feature: create audio translations entrypoint

Author: osugikoji

ychat/src/commonMain/kotlin/co/yml/ychat/YChat.kt

@@ -1,6 +1,7 @@
 package co.yml.ychat
 
 import co.yml.ychat.entrypoint.features.AudioTranscriptions
+import co.yml.ychat.entrypoint.features.AudioTranslations
 import co.yml.ychat.entrypoint.features.ChatCompletions
 import co.yml.ychat.entrypoint.features.Completion
 import co.yml.ychat.entrypoint.features.Edits
@@ -152,6 +153,20 @@ interface YChat {
      */
     fun audioTranscriptions(): AudioTranscriptions
 
+    /**
+     * The audioTranscriptions api is used to translates audio into English.
+     *
+     * You can configure the parameters before executing it. Example:
+     * ```
+     * val result = YChat.create(apiKey).audioTranslations()
+     *      .setTemperature(0.0)
+     *      .setResponseFormat("json")
+     *      .set...
+     *      .execute("file.mp4", byteArrayFile)
+     * ```
+     */
+    fun audioTranslations(): AudioTranslations
+
     /**
      * Callback is an interface used for handling the results of an operation.
      * It provides two methods, `onSuccess` and `onError`, for handling the success

ychat/src/commonMain/kotlin/co/yml/ychat/data/api/ChatGptApi.kt

@@ -29,4 +29,6 @@ internal interface ChatGptApi {
     suspend fun model(id: String): ApiResult<ModelDto>
 
     suspend fun audioTranscriptions(audioParamsDto: AudioParamsDto): ApiResult<AudioResultDto>
+
+    suspend fun audioTranslations(audioParamsDto: AudioParamsDto): ApiResult<AudioResultDto>
 }

ychat/src/commonMain/kotlin/co/yml/ychat/data/api/impl/ChatGptApiImpl.kt

@@ -76,6 +76,16 @@ internal class ChatGptApiImpl(private val apiExecutor: ApiExecutor) : ChatGptApi
         return apiBuilder.execute()
     }
 
+    override suspend fun audioTranslations(audioParamsDto: AudioParamsDto): ApiResult<AudioResultDto> {
+        val byteArray = audioParamsDto.byteArray.toByteArray()
+        val apiBuilder = apiExecutor
+            .setEndpoint("$VERSION/audio/translations")
+            .setHttpMethod(HttpMethod.Post)
+            .addFormPart("file", audioParamsDto.filename, byteArray)
+        audioParamsDto.getMap().forEach { apiBuilder.addFormPart(it.key, it.value) }
+        return apiBuilder.execute()
+    }
+
     companion object {
         private const val VERSION = "v1"
     }

ychat/src/commonMain/kotlin/co/yml/ychat/di/module/LibraryModule.kt

@@ -13,13 +13,15 @@ import co.yml.ychat.domain.usecases.ImageGenerationsUseCase
 import co.yml.ychat.domain.usecases.ListModelsUseCase
 import co.yml.ychat.domain.usecases.RetrieveModelUseCase
 import co.yml.ychat.entrypoint.features.AudioTranscriptions
+import co.yml.ychat.entrypoint.features.AudioTranslations
 import co.yml.ychat.entrypoint.features.ChatCompletions
 import co.yml.ychat.entrypoint.features.Completion
 import co.yml.ychat.entrypoint.features.Edits
 import co.yml.ychat.entrypoint.features.ImageGenerations
 import co.yml.ychat.entrypoint.features.ListModels
 import co.yml.ychat.entrypoint.features.RetrieveModel
 import co.yml.ychat.entrypoint.impl.AudioTranscriptionsImpl
+import co.yml.ychat.entrypoint.impl.AudioTranslationsImpl
 import co.yml.ychat.entrypoint.impl.ChatCompletionsImpl
 import co.yml.ychat.entrypoint.impl.CompletionImpl
 import co.yml.ychat.entrypoint.impl.EditsImpl
@@ -43,6 +45,7 @@ internal class LibraryModule(private val apiKey: String) {
         factory<ImageGenerations> { ImageGenerationsImpl(Dispatchers.Default, get()) }
         factory<Edits> { EditsImpl(Dispatchers.Default, get()) }
         factory<AudioTranscriptions> { AudioTranscriptionsImpl(Dispatchers.Default, get()) }
+        factory<AudioTranslations> { AudioTranslationsImpl(Dispatchers.Default, get()) }
     }
 
     private val domainModule = module {

ychat/src/commonMain/kotlin/co/yml/ychat/domain/usecases/AudioUseCase.kt

@@ -16,4 +16,14 @@ internal class AudioUseCase(private val chatGptApi: ChatGptApi) {
         val response = chatGptApi.audioTranscriptions(requestDto)
         return response.getBodyOrThrow().text
     }
+
+    suspend fun requestAudioTranslations(
+        filename: String,
+        audioFile: FileBytes,
+        params: AudioParams
+    ): String {
+        val requestDto = params.toAudioParamsDto(filename, audioFile)
+        val response = chatGptApi.audioTranslations(requestDto)
+        return response.getBodyOrThrow().text
+    }
 }

ychat/src/commonMain/kotlin/co/yml/ychat/entrypoint/features/AudioTranslations.kt

