smartpi 0.1.40__py3-none-any.whl → 0.1.42__py3-none-any.whl

smartpi/tencentcloud-speech-sdk-python/asr/speech_recognizer.py

@@ -0,0 +1,311 @@
+# -*- coding: utf-8 -*-
+import sys
+import hmac
+import hashlib
+import base64
+import time
+import json
+import threading
+import websocket
+import uuid
+import urllib
+from common.log import logger
+
+
+def is_python3():
+    if sys.version > '3':
+        return True
+    return False
+
+
+#实时识别语音使用
+class SpeechRecognitionListener():
+    '''
+    reponse:  
+    on_recognition_start的返回只有voice_id字段。
+    on_fail 只有voice_id、code、message字段。
+    on_recognition_complete没有result字段。
+    其余消息包含所有字段。
+    字段名	类型	
+    code	Integer	
+    message	String	
+    voice_id	String
+    message_id	String
+    result	Result	
+    final	Integer	
+
+    Result的结构体格式为:
+    slice_type	Integer	
+    index	Integer	
+    start_time	Integer	
+    end_time	Integer	
+    voice_text_str	String	
+    word_size	Integer	
+    word_list	Word Array
+
+    Word的类型为:
+    word    String 
+    start_time Integer 
+    end_time Integer 
+    stable_flag：Integer 
+    '''
+
+    def on_recognition_start(self, response):
+        pass
+
+    def on_sentence_begin(self, response):
+        pass
+
+    def on_recognition_result_change(self, response):
+        pass
+
+    def on_sentence_end(self, response):
+        pass
+
+    def on_recognition_complete(self, response):
+        pass
+
+    def on_fail(self, response):
+        pass
+
+
+NOTOPEN = 0
+STARTED = 1
+OPENED = 2
+FINAL = 3
+ERROR = 4
+CLOSED = 5
+
+#实时识别语音使用
+class SpeechRecognizer:
+
+    def __init__(self, appid, credential, engine_model_type, listener):
+        self.result = ""
+        self.credential = credential
+        self.appid = appid
+        self.engine_model_type = engine_model_type
+        self.status = NOTOPEN
+        self.ws = None
+        self.wst = None
+        self.voice_id = ""
+        self.new_start = 0
+        self.listener = listener
+        self.filter_dirty = 0
+        self.filter_modal = 0
+        self.filter_punc = 0
+        self.convert_num_mode = 0
+        self.word_info = 0
+        self.need_vad = 0
+        self.vad_silence_time = 0
+        self.hotword_id = ""
+        self.hotword_list = ""
+        self.reinforce_hotword = 0
+        self.noise_threshold = 0
+        self.voice_format = 4
+        self.nonce = ""
+        self.replace_text_id = ""
+
+    def set_filter_dirty(self, filter_dirty):
+        self.filter_dirty = filter_dirty
+
+    def set_filter_modal(self, filter_modal):
+        self.filter_modal = filter_modal
+
+    def set_filter_punc(self, filter_punc):
+        self.filter_punc = filter_punc
+
+    def set_convert_num_mode(self, convert_num_mode):
+        self.convert_num_mode = convert_num_mode
+
+    def set_word_info(self, word_info):
+        self.word_info = word_info
+
+    def set_need_vad(self, need_vad):
+        self.need_vad = need_vad
+
+    def set_vad_silence_time(self, vad_silence_time):
+        self.vad_silence_time = vad_silence_time
+
+    def set_hotword_id(self, hotword_id):
+        self.hotword_id = hotword_id
+
+    def set_hotword_list(self, hotword_list):
+        self.hotword_list = hotword_list
+
+    def set_voice_format(self, voice_format):
+        self.voice_format = voice_format
+
+    def set_nonce(self, nonce):
+        self.nonce = nonce
+
+    def set_reinforce_hotword(self, reinforce_hotword):
+        self.reinforce_hotword = reinforce_hotword
+
+    def set_noise_threshold(self, noise_threshold):
+        self.noise_threshold = noise_threshold
+
+    def set_replace_text_id(self, replace_text_id):
+        self.replace_text_id = replace_text_id
+
+    def format_sign_string(self, param):
+        signstr = "asr.cloud.tencent.com/asr/v2/"
+        for t in param:
+            if 'appid' in t:
+                signstr += str(t[1])
+                break
+        signstr += "?"
+        for x in param:
+            tmp = x
+            if 'appid' in x:
+                continue
+            for t in tmp:
+                signstr += str(t)
+                signstr += "="
+            signstr = signstr[:-1]
+            signstr += "&"
+        signstr = signstr[:-1]
+        return signstr
+
+    def create_query_string(self, param):
+        signstr = "wss://asr.cloud.tencent.com/asr/v2/"
+        for t in param:
+            if 'appid' in t:
+                signstr += str(t[1])
+                break
+        signstr += "?"
+        for x in param:
+            tmp = x
+            if 'appid' in x:
+                continue
+            for t in tmp:
+                signstr += str(t)
+                signstr += "="
+            signstr = signstr[:-1]
+            signstr += "&"
+        signstr = signstr[:-1]
+        return signstr
+
+    def sign(self, signstr, secret_key):
+        hmacstr = hmac.new(secret_key.encode('utf-8'),
+                           signstr.encode('utf-8'), hashlib.sha1).digest()
