resultFiles;\n\n /**\n * 微调作业的当前状态\n */\n private String status;\n\n /**\n * 此微调作业处理的计费令牌总数\n */\n @JsonProperty(\"trained_tokens\")\n private Integer trainedTokens;\n\n /**\n * 用于训练的文件 ID\n */\n @JsonProperty(\"training_file\")\n private String trainingFile;\n\n /**\n * 用于验证的文件 ID\n */\n @JsonProperty(\"validation_file\")\n private String validationFile;\n\n}\n"
},
{
"path": "ai-openai/src/main/java/com/ai/openai/endPoint/images/ImageObject.java",
"content": "package com.ai.openai.endPoint.images;\n\nimport com.fasterxml.jackson.annotation.JsonProperty;\nimport lombok.AllArgsConstructor;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\n\n@Data\n@NoArgsConstructor\n@AllArgsConstructor\npublic class ImageObject {\n\n private String url;\n @JsonProperty(\"b64_json\")\n private String b64Json;\n @JsonProperty(\"revised_prompt\")\n private String revisedPrompt;\n\n\n}\n"
},
{
"path": "ai-openai/src/main/java/com/ai/openai/endPoint/images/req/CreateImageRequest.java",
"content": "package com.ai.openai.endPoint.images.req;\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport com.fasterxml.jackson.annotation.JsonProperty;\nimport lombok.*;\n\nimport java.io.Serializable;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class CreateImageRequest implements Serializable {\n\n /**\n * 所需图像的文本描述。的最大长度为 1000 个字符，4000 个字符。\n */\n private String prompt;\n\n /**\n * 用于图像生成的模型\n * \n * 默认为 dall-e-2\n *\n * @see Model\n */\n @Builder.Default\n private String model = Model.DALL_E_2.getName();\n\n /**\n * 要生成的图像数。必须介于 1 和 10 之间，dall-e-3只能为1。\n */\n @Builder.Default\n private Integer n = 1;\n\n /**\n * 将生成的图像的质量。 创建具有更精细细节和更高一致性的图像。\n *\n * @see Quality\n */\n private String quality;\n\n /**\n * 返回生成的图像的格式：url、b64_json\n */\n @JsonProperty(\"response_format\")\n private String responseFormat;\n\n /**\n * 图片尺寸，默认值：1024x1024\n * dall-e-2支持：256x256, 512x512, or 1024x1024\n * dall-e-3支持：1024x1024, 1792x1024, or 1024x1792\n */\n private String size;\n\n /**\n * 生成的图像的样式。\n * 此参数仅仅dall-e-3,取值范围：vivid、natural\n * 默认值：vivid\n *\n * @see Style\n */\n private String style;\n\n /**\n * 代表最终用户的唯一标识符\n */\n private String user;\n\n /**\n * 构建基础请求内容\n *\n * @param prompt 提示词\n * @return 请求参数\n */\n public static CreateImageRequest baseBuild(String prompt) {\n return CreateImageRequest.builder().prompt(prompt).build();\n }\n\n /**\n * 图片生成模型\n */\n @Getter\n @AllArgsConstructor\n public enum Model {\n DALL_E_2(\"dall-e-2\"),\n DALL_E_3(\"dall-e-3\"),\n ;\n private final String name;\n }\n\n /**\n * 生成图片质量\n */\n @Getter\n @AllArgsConstructor\n public enum Quality {\n STANDARD(\"standard\"),\n HD(\"hd\"),\n ;\n private final String quality;\n }\n\n /**\n * 生成图片风格\n */\n @Getter\n @AllArgsConstructor\n public enum Style {\n VIVID(\"vivid\"),\n NATURAL(\"natural\"),\n ;\n private final String style;\n }\n\n @Getter\n @AllArgsConstructor\n public enum Format {\n URL(\"url\"),\n B64JSON(\"b64_json\"),\n ;\n private final String format;\n }\n\n\n}\n"
},
{
"path": "ai-openai/src/main/java/com/ai/openai/endPoint/images/req/ImageEditRequest.java",
"content": "package com.ai.openai.endPoint.images.req;\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport com.fasterxml.jackson.annotation.JsonProperty;\nimport lombok.*;\n\nimport java.io.Serializable;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class ImageEditRequest {\n\n /**\n * 必选项：描述文字，最多1000字符\n */\n @NonNull\n private String prompt;\n\n /**\n * 为每个提示生成的完成次数。\n */\n @Builder.Default\n private Integer n = 1;\n\n /**\n * 256x256\n * 512x512\n * 1024x1024\n */\n @Builder.Default\n private String size = SizeEnum.size_1024.getName();\n\n @JsonProperty(\"response_format\")\n @Builder.Default\n private String responseFormat = ResponseFormat.URL.getName();\n\n private String user;\n\n /**\n * 构造基础请求参数\n *\n * @param prompt 提示词\n * @return 请求参数\n */\n public static ImageEditRequest baseBuild(String prompt) {\n return ImageEditRequest.builder().prompt(prompt).build();\n }\n\n @Getter\n @AllArgsConstructor\n public enum SizeEnum implements Serializable {\n size_1024_1792(\"1024x1792\"),\n size_1792_1024(\"1792x1024\"),\n size_1024(\"1024x1024\"),\n size_512(\"512x512\"),\n size_256(\"256x256\"),\n ;\n private final String name;\n\n }\n\n @AllArgsConstructor\n @Getter\n public enum ResponseFormat implements Serializable {\n URL(\"url\"),\n B64_JSON(\"b64_json\"),\n ;\n\n private final String name;\n }\n\n}\n"
},
{
"path": "ai-openai/src/main/java/com/ai/openai/endPoint/images/req/ImageVariationRequest.java",
"content": "package com.ai.openai.endPoint.images.req;\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport com.fasterxml.jackson.annotation.JsonProperty;\nimport lombok.*;\nimport lombok.extern.slf4j.Slf4j;\n\nimport java.io.Serializable;\n\n@Slf4j\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class ImageVariationRequest {\n\n /**\n * 为每个提示生成的完成次数。\n */\n @Builder.Default\n private Integer n = 1;\n /**\n * 256x256\n * 512x512\n * 1024x1024\n */\n @Builder.Default\n private String size = SizeEnum.size_1024.getName();\n\n @JsonProperty(\"response_format\")\n @Builder.Default\n private String responseFormat = ResponseFormat.URL.getName();\n\n private String user;\n\n @Getter\n @AllArgsConstructor\n public enum SizeEnum implements Serializable {\n size_1024_1792(\"1024x1792\"),\n size_1792_1024(\"1792x1024\"),\n size_1024(\"1024x1024\"),\n size_512(\"512x512\"),\n size_256(\"256x256\"),\n ;\n private final String name;\n\n }\n\n @AllArgsConstructor\n @Getter\n public enum ResponseFormat implements Serializable {\n URL(\"url\"),\n B64_JSON(\"b64_json\"),\n ;\n\n private final String name;\n }\n\n\n}\n"
},
{
"path": "ai-openai/src/main/java/com/ai/openai/endPoint/images/resp/CreateImageResponse.java",
"content": "package com.ai.openai.endPoint.images.resp;\n\nimport com.ai.openai.endPoint.images.ImageObject;\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport lombok.AllArgsConstructor;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\n\nimport java.util.List;\n\n@Data\n@NoArgsConstructor\n@AllArgsConstructor\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class CreateImageResponse {\n private Long created;\n private List data;\n}\n"
},
{
"path": "ai-openai/src/main/java/com/ai/openai/endPoint/models/ModelObject.java",
"content": "package com.ai.openai.endPoint.models;\n\n\nimport com.fasterxml.jackson.annotation.JsonProperty;\nimport lombok.AllArgsConstructor;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\n\n@Data\n@AllArgsConstructor\n@NoArgsConstructor\npublic class ModelObject {\n\n /**\n * 模型标识符，可在 API 端点中引用。\n */\n private String id;\n\n /**\n * 对象类型，始终为“model”。\n */\n private String object;\n\n /**\n * 创建模型时的 Unix 时间戳（以秒为单位）。\n */\n private long created;\n\n /**\n * 拥有模型的组织。\n */\n @JsonProperty(\"owned_by\")\n private String ownedBy;\n\n}\n"
},
{
"path": "ai-openai/src/main/java/com/ai/openai/endPoint/models/resp/DeleteFineTuneModelResponse.java",
"content": "package com.ai.openai.endPoint.models.resp;\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport lombok.AllArgsConstructor;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\n\nimport java.io.Serializable;\n\n@Data\n@NoArgsConstructor\n@AllArgsConstructor\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class DeleteFineTuneModelResponse implements Serializable {\n\n private String id;\n private String object;\n private Boolean deleted;\n\n}\n"
},
{
"path": "ai-openai/src/main/java/com/ai/openai/endPoint/moderations/Categories.java",
"content": "package com.ai.openai.endPoint.moderations;\n\nimport com.fasterxml.jackson.annotation.JsonProperty;\nimport lombok.AllArgsConstructor;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\n\nimport java.io.Serializable;\n\n\n@Data\n@AllArgsConstructor\n@NoArgsConstructor\npublic class Categories implements Serializable {\n /**\n * 表达、煽动或宣扬基于种族、性别、民族、宗教、国籍、性取向、残疾状况或种姓的仇恨的内容。\n */\n private boolean hate;\n\n /**\n * 仇恨内容，还包括对目标群体的暴力或严重伤害。\n */\n @JsonProperty(\"hate/threatening\")\n private boolean hateThreatening;\n\n /**\n * 对任何目标表达、煽动或宣扬骚扰性语言的内容。\n */\n @JsonProperty(\"harassment\")\n private boolean harassment;\n\n /**\n * 骚扰内容还包括对任何目标的暴力或严重伤害。\n */\n @JsonProperty(\"harassment/threatening\")\n private boolean harassmentThreatening;\n\n /**\n * 宣扬、鼓励或描绘自残行为（例如自杀、割伤和饮食失调）的内容。\n */\n @JsonProperty(\"self-harm\")\n private boolean selfHarm;\n\n /**\n * 说话者表示他们正在或打算进行自残行为的内容，例如自杀、割伤和饮食失调。\n */\n @JsonProperty(\"self-harm/intent\")\n private boolean selfHarmIntent;\n\n /**\n * 鼓励进行自残行为（例如自杀、割伤和饮食失调）的内容，或者提供有关如何实施此类行为的说明或建议的内容。\n */\n @JsonProperty(\"self-harm/instructions\")\n private boolean selfHarmInstructions;\n\n /**\n * 旨在引起性兴奋的内容，例如对性活动的描述，或宣传性服务（不包括性教育和健康）的内容。\n */\n private boolean sexual;\n\n /**\n * 包含未满 18 周岁的个人的色情内容。\n */\n @JsonProperty(\"sexual/minors\")\n private boolean sexualMinors;\n\n /**\n * 宣扬或美化暴力或歌颂他人遭受苦难或羞辱的内容。\n */\n private boolean violence;\n\n /**\n * 以极端血腥细节描绘死亡、暴力或严重身体伤害的暴力内容。\n */\n @JsonProperty(\"violence/graphic\")\n private boolean violenceGraphic;\n\n\n}\n"
},
{
"path": "ai-openai/src/main/java/com/ai/openai/endPoint/moderations/CategoryScores.java",
