def Listen(self):
if self.__microphoneID is None:
raise Exception("Google: No microphone found - Exit")
with self.__microphone as source:
self.__audio = self.__recognizer.listen(source)
data = ""
try:
data = self.__recognizer.recognize_google(
self.__audio,
key=self.__apiKey,
language=self.__language_4letter_cc,
show_all=False)
except sr.UnknownValueError as e:
FileLogger().Warn(
"Google Line 75: Google Speech Recognition could not understand audio: {0}"
.format(e))
except sr.RequestError as e:
FileLogger().Warn(
"Google Line 77: Could not request results from Google Speech Recognition service: {0}"
.format(e))
except Exception as e:
FileLogger().Warn(
"Google Line 81: Error on executing Google Speech Recognition service: {0}"
.format(e))
return data
def AsyncCallback(self, recognizer, audio):
if self.__microphoneID is None:
raise Exception("Google: No microphone found - Exit")
data = ""
try:
data = self.__recognizer.recognize_google(
audio,
key=self.__apiKey,
language=self.__language_4letter_cc,
show_all=False)
except sr.UnknownValueError as e:
FileLogger().Warn(
"Google Line 83: Google Speech Recognition could not understand audio: {0}"
.format(e))
except sr.RequestError as e:
FileLogger().Warn(
"Google Line 85: Could not request results from Google Speech Recognition service: {0}"
.format(e))
except Exception as e:
FileLogger().Warn(
"Google Line 87: Error on executing Google Speech Recognition service: {0}"
.format(e))
if (len(data) > 0):
self.__asyncResultList.append(data)
def callback(data):
global GLOBAL_FileNamePublisher
dataParts = data.data.split("|")
if dataParts[0] != "TTS":
return
ttsProvider = TTSMemory().GetString("TTSProvider")
usePygame = TTSMemory().GetBoolean("UsePygame")
FileLogger().Info("TTS, callback(), Provider: {0}".format(ttsProvider))
try:
if usePygame and data == "TRIGGER_STOP_AUDIO":
SoundMixer().Stop()
return
if (ttsProvider.lower() == "google"):
data = Google().Speak(dataParts[1])
if (ttsProvider.lower() == "microsoft"):
data = Microsoft().Speak(dataParts[1])
if (ttsProvider.lower() == "ivona"):
data = Ivona().Speak(dataParts[1])
if (ttsProvider.lower() == "watson"):
data = Watson().Speak(dataParts[1])
try:
audio = MP3(data)
delay = Config().GetInt("TextToSpeech", "IntermediateAudioDelay")
TTSMemory().Set("TTS.Until",
(rospy.Time.now().to_sec() +
int(round(audio.info.length)) + delay))
except Exception as e:
FileLogger().Warn(
"TTS, callback() - Error on getting audio duration: {0}".
format(e))
if usePygame:
SoundMixer().Play(data)
else:
audioPlayer = Config().Get("TextToSpeech",
"AudioPlayer") + " '{0}'"
os.system(audioPlayer.format(data))
FileLogger().Info("TTS, callback(), Play Audio: {0}".format(data))
GLOBAL_FileNamePublisher.publish("TTS|{0}".format(data))
user = User().LoadObject()
if (user.GetName() is not None):
user.UpdateSpokenTo()
user.Update()
except Exception as e:
FileLogger().Error(
"TTS, callback(), Error on processing TTS data: {0}".format(e))
def __init__(self):
self.__language_4letter_cc = Config().Get("TextToSpeech",
"CountryCode4Letter")
self.__microphoneID = None
microphoneName = Config().Get("SpeechToText", "Microphone")
for i, microphone_name in enumerate(
sr.Microphone().list_microphone_names()):
if microphone_name == microphoneName:
self.__microphoneID = i
if self.__microphoneID is None:
FileLogger().Error("Wit Line 22: No microphone found - Exit")
raise Exception("Wit: No microphone found - Exit")
self.__recognizer = sr.Recognizer()
self.__microphone = sr.Microphone(device_index=self.__microphoneID)
with self.__microphone as source:
self.__recognizer.dynamic_energy_threshold = True
self.__recognizer.adjust_for_ambient_noise(source)
self.__apiKey = Config().Get("TextToSpeech", "WitAPIKey")