+        s = base64.b64encode(hmacstr)
+        s = s.decode('utf-8')
+        return s
+
+    def create_query_arr(self):
+        query_arr = dict()
+
+        query_arr['appid'] = self.appid
+        query_arr['sub_service_type'] = 1
+        query_arr['engine_model_type'] = self.engine_model_type
+        query_arr['filter_dirty'] = self.filter_dirty
+        query_arr['filter_modal'] = self.filter_modal
+        query_arr['filter_punc'] = self.filter_punc
+        query_arr['needvad'] = self.need_vad
+        query_arr['convert_num_mode'] = self.convert_num_mode
+        query_arr['word_info'] = self.word_info
+        if self.vad_silence_time != 0:
+            query_arr['vad_silence_time'] = self.vad_silence_time
+        if self.hotword_id != "":
+            query_arr['hotword_id'] = self.hotword_id
+        if self.hotword_list != "":
+            query_arr['hotword_list'] = self.hotword_list
+        if self.replace_text_id != "":
+            query_arr['replace_text_id'] = self.replace_text_id
+        query_arr['secretid'] = self.credential.secret_id
+        query_arr['voice_format'] = self.voice_format
+        query_arr['voice_id'] = self.voice_id
+        query_arr['timestamp'] = str(int(time.time()))
+        if self.nonce != "":
+            query_arr['nonce'] = self.nonce
+        else:
+            query_arr['nonce'] = query_arr['timestamp']
+        query_arr['expired'] = int(time.time()) + 24 * 60 * 60
+        query_arr['reinforce_hotword'] = self.reinforce_hotword
+        query_arr['noise_threshold'] = self.noise_threshold
+        return query_arr
+
+    def stop(self):
+        if self.status == OPENED: 
+            msg = {}
+            msg['type'] = "end"
+            text_str = json.dumps(msg)
+            self.ws.sock.send(text_str)
+        if self.ws:
+            if self.wst and self.wst.is_alive():
+                self.wst.join()
+        self.ws.close()
+        
+
+    def write(self, data):
+        while self.status == STARTED:
+            time.sleep(0.1)
+        if self.status == OPENED: 
+            self.ws.sock.send_binary(data)
+
+    def start(self):
+        def on_message(ws, message):
+            response = json.loads(message)
+            response['voice_id'] = self.voice_id
+            if response['code'] != 0:
+                logger.error("%s server recognition fail %s" %
+                             (response['voice_id'], response['message']))
+                self.listener.on_fail(response)
+                return
+            if "final" in response and response["final"] == 1:
+                self.status = FINAL
+                self.result = message
+                self.listener.on_recognition_complete(response)
+                logger.info("%s recognition complete" % response['voice_id'])
+                return
+            if "result" in response.keys():
+                if response["result"]['slice_type'] == 0:
+                    self.listener.on_sentence_begin(response)
+                    return
+                elif response["result"]["slice_type"] == 2:
+                    self.listener.on_sentence_end(response)
+                    return
+                elif response["result"]["slice_type"] == 1:
+                    self.listener.on_recognition_result_change(response)
+                    return
+
+        def on_error(ws, error):
+            if self.status == FINAL :
+                return
+            logger.error("websocket error %s  voice id %s" %
+                         (format(error), self.voice_id))
+            self.status = ERROR
+
+        def on_close(ws):
+            self.status = CLOSED
+            logger.info("websocket closed  voice id %s" %
+                          self.voice_id)
+
+        def on_open(ws):
+            self.status = OPENED
+
+        query_arr = self.create_query_arr()
+        if self.voice_id == "":
+            query_arr['voice_id'] = str(uuid.uuid1())
+            self.voice_id = query_arr['voice_id']
+        query = sorted(query_arr.items(), key=lambda d: d[0])
+        signstr = self.format_sign_string(query)
+
+        autho = self.sign(signstr, self.credential.secret_key)
+        requrl = self.create_query_string(query)
+        if is_python3():
+            autho = urllib.parse.quote(autho)
+        else:
+            autho = urllib.quote(autho)
+        requrl += "&signature=%s" % autho
+        self.ws = websocket.WebSocketApp(requrl,  None,
+                on_error=on_error, on_close=on_close, on_message=on_message)
+        self.ws.on_open = on_open
+        self.wst = threading.Thread(target=self.ws.run_forever)
+        self.wst.daemon = True
+        self.wst.start()
+        self.status = STARTED
+        response = {}
+        response['voice_id'] = self.voice_id
+        self.listener.on_recognition_start(response)
+        logger.info("%s recognition start" % response['voice_id'])