"content": "package com.ai.openai.endPoint.moderations;\n\nimport com.fasterxml.jackson.annotation.JsonProperty;\nimport lombok.AllArgsConstructor;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\n\nimport java.io.Serializable;\nimport java.math.BigDecimal;\n\n@Data\n@AllArgsConstructor\n@NoArgsConstructor\npublic class CategoryScores implements Serializable {\n\n private BigDecimal hate;\n\n @JsonProperty(\"hate/threatening\")\n private BigDecimal hateThreatening;\n\n /**\n * 对任何目标表达、煽动或宣扬骚扰性语言的内容。\n */\n @JsonProperty(\"harassment\")\n private BigDecimal harassment;\n\n /**\n * 骚扰内容还包括对任何目标的暴力或严重伤害。\n */\n @JsonProperty(\"harassment/threatening\")\n private BigDecimal harassmentThreatening;\n\n @JsonProperty(\"self-harm\")\n private BigDecimal selfHarm;\n\n /**\n * 说话者表示他们正在或打算进行自残行为的内容，例如自杀、割伤和饮食失调。\n */\n @JsonProperty(\"self-harm/intent\")\n private BigDecimal selfHarmIntent;\n\n /**\n * 鼓励进行自残行为（例如自杀、割伤和饮食失调）的内容，或者提供有关如何实施此类行为的说明或建议的内容。\n */\n @JsonProperty(\"self-harm/instructions\")\n private BigDecimal selfHarmInstructions;\n\n private BigDecimal sexual;\n\n @JsonProperty(\"sexual/minors\")\n private BigDecimal sexualMinors;\n\n private BigDecimal violence;\n\n @JsonProperty(\"violence/graphic\")\n private BigDecimal violenceGraphic;\n\n}\n"
},
{
"path": "ai-openai/src/main/java/com/ai/openai/endPoint/moderations/Result.java",
"content": "package com.ai.openai.endPoint.moderations;\n\nimport com.fasterxml.jackson.annotation.JsonIgnore;\nimport com.fasterxml.jackson.annotation.JsonProperty;\nimport lombok.AllArgsConstructor;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\n\nimport java.io.Serializable;\n\n\n@Data\n@AllArgsConstructor\n@NoArgsConstructor\npublic class Result implements Serializable {\n\n /**\n * 内容是否违反 OpenAI 的使用政策\n */\n private boolean flagged;\n\n /**\n * 类别列表，以及它们是否被标记\n */\n private Categories categories;\n\n /**\n * 类别列表及其按模型预测的分数\n */\n @JsonProperty(\"category_scores\")\n private CategoryScores categoryScores;\n\n /**\n * 原文内容\n */\n @JsonIgnore\n private String content;\n}\n"
},
{
"path": "ai-openai/src/main/java/com/ai/openai/endPoint/moderations/req/ModerationRequest.java",
"content": "package com.ai.openai.endPoint.moderations.req;\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport lombok.*;\n\nimport java.util.ArrayList;\nimport java.util.List;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class ModerationRequest {\n\n @NonNull\n private List input;\n\n @Builder.Default\n private String model = Model.TEXT_MODERATION_LATEST.getName();\n\n /**\n * 构造基础请求参数\n *\n * @param input 文本\n * @return 请求参数\n */\n public static ModerationRequest baseBuild(String input) {\n ArrayList list = new ArrayList<>();\n list.add(input);\n return baseBuild(list);\n }\n\n /**\n * 构造基础请求参数\n *\n * @param inputList 文本数组\n * @return 请求参数\n */\n public static ModerationRequest baseBuild(List inputList) {\n return ModerationRequest.builder().input(inputList).build();\n }\n\n @Getter\n @AllArgsConstructor\n public enum Model {\n TEXT_MODERATION_STABLE(\"text-moderation-stable\"),\n TEXT_MODERATION_LATEST(\"text-moderation-latest\"),\n ;\n\n private final String name;\n }\n\n}\n"
},
{
"path": "ai-openai/src/main/java/com/ai/openai/endPoint/moderations/resp/ModerationResponse.java",
"content": "package com.ai.openai.endPoint.moderations.resp;\n\nimport com.ai.openai.endPoint.moderations.Result;\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport lombok.AllArgsConstructor;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\n\nimport java.util.List;\n\n@Data\n@NoArgsConstructor\n@AllArgsConstructor\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class ModerationResponse {\n /**\n * 审核请求的唯一标识符\n */\n private String id;\n\n /**\n * 用于生成审核结果的模型\n */\n private String model;\n\n /**\n * 审核对象的列表\n */\n private List results;\n}\n"
},
{
"path": "ai-openai/src/main/java/com/ai/openai/interceptor/HeaderInterceptor.java",
"content": "package com.ai.openai.interceptor;\n\nimport cn.hutool.http.ContentType;\nimport cn.hutool.http.Header;\nimport com.ai.core.strategy.KeyStrategy;\nimport okhttp3.Interceptor;\nimport okhttp3.Request;\nimport okhttp3.Response;\nimport org.jetbrains.annotations.NotNull;\n\nimport java.io.IOException;\nimport java.util.List;\n\nimport static com.ai.core.exception.Constants.*;\n\n/**\n * @Description: OpenAI Key 拦截器\n **/\npublic class HeaderInterceptor implements Interceptor {\n\n /**\n * 系统设置的 openAI apiKey\n */\n private final List apiKeyBySystem;\n\n /**\n * 系统设置的 api 请求地址\n */\n private final String apiHostBySystem;\n\n /**\n * key 获取策略\n */\n private KeyStrategy, String> keyStrategy;\n\n public HeaderInterceptor(List apiKeyBySystem, String apiHostBySystem, KeyStrategy keyStrategy) {\n this.apiKeyBySystem = apiKeyBySystem;\n this.apiHostBySystem = apiHostBySystem;\n this.keyStrategy = keyStrategy;\n }\n\n @NotNull\n @Override\n public Response intercept(Chain chain) throws IOException {\n // 1. 获取原始 Request\n Request originalReq = chain.request();\n // 2. 读取 apiKey；优先使用用户传递的 apiKey\n String apiKeyByUser = originalReq.header(API_KEY);\n String apiHostByUser = originalReq.header(API_HOST);\n String apiUrlByUser = originalReq.header(URL);\n String apiKey = apiKeyByUser == NULL ? keyStrategy.apply(apiKeyBySystem) : apiKeyByUser;\n // 3. 读取 apiUrl 和 apiHost，apiUrl 优先级大于 apiHost\n String apiUrl = apiUrlByUser == NULL ? apiHostByUser == NULL ? String.valueOf(originalReq.url()) : apiHostByUser + originalReq.url().url().getPath() : apiUrlByUser;\n // 4. 构建 Request\n Request request = originalReq.newBuilder()\n .url(apiUrl)\n .header(Header.AUTHORIZATION.getValue(), \"Bearer \" + apiKey)\n .header(Header.CONTENT_TYPE.getValue(), ContentType.JSON.getValue())\n .method(originalReq.method(), originalReq.body())\n .build();\n // 5. 返回执行结果\n return chain.proceed(request);\n }\n}\n"
},
{
"path": "ai-openai/src/main/java/com/ai/openai/interceptor/ResponseInterceptor.java",
"content": "package com.ai.openai.interceptor;\n\nimport cn.hutool.json.JSONUtil;\nimport com.ai.core.exception.BaseException;\nimport com.ai.core.exception.Constants;\nimport com.ai.openai.common.CommonListResponse;\nimport lombok.extern.slf4j.Slf4j;\nimport okhttp3.Interceptor;\nimport okhttp3.Request;\nimport okhttp3.Response;\n\nimport java.io.IOException;\nimport java.util.Objects;\n\n@Slf4j\npublic class ResponseInterceptor implements Interceptor {\n @Override\n public Response intercept(Chain chain) throws IOException {\n // 1. 获取 req 和 resp\n Request original = chain.request();\n Response response = chain.proceed(original);\n // 2. 判断返回状态\n if (!response.isSuccessful() && response.body() != null) {\n // 2.1 获取返回的错误信息\n String errorMsg = response.body().string();\n CommonListResponse openAiResponse = JSONUtil.toBean(errorMsg, CommonListResponse.class);\n if (Constants.ERROR_MSG_MAP.containsKey(response.code())) {\n log.error(openAiResponse.getError().getMessage());\n throw new BaseException(openAiResponse.getError().getMessage());\n }\n log.error(\"--------> 请求异常：{}\", errorMsg);\n if (Objects.nonNull(openAiResponse.getError())) {\n log.error(openAiResponse.getError().getMessage());\n throw new BaseException(openAiResponse.getError().getMessage());\n }\n throw new BaseException(Constants.ErrorMsg.RETRY_ERROR);\n }\n return response;\n }\n}\n"
},
{
"path": "ai-openai/src/test/java/com/ai/openai/AudioApiTest.java",
"content": "package com.ai.openai;\n\nimport com.ai.core.strategy.impl.FirstKeyStrategy;\nimport com.ai.openai.achieve.Configuration;\nimport com.ai.openai.achieve.defaults.DefaultOpenAiSessionFactory;\nimport com.ai.openai.achieve.standard.session.AggregationSession;\nimport com.ai.openai.endPoint.audio.req.SttCompletionRequest;\nimport com.ai.openai.endPoint.audio.req.TtsCompletionRequest;\nimport com.ai.openai.endPoint.audio.resp.SttCompletionResponse;\nimport lombok.extern.slf4j.Slf4j;\nimport okhttp3.ResponseBody;\nimport org.junit.Before;\nimport org.junit.Test;\nimport retrofit2.Call;\nimport retrofit2.Callback;\nimport retrofit2.Response;\n\nimport java.io.*;\nimport java.net.InetSocketAddress;\nimport java.net.Proxy;\nimport java.util.Arrays;\nimport java.util.concurrent.CountDownLatch;\n\nimport static com.ai.core.exception.Constants.NULL;\n\n/**\n * @Description: 测试语音相关接口功能\n **/\n@Slf4j\npublic class AudioApiTest {\n\n private AggregationSession aggregationSession;\n\n @Before\n public void test_OpenAiSessionFactory() {\n // 1. 创建配置类\n Configuration configuration = new Configuration();\n // 2. 设置请求地址，若有代理商或者代理服务器，可填写为代理服务器的请求路径\n configuration.setApiHost(\"https://api.openai.com\");\n // 3. 设置鉴权所需的API Key,可设置多个。\n configuration.setKeyList(Arrays.asList(\"**************************\"));\n // 4. 设置请求时 key 的使用策略，默认实现了：随机获取 和 固定第一个Key 两种方式。\n configuration.setKeyStrategy(new FirstKeyStrategy());\n// configuration.setKeyStrategy(new RandomKeyStrategy());\n // 5. 设置代理，若不需要可不设置\n configuration.setProxy(new Proxy(Proxy.Type.HTTP, new InetSocketAddress(\"127.0.0.1\", 7890)));\n // 6. 