Revision f27f1860 tts_bridge/mary/mary_tts_bridge/MaryTTSClient.py
| tts_bridge/mary/mary_tts_bridge/MaryTTSClient.py | ||
|---|---|---|
| 26 | 26 |
""" |
| 27 | 27 |
|
| 28 | 28 |
import logging |
| 29 |
#try: |
|
| 30 |
# import rsb |
|
| 31 |
#except ImportError: |
|
| 32 |
# RSB_SUPPORT = False |
|
| 33 |
#else: |
|
| 34 |
# from MiddlewareRSB import * |
|
| 35 |
# RSB_SUPPORT = True |
|
| 36 |
|
|
| 37 |
#from MiddlewareROS import * |
|
| 38 | 29 |
import sys |
| 39 | 30 |
try: |
| 40 |
from http.client import HTTPConnection
|
|
| 41 |
from urllib.parse import urlencode
|
|
| 31 |
from http.client import HTTPConnection
|
|
| 32 |
from urllib.parse import urlencode
|
|
| 42 | 33 |
except ImportError: # Python 2 |
| 43 |
from httplib import HTTPConnection
|
|
| 44 |
from urllib import urlencode
|
|
| 34 |
from httplib import HTTPConnection
|
|
| 35 |
from urllib import urlencode
|
|
| 45 | 36 |
import wave |
| 46 | 37 |
import ctypes |
| 47 | 38 |
import wave |
| 48 | 39 |
import sys |
| 49 | 40 |
|
| 41 |
|
|
| 50 | 42 |
class MaryTTSClient: |
| 51 | 43 |
def __init__(self, voice="cmu-slt-hsmm", locale="en_US", tts_host="127.0.0.1", tts_port=59125, loglevel=logging.WARNING): |
| 52 |
"""initialise
|
|
| 53 |
:param loglevel: optional log level
|
|
| 54 |
"""
|
|
| 55 |
self.loglevel = loglevel
|
|
| 56 |
self.logger = logging.getLogger(__name__)
|
|
| 57 |
# create nice and actually usable formatter and add it to the handler
|
|
| 58 |
self.config_logger(loglevel)
|
|
| 44 |
"""initialise
|
|
| 45 |
:param loglevel: optional log level
|
|
| 46 |
"""
|
|
| 47 |
self.loglevel = loglevel
|
|
| 48 |
self.logger = logging.getLogger(__name__)
|
|
| 49 |
# create nice and actually usable formatter and add it to the handler
|
|
| 50 |
self.config_logger(loglevel)
|
|
| 59 | 51 |
|
| 60 |
self.logger.info("starting MaryTTSClient (voice="+voice+", locale="+locale+", host="+tts_host+", port="+str(tts_port))
|
|
| 52 |
self.logger.info("starting MaryTTSClient (voice="+voice+", locale="+locale+", host="+tts_host+", port="+str(tts_port))
|
|
| 61 | 53 |
|
| 62 |
self.tts_host = tts_host
|
|
| 63 |
self.tts_port = tts_port
|
|
| 64 |
self.locale = locale
|
|
| 65 |
self.voice = voice
|
|
| 54 |
self.tts_host = tts_host
|
|
| 55 |
self.tts_port = tts_port
|
|
| 56 |
self.locale = locale
|
|
| 57 |
self.voice = voice
|
|
| 66 | 58 |
|
| 67 | 59 |
def __del__(self): |
| 68 |
"""destructor
|
|
| 69 |
"""
|
|
| 70 |
self.logger.debug("destructor of MaryTTSClient called")
|
|
| 60 |
"""destructor
|
|
| 61 |
"""
|
|
| 62 |
self.logger.debug("destructor of MaryTTSClient called")
|
|
| 71 | 63 |
|
| 72 | 64 |
def config_logger(self, level): |
| 73 |
"""initialise a nice logger formatting
|
|
| 74 |
:param level: log level
|
|
| 75 |
"""
|
|
| 76 |
formatter = logging.Formatter('%(asctime)s %(name)-30s %(levelname)-8s > %(message)s')
|
|
| 77 |
ch = logging.StreamHandler()
|
|
| 78 |
#ch.setLevel(level)
|
|
| 79 |
ch.setFormatter(formatter)
|
|
| 80 |
self.logger.setLevel(level)
|
|
| 81 |
self.logger.addHandler(ch)
|
|
| 65 |
"""initialise a nice logger formatting
|
|
| 66 |
:param level: log level
|
|
| 67 |
"""
|
|
| 68 |