if (len(self.__apiKey) == 0):
self.__apiKey = None
def unknownPersonCallback(self, data):
if (not self.__CheckActive or not self.__TimeTable.IsActive()):
return
dataParts = data.data.split("|")
if (dataParts[0] == "CV" and self.__CVSURVOnly):
return
# No valid user within 5 Minutes
user = User().LoadObject(300)
if (user.GetName() is not None
and user.GetName().lower() != "unknown"):
return
timestamp = BrainMemory().GetInt(
"Brain.Trigger.UnknownPerson.Timestamp", self.__Delay * 3)
if (timestamp is None):
BrainMemory().Set("Brain.Trigger.UnknownPerson.Timestamp",
rospy.Time.now().to_sec())
return
if (rospy.Time.now().to_sec() - timestamp > self.__Delay):
response = ProcessTrigger().ProcessCategory("Intruder")
if (len(response) > 1):
FileLogger().Info(
"ActionTrigger, unknownPersonCallback(): {0}".format(
response))
self.__ResponsePublisher.publish("TTS|{0}".format(response))
self.__IFTTTWebhook.TriggerWebhook(self.__IFTTTIntruder)
self.__TriggerPublisher.publish("TRIGGER|Warning|Intruder")
def appCallback(self, data):
dataParts = data.data.split("|")
if dataParts[0] == "FACEAPP":
response = "TRIGGER_FACEAPP_{0}".format(
dataParts[1]) # == ON / OFF
FileLogger().Info(
"ActionTrigger, appCallback(): {0}".format(response))
self.__SpeechTriggerPublisher.publish("STT|{0}".format(response))
def Predict(self, image, predictionObjectList):
if (self.__FastDetection):
faces = self.DetectFaceFast(image)
else:
faces = self.DetectFaceBest(image)
result = []
if len(faces) > 0:
faceId = 1
for face in faces:
croppedImage = self.__cropImage(image, face)
resizedImage = cv2.resize(
self.__toGrayscale(croppedImage),
(self.__ResizeWidth, self.__ResizeHeight))
predictionResult = []
for predictionObject in predictionObjectList:
prediction = None
if predictionObject.Model != None:
prediction = predictionObject.Model.predict(
resizedImage)
try:
if prediction != None:
predictionResult.append({
'model':
predictionObject.Name,
'value':
predictionObject.Dictionary.keys()[
predictionObject.Dictionary.values().index(
prediction[0])],
'rawvalue':
prediction[0],
'distance':
prediction[1]
})
except Exception as e:
FileLogger().Error(
"ComputerVision: Value Error {0}".format(e))
result.append({
'face': {
'id': faceId,
'data': predictionResult,
'coords': {
'x': str(face[0]),
'y': str(face[1]),
'width': str(face[2]),
'height': str(face[3])
}
}
})
faceId += 1
return result, faces
def Process(self, PipelineArgs):
if not self.__trainData or not PipelineArgs.TrainConversation:
return False
context = ContextParameter().LoadObject()
if len(context.History) == 0:
return False
FileLogger().Info("Trainer, Process(), Train Sentence")
user = User().LoadObject()
try:
DialogTrainer().TrainSentence(context.History[-1]["Response"],
PipelineArgs.Normalized,
PipelineArgs.Language, user.Name)
except Exception as e:
FileLogger().Error(
"Trainer, Process(), Error on training sentence: {0}".format(
e))
return True
def EnsureModelUpdate():
moduleList = Config().GetList("ComputerVision", "Modules")
for moduleName in moduleList:
if (ModelMonitor().CompareHash(
moduleName,
ModelMonitor().GetStoredHash(moduleName))):
continue
FileLogger().Info(
"CV Model Rebuilder: Rebuild {0} Model".format(moduleName))
ModelMonitor().Rebuild(moduleName)
def __loadImages(self, datasetName, imageSize=None):
trainingData = []
trainingLabels = []
trainingLabelsDict = {}
for dirname, dirnames, _ in os.walk(
os.path.join(self.__DatasetBasePath, datasetName)):
for subdirname in dirnames:
if imageSize != None and not subdirname.startswith(imageSize):
continue
if subdirname == self.__DisabledFileFolder:
continue
subjectPath = os.path.join(dirname, subdirname)
for filename in os.listdir(subjectPath):
if (not filename.startswith('.')
