t in
                        cuesheet.track_numbers()]

    audiofile_size = audiofile.seconds_length()

    if cuesheet_lengths[-1] is None:
        # last track length unknown, so ensure there's enough room
        # based on the total size of the source
        last_track_size = (audiofile_size -
                           sum(cuesheet_lengths[0:-1]) -
                           cuesheet.pre_gap())
        if last_track_size >= 0:
            cuesheet_lengths.pop()
            cuesheet_lengths.append(last_track_size)
        else:
            msg.error(_.ERR_TRACKSPLIT_OVERLONG_CUESHEET)
            sys.exit(1)
    else:
        # last track length is known, so ensure the size
        # of all lengths and the pre-gap is the same size as the source
        if (cuesheet.pre_gap() + sum(cuesheet_lengths)) != audiofile_size:
            msg.error(_.ERR_TRACKSPLIT_OVERLONG_CUESHEET)
            sys.exit(1)

    output_track_count = len(cuesheet)

    # check whether pre-gap data should be preserved, if any
    if cuesheet.pre_gap() > 0:
        with audiotools.BufferedPCMReader(audiofile.to_pcm()) as r:
            # this could be optimized better
            # but it is a rare and unusual case
            preserve_pre_gap = set(r.read(int(cuesheet.pre_gap() *
                                              audiofile.sample_rate()))) != {0}

            if preserve_pre_gap:
                cuesheet_offsets.insert(0, Fraction(0, 1))
                cuesheet_lengths.insert(0, cuesheet.pre_gap())
    else:
        preserve_pre_gap = False

    # use cuesheet to query metadata services for metadata choices
    try:
        metadata_choices = audiotools.sheet_metadata_lookup(
            sheet=cuesheet,
            total_pcm_frames=audiofile.total_frames(),
            sample_rate=audiofile.sample_rate(),
            musicbrainz_server=options.musicbrainz_server,
            musicbrainz_port=options.musicbrainz_port,
            freedb_server=options.freedb_server,
            freedb_port=options.freedb_port,
            use_musicbrainz=options.use_musicbrainz,
            use_freedb=options.use_freedb)
    except KeyboardInterrupt:
        msg.ansi_clearline()
        msg.error(_.ERR_CANCELLED)
        sys.exit(1)

    if preserve_pre_gap:
        # prepend "track 0" track to start of list for each choice
        for choice in metadata_choices:
            track_0 = merge_metadatas(choice)
            track_0.track_number = 0
            choice.insert(0, track_0)

    # populate any empty Album-level metadata fields
    # with those from original file, if any
    album_metadata = audiofile.get_metadata()
    if album_metadata is not None:
        album_fields = {attr: field for (attr, field) in
                        album_metadata.filled_fields()
                        if (attr in {"album_name",
                                     "artist_name",
                                     "performer_name",
                                     "composer_name",
                                     "conductor_name",
                                     "media",
                                     "catalog",
                                     "copyright",
                                     "publisher",
                                     "year",
                                     "date",
                                     "album_number",
                                     "album_total",
                                     "comment"})}

        for c in metadata_choices:
            for m in c:
                for (attr, field) in m.empty_fields():
                    if attr in album_fields:
                        setattr(m, attr, album_fields[attr])

    # populate any remaining metadata fields
    # with those from cuesheet, if any
    cuesheet_metadata = cuesheet.get_metadata()
    for c in metadata_choices:
        for (track_metadata,
             sheet_track) in zip(c[1:] if preserve_pre_gap else c, cuesheet):
            sheet_track_metadata = sheet_track.get_metadata()
            for (attr, field) in track_metadata.empty_fields():
                if ((sheet_track_metadata is not None) and
                    (getattr(sheet_track_metadata, attr) is not None)):
                    setattr(track_metadata,
                            attr,
                            getattr(sheet_track_metadata, attr))
                elif ((cuesheet_metadata is not None) and
                      (getattr(cuesheet_metadata, attr) is not None)):
                    setattr(track_metadata,
                            attr,
                            getattr(cuesheet_metadata, attr))

    # update MetaData with command-line album-number/total, if given
    if options.album_number != 0:
        for c in metadata_choices:
            for m in c:
                m.album_number = options.album_number

    if options.album_total != 0:
        for c in metadata_choices:
            for m in c:
                m.album_total = options.album_total

    input_filenames = \
        [input_filename] * (output_track_count + (1 if preserve_pre_gap else 0))

    # decide which metadata and output options to use when splitting tracks
    if options.interactive:
        track_labels = [_.LAB_TRACK_X_OF_Y % (i, output_track_count)
                        for i in range(0 if preserve_pre_gap else 1,
                                       output_track_count + 1)]

