audiobook_split.py

#!/usr/bin/python
# -*- coding: utf-8 -*-

import re
import sys
import subprocess
import os
import math
import glob

if len(sys.argv)<2:
    print "pass arguments"
    sys.exit(0)

my_env = os.environ
my_env['PYTHONIOENCODING'] = 'utf-8'

def timeStrInt(time):
    m = re.search('([0-9]{2}):([0-9]{2}):([0-9]{2})',time)
    if m == None:
        return -1
    return int(m.group(1))*3600+int(m.group(2))*60+int(m.group(3))

def timeIntStr(time):
    h = math.floor(time/3600)
    time = time - h*3600
    m = math.floor(time/60)
    time = time - m*60
    s = time
    return "%02d:%02d:%02d" % (h,m,s)

def ffmpegDuration(file):
    proc = subprocess.Popen(['ffmpeg','-i',file],
                        stdout=subprocess.PIPE,
                        stderr=subprocess.PIPE,
                        )
    stdout_value, stderr_value = proc.communicate('')
    audioinfo = repr(stderr_value)
    m = re.search('Duration: ([0-9:]{8})',audioinfo)
    duration = timeStrInt(m.group(1))
    return duration

def ffmpegCut(filein,start,duration,artist,title,fileout):
    args = ['ffmpeg','-i',filein,'-ss',timeIntStr(start),'-t',timeIntStr(duration),'-acodec','copy','-metadata','artist="' + artist + '"','-metadata','title="' + title + '"',fileout]
    #print " ".join(args)
    proc = subprocess.call(args,stdout=subprocess.PIPE,stderr=subprocess.PIPE)

def getTitle(filename):
    filename = os.path.basename(filename).replace(".mp3","")
    filename1 = filename.split(" – ")
    filename2 = filename.split(" - ")
    if len(filename1) == 2:
        return filename1[1]
    if len(filename2) == 2:
        return filename2[1]
    return filename

def getArtist(filename):
    filename = os.path.basename(filename).replace(".mp3","")
    filename1 = filename.split(" – ")
    filename2 = filename.split(" - ")
    if len(filename1) == 2:
        return filename1[0]
    if len(filename2) == 2:
        return filename2[0]
    return ""

slice = 10 * 60
sync = 6
for i in range(len(sys.argv)-1):
    fileins = glob.glob(sys.argv[i+1]);
    for filein in fileins:
        print filein
        duration = ffmpegDuration(filein)
        m = re.search('(.*)\.mp3',filein)
        fileoutbase = m.group(1)
        for i in range(int(math.ceil(duration/slice))+1):
            fileout = fileoutbase + "-%02d.mp3" % (i+1)
            artist = getArtist(fileout)
            title = getTitle(fileout)
            start = i*slice
            duration = slice
            if i>0:
                start = start - sync / 2
                duration = duration + sync
            elif i==0:
                duration = duration + sync / 2
            #print title
            ffmpegCut(filein,start,duration,artist,title,fileout)