Convert XML Youtube subtitles to SubRip (srt) from YouTube ID or XML file

youtube2srt.rb

# Convert XML Youtube subtitles to SubRip (srt) format
# To download the subtitle in XML, put the code of the Youtube video
# at the end of the next url:
# http://video.google.com/timedtext?hl=en&lang=en&v=

# Usage:
# 
# $ ruby youtube2srt.rb [input_filename] [output_filename]
# 
# Where input_filename can be either the name of your xml file
# (probably timedtext.xml) or the hashid of your YouTube video.
# The output filename is optional.

require 'rubygems'
require 'nokogiri'
require 'uri'
require 'net/http'

BASE_URL = 'http://video.google.com/timedtext?hl=en&lang=en&v='
source_filename = ARGV[0]
output_filename = ARGV[1]
TIME_FORMAT = '%02H:%02M:%02S,%3N'

def create_srt output_filename, source_filename, source_file
  File.open(output_filename || source_filename.gsub('.xml', '').concat('.srt'), 'w+') do |srt_file|
    source_file.css('text').to_enum.with_index(1) do |sub, i|
      start_time = Time.at(sub['start'].to_f).utc
      end_time = start_time + sub['dur'].to_f
      srt_file.write(<<~CAPTION
        #{i}
        #{start_time.strftime(TIME_FORMAT)} --> #{end_time.strftime(TIME_FORMAT)}
        #{Nokogiri::HTML.parse(sub.text).text}

      CAPTION
      )
    end
  end
end

if source_filename =~ /\.xml$/i
  source_file = Nokogiri::XML(open(source_filename), &:noblanks)
  create_srt(output_filename, source_filename, source_file)
  puts "xml file #{source_filename} converted to srt"
else
  response = Net::HTTP.get_response URI.parse(BASE_URL+source_filename)
  if response.code_type.ancestors.include?(Net::HTTPSuccess)
    source_file = Nokogiri::XML(response.body, &:noblanks)
    create_srt(output_filename, source_filename, source_file)
    puts 'Google timedtext.xml converted to srt'
  else
    puts "Couldn't find a srt file for #{source_filename} at #{BASE_URL + source_filename}"
  end
end

Thanks, this helped me today!