创建 session 工厂，制造不同场景的 session\n DefaultOpenAiSessionFactory factory = new DefaultOpenAiSessionFactory(configuration);\n this.aggregationSession = factory.openAggregationSession();\n }\n\n /**\n * 测试文字转语音\n */\n @Test\n public void test_tts() throws InterruptedException {\n // 定义请求参数\n TtsCompletionRequest ttsCompletionRequest = TtsCompletionRequest.builder()\n .model(TtsCompletionRequest.Model.tts_1.getModuleName())// 设置使用的模型\n .input(\"你好，我是chatGPT\")\n .voice(TtsCompletionRequest.Voice.alloy.getVoiceName())// 设置声音的样式\n .build();\n // 回传文件存放的路径\n File file = new File(\"doc/test/test_tts.mp3\");\n // 添加回调函数，发送请求\n aggregationSession.getAudioSession().ttsCompletions(NULL, NULL, NULL, ttsCompletionRequest, new Callback() {\n @Override\n public void onResponse(Call call, Response response) {\n try (InputStream inputStream = response.body().byteStream();\n OutputStream os = new BufferedOutputStream(new FileOutputStream(file))) {\n // 创建文件\n if (!file.exists()) {\n if (!file.getParentFile().exists()) file.getParentFile().mkdir();\n file.createNewFile();\n }\n byte data[] = new byte[10240];\n int len;\n while ((len = inputStream.read(data, 0, 10240)) != -1) {\n os.write(data, 0, len);\n }\n } catch (IOException e) {\n e.printStackTrace();\n }\n }\n\n @Override\n public void onFailure(Call call, Throwable t) {\n t.printStackTrace();\n }\n }\n );\n // 阻塞等待\n new CountDownLatch(1).await();\n }\n\n /**\n * 测试语音转文字\n */\n @Test\n public void test_stt() {\n // 音频文件存放路径\n File file = new File(\"doc/test/test_tts.mp3\");\n SttCompletionRequest sttCompletionRequest = SttCompletionRequest.builder().file(file).build();\n SttCompletionResponse sttCompletionResponse = this.aggregationSession.getAudioSession().sttCompletions(NULL, NULL, NULL, sttCompletionRequest);\n log.info(\"测试结果：{}\", sttCompletionResponse);\n }\n\n /**\n * 测试音频文件转文字后翻译为英文\n */\n @Test\n public void test_translation() {\n // 音频文件存放路径\n File file = new File(\"doc/test/test_tts.mp3\");\n SttCompletionRequest sttCompletionRequest = SttCompletionRequest.builder().file(file).build();\n SttCompletionResponse sttCompletionResponse = this.aggregationSession.getAudioSession().translationCompletions(NULL, NULL, NULL, sttCompletionRequest);\n log.info(\"测试结果：{}\", sttCompletionResponse);\n }\n\n}\n"
},
{
"path": "ai-openai/src/test/java/com/ai/openai/ChatApiTest.java",
"content": "package com.ai.openai;\n\nimport cn.hutool.json.JSONObject;\nimport com.ai.core.exception.Constants;\nimport com.ai.core.strategy.impl.FirstKeyStrategy;\nimport com.ai.openai.achieve.Configuration;\nimport com.ai.openai.achieve.defaults.DefaultOpenAiSessionFactory;\nimport com.ai.openai.achieve.standard.session.AggregationSession;\nimport com.ai.openai.endPoint.chat.Parameters;\nimport com.ai.openai.endPoint.chat.msg.Content;\nimport com.ai.openai.endPoint.chat.msg.DefaultMessage;\nimport com.ai.openai.endPoint.chat.msg.ImgMessage;\nimport com.ai.openai.endPoint.chat.req.*;\nimport com.ai.openai.endPoint.chat.resp.ChatCompletionResponse;\nimport com.ai.openai.endPoint.chat.resp.QaCompletionResponse;\nimport com.ai.openai.endPoint.chat.tools.Tool;\nimport com.ai.openai.endPoint.chat.tools.ToolFunction;\nimport com.fasterxml.jackson.core.JsonProcessingException;\nimport lombok.extern.slf4j.Slf4j;\nimport okhttp3.Response;\nimport okhttp3.sse.EventSource;\nimport okhttp3.sse.EventSourceListener;\nimport org.junit.Before;\nimport org.junit.Test;\n\nimport java.net.InetSocketAddress;\nimport java.net.Proxy;\nimport java.util.Arrays;\nimport java.util.Collections;\nimport java.util.concurrent.CompletableFuture;\nimport java.util.concurrent.CountDownLatch;\nimport java.util.concurrent.ExecutionException;\n\nimport static com.ai.core.exception.Constants.NULL;\n\n/**\n * @Description: 测试聊天接口相关接口功能\n */\n@Slf4j\npublic class ChatApiTest {\n\n private AggregationSession aggregationSession;\n\n @Before\n public void test_OpenAiSessionFactory() {\n // 1. 创建配置类\n Configuration configuration = new Configuration();\n // 2. 设置请求地址，若有代理商或者代理服务器，可填写为代理服务器的请求路径\n configuration.setApiHost(\"https://api.openai.com\");\n // 3. 设置鉴权所需的API Key,可设置多个。\n configuration.setKeyList(Arrays.asList(\"填入你的API Key\"));\n // 4. 设置请求时 key 的使用策略，默认实现了：随机获取 和 固定第一个Key 两种方式。\n configuration.setKeyStrategy(new FirstKeyStrategy());\n// configuration.setKeyStrategy(new RandomKeyStrategy());\n // 5. 设置代理，若不需要可不设置\n configuration.setProxy(new Proxy(Proxy.Type.HTTP, new InetSocketAddress(\"127.0.0.1\", 7890)));\n // 6. 创建 session 工厂，制造不同场景的 session\n DefaultOpenAiSessionFactory factory = new DefaultOpenAiSessionFactory(configuration);\n this.aggregationSession = factory.openAggregationSession();\n }\n\n /**\n * 测试简单问答，相当于只有一轮问答。\n */\n @Test\n public void test_qa_completions() {\n QaCompletionRequest qaCompletionRequest = QaCompletionRequest.baseBuild(\"你是谁？\");\n QaCompletionResponse qaCompletionResponse = aggregationSession.getChatSession().qaCompletions(NULL, NULL, NULL, qaCompletionRequest);\n log.info(\"测试结果：{}\", qaCompletionResponse);\n }\n\n /**\n * 测试简单问答，流式返回结果。\n */\n @Test\n public void test_qa_completions_stream() throws InterruptedException, JsonProcessingException {\n QaCompletionRequest qaCompletionRequest = QaCompletionRequest.builder()\n .prompt(\"讲一个笑话\")\n .stream(true) // 设置流式返回\n .build();\n // 监听器监听返回的结果\n aggregationSession.getChatSession().qaCompletions(NULL, NULL, NULL, qaCompletionRequest, new EventSourceListener() {\n @Override\n public void onEvent(EventSource eventSource, String id, String type, String data) {\n log.info(\"测试结果 id:{} type:{} data:{}\", id, type, data);\n }\n\n @Override\n public void onFailure(EventSource eventSource, Throwable t, Response response) {\n log.error(\"失败 code:{} message:{}\", response.code(), response.message());\n }\n });\n // 阻塞等待\n new CountDownLatch(1).await();\n }\n\n /**\n * 测试多轮对话\n */\n @Test\n public void test_chat_completions() {\n // 创建参数，上下文对话。\n // 第一次的问题\n DefaultChatCompletionRequest defaultChatCompletionRequest = DefaultChatCompletionRequest.baseBuild(\"1+1=\");\n // 第一次的回复\n defaultChatCompletionRequest.addMessage(Constants.Role.ASSISTANT.getRoleName(), \"2\");\n // 第二次的问题\n defaultChatCompletionRequest.addMessage(Constants.Role.USER.getRoleName(), \"2+2=\");\n // 询问第二次的问题的结果\n ChatCompletionResponse chatCompletionResponse = aggregationSession.getChatSession().chatCompletions(NULL, NULL, NULL, defaultChatCompletionRequest);\n // 解析结果\n chatCompletionResponse.getChoices().forEach(e -> {\n log.info(\"测试结果：{}\", e.getMessage());\n });\n }\n\n /**\n * 测试函数对话，创建一个函数获取天气信息\n * 下面的请求参数根据官方案例转换而来\n */\n @Test\n public void test_func_chat_completions() {\n // 定义第一个属性，地址信息\n JSONObject location = new JSONObject();\n location.putOpt(\"type\", \"string\");\n location.putOpt(\"description\", \"The city and state, e.g. San Francisco, CA\");\n // 定义第二个属性，时间信息\n JSONObject unit = new JSONObject();\n unit.putOpt(\"type\", \"string\");\n unit.putOpt(\"enum\", Arrays.asList(\"celsius\", \"fahrenheit\"));\n // 定义 properties，及将函数属性组合起来\n JSONObject properties = new JSONObject();\n properties.putOpt(\"location\", location);\n properties.putOpt(\"unit\", unit);\n // 定义 parameters\n Parameters parameters = Parameters.builder().type(\"object\").properties(properties).required(Arrays.asList(\"location\")).build();\n // 构造函数信息\n ToolFunction toolFunction = ToolFunction.builder().name(\"get_current_weather\").description(\"Get the current weather in a given location\").parameters(parameters).build();\n // 构造工具\n Tool tool = Tool.builder().type(Tool.Type.FUNCTION.getName()).function(toolFunction).build();\n // 构造请求参数\n FuncChatCompletionRequest funcChatCompletionRequest = FuncChatCompletionRequest.baseBuild(\"What is the weather like in Boston?\");\n funcChatCompletionRequest.setTools(Arrays.asList(tool));\n funcChatCompletionRequest.setToolChoice(\"auto\");\n // 获取请求结果\n ChatCompletionResponse chatCompletionResponse = aggregationSession.getChatSession().chatCompletions(NULL, NULL, NULL, funcChatCompletionRequest);\n log.info(\"测试结果：{}\", chatCompletionResponse);\n }\n\n /**\n * 测试图片对话，需要GPT4权限\n */\n @Test\n public void test_img_chat_completions() {\n // 构造对话内容\n Content textContent = Content.BuildTextContent(\"这张图片当中有什么？\");\n Content imgContent = Content.BuildImageUrlContent(\"https://oaidalleapiprodscus.blob.core.windows.net/private/org-HL3RbCOW1GSH0YPFWak9m6be/user-AzfIxMQpzvc9raSA0TZ9sHOw/img-oft6JFrL4ilB4mmapSud8Vpy.png?st=2024-03-08T13%3A31%3A32Z&se=2024-03-08T15%3A31%3A32Z&sp=r&sv=2021-08-06&sr=b&rscd=inline&rsct=image/png&skoid=6aaadede-4fb3-4698-a8f6-684d7786b067&sktid=a48cca56-e6da-484e-a814-9c849652bcb3&skt=2024-03-07T18%3A11%3A55Z&ske=2024-03-08T18%3A11%3A55Z&sks=b&skv=2021-08-06&sig=wLcx9Uo6XiYT10GXIFcJwfd08BKoJ07k7cP7qJvYGx4%3D\");\n // 构造 msg 内容\n ImgMessage imgMessage = ImgMessage.builder().role(Constants.Role.USER.getRoleName()).content(Arrays.asList(textContent, imgContent)).build();\n // 构造请求参数，chatGPT 4 支持图片对话\n ImgChatCompletionRequest imgChatCompletionRequest = ImgChatCompletionRequest.builder().model(BaseChatCompletionRequest.Model.GPT_4_VISION_PREVIEW.getModuleName()).messages(Arrays.asList(imgMessage)).build();\n // 获取结果\n ChatCompletionResponse chatCompletionResponse = aggregationSession.getChatSession().chatCompletions(NULL, NULL, NULL, imgChatCompletionRequest);\n log.info(\"测试结果：{}\", chatCompletionResponse);\n }\n\n /**\n * 测试聊天对话，流式返回结果。