Create config.yaml

config.yaml

@@ -0,0 +1,54 @@
+server:
+  port: 8080
+  read_timeout: 21
+  write_timeout: 89
+  base_path: ""
+
+tts:
+  region: "eastasia"
+  default_voice: "zh-CN-XiaoxiaoNeural"
+  default_rate: "0"
+  default_pitch: "0"
+  default_format: "audio-24khz-48kbitrate-mono-mp3"
+  max_text_length: 4181 # 最大文本长度
+  request_timeout: 233
+  max_concurrent: 8
+  segment_threshold: 377 # 分段阈值，超过值时，会进行切分
+  min_sentence_length: 34 # 最小句子长度
+  max_sentence_length: 233 # 最大句子长度
+  api_key: ''
+
+  # OpenAI 到微软 TTS 中文语音的映射
+  voice_mapping:
+    alloy: "zh-CN-XiaoyiNeural"       # 中性女声
+    echo: "zh-CN-YunxiNeural"         # 年轻男声
+    fable: "zh-CN-XiaochenNeural"     # 儿童声
+    onyx: "zh-CN-YunjianNeural"       # 成熟男声
+    nova: "zh-CN-XiaohanNeural"       # 活力女声
+    shimmer: "zh-CN-XiaomoNeural"     # 温柔女声
+ssml:
+  preserve_tags:
+    - name: break
+      pattern: <break\s+[^>]*/>
+    - name: speak
+      pattern: <speak>|</speak>
+    - name: prosody
+      pattern: <prosody\s+[^>]*>|</prosody>
+    - name: emphasis
+      pattern: <emphasis\s+[^>]*>|</emphasis>
+    - name: voice
+      pattern: <voice\s+[^>]*>|</voice>
+    - name: say-as
+      pattern: <say-as\s+[^>]*>|</say-as>
+    - name: phoneme
+      pattern: <phoneme\s+[^>]*>|</phoneme>
+    - name: audio
+      pattern: <audio\s+[^>]*>|</audio>
+    - name: p
+      pattern: <p>|</p>
+    - name: s
+      pattern: <s>|</s>
+    - name: sub
+      pattern: <sub\s+[^>]*>|</sub>
+    - name: mstts
+      pattern: <mstts:[^>]*>|</mstts:[^>]*>