youtube-dl/youtube_dl/downloader/hls.py

from __future__ import unicode_literals

import os
import re
import subprocess

from .common import FileDownloader
from .fragment import FragmentFD

from ..compat import compat_urlparse
from ..postprocessor.ffmpeg import FFmpegPostProcessor
from ..utils import (
    encodeArgument,
    encodeFilename,
)


class HlsFD(FileDownloader):
    def real_download(self, filename, info_dict):
        url = info_dict['url']
        self.report_destination(filename)
        tmpfilename = self.temp_name(filename)

        ffpp = FFmpegPostProcessor(downloader=self)
        if not ffpp.available:
            self.report_error('m3u8 download detected but ffmpeg or avconv could not be found. Please install one.')
            return False
        ffpp.check_version()

        args = [
            encodeArgument(opt)
            for opt in (ffpp.executable, '-y', '-i', url, '-f', 'mp4', '-c', 'copy', '-bsf:a', 'aac_adtstoasc')]
        args.append(encodeFilename(tmpfilename, True))

        self._debug_cmd(args)

        retval = subprocess.call(args)
        if retval == 0:
            fsize = os.path.getsize(encodeFilename(tmpfilename))
            self.to_screen('\r[%s] %s bytes' % (args[0], fsize))
            self.try_rename(tmpfilename, filename)
            self._hook_progress({
                'downloaded_bytes': fsize,
                'total_bytes': fsize,
                'filename': filename,
                'status': 'finished',
            })
            return True
        else:
            self.to_stderr('\n')
            self.report_error('%s exited with code %d' % (ffpp.basename, retval))
            return False


class NativeHlsFD(FragmentFD):
    """ A more limited implementation that does not require ffmpeg """

    FD_NAME = 'hlsnative'

    def real_download(self, filename, info_dict):
        man_url = info_dict['url']
        self.to_screen('[%s] Downloading m3u8 manifest' % self.FD_NAME)
        manifest = self.ydl.urlopen(man_url).read()

        s = manifest.decode('utf-8', 'ignore')
        fragment_urls = []
        for line in s.splitlines():
            line = line.strip()
            if line and not line.startswith('#'):
                segment_url = (
                    line
                    if re.match(r'^https?://', line)
                    else compat_urlparse.urljoin(man_url, line))
                fragment_urls.append(segment_url)
                # We only download the first fragment during the test
                if self.params.get('test', False):
                    break

        ctx = {
            'filename': filename,
            'total_frags': len(fragment_urls),
        }

        self._prepare_and_start_frag_download(ctx)

        frags_filenames = []
        for i, frag_url in enumerate(fragment_urls):
            frag_filename = '%s-Frag%d' % (ctx['tmpfilename'], i)
            success = ctx['dl'].download(frag_filename, {'url': frag_url})
            if not success:
                return False
            with open(frag_filename, 'rb') as down:
                ctx['dest_stream'].write(down.read())
            frags_filenames.append(frag_filename)

        self._finish_frag_download(ctx)

        for frag_file in frags_filenames:
            os.remove(frag_file)

        return True