diff --git a/dsLightRag/KeDaXunFei/XunFeiAudioEvaluator.py b/dsLightRag/KeDaXunFei/XunFeiAudioEvaluator.py
index 9e6e5b57..0010199e 100644
--- a/dsLightRag/KeDaXunFei/XunFeiAudioEvaluator.py
+++ b/dsLightRag/KeDaXunFei/XunFeiAudioEvaluator.py
@@ -18,8 +18,8 @@ from Config.Config import XF_APPID, XF_APISECRET, XF_APIKEY
class XunFeiAudioEvaluator:
"""讯飞语音评测类"""
-
- def __init__(self, appid, api_key, api_secret, audio_file, language):
+
+ def __init__(self, appid, api_key, api_secret, audio_file, language, txt):
self.appid = appid
self.api_key = api_key
self.api_secret = api_secret
@@ -28,50 +28,51 @@ class XunFeiAudioEvaluator:
self.host_url = "wss://ise-api.xfyun.cn/v2/open-ise"
self.websocket_url = ""
self.evaluation_results = {}
-
+ self.txt = txt
+
def generate_auth_url(self):
"""生成鉴权URL"""
now_time = datetime.now()
now_date = format_date_time(mktime(now_time.timetuple()))
-
+
origin_base = "host: " + "ise-api.xfyun.cn" + "\n"
origin_base += "date: " + now_date + "\n"
origin_base += "GET " + "/v2/open-ise " + "HTTP/1.1"
-
+
# SHA256加密
signature_sha = hmac.new(
- self.api_secret.encode('utf-8'),
+ self.api_secret.encode('utf-8'),
origin_base.encode('utf-8'),
digestmod=hashlib.sha256
).digest()
signature_sha = base64.b64encode(signature_sha).decode(encoding='utf-8')
-
+
authorization_origin = f'api_key="{self.api_key}", algorithm="hmac-sha256", headers="host date request-line", signature="{signature_sha}"'
authorization = base64.b64encode(authorization_origin.encode('utf-8')).decode(encoding='utf-8')
-
+
dict_data = {
"authorization": authorization,
"date": now_date,
"host": "ise-api.xfyun.cn"
}
-
+
return self.host_url + '?' + urlencode(dict_data)
-
+
def on_message(self, ws, message):
"""WebSocket消息处理"""
print(f"Received message: {message}")
data = json.loads(message)
-
+
# 检查是否存在错误码
if data.get("code") != 0:
print(f"API错误: {data.get('message', '未知错误')}")
ws.close()
return
-
+
# 安全获取data字段
response_data = data.get("data", {})
status = response_data.get("status")
-
+
if status == 2 and response_data.get("data"):
# 解析评测结果
xml_data = base64.b64decode(response_data["data"])
@@ -106,7 +107,7 @@ class XunFeiAudioEvaluator:
"cmd": "ssb",
"auf": "audio/L16;rate=16000",
"aue": "lame",
- "text": '\uFEFF' + "[content]\nnice to meet you."