        # pick choice using interactive widget
        output_widget = audiotools.ui.OutputFiller(
            track_labels=track_labels,
            metadata_choices=metadata_choices,
            input_filenames=input_filenames,
            output_directory=options.dir,
            format_string=(options.format if
                           (options.format is not None) else
                           audiotools.FILENAME_FORMAT),
            output_class=AudioType,
            quality=options.quality,
            completion_label=_.LAB_TRACKSPLIT_APPLY)

        loop = audiotools.ui.urwid.MainLoop(
            output_widget,
            audiotools.ui.style(),
            screen=audiotools.ui.Screen(),
            unhandled_input=output_widget.handle_text,
            pop_ups=True)
        try:
            loop.run()
            msg.ansi_clearscreen()
        except (termios.error, IOError):
            msg.error(_.ERR_TERMIOS_ERROR)
            msg.info(_.ERR_TERMIOS_SUGGESTION)
            msg.info(audiotools.ui.xargs_suggestion(sys.argv))
            sys.exit(1)

        if not output_widget.cancelled():
            output_tracks = list(output_widget.output_tracks())
        else:
            sys.exit(0)
    else:
        # pick choice without using GUI
        try:
            output_tracks = list(
                audiotools.ui.process_output_options(
                    metadata_choices=metadata_choices,
                    input_filenames=input_filenames,
                    output_directory=options.dir,
                    format_string=options.format,
                    output_class=AudioType,
                    quality=options.quality,
                    msg=msg,
                    use_default=options.use_default))
        except audiotools.UnsupportedTracknameField as err:
            err.error_msg(msg)
            sys.exit(1)
        except (audiotools.InvalidFilenameFormat,
                audiotools.OutputFileIsInput,
                audiotools.DuplicateOutputFile) as err:
            msg.error(err)
            sys.exit(1)

    # perform actual track splitting and tagging
    jobs = zip(cuesheet_offsets, cuesheet_lengths, output_tracks)

    queue = audiotools.ExecProgressQueue(msg)

    if audiofile.seekable():
        for (offset, length, (output_class,
                              output_filename,
                              output_quality,
                              output_metadata)) in jobs:
            try:
                audiotools.make_dirs(str(output_filename))
            except OSError as err:
                msg.os_error(err)
                sys.exit(1)

            queue.execute(
                function=split,
                progress_text=output_filename.__unicode__(),
                completion_output=_.LAB_ENCODE % {
                    "source": audiotools.Filename(audiofile.filename),
                    "destination": output_filename},
                source_audiofile=audiofile,
                destination_filename=output_filename,
                destination_class=output_class,
                compression=output_quality,
                metadata=output_metadata,
                pcm_frames_offset=int(offset * audiofile.sample_rate()),
                total_pcm_frames=int(length * audiofile.sample_rate()))

        try:
            encoded_tracks = map(audiotools.open,
                                 queue.run(options.max_processes))
        except audiotools.EncodingError as err:
            msg.error(err)
            sys.exit(1)
        except KeyboardInterrupt:
            msg.error(_.ERR_CANCELLED)
            sys.exit(1)
    else:
        import tempfile

        # if input file isn't seekable

        # decode it into a single PCM blob of binary data
        temp_blob = tempfile.NamedTemporaryFile()
        cache_progress = audiotools.SingleProgressDisplay(
            msg, _.LAB_CACHING_FILE)
        try:
            audiotools.transfer_framelist_data(
                audiotools.PCMReaderProgress(
                    audiofile.to_pcm(),
                    audiofile.total_frames(),
                    cache_progress.update),
                temp_blob.write)
        except audiotools.DecodingError as err:
            cache_progress.clear_rows()
            msg.error(err)
            temp_blob.close()
            sys.exit(1)
        except KeyboardInterrupt:
            cache_progress.clear_rows()
            msg.error(_.ERR_CANCELLED)
            temp_blob.close()
            sys.exit(1)

        cache_progress.clear_rows()
        temp_blob.flush()

        # split the blob using multiple jobs
        for (offset, length, (output_class,
                              output_filename,
                              output_quality,
                              output_metadata)) in jobs:
            try:
                audiotools.make_dirs(str(output_filename))
            except OSError as err:
                msg.os_error(err)
                temp_blob.close()
                sys.exit(1)

            queue.execute(
                function=split_raw,
                progress_text=output_filename.__unicode__(),
                completion_output=_.LAB_ENCODE % {
                    "source": audiotools.Filename(audiofile.filename),
                    "destination": output_filename},
                source_filename=temp_blob.name,
                sample_rate=audiofile.sample_rate(),
                channels=audiofile.channels(),
                channel_mask=int(audiofile.channel_mask()),
                bits_per_sample=audiofile.bits_per_sample(),
                destination_filename=output_filename,
                destination_class=output_class,
                compression=output_quality,
                metadata=output_metadata,
                pcm_frames_offset=int(offset * audiofile.sample_rate()),
                total_pcm_frames=int(length * audiofile.sample_rate()))

        try:
            encoded_tracks = map(audiotools.open,
                                 queue.run(options.max_processes))
        except audiotools.EncodingError as err:
            msg.error(err)
            temp_blob.close()
            sys.exit(1)
        except KeyboardInterrupt:
            msg.error(_.ERR_CANCELLED)
            temp_blob.close()
            sys.exit(1)

        # then delete the blob when finished
        temp_blob.close()

    # apply ReplayGain to split tracks, if requested
    if (output_class.supports_replay_gain() and
        (options.add_replay_gain if options.add_replay_gain is not None else
         audiotools.ADD_REPLAYGAIN)):
        rg_progress = audiotools.ReplayGainProgressDisplay(msg)
        rg_progress.initial_message()
        try:
            # separate encoded files by album_name and album_number
            for album in audiotools.group_tracks(encoded_tracks):
                # add ReplayGain to groups of files
                # belonging to the same album

                audiotools.add_replay_gain(album, rg_progress.update)
        except ValueError as err:
            rg_progress.clear_rows()
            msg.error(err)
            sys.exit(1)
        except KeyboardInterrupt:
            rg_progress.clear_rows()
            msg.error(_.ERR_CANCELLED)
            sys.exit(1)
        rg_progress.final_message()