\n */\n @Test\n public void test_chat_completions_stream() throws InterruptedException, JsonProcessingException {\n // 建造者模式构造参数\n DefaultChatCompletionRequest defaultChatCompletionRequest = DefaultChatCompletionRequest.builder()\n .stream(true)// 开启流式返回\n .messages(Collections.singletonList(DefaultMessage\n .builder()\n .role(Constants.Role.USER.getRoleName())\n .content(\"讲一个笑话\")\n .build()))\n .model(BaseChatCompletionRequest.Model.GPT_3_5_TURBO.getModuleName())\n .build();\n\n aggregationSession.getChatSession().chatCompletions(NULL, NULL, NULL, defaultChatCompletionRequest, new EventSourceListener() {\n @Override\n public void onEvent(EventSource eventSource, String id, String type, String data) {\n log.info(\"测试结果 id:{} type:{} data:{}\", id, type, data);\n }\n\n @Override\n public void onFailure(EventSource eventSource, Throwable t, Response response) {\n log.error(\"失败 code:{} message:{}\", response.code(), response.message());\n }\n });\n // 阻塞等待\n new CountDownLatch(1).await();\n }\n\n /**\n * 测试 CompletableFuture 异步调用。\n */\n @Test\n public void test_chat_completions_future() throws JsonProcessingException, InterruptedException, ExecutionException {\n // 构造请求参数\n DefaultChatCompletionRequest defaultChatCompletionRequest = DefaultChatCompletionRequest.builder()\n .stream(true)\n .messages(Collections.singletonList(DefaultMessage\n .builder()\n .role(Constants.Role.USER.getRoleName())\n .content(\"1+1=\")\n .build()))\n .model(BaseChatCompletionRequest.Model.GPT_3_5_TURBO.getModuleName()).build();\n // 等待结果\n CompletableFuture future = aggregationSession.getChatSession().chatCompletionsFuture(NULL, NULL, NULL, defaultChatCompletionRequest);\n log.info(\"测试结果：{}\", future.get());\n }\n\n}\n"
},
{
"path": "ai-openai/src/test/java/com/ai/openai/EmbeddingApiTest.java",
"content": "package com.ai.openai;\n\nimport com.ai.core.strategy.impl.FirstKeyStrategy;\nimport com.ai.openai.achieve.Configuration;\nimport com.ai.openai.achieve.defaults.DefaultOpenAiSessionFactory;\nimport com.ai.openai.achieve.standard.session.AggregationSession;\nimport com.ai.openai.endPoint.embeddings.EmbeddingObject;\nimport com.ai.openai.endPoint.embeddings.req.EmbeddingCompletionRequest;\nimport com.ai.openai.endPoint.embeddings.resp.EmbeddingCompletionResponse;\nimport lombok.extern.slf4j.Slf4j;\nimport org.junit.Before;\nimport org.junit.Test;\n\nimport java.net.InetSocketAddress;\nimport java.net.Proxy;\nimport java.util.ArrayList;\nimport java.util.Arrays;\nimport java.util.List;\n\nimport static com.ai.core.exception.Constants.NULL;\n\n/**\n * @Description: 测试嵌入相关接口功能\n */\n@Slf4j\npublic class EmbeddingApiTest {\n\n private AggregationSession aggregationSession;\n\n @Before\n public void test_OpenAiSessionFactory() {\n // 1. 创建配置类\n Configuration configuration = new Configuration();\n // 2. 设置请求地址，若有代理商或者代理服务器，可填写为代理服务器的请求路径\n configuration.setApiHost(\"https://api.openai.com\");\n // 3. 设置鉴权所需的API Key,可设置多个。\n configuration.setKeyList(Arrays.asList(\"**************************\"));\n // 4. 设置请求时 key 的使用策略，默认实现了：随机获取 和 固定第一个Key 两种方式。\n configuration.setKeyStrategy(new FirstKeyStrategy());\n// configuration.setKeyStrategy(new RandomKeyStrategy());\n // 5. 设置代理，若不需要可不设置\n configuration.setProxy(new Proxy(Proxy.Type.HTTP, new InetSocketAddress(\"127.0.0.1\", 7890)));\n // 6. 创建 session 工厂，制造不同场景的 session\n DefaultOpenAiSessionFactory factory = new DefaultOpenAiSessionFactory(configuration);\n this.aggregationSession = factory.openAggregationSession();\n }\n\n /**\n * 测试单个文本\n */\n @Test\n public void test_embedding() {\n EmbeddingCompletionResponse embeddingCompletionResponse = aggregationSession.getEmbeddingSession().embeddingCompletions(NULL, NULL, NULL, \"你好\");\n for (EmbeddingObject embeddingObject : embeddingCompletionResponse.getData()) {\n System.out.println(embeddingObject.getEmbedding().length);\n }\n log.info(\"返回结果：{}\", embeddingCompletionResponse);\n log.info(\"返回结果：{}\", embeddingCompletionResponse.getData().size());\n }\n\n\n /**\n * 测试多个文本嵌入\n */\n @Test\n public void test_embedding_list() {\n List inputList = new ArrayList<>();\n inputList.add(\"你好\");\n inputList.add(\"世界\");\n EmbeddingCompletionResponse embeddingCompletionResponse = aggregationSession.getEmbeddingSession().embeddingCompletions(NULL, NULL, NULL, inputList);\n log.info(\"返回结果：{}\", embeddingCompletionResponse);\n }\n\n @Test\n public void test_embedding_req() {\n List inputList = new ArrayList<>();\n inputList.add(\"你好\");\n inputList.add(\"世界\");\n EmbeddingCompletionRequest embeddingCompletionRequest = EmbeddingCompletionRequest.baseBuild(inputList);\n EmbeddingCompletionResponse embeddingCompletionResponse = aggregationSession.getEmbeddingSession().embeddingCompletions(NULL, NULL, NULL, embeddingCompletionRequest);\n log.info(\"返回结果：{}\", embeddingCompletionResponse);\n }\n\n}\n"
},
{
"path": "ai-openai/src/test/java/com/ai/openai/FilesApiTest.java",
"content": "package com.ai.openai;\n\nimport com.ai.core.strategy.impl.FirstKeyStrategy;\nimport com.ai.openai.achieve.Configuration;\nimport com.ai.openai.achieve.defaults.DefaultOpenAiSessionFactory;\nimport com.ai.openai.achieve.standard.session.AggregationSession;\nimport com.ai.openai.endPoint.files.FileObject;\nimport com.ai.openai.endPoint.files.resp.DeleteFileResponse;\nimport lombok.extern.slf4j.Slf4j;\nimport okhttp3.ResponseBody;\nimport org.junit.Before;\nimport org.junit.Test;\n\nimport java.io.File;\nimport java.net.InetSocketAddress;\nimport java.net.Proxy;\nimport java.util.Arrays;\nimport java.util.List;\n\nimport static com.ai.core.exception.Constants.NULL;\n\n/**\n * @Description: 测试文件相关接口功能\n **/\n@Slf4j\npublic class FilesApiTest {\n\n private AggregationSession aggregationSession;\n\n @Before\n public void test_OpenAiSessionFactory() {\n // 1. 创建配置类\n Configuration configuration = new Configuration();\n // 2. 设置请求地址，若有代理商或者代理服务器，可填写为代理服务器的请求路径\n configuration.setApiHost(\"https://api.openai.com\");\n // 3. 设置鉴权所需的API Key,可设置多个。\n configuration.setKeyList(Arrays.asList(\"填入你的API Key\"));\n // 4. 设置请求时 key 的使用策略，默认实现了：随机获取 和 固定第一个Key 两种方式。\n configuration.setKeyStrategy(new FirstKeyStrategy());\n// configuration.setKeyStrategy(new RandomKeyStrategy());\n // 5. 设置代理，若不需要可不设置\n configuration.setProxy(new Proxy(Proxy.Type.HTTP, new InetSocketAddress(\"127.0.0.1\", 7890)));\n // 6. 创建 session 工厂，制造不同场景的 session\n DefaultOpenAiSessionFactory factory = new DefaultOpenAiSessionFactory(configuration);\n this.aggregationSession = factory.openAggregationSession();\n }\n\n /**\n * 测试列出文件\n */\n @Test\n public void test_list_file() {\n List fileObjects = aggregationSession.getFilesSession().listFilesCompletions(NULL, NULL, NULL);\n log.info(\"测试结果：{}\", fileObjects);\n }\n\n /**\n * 测试上传文件\n */\n @Test\n public void test_upload_file() {\n File file = new File(\"src/main/resources/test1.txt\");\n FileObject fileObject = aggregationSession.getFilesSession().uploadFileCompletions(NULL, NULL, NULL, file, \"fine-tune\");\n log.info(\"测试结果：{}\", fileObject);\n }\n\n /**\n * 测试删除文件\n */\n @Test\n public void test_delete_file() {\n DeleteFileResponse deleteFileResponse = aggregationSession.getFilesSession().deleteFileCompletions(NULL, NULL, NULL, \"file-B3CAfSS2ibv7cFmSbl5m1CPI\");\n log.info(\"测试结果：{}\", deleteFileResponse);\n }\n\n /**\n * 测试检索文件\n */\n @Test\n public void test_retrieve_file() {\n FileObject fileObject = aggregationSession.getFilesSession().retrieveFileCompletions(NULL, NULL, NULL, \"file-B3CAfSS2ibv7cFmSbl5m1CPI\");\n log.info(\"测试结果：{}\", fileObject);\n }\n\n /**\n * 测试获取检索的文件内容\n */\n @Test\n public void test_retrieve_file_context() {\n ResponseBody responseBody = aggregationSession.getFilesSession().retrieveFileContextCompletions(NULL, NULL, NULL, \"file-B3CAfSS2ibv7cFmSbl5m1CPI\");\n log.info(\"测试结果：{}\", responseBody);\n }\n\n}\n"
},
{
"path": "ai-openai/src/test/java/com/ai/openai/FineTuningApiTest.java",