+ "text": '\uFEFF' + "[content]\n" + self.txt
},
"data": {
"status": 0,
@@ -125,19 +126,19 @@ class XunFeiAudioEvaluator:
if total_sent > 0:
my_dict = {
"business": {
- "cmd": "auw",
- "aus": 4,
+ "cmd": "auw",
+ "aus": 4,
"aue": "lame"
},
"data": {
- "status": 2,
+ "status": 2,
"data": "" # 空字符串表示结束
}
}
ws.send(json.dumps(my_dict))
time.sleep(1)
break
-
+
# 发送中间音频帧
send_dict = {
"business": {
@@ -155,12 +156,12 @@ class XunFeiAudioEvaluator:
ws.send(json.dumps(send_dict))
total_sent += len(buffer)
time.sleep(0.04)
-
+
def parse_evaluation_results(self, xml_content):
"""解析评测结果XML并提取得分信息"""
try:
root = ET.fromstring(xml_content)
-
+
# 查找read_chapter节点
read_chapter = root.find('.//read_chapter')
if read_chapter is not None:
@@ -174,7 +175,7 @@ class XunFeiAudioEvaluator:
'word_count': int(read_chapter.get('word_count', 0)),
'is_rejected': read_chapter.get('is_rejected', 'false') == 'true'
}
-
+
# 提取句子级别得分
sentence = read_chapter.find('.//sentence')
if sentence is not None:
@@ -183,7 +184,7 @@ class XunFeiAudioEvaluator:
'fluency_score': float(sentence.get('fluency_score', 0)),
'total_score': float(sentence.get('total_score', 0))
}
-
+
# 提取单词级别得分
words = []
for word in read_chapter.findall('.//word'):
@@ -193,17 +194,17 @@ class XunFeiAudioEvaluator:
'dp_message': int(word.get('dp_message', 0))
}
words.append(word_data)
-
+
self.evaluation_results['words'] = words
-
+
except ET.ParseError as e:
print(f"XML解析错误: {e}")
-
+
def get_evaluation_summary(self):
"""获取评测结果摘要"""
if not self.evaluation_results:
return "暂无评测结果"
-
+
summary = "=== 语音评测结果摘要 ===\n"
summary += f"总得分: {self.evaluation_results.get('total_score', 0):.4f}\n"
summary += f"准确度得分: {self.evaluation_results.get('accuracy_score', 0):.4f}\n"
@@ -213,53 +214,53 @@ class XunFeiAudioEvaluator:
summary += f"标准度得分: {self.evaluation_results.get('standard_score', 0):.4f}\n"
summary += f"单词数量: {self.evaluation_results.get('word_count', 0)}\n"
summary += f"是否被拒绝: {'是' if self.evaluation_results.get('is_rejected', False) else '否'}\n"
-
+
if 'sentence' in self.evaluation_results:
sentence = self.evaluation_results['sentence']
summary += f"\n=== 句子级别得分 ===\n"
summary += f"句子准确度: {sentence.get('accuracy_score', 0):.4f}\n"
summary += f"句子流畅度: {sentence.get('fluency_score', 0):.4f}\n"
summary += f"句子总分: {sentence.get('total_score', 0):.4f}\n"
-
+
if 'words' in self.evaluation_results:
summary += f"\n=== 单词级别得分 ===\n"
for i, word in enumerate(self.evaluation_results['words']):
dp_msg = self._get_dp_message_description(word['dp_message'])
- summary += f"{i+1}. {word['content']}: {word['total_score']:.4f} ({dp_msg})\n"
-
+ summary += f"{i + 1}. {word['content']}: {word['total_score']:.4f} ({dp_msg})\n"
+
return summary
-
+
def _get_dp_message_description(self, dp_message):
"""获取dp_message描述"""
descriptions = {
0: "正常",
16: "漏读",
- 32: "增读",
+ 32: "增读",
64: "回读",
128: "替换"
}
return descriptions.get(dp_message, f"未知({dp_message})")
-
+
def run_evaluation(self):
"""运行评测"""
start_time = datetime.now()
websocket.enableTrace(False)
-
+
ws_url = self.generate_auth_url()
ws_entity = websocket.WebSocketApp(
- ws_url,
- on_message=self.on_message,
- on_error=self.on_error,
+ ws_url,
+ on_message=self.on_message,
+ on_error=self.on_error,
on_close=self.on_close,
on_open=self.on_open
)
-
+
ws_entity.run_forever(sslopt={"cert_reqs": ssl.CERT_NONE})
-
+
end_time = datetime.now()
evaluation_time = end_time - start_time
print(f"评测耗时： {evaluation_time}")
-
+
return self.evaluation_results, evaluation_time
@@ -272,16 +273,16 @@ if __name__ == '__main__':
appid = XF_APPID
api_secret = XF_APISECRET
api_key = XF_APIKEY
- #audio_file = "./1.mp3"
- audio_file=r'D:\dsWork\dsProject\dsLightRag\static\audio\audio_1f0a8c47db2d4f9ba7674055a352cab8.wav'
+ # audio_file = "./1.mp3"
+ audio_file = r'D:\dsWork\dsProject\dsLightRag\static\audio\audio_1f0a8c47db2d4f9ba7674055a352cab8.wav'
# 创建评测器实例
- evaluator = XunFeiAudioEvaluator(appid, api_key, api_secret, audio_file, "english")
-
+ evaluator = XunFeiAudioEvaluator(appid, api_key, api_secret, audio_file, "english","nice to meet you.")
