Converts a folder of srt files into a single srt

dir_to_srt.py

# -*- coding: utf-8 -*-
'''
   Quick script to embed into Automator to concat a folder of srt files into
   a single srt.

   Note: This also removes extra blanks lines, to allow pycaption to process
   the files. This won't be needed if https://github.com/pbs/pycaption/pull/39
   is merged.
'''

import argparse
import codecs
import re
import traceback

from os import walk


def detect_by_bom(path, default):
    print path, default
    with open(path, 'rb') as f:
        raw = f.read(4)

    for enc, boms in \
            ('utf-8-sig', (codecs.BOM_UTF8,)),\
            ('utf-16', (codecs.BOM_UTF16_LE, codecs.BOM_UTF16_BE)),\
            ('utf-32', (codecs.BOM_UTF32_LE, codecs.BOM_UTF32_BE)):
        if any(raw.startswith(bom) for bom in boms):
            return enc

    return default


def srt_it(srt_folder, srt_out):
    srt_file = codecs.open(srt_out, "w", "utf-8")

    files = []
    for (_, _, filenames) in walk(srt_folder):
        files.extend([x for x in filenames
                     if x.lower().rsplit('.', 1)[1] is 'srt'])
        break

    counter = 0
    last_blank = False
    for f in files:
        filename = "{}/{}".format(srt_folder, f)
        encoding = detect_by_bom(filename, 'utf-8')
        print encoding

        for l in codecs.open(filename, "r", encoding):
            # Convert Line Ending
            l = l.rstrip() + '\n'
            # For some reason the first line of a files isn't catching so
            #  adding a . seems to catch it...
            if re.match(r"\d+$", l):
                last_blank = False
                counter += 1
                srt_file.write("{}\n".format(counter))
            elif l == "\n":
                if not last_blank:
                    srt_file.write(l)
                last_blank = True
            else:
                last_blank = False
                srt_file.write(l)

    srt_file.close()

if __name__ == "__main__":
    try:
        parser = argparse.ArgumentParser(description='Merge Folder of SRTs.')
        parser.add_argument(dest='srt_folders', metavar='file', nargs='+',
                            help='Folders full of srt files.')

        args = parser.parse_args()

        for srt_folder in args.srt_folders:
            if srt_folder[-1] in ['\\', '/']:
                srt_folder = srt_folder[:-1]
            srt_out = "{}.srt".format(srt_folder)

            srt_it(srt_folder, srt_out)
    except Exception as e:
            print ("Connection failed:\n{}\n{}\n{}".
                   format(str(e),
                          type(e).__name__,
                          traceback.format_exc()))

Note: This script is embedded into an automator program so added manual print out of exception, to allow debugging.