radio/util/tts.py

   1 #!/usr/bin/env python
   2 # -*- coding: utf-8 -*-
   3
   4
   5 # To use this script you need the following installed within python
   6 #  Python 2.5  http://www.python.org/download/releases/2.5.4/
   7 #  Python 2.7   http://www.python.org/download/releases/2.7.3/
   8 #  PyTTS  http://pypi.python.org/pypi/pyTTS
   9 #  PyWin32 http://sourceforge.net/projects/pywin32/files/pywin32/
  10 #
  11 #  Note
  12 #  At the moment, pyTTS is only available for Python 2.3, 2.4 and 2.5. To use it for later versions without having to
  13 #  recompile it, a quick and dirty solution is to:
  14 #  copy the entire pyTTS directory from Python25\Lib\site-packages to Python26 or Python27
  15 #  replace TTSFast.py with an empty file. This way the version-dependent pyd file isn't loaded.
  16
  17 # in addition you will need some tools.
  18 #  ffmpeg, sox, adconvertor ttscmd (2cnd speach centre)
  19 # have fun!
  20 # Sound pack maintainers (incomplete list) by language alphabetical order
  21 # Czech      : Martin Hotar
  22 # French     : Bertrand Songis & André Bernet
  23 # English    : Rob Thompson & Martin Hotar
  24 # German     : Romolo Manfredini (Some corrections by Peer)
  25 # Italian    : Romolo Manfredini
  26 # Portuguese : Romolo Manfredini
  27 # Spanish    : Romolo Manfredini (With the help of Jose Moreno)
  28
  29 # from __future__ import print_function
  30
  31 import os
  32 import sys
  33 import subprocess
  34 import zipfile
  35 from tts_common import *
  36
  37 try:
  38     # Python 3
  39     from http.client import HTTPConnection
  40     from urllib.parse import urlencode
  41 except ImportError:
  42     # Python 2
  43     from httplib import HTTPConnection
  44     from urllib import urlencode
  45
  46
  47 def wavstrip(filename):
  48     output = "_" + filename
  49     subprocess.Popen(["sox", filename, output, "silence", "1", "0.1", "0.1%", "reverse", "silence", "1", "0.1", "0.1%", "reverse"], stdout=subprocess.PIPE, stderr=subprocess.PIPE).wait()
  50     os.remove(filename)
  51     os.rename(output, filename)
  52
  53
  54 def generate(str, filename):
  55     print(filename, str)
  56
  57     if not str:
  58         str = " !"  # this is so blank wav files never exist!
  59
  60     if "speak" in sys.argv:
  61         if "sapi" in sys.argv:
  62             tts.Speak(str)
  63         elif "espeak" in sys.argv:
  64             subprocess.Popen(["espeak", "-v", espeakVoice, "-s", espeakspeed, "-z", str.encode("utf-8")], stdout=subprocess.PIPE, stderr=subprocess.PIPE).wait()
  65         elif "google" in sys.argv:
  66             "speak not implemented with google tts engine"
  67             exit()
  68         else:
  69             print("which speach engine?")
  70             exit()
  71     else:
  72         if "sapi" in sys.argv:
  73             ttsfilename = "ttsfile.wav"
  74             tts.SpeakToWave(ttsfilename, str)
  75         elif "sapi2" in sys.argv:
  76             ttsfilename = "ttsfile.wav"
  77             subprocess.Popen(["ttscmd", "/ttw", str.encode("utf-8"), ttsfilename, "-v", voiceID, "-b", "32", "-s", "\"-3\"", "-w", "32", "-f", "47"], stdout=subprocess.PIPE, stderr=subprocess.PIPE).wait()
  78         elif "espeak" in sys.argv:
  79             ttsfilename = "ttsfile.wav"
  80             subprocess.Popen(["espeak", "-v", espeakVoice, "-s", espeakspeed, "-z", "-w", ttsfilename, str.encode("utf-8")], stdout=subprocess.PIPE, stderr=subprocess.PIPE).wait()
  81         elif "google" in sys.argv:
  82             ttsmp3 = "ttsfile.mp3"
  83             ttsfilename = "ttsfile.wav"
  84             conn = HTTPConnection("translate.google.com")
  85             params = urlencode({'ie': "UTF-8", 'tl': directory, 'q': str.encode("utf-8")})
  86             headers = {"User-Agent": "Mozilla"}
  87             conn.request("GET", u"/translate_tts?%s" % params, headers=headers)
  88             # conn.request("GET", "/translate_tts?ie=UTF-8&tl=%s&q=%s" % (directory, urllib.urlencode(str)), headers={"User-Agent": "Mozilla"})
  89             resp = conn.getresponse()
  90             with open(ttsmp3, "wb") as f:
  91                 f.write(resp.read())
  92             subprocess.Popen(["ffmpeg", "-y", "-i", ttsmp3, "-acodec", "pcm_s16le", ttsfilename], stdout=subprocess.PIPE, stderr=subprocess.PIPE).wait()
  93             conn.close()
  94
  95         else:
  96             print("which speach engine?")