smartpi/tencentcloud-speech-sdk-python/common/__init__.py

@@ -0,0 +1 @@
+#!#-*-coding:utf-8 -*-

smartpi/tencentcloud-speech-sdk-python/common/credential.py

@@ -0,0 +1,6 @@
+# -*- coding: utf-8 -*-
+class Credential:
+    def __init__(self, secret_id, secret_key, token=""):
+        self.secret_id = secret_id
+        self.secret_key = secret_key
+        self.token = token

smartpi/tencentcloud-speech-sdk-python/common/log.py

@@ -0,0 +1,16 @@
+import logging
+import logging.handlers
+
+
+
+FORMAT = '%(asctime)15s %(name)s-%(levelname)s  %(funcName)s:%(lineno)s %(message)s'
+logging.basicConfig(level=logging.DEBUG, format=FORMAT)
+logger = logging.getLogger('tencent_speech.log')
+
+handler = logging.handlers.RotatingFileHandler('tencent_speech.log', maxBytes=1024 * 1024,
+                                               backupCount=5, encoding='utf-8')
+handler.setLevel(logging.DEBUG)
+handler.setFormatter(logging.Formatter(FORMAT))
+logger.addHandler(handler)
+logger.setLevel('INFO')
+

smartpi/tencentcloud-speech-sdk-python/common/utils.py

@@ -0,0 +1,7 @@
+# -*- coding: utf-8 -*-
+import sys
+
+def is_python3():
+    if sys.version > '3':
+        return True
+    return False

smartpi/tencentcloud-speech-sdk-python/examples/tts/tts_text.txt

@@ -0,0 +1,60 @@
+床前明月光，疑是地上霜。
+举头望明月，低头思故乡。
+红豆生南国，春来发几枝。
+君自故乡来，应知故乡事。
+春眠不觉晓，处处闻啼鸟。
+欲穷千里目，更上一层楼。
+黄河远上白云间，一片孤城万仞山。
+两岸猿声啼不住，轻舟已过万重山。
+独在异乡为异客，每逢佳节倍思亲。
+春城无处不飞花，寒食东风御柳斜。
+月落乌啼霜满天，江枫渔火对愁眠。
+春潮带雨晚来急，野渡无人舟自横。
+洛阳亲友如相问，一片冰心在玉壶。
+空山新雨后，天气晚来秋。
+千里莺啼绿映红，水村山郭酒旗风。
+朝辞白帝彩云间，千里江陵一日还。
+会当凌绝顶，一览众山小。
+野旷天低树，江清月近人。
+国破山河在，城春草木深。
+感时花溅泪，恨别鸟惊心。
+白日依山尽，黄河入海流。
+红豆不堪看，满眼相思泪。
+独怜幽草涧边生，上有黄鹂深树鸣。
+云想衣裳花想容，春风拂槛露华浓。
+夜来风雨声，花落知多少。
+江水流春去欲尽，江潭落月复西斜。
+独在异乡为异客，每逢佳节倍思亲。
+春江潮水连海平，海上明月共潮生。
+此夜曲中闻折柳，何人不起故园情。
+无边落木萧萧下，不尽长江滚滚来。
+春心莫共花争发，一寸相思一寸灰。
+月出惊山鸟，时鸣春涧中。
+江畔何人初见月，江月何年初照人。
+谁家玉笛暗飞声，散入春风满洛城。
+草长莺飞二月天，拂堤杨柳醉春烟。
+春风又绿江南岸，明月何时照我还。
+露从今夜白，月是故乡明。
+君问归期未有期，巴山夜雨涨秋池。
+秋水共长天一色，落霞与孤鹜齐飞。
+月黑雁飞高，单于夜遁逃。
+风急天高猿啸哀，渚清沙白鸟飞回。
+星垂平野阔，月涌大江流。
+秦时明月汉时关，万里长征人未还。
+沧海月明珠有泪，蓝田日暖玉生烟。
+独上江楼思渺然，月光如水水如天。
+相看两不厌，只有敬亭山。
+桃花流水窅然去，别有天地非人间。
+黄鹤一去不复返，白云千载空悠悠。
+朱雀桥边野草花，乌衣巷口夕阳斜。
+绿蚁新醅酒，红泥小火炉。
+姑苏城外寒山寺，夜半钟声到客船。
+银烛秋光冷画屏，轻罗小扇扑流萤。
+碧玉妆成一树高，万条垂下绿丝绦。
+渭城朝雨浥轻尘，客舍青青柳色新。
+忽如一夜春风来，千树万树梨花开。
+故人西辞黄鹤楼，烟花三月下扬州。
+洛阳城里春光好，洛阳才子他乡老。
+江南好，风景旧曾谙。
+江南可采莲，莲叶何田田。
+采得百花成蜜后，为谁辛苦为谁甜。

