epicyon/speaker_client.py

__filename__ = "speaker_client.py"
__author__ = "Bob Mottram"
__license__ = "AGPL3+"
__version__ = "1.2.0"
__maintainer__ = "Bob Mottram"
__email__ = "bob@freedombone.net"
__status__ = "Production"

import os
import html
import time
from session import createSession
from speaker import getSpeakerFromServer
from speaker import getSpeakerPitch
from speaker import getSpeakerRate
from speaker import getSpeakerRange
import signal


def _waitForKeypress(timeout: int, debug: bool) -> str:
    """Waits for a keypress with a timeout
    Returns the key pressed, or None on timeout
    """
    def _nothing(sig, frame): pass
    signal.signal(signal.SIGALRM, _nothing)
    signal.alarm(timeout)
    keyPress = None
    try:
        keyPress = input()
        signal.alarm(0)
    except (IOError, EOFError):
        if debug:
            print('Keypress Timeout')
        return None
    return keyPress


def _speakerEspeak(espeak, pitch: int, rate: int, srange: int,
                   sayText: str) -> None:
    """Speaks the given text with espeak
    """
    espeak.set_parameter(espeak.Parameter.Pitch, pitch)
    espeak.set_parameter(espeak.Parameter.Rate, rate)
    espeak.set_parameter(espeak.Parameter.Range, srange)
    espeak.synth(html.unescape(sayText))


def _speakerPicospeaker(pitch: int, rate: int, systemLanguage: str,
                        sayText: str) -> None:
    speakerLang = 'en-GB'
    if systemLanguage:
        if systemLanguage.startswith('fr'):
            speakerLang = 'fr-FR'
        elif systemLanguage.startswith('es'):
            speakerLang = 'es-ES'
        elif systemLanguage.startswith('de'):
            speakerLang = 'de-DE'
        elif systemLanguage.startswith('it'):
            speakerLang = 'it-IT'
    speakerCmd = 'picospeaker ' + \
        '-l ' + speakerLang + \
        ' -r ' + str(rate) + \
        ' -p ' + str(pitch) + ' "' + \
        html.unescape(sayText) + '"'
    # print(speakerCmd)
    os.system(speakerCmd)


def runSpeakerClient(baseDir: str, proxyType: str, httpPrefix: str,
                     nickname: str, domain: str, port: int, password: str,
                     screenreader: str, systemLanguage: str,
                     debug: bool) -> None:
    """Runs the screen reader client, which announces new inbox items via TTS
    """
    if screenreader == 'espeak':
        print('Setting up espeak')
        from espeak import espeak
    elif screenreader != 'picospeaker':
        print(screenreader + ' is not a supported TTS system')
        return

    print('Running ' + screenreader + ' for ' + nickname + '@' + domain)

    prevSay = ''
    while (1):
        session = createSession(proxyType)
        speakerJson = \
            getSpeakerFromServer(baseDir, session, nickname, password,
                                 domain, port, httpPrefix, True, __version__)
        if speakerJson:
            if speakerJson['say'] != prevSay:
                if speakerJson.get('name'):
                    nameStr = speakerJson['name']
                    gender = 'They/Them'
                    if speakerJson.get('gender'):
                        gender = speakerJson['gender']

                    # get the speech parameters
                    pitch = getSpeakerPitch(nameStr, screenreader, gender)
                    rate = getSpeakerRate(nameStr, screenreader)
                    srange = getSpeakerRange(nameStr)

                    # say the speaker's name
                    if screenreader == 'espeak':
                        _speakerEspeak(espeak, pitch, rate, srange, nameStr)
                    elif screenreader == 'picospeaker':
                        _speakerPicospeaker(pitch, rate,
                                            systemLanguage, nameStr)
                    time.sleep(2)

                    # append image description if needed
                    if not speakerJson.get('imageDescription'):
                        sayStr = speakerJson['say']
                        # echo spoken text to the screen
                        print(html.unescape(nameStr) + ': ' +
                              html.unescape(speakerJson['say']) + '\n')
                    else:
                        sayStr = speakerJson['say'] + '. ' + \
                            speakerJson['imageDescription']
                        # echo spoken text to the screen
                        print(html.unescape(nameStr) + ': ' +
                              html.unescape(speakerJson['say']) + '\n' +
                              html.unescape(speakerJson['imageDescription']))

                    # speak the post content
                    if screenreader == 'espeak':
                        _speakerEspeak(espeak, pitch, rate, srange, sayStr)
                    elif screenreader == 'picospeaker':
                        _speakerPicospeaker(pitch, rate,
                                            systemLanguage, sayStr)

                prevSay = speakerJson['say']

        # wait for a while, or until a key is pressed
        if _waitForKeypress(30, debug):
            break