and filename != self.__DisabledFileFolder):
try:
image = cv2.imread(
os.path.join(subjectPath, filename),
cv2.IMREAD_GRAYSCALE)
trainingData.append(image)
trimmedSubdirname = subdirname.replace(
imageSize, "")
if (trimmedSubdirname not in trainingLabelsDict):
trainingLabelsDict[trimmedSubdirname] = len(
trainingLabelsDict)
labelID = trainingLabelsDict[trimmedSubdirname]
trainingLabels.append(labelID)
except IOError, (errno, strerror):
FileLogger().Error(
"ComputerVision: IO Exception: {0}{1}".format(
errno, strerror))
except Exception as e:
FileLogger().Error(
"ComputerVision: Exception: {0}".format(e))
def Listen(self):
if self.__microphoneID is None:
raise Exception("Wit: No microphone found - Exit")
with self.__microphone as source:
self.__audio = self.__recognizer.listen(source)
data = ""
try:
data = self.__recognizer.recognize_wit(self.__audio,
key=self.__apiKey,
show_all=False)
except sr.UnknownValueError as e:
FileLogger().Warn(
"Wit.ai Line 47: Could not understand audio: {0}".format(
e))
except sr.RequestError as e:
FileLogger().Warn(
"Wit.ai Line 49: Could not request results from Wit.ai service: {0}"
.format(e))
return data
def RunSTT(printData):
pub = rospy.Publisher('/emerald_ai/io/speech_to_text',
String,
queue_size=10)
rospy.Subscriber("/emerald_ai/io/hardware_trigger", String, callback)
rospy.init_node('STT_node', anonymous=True)
rospy.Rate(10) # 10hz
useTrigger = Config().GetBoolean("Application.SpeechToText", "Trigger")
triggerTime = Config().GetInt("Application.SpeechToText", "TriggerTime")
sttProvider = Config().Get("SpeechToText", "Provider") # Google
if (sttProvider.lower() == "google"):
provider = Google()
if (sttProvider.lower() == "microsoft"):
provider = Microsoft()
if (sttProvider.lower() == "wit"):
provider = Wit()
if (sttProvider.lower() == "watson"):
return
print sttProvider.lower()
while True:
#rate.sleep()
if (useTrigger
and (STTMemory().GetFloat("TriggerTimestamp") + triggerTime) <
time.time()):
time.sleep(1)
continue
data = provider.Listen()
if (len(data) == 0):
if (printData):
print "None"
continue
if (printData):
print "We got:", data
FileLogger().Info("STT, RunSTT(), Input Data: {0}".format(data))
rospy.loginfo("STT|{0}".format(data))
pub.publish("STT|{0}".format(data))
def GetSummary(self, term, fallback=True, trimBrackets=True):
summary = None
try:
try:
#wikipedia.summary(query, sentences=0, chars=0, auto_suggest=True, redirect=True)
summary = wikipedia.summary(term.title(), 0, 0, False, True)
except wikipedia.exceptions.DisambiguationError as e:
FileLogger().Error(
"Wikipedia Line 22: DisambiguationError: {0}".format(e))
if fallback:
topics = wikipedia.search(e.options[0])
for _, topic in enumerate(topics):
summary = wikipedia.summary(topic)
break
if summary is None or len(summary) < 5:
return None
if (trimBrackets):
summary = re.sub(r"[\(\[].*?[\)\]][,.;\s]", "", summary)
return summary
except Exception as e:
FileLogger().Error("Wikipedia Line 36: Exception: {0}".format(e))
return None
def TrainSentence(self, OutputSentence, ResponseSentence, Language,
UserName):
# Train Keywords of both sentences
outputKeywords = self.SaveKeywordsFromSentence(OutputSentence,
Language)
self.SaveKeywordsFromSentence(ResponseSentence, Language)
# save sentence
sentenceID = self.SaveSentence(ResponseSentence, Language, UserName)
# link keywords to sentence
self.LinkKeywordAndSentence(outputKeywords, Language, sentenceID)
FileLogger().Info("DialogTrainer: User sentence trained: {0}".format(
ResponseSentence))
def Listen(self):
if self.__microphoneID is None:
raise Exception("Microsoft: No microphone found - Exit")
with self.__microphone as source:
self.__audio = self.__recognizer.listen(source)
data = ""
try:
data = self.__recognizer.recognize_bing(
self.__audio,
key=self.__apiKey,
language=self.__language_4letter_cc,
show_all=False)
except sr.UnknownValueError as e:
FileLogger().Warn(