"content": "package com.ai.openai;\n\nimport com.ai.core.strategy.impl.FirstKeyStrategy;\nimport com.ai.openai.achieve.Configuration;\nimport com.ai.openai.achieve.defaults.DefaultOpenAiSessionFactory;\nimport com.ai.openai.achieve.standard.session.AggregationSession;\nimport com.ai.openai.common.CommonListResponse;\nimport com.ai.openai.endPoint.fineTuning.FineTuningEvent;\nimport com.ai.openai.endPoint.fineTuning.req.FineTuningRequest;\nimport com.ai.openai.endPoint.fineTuning.req.ListFineTuningRequest;\nimport com.ai.openai.endPoint.fineTuning.resp.FineTuningResponse;\nimport lombok.extern.slf4j.Slf4j;\nimport org.junit.Before;\nimport org.junit.Test;\n\nimport java.net.InetSocketAddress;\nimport java.net.Proxy;\nimport java.util.Arrays;\n\nimport static com.ai.core.exception.Constants.NULL;\n\n/**\n * @Description: 测试微调相关接口功能\n **/\n@Slf4j\npublic class FineTuningApiTest {\n\n private AggregationSession aggregationSession;\n\n @Before\n public void test_OpenAiSessionFactory() {\n // 1. 创建配置类\n Configuration configuration = new Configuration();\n // 2. 设置请求地址，若有代理商或者代理服务器，可填写为代理服务器的请求路径\n configuration.setApiHost(\"https://api.openai.com\");\n // 3. 设置鉴权所需的API Key,可设置多个。\n configuration.setKeyList(Arrays.asList(\"填入你的API Key\"));\n // 4. 设置请求时 key 的使用策略，默认实现了：随机获取 和 固定第一个Key 两种方式。\n configuration.setKeyStrategy(new FirstKeyStrategy());\n// configuration.setKeyStrategy(new RandomKeyStrategy());\n // 5. 设置代理，若不需要可不设置\n configuration.setProxy(new Proxy(Proxy.Type.HTTP, new InetSocketAddress(\"127.0.0.1\", 7890)));\n // 6. 创建 session 工厂，制造不同场景的 session\n DefaultOpenAiSessionFactory factory = new DefaultOpenAiSessionFactory(configuration);\n this.aggregationSession = factory.openAggregationSession();\n }\n\n /**\n * 测试创建微调\n */\n @Test\n public void test_create_fine_tuning() {\n FineTuningRequest fineTuningRequest = FineTuningRequest\n .builder()\n .trainingFile(\"file-BK7bzQj3FfZFXr7DbL6xJwfo\")\n .build();\n FineTuningResponse fineTuningResponse = aggregationSession.getFineTuningSession().createFineTuningJobCompletions(NULL, NULL, NULL, fineTuningRequest);\n log.info(\"返回结果：{}\", fineTuningResponse);\n }\n\n /**\n * 测试列出微调作业\n */\n @Test\n public void test_list_fine_tuning_1() {\n ListFineTuningRequest listFineTuningRequest = ListFineTuningRequest.builder().build();\n CommonListResponse fineTuneJobListFineTuningResponse = aggregationSession.getFineTuningSession().listFineTuningJobsCompletions(NULL, NULL, NULL, listFineTuningRequest);\n log.info(\"返回结果：{}\", fineTuneJobListFineTuningResponse);\n }\n\n /**\n * 测试列出微调作业\n */\n @Test\n public void test_list_fine_tuning_2() {\n CommonListResponse fineTuneJobListFineTuningResponse = aggregationSession.getFineTuningSession().listFineTuningJobsCompletions(NULL, NULL, NULL, null, null);\n log.info(\"返回结果：{}\", fineTuneJobListFineTuningResponse);\n }\n\n /**\n * 测试检索微调作业\n */\n @Test\n public void test_retrieve_fine_tuning() {\n FineTuningResponse fineTuningResponse = aggregationSession.getFineTuningSession().retrieveFineTuningJobCompletions(NULL, NULL, NULL, \"ft-AF1WoRqd3aJAHsqc9NY7iL8F\");\n log.info(\"返回结果：{}\", fineTuningResponse);\n }\n\n /**\n * 测试关闭微调作业\n */\n @Test\n public void test_cancel_fine_tuning() {\n FineTuningResponse fineTuningResponse = aggregationSession.getFineTuningSession().cancelFineTuningJobCompletions(NULL, NULL, NULL, \"ft-AF1WoRqd3aJAHsqc9NY7iL8F\");\n log.info(\"返回结果：{}\", fineTuningResponse);\n }\n\n /**\n * 测试列出微调事件\n */\n @Test\n public void test_list_fine_tuning_events() {\n CommonListResponse fineTuningEventListFineTuningResponse = aggregationSession.getFineTuningSession().listFineTuningEventsCompletions(NULL, NULL, NULL, \"ftjob-abc123\");\n log.info(\"返回结果：{}\", fineTuningEventListFineTuningResponse);\n\n }\n\n}\n"
},
{
"path": "ai-openai/src/test/java/com/ai/openai/ImageApiTest.java",
"content": "package com.ai.openai;\n\nimport com.ai.common.utils.ImageUtils;\nimport com.ai.core.strategy.impl.FirstKeyStrategy;\nimport com.ai.openai.achieve.Configuration;\nimport com.ai.openai.achieve.defaults.DefaultOpenAiSessionFactory;\nimport com.ai.openai.achieve.standard.session.AggregationSession;\nimport com.ai.openai.endPoint.images.ImageObject;\nimport com.ai.openai.endPoint.images.req.CreateImageRequest;\nimport com.ai.openai.endPoint.images.req.ImageEditRequest;\nimport com.ai.openai.endPoint.images.req.ImageVariationRequest;\nimport lombok.extern.slf4j.Slf4j;\nimport org.junit.Before;\nimport org.junit.Test;\n\nimport java.io.File;\nimport java.net.InetSocketAddress;\nimport java.net.Proxy;\nimport java.util.Arrays;\nimport java.util.List;\n\nimport static com.ai.core.exception.Constants.NULL;\n\n/**\n * @Description: 测试图片相关接口功能\n **/\n@Slf4j\npublic class ImageApiTest {\n private AggregationSession aggregationSession;\n\n @Before\n public void test_OpenAiSessionFactory() {\n // 1. 创建配置类\n Configuration configuration = new Configuration();\n // 2. 设置请求地址，若有代理商或者代理服务器，可填写为代理服务器的请求路径\n configuration.setApiHost(\"https://api.openai.com\");\n // 3. 设置鉴权所需的API Key,可设置多个。\n configuration.setKeyList(Arrays.asList(\"填入你的API Key\"));\n // 4. 设置请求时 key 的使用策略，默认实现了：随机获取 和 固定第一个Key 两种方式。\n configuration.setKeyStrategy(new FirstKeyStrategy());\n// configuration.setKeyStrategy(new RandomKeyStrategy());\n // 5. 设置代理，若不需要可不设置\n configuration.setProxy(new Proxy(Proxy.Type.HTTP, new InetSocketAddress(\"127.0.0.1\", 7890)));\n // 6. 创建 session 工厂，制造不同场景的 session\n DefaultOpenAiSessionFactory factory = new DefaultOpenAiSessionFactory(configuration);\n this.aggregationSession = factory.openAggregationSession();\n }\n\n /**\n * 测试图片生成，返回B64Json格式\n */\n @Test\n public void test_create_image_b64json() {\n CreateImageRequest createImageRequest = CreateImageRequest.baseBuild(\"画一个花园，花园里面有蝴蝶。\");\n createImageRequest.setResponseFormat(CreateImageRequest.Format.B64JSON.getFormat());\n List imageObjectList = aggregationSession.getImageSession().createImageCompletions(NULL, NULL, NULL, createImageRequest);\n for (int i = 0; i < imageObjectList.size(); i++) {\n ImageUtils.convertBase64StrToImage(imageObjectList.get(i).getB64Json(), \"D:\\\\chatGPT-api\\\\AI-java\\\\doc\\\\test\\\\test_openai_create_image_\" + i + \".png\");\n }\n }\n\n /**\n * 测试图片生成，返回URL格式\n */\n @Test\n public void test_create_image_url() {\n CreateImageRequest createImageRequest = CreateImageRequest.baseBuild(\"画一个花园，花园里面有蝴蝶。\");\n createImageRequest.setResponseFormat(CreateImageRequest.Format.URL.getFormat());\n List imageObjectList = aggregationSession.getImageSession().createImageCompletions(NULL, NULL, NULL, createImageRequest);\n for (int i = 0; i < imageObjectList.size(); i++) {\n System.out.println(imageObjectList.get(i).getUrl());\n }\n }\n\n /**\n * 测试编辑图片\n */\n @Test\n public void test_edit_image() {\n File file = new File(\"D:\\\\chatGPT-api\\\\AI-java\\\\doc\\\\test\\\\test_edit_image.png\");\n ImageEditRequest imageEditRequest = ImageEditRequest.baseBuild(\"给小熊的背后加上一只梅花鹿。\");\n List imageObjectList = aggregationSession.getImageSession().editImageCompletions(NULL, NULL, NULL, file, null, imageEditRequest);\n log.info(\"测试结果：{}\", imageObjectList);\n }\n\n /**\n * 测试创建图片变体\n */\n @Test\n public void test_variation_image() {\n File file = new File(\"D:\\\\chatGPT-api\\\\AI-java\\\\doc\\\\test\\\\test_edit_image.png\");\n ImageVariationRequest imageVariationRequest = ImageVariationRequest.builder().build();\n List imageObjectList = aggregationSession.getImageSession().variationImageCompletions(NULL, NULL, NULL, file, imageVariationRequest);\n log.info(\"测试结果：{}\", imageObjectList);\n }\n}\n"
},
{
"path": "ai-openai/src/test/java/com/ai/openai/ModelApiTest.java",
"content": "package com.ai.openai;\n\nimport com.ai.core.strategy.impl.FirstKeyStrategy;\nimport com.ai.openai.achieve.Configuration;\nimport com.ai.openai.achieve.defaults.DefaultOpenAiSessionFactory;\nimport com.ai.openai.achieve.standard.session.AggregationSession;\nimport com.ai.openai.endPoint.models.ModelObject;\nimport com.ai.openai.endPoint.models.resp.DeleteFineTuneModelResponse;\nimport lombok.extern.slf4j.Slf4j;\nimport org.junit.Before;\nimport org.junit.Test;\n\nimport java.net.InetSocketAddress;\nimport java.net.Proxy;\nimport java.util.Arrays;\nimport java.util.List;\n\nimport static com.ai.core.exception.Constants.NULL;\n\n/**\n * @Description: 测试模型相关接口功能\n **/\n@Slf4j\npublic class ModelApiTest {\n\n private AggregationSession aggregationSession;\n\n @Before\n public void test_OpenAiSessionFactory() {\n // 1. 创建配置类\n Configuration configuration = new Configuration();\n // 2. 设置请求地址，若有代理商或者代理服务器，可填写为代理服务器的请求路径\n configuration.setApiHost(\"https://api.openai.com\");\n // 3. 设置鉴权所需的API Key,可设置多个。\n configuration.setKeyList(Arrays.asList(\"填入你的API Key\"));\n // 4. 设置请求时 key 的使用策略，默认实现了：随机获取 和 固定第一个Key 两种方式。\n configuration.setKeyStrategy(new FirstKeyStrategy());\n// configuration.setKeyStrategy(new RandomKeyStrategy());\n // 5. 设置代理，若不需要可不设置\n configuration.setProxy(new Proxy(Proxy.Type.HTTP, new InetSocketAddress(\"127.0.0.1\", 7890)));\n // 6. 创建 session 工厂，制造不同场景的 session\n DefaultOpenAiSessionFactory factory = new DefaultOpenAiSessionFactory(configuration);\n this.aggregationSession = factory.openAggregationSession();\n }\n\n /**\n * 测试列出模型接口\n */\n @Test\n public void test_list_model() {\n List modelObjects = aggregationSession.getModelSession().listModelCompletions(NULL, NULL, NULL);\n log.info(\"返回结果：{}\", modelObjects);\n }\n\n /**\n * 测试检索模型接口\n */\n @Test\n public void test_retrieve_model() {\n ModelObject modelObject = aggregationSession.getModelSession().retrieveModelCompletions(NULL, NULL, NULL, \"gpt-3.5-turbo-instruct\");\n log.info(\"返回结果：{}\", modelObject);\n }\n\n /**\n * 测试删除微调模型接口\n */\n @Test\n public void test_delete_fine_tune_model() {\n DeleteFineTuneModelResponse deleteFineTuneModelResponse = aggregationSession.getModelSession().deleteFineTuneModelCompletions(NULL, NULL, NULL, \"gpt-3.5-turbo-instruct\");\n log.info(\"返回结果：{}\", deleteFineTuneModelResponse);\n }\n\n\n}\n"
},
{
"path": "ai-openai/src/test/java/com/ai/openai/ModerationApiTest.java",