smartpi/tencentcloud-speech-sdk-python/soe/speaking_assessment.py

@@ -0,0 +1,276 @@
+# -*- coding: utf-8 -*-
+import sys
+import hmac
+import hashlib
+import base64
+import time
+import json
+import threading
+import urllib
+
+import websocket
+import uuid
+from urllib.parse import quote
+from common.log import logger
+
+
+def is_python3():
+    if sys.version > '3':
+        return True
+    return False
+
+
+# 实时识别语音使用
+class SpeakingAssessmentListener():
+    '''
+    reponse:  
+    on_recognition_start的返回只有voice_id字段。
+    on_fail 只有voice_id、code、message字段。
+    on_recognition_complete没有result字段。
+    其余消息包含所有字段。
+    字段名	类型	
+    code	Integer	
+    message	String	
+    voice_id	String
+    message_id	String
+    result
+    final	Integer	
+
+    # Result的结构体格式为:
+    # slice_type	Integer
+    # index	Integer
+    # start_time	Integer
+    # end_time	Integer
+    # voice_text_str	String
+    # word_size	Integer
+    # word_list	Word Array
+    #
+    # Word的类型为:
+    # word    String
+    # start_time Integer
+    # end_time Integer
+    # stable_flag：Integer
+    '''
+
+    def on_recognition_start(self, response):
+        pass
+
+    def on_intermediate_result(self, response):
+        pass
+
+    def on_recognition_complete(self, response):
+        pass
+
+    def on_fail(self, response):
+        pass
+
+
+NOTOPEN = 0
+STARTED = 1
+OPENED = 2
+FINAL = 3
+ERROR = 4
+CLOSED = 5
+
+
+def quote_autho(autho):
+    if sys.version_info >= (3, 0):
+        import urllib.parse as urlparse
+        return urlparse.quote(autho)
+    else:
+        return urllib.quote(autho)
+
+
+# 实时识别使用
+class SpeakingAssessment:
+
+    def __init__(self, appid, credential, engine_model_type, listener):
+        self.result = ""
+        self.credential = credential
+        self.appid = appid
+        self.server_engine_type = engine_model_type
+        self.status = NOTOPEN
+        self.ws = None
+        self.wst = None
+        self.voice_id = ""
+        self.new_start = 0
+        self.listener = listener
+        self.text_mode = 0
+        self.ref_text = ""
+        self.keyword = ""
+        self.eval_mode = 0
+        self.score_coeff = 1.0
+        self.sentence_info_enabled = 0
+        self.voice_format = 0
+        self.nonce = ""
+        self.rec_mode = 0
+
+    def set_text_mode(self, text_mode):
+        self.text_mode = text_mode
+    
+    def set_rec_mode(self, rec_mode):
+        self.rec_mode = rec_mode
+
+    def set_ref_text(self, ref_text):
+        self.ref_text = ref_text
+
+    def set_keyword(self, keyword):
+        self.keyword = keyword
+
+    def set_eval_mode(self, eval_mode):
+        self.eval_mode = eval_mode
+
+    def set_sentence_info_enabled(self, sentence_info_enabled):
+        self.sentence_info_enabled = sentence_info_enabled
+
+    def set_voice_format(self, voice_format):
+        self.voice_format = voice_format
+
+    def set_nonce(self, nonce):
+        self.nonce = nonce
+
+    def format_sign_string(self, param):
+        signstr = "soe.cloud.tencent.com/soe/api/"
+        for t in param:
+            if 'appid' in t:
+                signstr += str(t[1])
+                break
+        signstr += "?"
+        for x in param:
+            tmp = x
+            if 'appid' in x:
+                continue
+            for t in tmp:
+                signstr += str(t)
+                signstr += "="
+            signstr = signstr[:-1]
+            signstr += "&"
+        signstr = signstr[:-1]
+        return signstr
+
+    def create_query_string(self, param):
+        signstr = ""
+        for key, value in param.items():
+            if key == 'appid':
+                signstr += str(value)
+                break
+        signstr += "?"
+        for key, value in param.items():
+            if key == 'appid':
+                continue
+            value = quote_autho(str(value))
+            signstr += str(key) + "=" + str(value) + "&"
+        signstr = signstr[:-1]
+        return "wss://soe.cloud.tencent.com/soe/api/" + signstr
+
+    def sign(self, signstr, secret_key):
+        hmacstr = hmac.new(secret_key.encode('utf-8'),
+                           signstr.encode('utf-8'), hashlib.sha1).digest()
+        s = base64.b64encode(hmacstr)