"Microsoft Line 119: Microsoft Bing Voice Recognition could not understand audio: {0}"
.format(e))
except sr.RequestError as e:
FileLogger().Warn(
"Microsoft Line 121: Could not request results from Microsoft Bing Voice Recognition service: {0}"
.format(e))
return data
def __init__(self):
self.CHUNK = 1024
self.BUF_MAX_SIZE = self.CHUNK * 10
self.q = Queue(maxsize=int(round(self.BUF_MAX_SIZE / self.CHUNK)))
self.audio_source = AudioSource(self.q, True, True)
self.FORMAT = pyaudio.paInt16
self.CHANNELS = 1
self.RATE = 44100
self.__apikey_stt = Config().Get("SpeechToText", "WatsonSTTAPIKey")
self.__url_stt = Config().Get("SpeechToText", "WatsonSTTUrl")
self.__apikey_tts = Config().Get("TextToSpeech", "WatsonTTSAPIKey")
self.__url_tts = Config().Get("TextToSpeech", "WatsonTTSUrl")
self.__voiceName = Config().Get("TextToSpeech", "WatsonVoiceName")
self.__language_2letter_cc = Config().Get("SpeechToText",
"CountryCode2Letter")
self.__language_4letter_cc = Config().Get("SpeechToText",
"CountryCode4Letter")
self.__audioPlayer = Config().Get("TextToSpeech",
"AudioPlayer") + " '{0}'"
self.text_to_speech = TextToSpeechV1(url=self.__url_tts,
iam_apikey=self.__apikey_tts)
self.text_to_speech.set_default_headers(
{'x-watson-learning-opt-out': "true"})
self.speech_to_text = SpeechToTextV1(url=self.__url_stt,
iam_apikey=self.__apikey_stt)
self.speech_to_text.set_default_headers(
{'x-watson-learning-opt-out': "true"})
self.audio = pyaudio.PyAudio()
# open stream using callback
self.stream = self.audio.open(format=self.FORMAT,
channels=self.CHANNELS,
rate=self.RATE,
input=True,
frames_per_buffer=self.CHUNK,
stream_callback=self.pyaudio_callback,
start=False)
try:
rospy.init_node('STT_watson_node', anonymous=True)
except:
FileLogger().Info('already initialized')
def __init__(self):
self.__language_2letter_cc = Config().Get("TextToSpeech",
"CountryCode2Letter")
self.__language_4letter_cc = Config().Get("TextToSpeech",
"CountryCode4Letter")
self.__audioPlayer = Config().Get("TextToSpeech",
"AudioPlayer") + " '{0}'"
self.__voiceGender = Config().Get("TextToSpeech",
"MicrosoftVoiceGender")
self.__voiceName = Config().Get("TextToSpeech", "MicrosoftVoiceName")
self.__apiKey = Config().Get("TextToSpeech", "MicrosoftAPIKey")
params = ""
headers = {"Ocp-Apim-Subscription-Key": self.__apiKey}
__AccessTokenHost = "api.cognitive.microsoft.com"
path = "/sts/v1.0/issueToken"
conn = httplib.HTTPSConnection(__AccessTokenHost)
conn.request("POST", path, params, headers)
response = conn.getresponse()
data = response.read()
conn.close()
self.__accesstoken = data.decode("UTF-8")
self.__microphoneID = None
microphoneName = Config().Get("SpeechToText", "Microphone")
for i, microphone_name in enumerate(
sr.Microphone().list_microphone_names()):
if microphone_name == microphoneName:
self.__microphoneID = i
if self.__microphoneID is None:
FileLogger().Error(
"Microsoft Line 44: No microphone found - skip listen initialisation"
)
return
self.__recognizer = sr.Recognizer()
self.__microphone = sr.Microphone(device_index=self.__microphoneID)
with self.__microphone as source:
self.__recognizer.dynamic_energy_threshold = True
self.__recognizer.adjust_for_ambient_noise(source)
def knownPersonCallback(self, data):
dataParts = data.data.split("|")
if (dataParts[0] == "PERSON"
and dataParts[1] != self.__UnknownUserTag):
if (self.__LastTriggerPerson == dataParts[1]):
self.__TriggeredCounter += 1
else:
self.__TriggeredCounter = 1
self.__LastTriggerPerson = dataParts[1]
if (self.__TriggeredCounter >= self.__MinTriggered):
User().SetUserByCVTag(dataParts[1])
# Greeting
initGreeting = False
try:
lastSpokenToDate = datetime.strptime(
User().LastSpokenTo, "%Y-%m-%d %H:%M:%S")
initGreeting = (lastSpokenToDate.date() <
datetime.today().date())
except:
initGreeting = True
if (initGreeting):
response = ProcessTrigger().ProcessCategory(
"Greeting", User())
lastAudioTimestamp = BrainMemory().GetString(