"content": "package com.ai.openai;\n\nimport com.ai.core.strategy.impl.FirstKeyStrategy;\nimport com.ai.openai.achieve.Configuration;\nimport com.ai.openai.achieve.defaults.DefaultOpenAiSessionFactory;\nimport com.ai.openai.achieve.standard.session.AggregationSession;\nimport com.ai.openai.endPoint.moderations.Result;\nimport com.ai.openai.endPoint.moderations.req.ModerationRequest;\nimport com.ai.openai.endPoint.moderations.resp.ModerationResponse;\nimport lombok.extern.slf4j.Slf4j;\nimport org.junit.Before;\nimport org.junit.Test;\n\nimport java.net.InetSocketAddress;\nimport java.net.Proxy;\nimport java.util.ArrayList;\nimport java.util.Arrays;\n\nimport static com.ai.core.exception.Constants.NULL;\n\n/**\n * @Description: 测试审核相关接口功能\n **/\n@Slf4j\npublic class ModerationApiTest {\n\n private AggregationSession aggregationSession;\n\n @Before\n public void test_OpenAiSessionFactory() {\n // 1. 创建配置类\n Configuration configuration = new Configuration();\n // 2. 设置请求地址，若有代理商或者代理服务器，可填写为代理服务器的请求路径\n configuration.setApiHost(\"https://api.openai.com\");\n // 3. 设置鉴权所需的API Key,可设置多个。\n configuration.setKeyList(Arrays.asList(\"**************************\"));\n // 4. 设置请求时 key 的使用策略，默认实现了：随机获取 和 固定第一个Key 两种方式。\n configuration.setKeyStrategy(new FirstKeyStrategy());\n// configuration.setKeyStrategy(new RandomKeyStrategy());\n // 5. 设置代理，若不需要可不设置\n configuration.setProxy(new Proxy(Proxy.Type.HTTP, new InetSocketAddress(\"127.0.0.1\", 7890)));\n // 6. 创建 session 工厂，制造不同场景的 session\n DefaultOpenAiSessionFactory factory = new DefaultOpenAiSessionFactory(configuration);\n this.aggregationSession = factory.openAggregationSession();\n }\n\n /**\n * 测试审核\n */\n @Test\n public void test_moderation() {\n ArrayList list = new ArrayList<>();\n list.add(\"你好\");\n list.add(\"我要杀了你\");\n ModerationRequest moderationRequest = ModerationRequest.builder().input(list).build();\n ModerationResponse moderationResponse = aggregationSession.getModerationSession().moderationCompletions(NULL, NULL, NULL, moderationRequest);\n for (Result result : moderationResponse.getResults()) {\n System.out.println(result.getContent() + \" \" + result.isFlagged());\n }\n }\n\n}\n"
},
{
"path": "ai-spark/pom.xml",
"content": "\n\n \n AI-java\n com.ai\n 1.0\n \n 4.0.0\n ai-spark\n\n \n UTF-8\n UTF-8\n 1.8\n 1.8\n 1.8\n 4.13.2\n \n\n \n \n com.ai\n ai-common\n 1.0\n \n \n com.ai\n ai-core\n 1.0\n \n \n junit\n junit\n ${junit.version}\n test\n \n \n\n\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/achieve/ApiData.java",
"content": "package com.ai.spark.achieve;\n\n\nimport lombok.AllArgsConstructor;\nimport lombok.Builder;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\n\n/**\n * 记录用户API信息\n */\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\npublic class ApiData {\n\n private String appId;\n private String apiKey;\n private String apiSecret;\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/achieve/Configuration.java",
"content": "package com.ai.spark.achieve;\n\nimport com.ai.core.config.BaseConfiguration;\nimport com.ai.spark.achieve.standard.api.SparkApiServer;\nimport lombok.*;\nimport org.jetbrains.annotations.NotNull;\n\nimport java.util.List;\n\n/**\n * @Description: 配置信息\n **/\n@Data\n@ToString(callSuper = true)\n@EqualsAndHashCode(callSuper = true)\n@NoArgsConstructor\n@AllArgsConstructor\npublic class Configuration extends BaseConfiguration {\n\n /**\n * api 服务提供者\n */\n private SparkApiServer sparkApiServer;\n\n /**\n * api Key 集合\n */\n @NotNull\n private List keyList;\n\n public ApiData getSystemApiData() {\n return (ApiData) this.getKeyStrategy().apply(keyList);\n }\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/achieve/defaults/DefaultSparkSessionFactory.java",
"content": "package com.ai.spark.achieve.defaults;\n\nimport com.ai.core.factory.SessionFactory;\nimport com.ai.spark.achieve.Configuration;\nimport com.ai.spark.achieve.defaults.session.DefaultAggregationSession;\nimport com.ai.spark.achieve.standard.api.SparkApiServer;\nimport com.ai.spark.achieve.standard.session.AggregationSession;\nimport com.ai.spark.interceptor.BaseUrlInterceptor;\nimport com.ai.spark.interceptor.ResponseInterceptor;\nimport okhttp3.OkHttpClient;\nimport okhttp3.logging.HttpLoggingInterceptor;\nimport retrofit2.Retrofit;\nimport retrofit2.adapter.rxjava2.RxJava2CallAdapterFactory;\nimport retrofit2.converter.jackson.JacksonConverterFactory;\n\nimport java.util.concurrent.TimeUnit;\n\nimport static com.ai.common.utils.ValidationUtils.ensureNotNull;\n\npublic class DefaultSparkSessionFactory implements SessionFactory {\n\n private final Configuration configuration;\n\n public DefaultSparkSessionFactory(Configuration configuration) {\n this.configuration = ensureNotNull(configuration, \"configuration\");\n }\n\n @Override\n public OkHttpClient createHttpClient() {\n // 1. 日志配置\n HttpLoggingInterceptor httpLoggingInterceptor = new HttpLoggingInterceptor();\n httpLoggingInterceptor.setLevel(HttpLoggingInterceptor.Level.NONE);\n // 2. 开启 Http 客户端\n OkHttpClient.Builder builder = new OkHttpClient.Builder()\n .addInterceptor(httpLoggingInterceptor)\n .addInterceptor(new BaseUrlInterceptor())\n .addInterceptor(new ResponseInterceptor())// 设置返回信息拦截器\n .connectTimeout(450, TimeUnit.SECONDS)\n .writeTimeout(450, TimeUnit.SECONDS)\n .readTimeout(450, TimeUnit.SECONDS);\n // 3. 检查是否需要代理\n if (configuration.getProxy() != null) {\n builder.proxy(configuration.getProxy());\n }\n return builder.build();\n }\n\n @Override\n public SparkApiServer createApiServer(OkHttpClient okHttpClient) {\n return new Retrofit.Builder()\n .baseUrl(configuration.getApiHost())\n .client(okHttpClient)\n .addCallAdapterFactory(RxJava2CallAdapterFactory.create())\n .addConverterFactory(JacksonConverterFactory.create())\n .build().create(SparkApiServer.class);\n }\n\n @Override\n public AggregationSession openAggregationSession() {\n OkHttpClient okHttpClient = createHttpClient();\n configuration.setOkHttpClient(okHttpClient);\n configuration.setSparkApiServer(createApiServer(okHttpClient));\n return new DefaultAggregationSession(configuration);\n }\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/achieve/defaults/listener/BaseListener.java",
"content": "package com.ai.spark.achieve.defaults.listener;\n\nimport lombok.Data;\nimport lombok.extern.slf4j.Slf4j;\nimport okhttp3.Response;\nimport okhttp3.WebSocket;\nimport okhttp3.WebSocketListener;\nimport okio.ByteString;\nimport org.jetbrains.annotations.NotNull;\nimport org.jetbrains.annotations.Nullable;\n\n@Slf4j\n@Data\npublic abstract class BaseListener extends WebSocketListener {\n\n /**\n * WebSocket服务发生异常的回调，可以覆盖重写。\n * 默认抛出异常\n *\n * @param t 异常\n * @param response 返回值\n */\n public void onWebSocketError(Throwable t, Response response) {\n log.error(\"调用星火模型时，WebSocket发生异常:{}\", response);\n t.printStackTrace();\n }\n\n /**\n * 星火大模型发生异常\n *\n * @param resp 大模型返回值\n */\n public abstract void onChatError(RESP resp);\n\n /**\n * 星火大模型正常返回信息\n *\n * @param resp 大模型返回值\n */\n public abstract void onChatOutput(RESP resp);\n\n /**\n * 星火大模型返回信息结束回调\n */\n public abstract void onChatEnd();\n\n @Override\n public void onClosed(@NotNull WebSocket webSocket, int code, @NotNull String reason) {\n super.onClosed(webSocket, code, reason);\n }\n\n @Override\n public void onClosing(@NotNull WebSocket webSocket, int code, @NotNull String reason) {\n super.onClosing(webSocket, code, reason);\n }\n\n @Override\n public void onFailure(@NotNull WebSocket webSocket, @NotNull Throwable t, @Nullable Response response) {\n webSocket.close(1000, \"\");\n this.onWebSocketError(t, response);\n }\n\n @Override\n public abstract void onMessage(@NotNull WebSocket webSocket, @NotNull String text);\n\n @Override\n public void onMessage(@NotNull WebSocket webSocket, @NotNull ByteString bytes) {\n super.onMessage(webSocket, bytes);\n }\n\n @Override\n public void onOpen(@NotNull WebSocket webSocket, @NotNull Response response) {\n super.onOpen(webSocket, response);\n }\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/achieve/defaults/listener/ChatListener.java",