+        s = s.decode('utf-8')
+        return s
+
+    def create_query_arr(self):
+        query_arr = dict()
+
+        query_arr['appid'] = self.appid
+        query_arr['server_engine_type'] = self.server_engine_type
+        query_arr['text_mode'] = self.text_mode
+        query_arr['rec_mode'] = self.rec_mode
+        query_arr['ref_text'] = self.ref_text
+        query_arr['keyword'] = self.keyword
+        query_arr['eval_mode'] = self.eval_mode
+        query_arr['score_coeff'] = self.score_coeff
+        query_arr['sentence_info_enabled'] = self.sentence_info_enabled
+        query_arr['secretid'] = self.credential.secret_id
+        if self.credential.token != "":
+            query_arr['token'] = self.credential.token
+        query_arr['voice_format'] = self.voice_format
+        query_arr['voice_id'] = self.voice_id
+        query_arr['timestamp'] = str(int(time.time()))
+        if self.nonce != "":
+            query_arr['nonce'] = self.nonce
+        else:
+            query_arr['nonce'] = query_arr['timestamp']
+        query_arr['expired'] = int(time.time()) + 24 * 60 * 60
+        return query_arr
+
+    def stop(self):
+        if self.status == OPENED:
+            msg = {'type': "end"}
+            text_str = json.dumps(msg)
+            self.ws.sock.send(text_str)
+        if self.ws:
+            if self.wst and self.wst.is_alive():
+                self.wst.join()
+        self.ws.close()
+
+    def write(self, data):
+        while self.status == STARTED:
+            time.sleep(0.1)
+        if self.status == OPENED:
+            self.ws.sock.send_binary(data)
+
+    def start(self):
+        def on_message(ws, message):
+            # print(message)
+            response = json.loads(message)
+            response['voice_id'] = self.voice_id
+            if response['code'] != 0:
+                logger.error("%s server recognition fail %s" %
+                             (response['voice_id'], response['message']))
+                self.listener.on_fail(response)
+                return
+            if "final" in response and response["final"] == 1:
+                self.status = FINAL
+                self.result = message
+                self.listener.on_recognition_complete(response)
+                logger.info("%s recognition complete" % response['voice_id'])
+                self.ws.close()
+                return
+            else:
+                if response["result"] is not None:
+                    self.listener.on_intermediate_result(response)
+                    logger.info("%s recognition doing" % response['voice_id'])
+                    return
+
+        def on_error(ws, error):
+            if self.status == FINAL:
+                return
+            logger.error("websocket error %s  voice id %s" %
+                         (format(error), self.voice_id))
+            self.status = ERROR
+
+        def on_close(ws):
+            self.status = CLOSED
+            logger.info("websocket closed  voice id %s" %
+                        self.voice_id)
+
+        def on_open(ws):
+            self.status = OPENED
+
+        query_arr = self.create_query_arr()
+        if self.voice_id == "":
+            query_arr['voice_id'] = str(uuid.uuid1())
+            self.voice_id = query_arr['voice_id']
+        query = sorted(query_arr.items(), key=lambda d: d[0])
+        signstr = self.format_sign_string(query)
+        autho = self.sign(signstr, self.credential.secret_key)
+        requrl = self.create_query_string(query_arr)
+        print(requrl)
+        if is_python3():
+            autho = urllib.parse.quote(autho)
+        else:
+            autho = urllib.quote(autho)
+        requrl += "&signature=%s" % autho
+        print(requrl)
+        self.ws = websocket.WebSocketApp(requrl, None,
+                                         on_error=on_error, on_close=on_close, on_message=on_message)
+        self.ws.on_open = on_open
+        self.wst = threading.Thread(target=self.ws.run_forever)
+        self.wst.daemon = True
+        self.wst.start()
+        self.status = STARTED
+        response = {'voice_id': self.voice_id}
+        self.listener.on_recognition_start(response)
+        logger.info("%s recognition start" % response['voice_id'])