Class: SmartPrompt::TTSAdapter
- Inherits:
-
LLMAdapter
- Object
- LLMAdapter
- SmartPrompt::TTSAdapter
- Defined in:
- lib/smart_prompt/tts_adapter.rb
Constant Summary collapse
- PREDEFINED_VOICES =
Predefined voice options
{ "alloy" => "沉稳男声alex", "echo" => "温柔女声claire", "fable" => "活泼女声fable", "onyx" => "磁性男声onyx", "nova" => "甜美女声nova", "shimmer" => "优雅女声shimmer" }
- SUPPORTED_LANGUAGES =
Supported languages
%w[zh en ja ko]
- SUPPORTED_FORMATS =
Supported output formats
%w[mp3 opus wav pcm]
Instance Attribute Summary
Attributes inherited from LLMAdapter
Instance Method Summary collapse
-
#available_voices ⇒ Object
Get available voices.
-
#create_custom_voice(name, reference_audio_file, description: nil) ⇒ Object
Create custom voice from reference audio.
-
#delete_custom_voice(voice_id) ⇒ Object
Delete custom voice.
-
#initialize(config) ⇒ TTSAdapter
constructor
A new instance of TTSAdapter.
-
#list_custom_voices ⇒ Object
List custom voices.
-
#synthesize_speech(text, voice: "alloy", model: nil, speed: 1.0, response_format: "mp3", language: nil) ⇒ Object
Text-to-speech synthesis.
-
#synthesize_to_file(text, output_path, voice: "alloy", model: nil, speed: 1.0, response_format: "mp3", language: nil) ⇒ Object
Synthesize speech and save to file.
Constructor Details
#initialize(config) ⇒ TTSAdapter
Returns a new instance of TTSAdapter.
24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 |
# File 'lib/smart_prompt/tts_adapter.rb', line 24 def initialize(config) super api_key = @config["api_key"] if api_key.is_a?(String) && api_key.start_with?("ENV[") && api_key.end_with?("]") api_key = eval(api_key) end begin @client = OpenAI::Client.new( access_token: api_key, uri_base: @config["url"], request_timeout: 120, ) rescue OpenAI::ConfigurationError => e SmartPrompt.logger.error "Failed to initialize TTS client: #{e.}" raise LLMAPIError, "Invalid TTS configuration: #{e.}" rescue OpenAI::Error => e SmartPrompt.logger.error "Failed to initialize TTS client: #{e.}" raise LLMAPIError, "TTS authentication failed: #{e.}" rescue SocketError => e SmartPrompt.logger.error "Failed to initialize TTS client: #{e.}" raise LLMAPIError, "Network error: Unable to connect to TTS API" rescue => e SmartPrompt.logger.error "Failed to initialize TTS client: #{e.}" raise Error, "Unexpected error initializing TTS client: #{e.}" ensure SmartPrompt.logger.info "Successfully created a TTS client." end end |
Instance Method Details
#available_voices ⇒ Object
Get available voices
143 144 145 |
# File 'lib/smart_prompt/tts_adapter.rb', line 143 def available_voices PREDEFINED_VOICES.dup end |
#create_custom_voice(name, reference_audio_file, description: nil) ⇒ Object
Create custom voice from reference audio
148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 |
# File 'lib/smart_prompt/tts_adapter.rb', line 148 def create_custom_voice(name, reference_audio_file, description: nil) SmartPrompt.logger.info "TTSAdapter: Creating custom voice" begin unless File.exist?(reference_audio_file) raise Error, "Reference audio file not found: #{reference_audio_file}" end # Check audio file size (should be less than 30 seconds) file_size = File.size(reference_audio_file) if file_size > 5 * 1024 * 1024 # 5MB limit raise Error, "Reference audio file too large (max 5MB)" end # Convert audio to base64 audio_data = File.binread(reference_audio_file) base64_audio = Base64.strict_encode64(audio_data) parameters = { name: name, audio: base64_audio } parameters[:description] = description if description SmartPrompt.logger.info "Creating custom voice: #{name}" # Custom implementation for voice creation response = create_custom_voice_request(parameters) @last_response = response if response["voice_id"] voice_data = { voice_id: response["voice_id"], name: response["name"], status: response["status"], created_at: response["created_at"] } SmartPrompt.logger.info "Custom voice created successfully: #{voice_data[:voice_id]}" return voice_data else SmartPrompt.logger.error "Failed to create custom voice" raise LLMAPIError, "Failed to create custom voice" end rescue => e SmartPrompt.logger.error "Error creating custom voice: #{e.}" raise Error, "Error creating custom voice: #{e.}" end end |
#delete_custom_voice(voice_id) ⇒ Object
Delete custom voice