"content": "package com.ai.spark.achieve.defaults.listener;\n\nimport com.ai.common.utils.JsonUtils;\nimport com.ai.spark.common.Usage;\nimport com.ai.spark.endPoint.chat.ChatHeader;\nimport com.ai.spark.endPoint.chat.Choice;\nimport com.ai.spark.endPoint.chat.req.ChatRequest;\nimport com.ai.spark.endPoint.chat.resp.ChatResponse;\nimport lombok.Data;\nimport lombok.extern.slf4j.Slf4j;\nimport okhttp3.Response;\nimport okhttp3.WebSocket;\nimport okio.ByteString;\nimport org.jetbrains.annotations.NotNull;\nimport org.jetbrains.annotations.Nullable;\n\nimport static com.ai.common.utils.ValidationUtils.ensureNotNull;\n\n/**\n * 星火大模型流式返回监听器\n */\n@Slf4j\n@Data\npublic abstract class ChatListener extends BaseListener {\n\n private ChatRequest chatRequest;\n\n public ChatListener(ChatRequest chatRequest) {\n this.chatRequest = ensureNotNull(chatRequest, \"chatRequest\");\n }\n\n /**\n * WebSocket服务发生异常的回调，可以覆盖重写。\n * 默认抛出异常\n *\n * @param t 异常\n * @param response 返回值\n */\n public void onWebSocketError(Throwable t, Response response) {\n log.error(\"调用星火模型时，WebSocket发生异常:{}\", response);\n t.printStackTrace();\n }\n\n /**\n * 星火大模型发生异常\n *\n * @param chatResponse 大模型返回值\n */\n public abstract void onChatError(ChatResponse chatResponse);\n\n /**\n * 星火大模型正常返回信息\n *\n * @param chatResponse 大模型返回值\n */\n public abstract void onChatOutput(ChatResponse chatResponse);\n\n /**\n * 星火大模型返回信息结束回调\n */\n public abstract void onChatEnd();\n\n @Override\n public void onClosed(@NotNull WebSocket webSocket, int code, @NotNull String reason) {\n super.onClosed(webSocket, code, reason);\n }\n\n @Override\n public void onClosing(@NotNull WebSocket webSocket, int code, @NotNull String reason) {\n super.onClosing(webSocket, code, reason);\n }\n\n @Override\n public void onFailure(@NotNull WebSocket webSocket, @NotNull Throwable t, @Nullable Response response) {\n webSocket.close(1000, \"\");\n this.onWebSocketError(t, response);\n }\n\n @Override\n public void onMessage(@NotNull WebSocket webSocket, @NotNull String text) {\n ChatResponse chatResponse = JsonUtils.fromJson(text, ChatResponse.class);\n\n if (ChatHeader.Code.SUCCESS.getValue() != chatResponse.getChatHeader().getCode()) {\n log.warn(\"调用星火模型发生错误，错误码为：{}，请求的sid为：{}\", chatResponse.getChatHeader().getCode(), chatResponse.getChatHeader().getSid());\n webSocket.close(1000, \"星火模型调用异常\");\n this.onChatError(chatResponse);\n return;\n }\n\n this.onChatOutput(chatResponse);\n\n if (Choice.Status.END.getValue() == chatResponse.getChatHeader().getStatus()) {\n // 可以关闭连接，释放资源\n webSocket.close(1000, \"星火模型返回结束\");\n Usage usage = chatResponse.getChatPayload().getUsage();\n this.onChatEnd();\n }\n }\n\n @Override\n public void onMessage(@NotNull WebSocket webSocket, @NotNull ByteString bytes) {\n super.onMessage(webSocket, bytes);\n }\n\n @Override\n public void onOpen(@NotNull WebSocket webSocket, @NotNull Response response) {\n super.onOpen(webSocket, response);\n webSocket.send(JsonUtils.toJson(chatRequest));\n }\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/achieve/defaults/listener/DocumentChatListener.java",
"content": "package com.ai.spark.achieve.defaults.listener;\n\nimport com.ai.common.utils.JsonUtils;\nimport com.ai.spark.endPoint.chat.req.DocumentChatRequest;\nimport com.ai.spark.endPoint.chat.resp.DocumentChatResponse;\nimport lombok.Data;\nimport lombok.extern.slf4j.Slf4j;\nimport okhttp3.Response;\nimport okhttp3.WebSocket;\nimport okio.ByteString;\nimport org.jetbrains.annotations.NotNull;\nimport org.jetbrains.annotations.Nullable;\n\nimport static com.ai.common.utils.ValidationUtils.ensureNotNull;\n\n/**\n * 星火大模型文档对话流式返回监听器\n */\n@Slf4j\n@Data\npublic abstract class DocumentChatListener extends BaseListener {\n\n private DocumentChatRequest documentChatRequest;\n\n /**\n * 构造方法，传入大模型参数\n *\n * @param documentChatRequest 大模型参数\n */\n public DocumentChatListener(DocumentChatRequest documentChatRequest) {\n this.documentChatRequest = ensureNotNull(documentChatRequest, \"documentChatRequest\");\n }\n\n /**\n * WebSocket服务发生异常的回调，可以覆盖重写。\n * 默认抛出异常\n *\n * @param t 异常\n * @param response 返回值\n */\n public void onWebSocketError(Throwable t, Response response) {\n log.error(\"调用星火模型时，WebSocket发生异常:{}\", response);\n t.printStackTrace();\n }\n\n /**\n * 星火大模型发生异常\n *\n * @param documentChatResponse 大模型返回值\n */\n public abstract void onChatError(DocumentChatResponse documentChatResponse);\n\n /**\n * 星火大模型正常返回信息\n *\n * @param documentChatResponse 大模型返回值\n */\n public abstract void onChatOutput(DocumentChatResponse documentChatResponse);\n\n /**\n * 星火大模型返回信息结束回调\n */\n public abstract void onChatEnd();\n\n @Override\n public void onClosed(@NotNull WebSocket webSocket, int code, @NotNull String reason) {\n super.onClosed(webSocket, code, reason);\n }\n\n @Override\n public void onClosing(@NotNull WebSocket webSocket, int code, @NotNull String reason) {\n super.onClosing(webSocket, code, reason);\n }\n\n @Override\n public void onFailure(@NotNull WebSocket webSocket, @NotNull Throwable t, @Nullable Response response) {\n webSocket.close(1000, \"\");\n this.onWebSocketError(t, response);\n }\n\n @Override\n public void onMessage(@NotNull WebSocket webSocket, @NotNull String text) {\n DocumentChatResponse documentChatResponse = JsonUtils.fromJson(text, DocumentChatResponse.class);\n\n if (DocumentChatResponse.Code.SUCCESS.getValue() != documentChatResponse.getCode()) {\n log.warn(\"调用星火模型文档对话发生错误，错误码为：{}，请求的sid为：{}\", documentChatResponse.getCode(), documentChatResponse.getSid());\n webSocket.close(1000, \"星火模型调用异常\");\n this.onChatError(documentChatResponse);\n return;\n }\n\n this.onChatOutput(documentChatResponse);\n\n if (DocumentChatResponse.Status.END.getValue() == documentChatResponse.getStatus()) {\n // 可以关闭连接，释放资源\n webSocket.close(1000, \"星火模型返回结束\");\n this.onChatEnd();\n }\n }\n\n @Override\n public void onMessage(@NotNull WebSocket webSocket, @NotNull ByteString bytes) {\n super.onMessage(webSocket, bytes);\n }\n\n @Override\n public void onOpen(@NotNull WebSocket webSocket, @NotNull Response response) {\n super.onOpen(webSocket, response);\n webSocket.send(JsonUtils.toJson(documentChatRequest));\n }\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/achieve/defaults/listener/ImageUnderstandingListener.java",
"content": "package com.ai.spark.achieve.defaults.listener;\n\nimport com.ai.common.utils.JsonUtils;\nimport com.ai.spark.endPoint.chat.ChatHeader;\nimport com.ai.spark.endPoint.chat.Choice;\nimport com.ai.spark.endPoint.chat.resp.DocumentChatResponse;\nimport com.ai.spark.endPoint.images.ImageHeader;\nimport com.ai.spark.endPoint.images.req.ImageUnderstandingRequest;\nimport com.ai.spark.endPoint.images.resp.ImageUnderstandingResponse;\nimport lombok.Data;\nimport lombok.extern.slf4j.Slf4j;\nimport okhttp3.Response;\nimport okhttp3.WebSocket;\nimport okio.ByteString;\nimport org.jetbrains.annotations.NotNull;\nimport org.jetbrains.annotations.Nullable;\n\nimport static com.ai.common.utils.ValidationUtils.ensureNotNull;\n\n/**\n * 图片理解对话监听器\n */\n@Slf4j\n@Data\npublic abstract class ImageUnderstandingListener extends BaseListener {\n\n private ImageUnderstandingRequest imageUnderstandingRequest;\n\n /**\n * 构造方法，传入大模型参数\n *\n * @param imageUnderstandingRequest 大模型参数\n */\n public ImageUnderstandingListener(ImageUnderstandingRequest imageUnderstandingRequest) {\n this.imageUnderstandingRequest = ensureNotNull(imageUnderstandingRequest, \"documentChatRequest\");\n }\n\n /**\n * WebSocket服务发生异常的回调，可以覆盖重写。\n * 默认抛出异常\n *\n * @param t 异常\n * @param response 返回值\n */\n public void onWebSocketError(Throwable t, Response response) {\n log.error(\"调用星火模型时，WebSocket发生异常:{}\", response);\n t.printStackTrace();\n }\n\n /**\n * 星火大模型发生异常\n *\n * @param imageUnderstandingResponse 大模型返回值\n */\n public abstract void onChatError(ImageUnderstandingResponse imageUnderstandingResponse);\n\n /**\n * 星火大模型正常返回信息\n *\n * @param imageUnderstandingResponse 大模型返回值\n */\n public abstract void onChatOutput(ImageUnderstandingResponse imageUnderstandingResponse);\n\n /**\n * 星火大模型返回信息结束回调\n */\n public abstract void onChatEnd();\n\n @Override\n public void onClosed(@NotNull WebSocket webSocket, int code, @NotNull String reason) {\n super.onClosed(webSocket, code, reason);\n }\n\n @Override\n public void onClosing(@NotNull WebSocket webSocket, int code, @NotNull String reason) {\n super.onClosing(webSocket, code, reason);\n }\n\n @Override\n public void onFailure(@NotNull WebSocket webSocket, @NotNull Throwable t, @Nullable Response response) {\n webSocket.close(1000, \"\");\n this.onWebSocketError(t, response);\n }\n\n @Override\n public void onMessage(@NotNull WebSocket webSocket, @NotNull String text) {\n ImageUnderstandingResponse imageUnderstandingResponse = JsonUtils.fromJson(text, ImageUnderstandingResponse.class);\n\n ImageHeader imageHeader = imageUnderstandingResponse.getImageHeader();\n if (imageHeader.getCode() != ChatHeader.Code.SUCCESS.getValue()) {\n log.warn(\"调用星火模型文档对话发生错误，错误码为：{}，请求的sid为：{}\", imageHeader.getCode(), imageHeader.getSid());\n webSocket.close(1000, \"星火模型调用异常\");\n this.onChatError(imageUnderstandingResponse);\n return;\n }\n\n this.onChatOutput(imageUnderstandingResponse);\n\n if (DocumentChatResponse.Status.END.getValue() == Choice.Status.END.getValue()) {\n // 可以关闭连接，释放资源\n webSocket.close(1000, \"星火模型返回结束\");\n this.onChatEnd();\n }\n }\n\n @Override\n public void onMessage(@NotNull WebSocket webSocket, @NotNull ByteString bytes) {\n super.onMessage(webSocket, bytes);\n }\n\n @Override\n public void onOpen(@NotNull WebSocket webSocket, @NotNull Response response) {\n super.onOpen(webSocket, response);\n webSocket.send(JsonUtils.toJson(imageUnderstandingRequest));\n }\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/achieve/defaults/session/DefaultAggregationSession.java",
