Add support for C-SPAN subtitles

[lfl2]

Most of C-SPAN videos, contain subtitles, it is presented in DFXP format, and loaded from URL like this:
http://data.c-spanvideo.org/Programs/aaa/bbb/ccc.dfxp

DFXP can be easily converted to SRT format:

#!/usr/bin/python
# Usage: python tt2srt.py source.xml output.srt

from xml.dom.minidom import parse
import sys
i=1
dom = parse(sys.argv[1])
out = open(sys.argv[2], 'w')
body = dom.getElementsByTagName("body")[0]
paras = body.getElementsByTagName("p")
for para in paras:
    out.write(str(i) + "\n")
    try:
        a=float(para.attributes['begin'].value)
    except ValueError:
        a=0
    out.write('%02d' %(int(a/3600)))
    out.write(":")
    out.write('%02d' % (int(a/60)-60*(int(a/3600))))
    out.write(':')
    out.write('%02d' % (a%60))
    out.write(',')
    out.write('%03d' % (a%60.0-a%60))
    out.write(' --> ')

    a= float(para.attributes['end'].value)

    out.write('%02d' %(int(a/3600)))
    out.write(":")
    out.write('%02d' % (int(a/60)-60*(int(a/3600))))
    out.write(':')
    out.write('%02d' % (a%60))
    out.write(',')
    out.write('%03d' % (a%60.0-a%60))

    out.write("\n")
    for child in para.childNodes:
        if child.nodeName == 'br':
            out.write("\n")
        elif child.nodeName == '#text':
            out.write(unicode(child.data).encode('utf=8'))
    out.write("\n\n")
    i += 1

[jaimeMF]

For reference, our test video http://www.c-span.org/video/?313572-1/attorney-general-eric-holder-voting-rights-act-decision has as subtitle url http://data.c-spanvideo.org/Programs/315/315139/1421809454.dfxp

[yan12125]

Thanks @lfl2. I've adopted your implementation and do some modifications so that is can be integrated better with youtube-dl. Now dfxp/TTML subtitles are supported.