"Brain.AudioTimestamp", 20)
lastTriggerTimestamp = BrainMemory().GetString(
"Brain.TriggerTimestamp", 20)
if (lastAudioTimestamp is None
and lastTriggerTimestamp is None
and len(response) > 1):
FileLogger().Info(
"ActionTrigger, knownPersonCallback(): {0}".format(
response))
self.__ResponsePublisher.publish(
"TTS|{0}".format(response))
self.__IFTTTWebhook.TriggerWebhook(
self.__IFTTTGreeting,
User().FullName, response)
self.__TriggerPublisher.publish(
"TRIGGER|Info|Greeting")
BrainMemory().Set("Brain.TriggerTimestamp",
rospy.Time.now().to_sec())
def CallFunction(moduleName,
className,
functionName,
arg1=None,
arg2=None,
arg3=None):
FileLogger().Info("Action called: {0}, {1}, {2}".format(
moduleName, className, functionName))
instance = CreateClass(moduleName, className)
method = getattr(instance, functionName)
if arg3 != None:
return method(arg1, arg2, arg3)
if arg2 != None:
return method(arg1, arg2)
if arg1 != None:
return method(arg1)
return method()
def TrainModel(self, datasetName, imageSize=None):
if imageSize is None:
imageSize = "{0}x{1}".format(self.__ResizeWidth,
self.__ResizeHeight)
images, labels, labelDict = self.__loadImages(datasetName, imageSize)
if len(images) == 0 or len(labels) == 0:
FileLogger().Error("ComputerVision: No Data given")
return
self.__RecognizerModel.train(images, labels)
self.__RecognizerDictionary = labelDict
path = os.path.join(self.__DatasetBasePath, datasetName)
try:
self.__RecognizerModel.save(
os.path.join(path, self.__ModelFile.format(imageSize)))
except:
self.__RecognizerModel.write(
os.path.join(path, self.__ModelFile.format(imageSize)))
np.save(os.path.join(path, self.__DictionaryFile.format(imageSize)),
labelDict)
def ProcessSpeech(self, sentence):
if(not BrainMemory().GetBoolean("Listen") or self.__TTSActive()):
return
cancelSpeech = False
stopwordList = Config().GetList("Bot", "StoppwordList")
if(sentence in stopwordList):
cancelSpeech = True
self.__ResponsePublisher.publish("TTS|TRIGGER_STOP_AUDIO")
if self.Pipeline is None:
self.Pipeline = PipelineArgs()
BrainMemory().Set("Brain.AudioTimestamp", rospy.Time.now().to_sec())
self.Pipeline.AddSentence(sentence)
self.Pipeline = AnalyzeScope().Process(self.Pipeline)
self.Pipeline = ProcessResponse().Process(self.Pipeline)
if(not cancelSpeech and not BrainMemory().GetBoolean("Mute")):
self.ProcessAnimation(self.Pipeline.Animation)
if(self.Pipeline.ResponseFound):
FileLogger().Info("Brain STT, ProcessSpeech(): {0}".format(self.Pipeline.Response))
self.__ResponsePublisher.publish("TTS|{0}".format(self.Pipeline.Response))
trainerResult = Trainer().Process(self.Pipeline)
contextParameter = ContextParameter().LoadObject(240)
contextParameter.AppendHistory(self.Pipeline)
contextParameter.SaveObject()
#print "Pipeline Args", self.Pipeline.toJSON()
#print "Context Parameter", contextParameter.toJSON()
#print "Trainer Result: ", trainerResult
print "Input: ", sentence
print "Response: ", self.Pipeline.Response
self.Pipeline = None
def __init__(self):
self.__language_2letter_cc = Config().Get("TextToSpeech",
"CountryCode2Letter")
self.__language_4letter_cc = Config().Get("TextToSpeech",
"CountryCode4Letter")
self.__audioPlayer = Config().Get("TextToSpeech",
"AudioPlayer") + " '{0}'"
self.__asyncInit = False
self.__apiKey = Config().Get("TextToSpeech", "GoogleAPIKey")
if (len(self.__apiKey) == 0):
self.__apiKey = None
self.__microphoneID = None
microphoneName = Config().Get("SpeechToText", "Microphone")
for i, microphone_name in enumerate(
sr.Microphone().list_microphone_names()):
if microphone_name == microphoneName:
self.__microphoneID = i
if self.__microphoneID is None:
FileLogger().Error(
"Google Line 38: No microphone found - skip listen initialisation"
)
return
self.__recognizer = sr.Recognizer()
#Represents the minimum length of silence (in seconds) that will register as the
#end of a phrase. Can be changed.