@@ -0,0 +1,63 @@
+package co.yml.ychat.entrypoint.features
+
+import co.yml.ychat.YChat
+import co.yml.ychat.data.exception.ChatGptException
+import co.yml.ychat.domain.model.FileBytes
+import kotlin.coroutines.cancellation.CancellationException
+
+interface AudioTranslations {
+
+    /**
+     * Sets the ID of the model to use. The default value is "whisper-1".
+     * @param model The ID of the model to use.
+     * @return The updated [AudioTranslations] object with the new model ID.
+     */
+    fun setModel(model: String): AudioTranslations
+
+    /**
+     * Sets an optional text to guide the model's style or continue a previous audio segment.
+     * The [prompt] should be in English.
+     * @param prompt The prompt to use.
+     * @return The updated [AudioTranslations] object with the new prompt.
+     */
+    fun setPrompt(prompt: String): AudioTranslations
+
+    /**
+     * Sets the format of the transcript output.
+     * @param format The format of the transcript output, in one of these options:
+     * json, text, srt, verbose_json, or vtt.
+     * @return The updated [AudioTranslations] object with the new response format.
+     */
+    fun setResponseFormat(format: String): AudioTranslations
+
+    /**
+     * Sets the sampling temperature, between 0 and 1. Higher values like 0.8 will make the output
+     * more random, while lower values like 0.2 will make it more focused and deterministic.
+     * If set to 0, the model will use log probability to automatically increase the temperature
+     * until certain thresholds are hit.
+     * @param temperature The sampling temperature to use.
+     * @return The updated [AudioTranslations] object with the new temperature.
+     */
+    fun setTemperature(temperature: Double): AudioTranslations
+
+    /**
+     * Translates audio into English.
+     * @param filename The name of the audio file.
+     * @param audioFile The audio file to translate, in one of these formats: mp3, mp4, mpeg,
+     * mpga, m4a, wav, or webm.
+     * @return The transcript output in the specified format.
+     * @throws CancellationException if the operation is cancelled.
+     * @throws ChatGptException if there is an error transcribing the audio file.
+     */
+    @Throws(CancellationException::class, ChatGptException::class)
+    suspend fun execute(filename: String, audioFile: FileBytes): String
+
+    /**
+     * Translates audio into English and passes it to the provided [callback].
+     * @param filename The name of the audio file.
+     * @param audioFile The audio file to translate, in one of these formats:
+     * mp3, mp4, mpeg, mpga, m4a, wav, or webm.
+     * @param callback The callback to receive the transcript output in the specified format.
+     */
+    fun execute(filename: String, audioFile: FileBytes, callback: YChat.Callback<String>)
+}

ychat/src/commonMain/kotlin/co/yml/ychat/entrypoint/impl/AudioTranslationsImpl.kt

@@ -0,0 +1,53 @@
+package co.yml.ychat.entrypoint.impl
+
+import co.yml.ychat.YChat
+import co.yml.ychat.domain.model.AudioParams
+import co.yml.ychat.domain.model.FileBytes
+import co.yml.ychat.domain.usecases.AudioUseCase
+import co.yml.ychat.entrypoint.features.AudioTranslations
+import kotlinx.coroutines.CoroutineDispatcher
+import kotlinx.coroutines.CoroutineScope
+import kotlinx.coroutines.SupervisorJob
+import kotlinx.coroutines.launch
+
+internal class AudioTranslationsImpl(
+    private val dispatcher: CoroutineDispatcher,
+    private val audioUseCase: AudioUseCase,
+) : AudioTranslations {
+
+    private val scope by lazy { CoroutineScope(SupervisorJob() + dispatcher) }
+
+    private val params: AudioParams = AudioParams()
+
+    override fun setModel(model: String): AudioTranslations {
+        params.model = model
+        return this
+    }
+
+    override fun setPrompt(prompt: String): AudioTranslations {
+        params.prompt = prompt
+        return this
+    }
+
+    override fun setResponseFormat(format: String): AudioTranslations {
+        params.responseFormat = format
+        return this
+    }
+
+    override fun setTemperature(temperature: Double): AudioTranslations {
+        params.temperature = temperature
+        return this
+    }
+
+    override suspend fun execute(filename: String, audioFile: FileBytes): String {
+        return audioUseCase.requestAudioTranslations(filename, audioFile, params)
+    }
+
+    override fun execute(filename: String, audioFile: FileBytes, callback: YChat.Callback<String>) {
+        scope.launch {
+            runCatching { execute(filename, audioFile) }
+                .onSuccess { callback.onSuccess(it) }
+                .onFailure { callback.onError(it) }
+        }
+    }
+}

ychat/src/commonMain/kotlin/co/yml/ychat/entrypoint/impl/YChatImpl.kt

@@ -3,6 +3,7 @@ package co.yml.ychat.entrypoint.impl
 import co.yml.ychat.YChat
 import co.yml.ychat.di.module.LibraryModule
 import co.yml.ychat.entrypoint.features.AudioTranscriptions
+import co.yml.ychat.entrypoint.features.AudioTranslations
 import co.yml.ychat.entrypoint.features.ChatCompletions
 import co.yml.ychat.entrypoint.features.Completion
 import co.yml.ychat.entrypoint.features.Edits
@@ -47,4 +48,8 @@ internal class YChatImpl(apiKey: String) : YChat {
     override fun audioTranscriptions(): AudioTranscriptions {
         return koinApp.koin.get()
     }
+
+    override fun audioTranslations(): AudioTranslations {
+        return koinApp.koin.get()
+    }
 }