formatter = logging.Formatter('%(asctime)s %(name)-30s %(levelname)-8s > %(message)s')
|
|
| 69 |
ch = logging.StreamHandler()
|
|
| 70 |
# ch.setLevel(level)
|
|
| 71 |
ch.setFormatter(formatter)
|
|
| 72 |
self.logger.setLevel(level)
|
|
| 73 |
self.logger.addHandler(ch)
|
|
| 82 | 74 |
|
| 83 | 75 |
def generate_audio(self, message): |
| 84 |
"""generate audio from text
|
|
| 85 |
:param message: text to synthesize
|
|
| 86 |
"""
|
|
| 87 |
return self.generate(message, "AUDIO")
|
|
| 76 |
"""generate audio from text
|
|
| 77 |
:param message: text to synthesize
|
|
| 78 |
"""
|
|
| 79 |
return self.generate(message, "AUDIO")
|
|
| 88 | 80 |
|
| 89 | 81 |
def generate_phonemes(self, message): |
| 90 |
"""generate phoneme list from text
|
|
| 91 |
:param message: text to synthesize
|
|
| 92 |
"""
|
|
| 93 |
return self.generate(message, "REALISED_DURATIONS")
|
|
| 82 |
"""generate phoneme list from text
|
|
| 83 |
:param message: text to synthesize
|
|
| 84 |
"""
|
|
| 85 |
return self.generate(message, "REALISED_DURATIONS")
|
|
| 94 | 86 |
|
| 95 | 87 |
def generate(self, message, output_type): |
| 96 |
"""generate requested data object from text
|
|
| 97 |
:param message: text to synthesize
|
|
| 98 |
"""
|
|
| 99 |
|
|
| 100 |
raw_params = {
|
|
| 101 |
"INPUT_TEXT": message,
|
|
| 102 |
"INPUT_TYPE": "RAWMARYXML",
|
|
| 103 |
"OUTPUT_TYPE": output_type,
|
|
| 104 |
"LOCALE": self.locale,
|
|
| 105 |
"AUDIO": "WAVE_FILE",
|
|
| 106 |
"VOICE": self.voice,
|
|
| 107 |
}
|
|
| 108 |
|
|
| 109 |
params = urlencode(raw_params)
|
|
| 110 |
headers = {}
|
|
| 111 |
|
|
| 112 |
#conn.set_debuglevel(5)
|
|
| 113 |
#open connection to mary server
|
|
| 88 |
"""generate requested data object from text
|
|
| 89 |
:param message: text to synthesize
|
|
| 90 |
"""
|
|
| 91 |
|
|
| 92 |
raw_params = {
|
|
| 93 |
"INPUT_TEXT": message,
|
|
| 94 |
"INPUT_TYPE": "RAWMARYXML",
|
|
| 95 |
"OUTPUT_TYPE": output_type,
|
|
| 96 |
"LOCALE": self.locale,
|
|
| 97 |
"AUDIO": "WAVE_FILE",
|
|
| 98 |
"VOICE": self.voice,
|
|
| 99 |
}
|
|
| 100 |
|
|
| 101 |
params = urlencode(raw_params)
|
|
| 102 |
headers = {}
|
|
| 103 |
|
|
| 104 |
# conn.set_debuglevel(5)
|
|
| 105 |
# open connection to mary server
|
|
| 114 | 106 |
conn = HTTPConnection(self.tts_host, self.tts_port) |
| 115 | 107 |
|
| 116 |
conn.request("POST", "/process", params, headers)
|
|
| 117 |
response = conn.getresponse()
|
|
| 108 |
conn.request("POST", "/process", params, headers)
|
|
| 109 |
response = conn.getresponse()
|
|
| 118 | 110 |
|
| 119 |
if response.status != 200:
|
|
| 120 |
print(response.getheaders())
|
|
| 111 |
if response.status != 200:
|
|
| 112 |
print(response.getheaders())
|
|
| 121 | 113 |
conn.close() |
| 122 |
raise RuntimeError("{0}: {1}".format(response.status,response.reason))
|
|
| 123 |
return response.read() |
|
| 114 |
raise RuntimeError("{0}: {1}".format(response.status,response.reason))
|
|
| 115 |
return response.read() |
|
| 116 |
|
|
| 124 | 117 |
|
| 125 |
#test code |
|
| 118 |
# test code
|
|
| 126 | 119 |
if __name__ == "__main__": |
| 127 | 120 |
client = MaryTTSClient() |
| 128 | 121 |
audio = client.generate_phonemes("test 1 2 3 4 5 6 7 8 9 10")
|
Also available in: Unified diff