#Smaller values result in the recognition completing more quickly, but might result
#in slower speakers being cut off.
self.__recognizer.pause_threshold = 0.5
self.__recognizer.operation_timeout = 3
self.__microphone = sr.Microphone(device_index=self.__microphoneID)
with self.__microphone as source:
self.__recognizer.dynamic_energy_threshold = True
self.__recognizer.adjust_for_ambient_noise(source)
def LoadModel(self, datasetName, imageSize=None):
if imageSize is None:
imageSize = "{0}x{1}".format(self.__ResizeWidth,
self.__ResizeHeight)
path = os.path.join(self.__DatasetBasePath, datasetName)
try:
try:
self.__RecognizerModel.load(
os.path.join(path, self.__ModelFile.format(imageSize)))
except:
self.__RecognizerModel.read(
os.path.join(path, self.__ModelFile.format(imageSize)))
self.__RecognizerDictionary = np.load(
os.path.join(path,
self.__DictionaryFile.format(imageSize))).item()
return self.__RecognizerModel, self.__RecognizerDictionary
except Exception as e:
FileLogger().Error(
"ComputerVision: Exception: Error while opening File {0}".
format(e))
return None, None
def CalculateRequirement(self, sentenceList, parameterList, delete=True):
query = """SELECT Conversation_Sentence_Requirement.Comparison,
Conversation_Sentence_Requirement.Value, Conversation_Requirement.Name
FROM Conversation_Sentence_Requirement, Conversation_Requirement
WHERE Conversation_Sentence_Requirement.RequirementID = Conversation_Requirement.ID
AND Conversation_Sentence_Requirement.SentenceID='{0}'
GROUP BY Conversation_Sentence_Requirement.SentenceID,
Conversation_Sentence_Requirement.RequirementID,
Conversation_Sentence_Requirement.Comparison
"""
deleteList = []
for sentenceID in sentenceList.iterkeys():
sqlResult = db().Fetchall(query.format(sentenceID))
for r in sqlResult:
requirementName = r[2].title()
if requirementName not in parameterList and r[1].lower(
) != self.__NoneTag.lower():
FileLogger().Error(
"SentenceResolver Line 171: Requirement missing in parameter list: {0}"
.format(requirementName))
deleteList.append(sentenceID)
continue
if r[0] is None:
if type(parameterList[requirementName]) == list and r[
1].lower() not in parameterList[requirementName]:
deleteList.append(sentenceID)
elif type(
parameterList[requirementName]
) == str and parameterList[requirementName].lower(
) != r[1].lower():
deleteList.append(sentenceID)
else:
sentenceList[sentenceID].AddPriority(
self.__RequirementBonus)
continue
else:
if r[1].lower() == self.__NoneTag.lower():
if (r[0] == "eq" and
(requirementName in parameterList
and parameterList[requirementName] is not None)):
deleteList.append(sentenceID)
continue
if (r[0] == "ne"
and not (requirementName in parameterList
and parameterList[requirementName]
is not None)):
deleteList.append(sentenceID)
continue
sentenceList[sentenceID].AddPriority(
self.__RequirementBonus)
continue
if r[0] == "lt" and not parameterList[requirementName] < r[
1]:
deleteList.append(sentenceID)
continue
if r[0] == "le" and not parameterList[
requirementName] <= r[1]:
deleteList.append(sentenceID)
continue
if r[0] == "eq" and not parameterList[
requirementName] == r[1]:
deleteList.append(sentenceID)
continue
if r[0] == "ne" and not parameterList[
requirementName] != r[1]:
deleteList.append(sentenceID)
continue
if r[0] == "ge" and not parameterList[
requirementName] >= r[1]:
deleteList.append(sentenceID)
continue