235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 |
# File 'lib/smart_prompt/tts_adapter.rb', line 235 def delete_custom_voice(voice_id) SmartPrompt.logger.info "TTSAdapter: Deleting custom voice" begin response = delete_custom_voice_request(voice_id) @last_response = response if response["deleted"] SmartPrompt.logger.info "Custom voice deleted successfully: #{voice_id}" return { deleted: true, voice_id: voice_id } else SmartPrompt.logger.error "Failed to delete custom voice" raise LLMAPIError, "Failed to delete custom voice" end rescue => e SmartPrompt.logger.error "Error deleting custom voice: #{e.}" raise Error, "Error deleting custom voice: #{e.}" end end |
#list_custom_voices ⇒ Object
List custom voices
202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 |
# File 'lib/smart_prompt/tts_adapter.rb', line 202 def list_custom_voices SmartPrompt.logger.info "TTSAdapter: Listing custom voices" begin response = list_custom_voices_request @last_response = response if response["voices"] voices = response["voices"].map do |voice| { voice_id: voice["id"], name: voice["name"], description: voice["description"], status: voice["status"], created_at: voice["created_at"] } end SmartPrompt.logger.info "Found #{voices.size} custom voices" return voices else SmartPrompt.logger.error "No custom voices found" return [] end rescue => e SmartPrompt.logger.error "Error listing custom voices: #{e.}" raise Error, "Error listing custom voices: #{e.}" end end |
#synthesize_speech(text, voice: "alloy", model: nil, speed: 1.0, response_format: "mp3", language: nil) ⇒ Object
Text-to-speech synthesis
54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 |
# File 'lib/smart_prompt/tts_adapter.rb', line 54 def synthesize_speech(text, voice: "alloy", model: nil, speed: 1.0, response_format: "mp3", language: nil) SmartPrompt.logger.info "TTSAdapter: Synthesizing speech from text" model_name = model || @config["model"] # Validate parameters validate_tts_parameters(text, voice, speed, response_format, language) begin # Map voice name if it's a predefined voice voice_name = PREDEFINED_VOICES[voice] || voice parameters = { model: model_name, input: text, voice: voice_name, speed: speed, response_format: response_format } # Add language parameter if specified parameters[:language] = language if language SmartPrompt.logger.info "TTS parameters: #{parameters.except(:input)}" # Custom implementation for TTS since OpenAI gem doesn't support audio endpoints response = submit_tts_request(parameters) @last_response = response # Process response if response.is_a?(String) && response.start_with?("data:audio/") # Base64 encoded audio data audio_data = { audio_data: response, format: response_format, text_length: text.length, voice: voice_name } SmartPrompt.logger.info "TTS synthesis successful, generated #{text.length} characters" return audio_data else SmartPrompt.logger.error "Invalid TTS response format" raise LLMAPIError, "Invalid TTS response format" end rescue OpenAI::Error => e SmartPrompt.logger.error "TTS API error: #{e.}" raise LLMAPIError, "TTS API error: #{e.}" rescue => e SmartPrompt.logger.error "Unexpected error during TTS synthesis: #{e.}" raise Error, "Unexpected error during TTS synthesis: #{e.}" end end |
#synthesize_to_file(text, output_path, voice: "alloy", model: nil, speed: 1.0, response_format: "mp3", language: nil) ⇒ Object
Synthesize speech and save to file
111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 |
# File 'lib/smart_prompt/tts_adapter.rb', line 111 def synthesize_to_file(text, output_path, voice: "alloy", model: nil, speed: 1.0, response_format: "mp3", language: nil) SmartPrompt.logger.info "TTSAdapter: Synthesizing speech to file" begin # Synthesize speech audio_data = synthesize_speech( text, voice: voice, model: model, speed: speed, response_format: response_format, language: language ) # Save to file save_audio_to_file(audio_data[:audio_data], output_path, response_format) SmartPrompt.logger.info "TTS audio saved to: #{output_path}" return { file_path: output_path, text_length: audio_data[:text_length], voice: audio_data[:voice], format: response_format } rescue => e SmartPrompt.logger.error "Error synthesizing to file: #{e.}" raise Error, "Error synthesizing to file: #{e.}" end end |