+
# 运行评测
results, eval_time = evaluator.run_evaluation()
-
+
# 输出评测结果摘要
- print("\n" + "="*50)
+ print("\n" + "=" * 50)
print(evaluator.get_evaluation_summary())
print(f"总评测时间: {eval_time}")
- print("="*50)
+ print("=" * 50)
diff --git a/dsLightRag/KeDaXunFei/__pycache__/XunFeiAudioEvaluator.cpython-310.pyc b/dsLightRag/KeDaXunFei/__pycache__/XunFeiAudioEvaluator.cpython-310.pyc
index d5f16b6d..159a75fe 100644
Binary files a/dsLightRag/KeDaXunFei/__pycache__/XunFeiAudioEvaluator.cpython-310.pyc and b/dsLightRag/KeDaXunFei/__pycache__/XunFeiAudioEvaluator.cpython-310.pyc differ
diff --git a/dsLightRag/Routes/XunFeiRoute.py b/dsLightRag/Routes/XunFeiRoute.py
index 7701bea2..ae09e991 100644
--- a/dsLightRag/Routes/XunFeiRoute.py
+++ b/dsLightRag/Routes/XunFeiRoute.py
@@ -4,8 +4,7 @@ import uuid
import tempfile
import shutil
import sys
-import logging
-from fastapi import APIRouter, UploadFile, File
+from fastapi import APIRouter, UploadFile, File, Form
# 配置日志
logger = logging.getLogger(__name__)
@@ -20,7 +19,7 @@ sys.path.append(os.path.join(os.path.dirname(__file__), ".."))
from Config.Config import XF_APPID, XF_APIKEY, XF_APISECRET
@router.post("/save-audio")
-async def save_audio(audio: UploadFile = File(...)):
+async def save_audio(audio: UploadFile = File(...), txt: str = Form(...)):
"""保存音频文件并评分"""
temp_file = None
try:
@@ -45,7 +44,8 @@ async def save_audio(audio: UploadFile = File(...)):
api_key=XF_APIKEY,
api_secret=XF_APISECRET,
audio_file=temp_file,
- language="english"
+ language="english",
+ txt=txt
)
results, eval_time = evaluator.run_evaluation()
print(evaluator.get_evaluation_summary())
diff --git a/dsLightRag/Routes/__pycache__/XunFeiRoute.cpython-310.pyc b/dsLightRag/Routes/__pycache__/XunFeiRoute.cpython-310.pyc
index 2695cba8..fef601ed 100644
Binary files a/dsLightRag/Routes/__pycache__/XunFeiRoute.cpython-310.pyc and b/dsLightRag/Routes/__pycache__/XunFeiRoute.cpython-310.pyc differ
diff --git a/dsLightRag/static/XunFei/audio_evaluation.html b/dsLightRag/static/XunFei/audio_evaluation.html
index e13e553e..f26c2ce6 100644
--- a/dsLightRag/static/XunFei/audio_evaluation.html
+++ b/dsLightRag/static/XunFei/audio_evaluation.html
@@ -450,7 +450,11 @@
-
+
+
+
📝 请朗读以下文本
+
+
@@ -760,7 +764,9 @@
const fileName = `recording_${Date.now()}.${extensions[format]}`;
formData.append('audio', this.recordedBlob, fileName);
-
+ // 添加朗读文本到表单数据
+ const readingText = document.getElementById('readingText').value;
+ formData.append('txt', readingText);
try {
this.simulateUploadProgress();
diff --git a/dsLightRag/static/audio/audio_1013c986df5245f0bdf921aa5b198149.wav b/dsLightRag/static/audio/audio_1013c986df5245f0bdf921aa5b198149.wav
new file mode 100644
index 00000000..555ca7f4
Binary files /dev/null and b/dsLightRag/static/audio/audio_1013c986df5245f0bdf921aa5b198149.wav differ
diff --git a/dsLightRag/static/audio/audio_8357cf7e4b754d118232f3bd08239ad4.wav b/dsLightRag/static/audio/audio_8357cf7e4b754d118232f3bd08239ad4.wav
new file mode 100644
index 00000000..0385d9e9
Binary files /dev/null and b/dsLightRag/static/audio/audio_8357cf7e4b754d118232f3bd08239ad4.wav differ