Ejemplos de Amazon Polly con SDK for Ruby

require 'aws-sdk-polly' # In v2: require 'aws-sdk'

begin
  # Create an Amazon Polly client using
  # credentials from the shared credentials file ~/.aws/credentials
  # and the configuration (region) from the shared configuration file ~/.aws/config
  polly = Aws::Polly::Client.new

  # Get US English voices
  resp = polly.describe_voices(language_code: 'en-US')

  resp.voices.each do |v|
    puts v.name
    puts "  #{v.gender}"
    puts
  end
rescue StandardError => e
  puts 'Could not get voices'
  puts 'Error message:'
  puts e.message
end

require 'aws-sdk-polly' # In v2: require 'aws-sdk'

begin
  # Create an Amazon Polly client using
  # credentials from the shared credentials file ~/.aws/credentials
  # and the configuration (region) from the shared configuration file ~/.aws/config
  polly = Aws::Polly::Client.new

  resp = polly.list_lexicons

  resp.lexicons.each do |l|
    puts l.name
    puts "  Alphabet:#{l.attributes.alphabet}"
    puts "  Language:#{l.attributes.language}"
    puts
  end
rescue StandardError => e
  puts 'Could not get lexicons'
  puts 'Error message:'
  puts e.message
end

require 'aws-sdk-polly' # In v2: require 'aws-sdk'

begin
  # Get the filename from the command line
  if ARGV.empty?
    puts 'You must supply a filename'
    exit 1
  end

  filename = ARGV[0]

  # Open file and get the contents as a string
  if File.exist?(filename)
    contents = IO.read(filename)
  else
    puts "No such file: #{filename}"
    exit 1
  end

  # Create an Amazon Polly client using
  # credentials from the shared credentials file ~/.aws/credentials
  # and the configuration (region) from the shared configuration file ~/.aws/config
  polly = Aws::Polly::Client.new

  resp = polly.synthesize_speech({
                                   output_format: 'mp3',
                                   text: contents,
                                   voice_id: 'Joanna'
                                 })

  # Save output
  # Get just the file name
  #  abc/xyz.txt -> xyx.txt
  name = File.basename(filename)

  # Split up name so we get just the xyz part
  parts = name.split('.')
  first_part = parts[0]
  mp3_file = "#{first_part}.mp3"

  IO.copy_stream(resp.audio_stream, mp3_file)

  puts "Wrote MP3 content to: #{mp3_file}"
rescue StandardError => e
  puts 'Got error:'
  puts 'Error message:'
  puts e.message
end

Esta aplicación de ejemplo analiza y almacena las tarjetas de comentarios de los clientes. Concretamente, satisface la necesidad de un hotel ficticio en la ciudad de Nueva York. El hotel recibe comentarios de los huéspedes en varios idiomas en forma de tarjetas de comentarios físicas. Esos comentarios se cargan en la aplicación a través de un cliente web. Una vez cargada la imagen de una tarjeta de comentarios, se llevan a cabo los siguientes pasos:

La aplicación completa se puede implementar con AWS CDK. Para ver el código fuente y las instrucciones de implementación, consulta el proyecto en GitHub.