  97             exit()
  98
  99         wavstrip(ttsfilename)
 100
 101         if board in ('sky9x', 'taranis'):
 102             if 'sox' in sys.argv:
 103                 maxvolume = subprocess.Popen(["sox", ttsfilename, "-n", "stat", "-v"], stdout=subprocess.PIPE, stderr=subprocess.PIPE).communicate()[1]
 104                 if "not sound" in maxvolume:
 105                     subprocess.Popen(["sox", "--show-progress", filename, ttsfilename], stdout=subprocess.PIPE).communicate()[0]
 106                 else:
 107                     subprocess.Popen(["sox", "--show-progress", "-v", maxvolume, filename, ttsfilename], stdout=subprocess.PIPE).communicate()[0]
 108                     if board == 'sky9x':
 109                         subprocess.Popen(["sox", "-twav", ttsfilename, "-b1600", "-c1", "-e", "a-law", filename], stdout=subprocess.PIPE, stderr=subprocess.PIPE).wait()
 110                     else:
 111                         subprocess.Popen(["sox", "-twav", ttsfilename, "-b32000", "-c1", "-e", "a-law", filename], stdout=subprocess.PIPE, stderr=subprocess.PIPE).wait()
 112             else:
 113                 if board == 'sky9x':
 114                     subprocess.Popen(["ffmpeg", "-y", "-i", ttsfilename, "-acodec", defaultcodec, "-ar", "16000", filename], stdout=subprocess.PIPE, stderr=subprocess.PIPE).wait()
 115                 else:
 116                     subprocess.Popen(["ffmpeg", "-y", "-i", ttsfilename, "-acodec", defaultcodec, "-ar", "32000", filename], stdout=subprocess.PIPE, stderr=subprocess.PIPE).wait()
 117         elif board == 'gruvin9x':
 118             subprocess.Popen(["AD4CONVERTER", "-E4", ttsfilename], stdout=subprocess.PIPE, stderr=subprocess.PIPE).wait()
 119             os.rename(ttsfilename.replace(".wav", ".ad4"), filename)
 120         else:
 121             subprocess.Popen(["ffmpeg", "-y", "-i", ttsfilename, "-acodec", "pcm_u8", "-ar", "16000", filename], stdout=subprocess.PIPE, stderr=subprocess.PIPE).wait()
 122         os.remove(ttsfilename)
 123
 124 ################################################################
 125
 126 if __name__ == "__main__":
 127
 128
 129
 130     if "sapi" in sys.argv:
 131         import pyTTS
 132         tts = pyTTS.Create()
 133         # tts.SetOutputFormat(16, 16, 1)
 134         # tts.Volume = 40
 135         # tts.SetRate(1)
 136         if "list" in sys.argv:
 137             print(tts.GetVoiceNames())
 138             exit()
 139
 140     if "mulaw" in sys.argv:
 141         defaultcodec = "pcm_mulaw"
 142     else:
 143         defaultcodec = "pcm_alaw"
 144
 145     if "en" in sys.argv:
 146         from tts_en import systemSounds, sounds
 147
 148         directory = "en"
 149         voice = "english"
 150         if "sapi" in sys.argv:
 151             if "scottish" in sys.argv:
 152                 tts.SetVoiceByName("ScanSoftFiona_Full_22kHz")
 153                 voice = "english-scottish"
 154             elif "american" in sys.argv:
 155                 tts.SetVoiceByName("ScanSoftJennifer_Full_22kHz")
 156                 voice = "english-american"
 157             elif "australian" in sys.argv:
 158                 tts.SetVoiceByName("ScanSoftKaren_Full_22kHz")
 159                 voice = "english-australian"
 160             elif "irish" in sys.argv:
 161                 tts.SetVoiceByName("ScanSoftMoira_Full_22kHz")
 162                 voice = "english-irish"
 163             else:
 164                 tts.SetVoiceByName("ScanSoftFiona_Full_22kHz")
 165                 voice = "english-english"
 166         elif "sapi2" in sys.argv:
 167             if "scottish" in sys.argv:
 168                 voiceID = "17"
 169                 voice = "english-scottish"
 170             elif "american" in sys.argv:
 171                 voiceID = "18"
 172                 voice = "english-american"
 173             elif "australian" in sys.argv:
 174                 voiceID = "20"
 175                 voice = "english-australian"
 176             elif "irish" in sys.argv:
 177                 voiceID = "21"
 178                 voice = "english-irish"
 179             elif "french" in sys.argv:
 180                 voiceID = "19"
 181                 voice = "english-french"
 182             elif "german" in sys.argv:
 183                 voiceID = "22"
 184                 voice = "english-german"
 185             else:
 186                 voiceID = "17"
 187                 voice = "english-english"
 188
 189         elif "espeak" in sys.argv:
 190             espeakVoice = "mb-us1"
 191             espeakspeed = "150"
 192
 193     elif "fr" in sys.argv:
 194         from tts_fr import systemSounds, sounds
 195
 196         directory = "fr"
 197         voice = "french"
 198         if "sapi" in sys.argv:
 199             tts.SetVoiceByName("ScanSoftVirginie_Full_22kHz")
 200         elif "espeak" in sys.argv:
 201             espeakVoice = "mb-fr4+f4"
 202             espeakspeed = "140"
 203
 204     elif "it" in sys.argv:
 205         from tts_it import systemSounds, sounds
 206
 207         directory = "it"
 208         voice = "italian"
 209         if "sapi" in sys.argv:
 210             tts.SetVoiceByName("ScanSoftVirginie_Full_22kHz")
 211         elif "espeak" in sys.argv:
 212             espeakVoice = "mb-it4"
 213             espeakspeed = "160"
 214
 215     elif "de" in sys.argv:
 216         from tts_de import systemSounds, sounds
 217
 218         directory = "de"
 219         voice = "german"
 220         if "sapi" in sys.argv:
 221             tts.SetVoiceByName("ScanSoftVirginie_Full_22kHz")
 222         elif "espeak" in sys.argv:
 223             espeakVoice = "mb-de4"
 224             espeakspeed = "160"
 225
 226     elif "pt" in sys.argv:
 227         from tts_pt import systemSounds, sounds
 228
 229         directory = "pt"
 230         voice = "portuguese"
 231         if "sapi" in sys.argv:
 232             tts.SetVoiceByName("ScanSoftVirginie_Full_22kHz")
 233         elif "espeak" in sys.argv:
 234             espeakVoice = "mb-pt1+f1"
 235             espeakspeed = "160"
 236
 237     elif "es" in sys.argv:
 238         from tts_es import systemSounds, sounds
 239
 240         directory = "es"
 241         voice = "spanish"
 242         if "sapi" in sys.argv:
 243             tts.SetVoiceByName("ScanSoftVirginie_Full_22kHz")
 244         elif "espeak" in sys.argv:
 245             espeakVoice = "mb-es1+f1"
 246             espeakspeed = "160"
 247
 248     elif "cz" in sys.argv:
 249         from tts_cz import systemSounds, sounds
 250
 251         directory = "cz"
 252         voice = "czech"
 253         if "sapi" in sys.argv:
 254             tts.SetVoiceByName("ScanSoftZuzana_Full_22kHz")
 255         elif "espeak" in sys.argv:
 256             espeakVoice = "mb-cz2"
 257             espeakspeed = "150"
 258
 259     else:
 260         print("which language?")
 261         exit()
 262
 263     if "csv" in sys.argv:
 264         with open("%s-%s.csv" % (voice, board), "w") as csvFile:
 265             for s, f in systemSounds:
 266                 if s and f:
 267                     l = u""
 268                     if board in ("sky9x", "taranis"):
 269                         l += u"SOUNDS/%s/SYSTEM;" % directory
 270                     l += f + u";" + s + u"\n"
 271                     csvFile.write(l.encode("utf-8"))
 272             for s, f in sounds:
 273                 if s and f:
 274                     l = u""
 275                     if board in ("sky9x", "taranis"):
 276                         l += u"SOUNDS/%s;" % directory
 277                     l += f + u";" + s + u"\n"
 278                     csvFile.write(l.encode("utf-8"))
 279
 280     if "zip" in sys.argv:
 281         zip_name = "%s-%s.zip" % (voice, board)
 282         with zipfile.ZipFile(zip_name, "w", zipfile.ZIP_DEFLATED) as zip:
 283             for s, f in systemSounds:
 284                 if s and f:
 285                     generate(s, f)
 286                     if board in ("sky9x", "taranis"):
 287                         zip.write(f, "SOUNDS/%s/SYSTEM/" % directory + f)
 288                     else:
 289                         zip.write(f, f)
 290                     os.remove(f)
 291             for s, f in sounds:
 292                 if s and f:
 293                     generate(s, f)
 294                     if board in ("sky9x", "taranis"):
 295                         zip.write(f, "SOUNDS/%s/" % directory + f)
 296                     else:
 297                         zip.write(f, f)
 298                     os.remove(f)