if r[0] == "gt" and not parameterList[requirementName] > r[
1]:
deleteList.append(sentenceID)
continue
sentenceList[sentenceID].AddPriority(
self.__RequirementBonus)
if delete:
for d in list(set(deleteList)):
del sentenceList[d]
return {'sentenceList': sentenceList, 'deleteList': deleteList}
def on_close(self):
FileLogger().Warn("Connection closed")
def on_listening(self):
FileLogger().Info('Service is listening')
def on_inactivity_timeout(self, error):
FileLogger().Error('Inactivity timeout: {}'.format(error))
def on_error(self, error):
FileLogger().Error('Error received: {}'.format(error))
def on_connected(self):
FileLogger().Info('Connection was successful')
def Process(self, PipelineArgs):
sentence = PipelineArgs.GetRandomSentenceWithHighestValue()
FileLogger().Info("ProcessResponse, Process(), Sentence: {0}".format(sentence))
responseFound = True
if sentence is None or sentence.Rating < self.__sentenceRatingThreshold:
responseFound = False
if responseFound:
user = User().LoadObject()
PipelineArgs.ResponseRaw = sentence.GetSentenceString(user.Formal)
PipelineArgs.Response = PipelineArgs.ResponseRaw
PipelineArgs.ResponseID = sentence.ID
PipelineArgs.Animation = sentence.GetAnimation()
PipelineArgs.ResponseFound = True
PipelineArgs.BasewordTrimmedInput = NLP.TrimBasewords(PipelineArgs)
PipelineArgs.FullyTrimmedInput = NLP.TrimStopwords(PipelineArgs.BasewordTrimmedInput, PipelineArgs.Language)
contextParameter = ContextParameter().LoadObject(240)
if sentence.HasInteraction():
contextParameter.InteractionName = sentence.InteractionName
sentenceAction = sentence.GetAction()
if sentenceAction != None and len(sentenceAction["Module"]) > 0:
FileLogger().Info("ProcessResponse, Process(), Call Action: {0}, {1}, {2}".format(sentenceAction["Module"], sentenceAction["Class"], sentenceAction["Function"]))
actionResult = Action.CallFunction(sentenceAction["Module"], sentenceAction["Class"], sentenceAction["Function"], PipelineArgs)
if actionResult["ResultType"].title() is "Error":
PipelineArgs.Response = sentence.GetActionErrorResponse(PipelineArgs.Language, user.Formal)
PipelineArgs.ResponseRaw = None
PipelineArgs.Error.append("ProcessResponse - Action Error")
else:
contextParameter.SetInput(actionResult["Input"])
contextParameter.SetResult(actionResult["Result"])
contextParameter.SaveObject()
contextParameter.ResetInteraction()
contextParameterDict = contextParameter.GetParameterDictionary()
keywords = re.findall(r"\{(.*?)\}", PipelineArgs.Response)
for keyword in keywords:
if keyword.title() in contextParameterDict:
replaceword = contextParameterDict[keyword.title()]
if replaceword is None or replaceword == "Unknown":
replaceword = ""
else:
replaceword = "'{0}'".format(replaceword)
PipelineArgs.Response = PipelineArgs.Response.replace("{{{0}}}".format(keyword.lower()), str(replaceword))
else:
PipelineArgs.Response = PipelineArgs.Response.replace("{{{0}}}".format(keyword.lower()), "")
FileLogger().Error("ProcessResponse Line 63: Parameter missing: '{0}'".format(keyword))
contextParameter.UnsetInputAndResult()
contextParameter.SaveObject()
elif not responseFound and self.__aliceAsFallback:
PipelineArgs.Response = self.__alice.GetResponse(PipelineArgs.Input)
PipelineArgs.ResponseFound = True
PipelineArgs.TrainConversation = False
FileLogger().Info("ProcessResponse, Process(), Response: {0}".format(PipelineArgs.Response))
return PipelineArgs