"content": "package com.ai.spark.achieve.defaults.session;\n\nimport com.ai.spark.achieve.Configuration;\nimport com.ai.spark.achieve.standard.session.*;\n\nimport static com.ai.common.utils.ValidationUtils.ensureNotNull;\n\n/**\n * 聚合各个场景的session\n */\npublic class DefaultAggregationSession implements AggregationSession {\n\n private Configuration configuration;\n\n private volatile ChatSession chatSession;\n\n private volatile DocumentSession documentSession;\n\n private volatile EmbeddingSession embeddingSession;\n\n private volatile ImageSession imageSession;\n\n private volatile AudioSession audioSession;\n\n public DefaultAggregationSession(Configuration configuration) {\n this.configuration = ensureNotNull(configuration, \"configuration\");\n }\n\n @Override\n public ChatSession getChatSession() {\n if (chatSession == null) {\n synchronized (this) {\n if (chatSession == null) {\n chatSession = new DefaultChatSession(configuration);\n }\n }\n }\n return chatSession;\n }\n\n @Override\n public DocumentSession getDocumentSession() {\n if (documentSession == null) {\n synchronized (this) {\n if (documentSession == null) {\n documentSession = new DefaultDocumentSession(configuration);\n }\n }\n }\n return documentSession;\n }\n\n @Override\n public EmbeddingSession getEmbeddingSession() {\n if (embeddingSession == null) {\n synchronized (this) {\n if (embeddingSession == null) {\n embeddingSession = new DefaultEmbeddingSession(configuration);\n }\n }\n }\n return embeddingSession;\n }\n\n @Override\n public ImageSession getImageSession() {\n if (imageSession == null) {\n synchronized (this) {\n if (imageSession == null) {\n imageSession = new DefaultImageSession(configuration);\n }\n }\n }\n return imageSession;\n }\n\n @Override\n public AudioSession getAudioSession() {\n if (audioSession == null) {\n synchronized (this) {\n if (audioSession == null) {\n audioSession = new DefaultAudioSession(configuration);\n }\n }\n }\n return audioSession;\n }\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/achieve/defaults/session/DefaultAudioSession.java",
"content": "package com.ai.spark.achieve.defaults.session;\n\nimport com.ai.spark.achieve.Configuration;\nimport com.ai.spark.achieve.standard.session.AudioSession;\nimport lombok.Data;\nimport lombok.EqualsAndHashCode;\nimport lombok.ToString;\n\nimport static com.ai.common.utils.ValidationUtils.ensureNotNull;\n\n@Data\n@ToString(callSuper = true)\n@EqualsAndHashCode(callSuper = true)\npublic class DefaultAudioSession extends Session implements AudioSession {\n\n public DefaultAudioSession(Configuration configuration) {\n this.setConfiguration(ensureNotNull(configuration, \"configuration\"));\n this.setSparkApiServer(ensureNotNull(configuration.getSparkApiServer(), \"sparkApiServer\"));\n }\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/achieve/defaults/session/DefaultChatSession.java",
"content": "package com.ai.spark.achieve.defaults.session;\n\n\nimport cn.hutool.core.date.DateUtil;\nimport com.ai.spark.achieve.ApiData;\nimport com.ai.spark.achieve.Configuration;\nimport com.ai.spark.achieve.defaults.listener.ChatListener;\nimport com.ai.spark.achieve.defaults.listener.DocumentChatListener;\nimport com.ai.spark.achieve.standard.session.ChatSession;\nimport com.ai.spark.common.SparkApiUrl;\nimport com.ai.spark.common.utils.AuthUtils;\nimport lombok.Data;\nimport lombok.EqualsAndHashCode;\nimport lombok.SneakyThrows;\nimport lombok.ToString;\nimport okhttp3.Request;\nimport okhttp3.WebSocket;\n\nimport static com.ai.common.utils.ValidationUtils.ensureNotNull;\nimport static com.ai.spark.common.SparkApiUrl.DOCUMENT_CHAT;\n\n@Data\n@ToString(callSuper = true)\n@EqualsAndHashCode(callSuper = true)\npublic class DefaultChatSession extends Session implements ChatSession {\n\n public DefaultChatSession(Configuration configuration) {\n this.setConfiguration(ensureNotNull(configuration, \"configuration\"));\n this.setSparkApiServer(ensureNotNull(configuration.getSparkApiServer(), \"sparkApiServer\"));\n }\n\n @Override\n @SneakyThrows\n public WebSocket chat(T chatListener) {\n // 默认情况下根据apiData获取策略得到创建时设置的参数\n ApiData apiData = this.getConfiguration().getSystemApiData();\n return this.chat(apiData.getApiKey(), apiData.getApiSecret(), chatListener);\n }\n\n @Override\n @SneakyThrows\n public WebSocket chat(String apiKey, String apiSecret, T chatListener) {\n // 获取到对应访问的domain，根据domain获取对应的请求地址\n String domain = chatListener.getChatRequest().getChatParameter().getChat().getDomain();\n // 生成请求的URL\n String url = AuthUtils.replaceAllHttp(\n AuthUtils.getAuthUrl(AuthUtils.RequestMethod.GET.getMethod(), SparkApiUrl.getUrl(domain), apiKey, apiSecret)\n );\n // 发起请求返回结果\n return this.getConfiguration().getOkHttpClient().newWebSocket(new Request.Builder().url(url).build(), chatListener);\n }\n\n @Override\n public WebSocket documentChat(T documentChatListener) {\n ApiData apiData = this.getConfiguration().getSystemApiData();\n return documentChat(apiData.getAppId(), apiData.getApiSecret(), documentChatListener);\n }\n\n @Override\n public WebSocket documentChat(String appId, String apiSecret, T documentChatListener) {\n // 得当当前时间戳，按秒计算\n long ts = DateUtil.currentSeconds();\n // 进行签名设置\n String url = SparkApiUrl.getUrl(DOCUMENT_CHAT) + \"?\"\n + \"appId=\" + appId\n + \"×tamp=\" + ts\n + \"&signature=\" + AuthUtils.getSignature(appId, apiSecret, ts);\n // 发起请求返回结果\n return this.getConfiguration().getOkHttpClient().newWebSocket(new Request.Builder().url(url).build(), documentChatListener);\n }\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/achieve/defaults/session/DefaultDocumentSession.java",
"content": "package com.ai.spark.achieve.defaults.session;\n\nimport cn.hutool.core.date.DateUtil;\nimport cn.hutool.core.util.StrUtil;\nimport com.ai.spark.achieve.ApiData;\nimport com.ai.spark.achieve.Configuration;\nimport com.ai.spark.achieve.standard.session.DocumentSession;\nimport com.ai.spark.common.SparkApiUrl;\nimport com.ai.spark.common.utils.AuthUtils;\nimport com.ai.spark.endPoint.document.req.FileUploadRequest;\nimport com.ai.spark.endPoint.document.resp.DocumentSummaryResponse;\nimport com.ai.spark.endPoint.document.resp.FileUploadResponse;\nimport lombok.Data;\nimport lombok.EqualsAndHashCode;\nimport lombok.ToString;\nimport okhttp3.MediaType;\nimport okhttp3.MultipartBody;\nimport okhttp3.RequestBody;\n\nimport java.util.HashMap;\nimport java.util.Map;\n\nimport static com.ai.common.utils.ValidationUtils.ensureNotNull;\n\n@Data\n@ToString(callSuper = true)\n@EqualsAndHashCode(callSuper = true)\npublic class DefaultDocumentSession extends Session implements DocumentSession {\n\n public DefaultDocumentSession(Configuration configuration) {\n this.setConfiguration(ensureNotNull(configuration, \"configuration\"));\n this.setSparkApiServer(ensureNotNull(configuration.getSparkApiServer(), \"sparkApiServer\"));\n }\n\n @Override\n public FileUploadResponse fileUpload(FileUploadRequest fileUploadRequest) {\n ApiData apiData = this.getConfiguration().getSystemApiData();\n return this.fileUpload(apiData.getAppId(), apiData.getApiSecret(), fileUploadRequest);\n }\n\n @Override\n public FileUploadResponse fileUpload(String appId, String apiSecret, FileUploadRequest fileUploadRequest) {\n // 得到当前时间戳，按秒计算\n long ts = DateUtil.currentSeconds();\n // 设置文件\n RequestBody fileBody = RequestBody.create(MediaType.parse(\"multipart/form-data\"), fileUploadRequest.getFile());\n MultipartBody.Part multipartBody = MultipartBody.Part.createFormData(\"file\", fileUploadRequest.getFile().getName(), fileBody);\n // 设置其余参数\n Map requestBodyMap = new HashMap<>();\n if (StrUtil.isNotBlank(fileUploadRequest.getUrl())) {\n requestBodyMap.put(FileUploadRequest.Fields.url, RequestBody.create(MediaType.parse(\"multipart/form-data\"), fileUploadRequest.getUrl()));\n }\n if (StrUtil.isNotBlank(fileUploadRequest.getFileName())) {\n requestBodyMap.put(FileUploadRequest.Fields.fileName, RequestBody.create(MediaType.parse(\"multipart/form-data\"), fileUploadRequest.getFileName()));\n }\n if (StrUtil.isNotBlank(fileUploadRequest.getFileType())) {\n requestBodyMap.put(FileUploadRequest.Fields.fileType, RequestBody.create(MediaType.parse(\"multipart/form-data\"), fileUploadRequest.getFileType()));\n }\n if (StrUtil.isNotBlank(fileUploadRequest.getCallbackUrl())) {\n requestBodyMap.put(FileUploadRequest.Fields.callbackUrl, RequestBody.create(MediaType.parse(\"multipart/form-data\"), fileUploadRequest.getCallbackUrl()));\n }\n // 发起请求返回结果\n return this.getSparkApiServer().fileUpload(appId, String.valueOf(ts), AuthUtils.getSignature(appId, apiSecret, ts), multipartBody, requestBodyMap).blockingGet();\n }\n\n @Override\n public DocumentSummaryResponse documentSummaryStart(String fileId) {\n ApiData apiData = this.getConfiguration().getSystemApiData();\n return this.documentSummaryStart(apiData.getAppId(), apiData.getApiSecret(), fileId);\n }\n\n @Override\n public DocumentSummaryResponse documentSummaryStart(String appId, String apiSecret, String fileId) {\n return this.documentSummary(SparkApiUrl.ApiUrl.documentSummaryStart.getUrl(), appId, apiSecret, fileId);\n }\n\n @Override\n public DocumentSummaryResponse documentSummaryQuery(String fileId) {\n ApiData apiData = this.getConfiguration().getSystemApiData();\n return this.documentSummaryQuery(apiData.getAppId(), apiData.getApiSecret(), fileId);\n }\n\n @Override\n public DocumentSummaryResponse documentSummaryQuery(String appId, String apiSecret, String fileId) {\n return this.documentSummary(SparkApiUrl.ApiUrl.documentSummaryQuery.getUrl(), appId, apiSecret, fileId);\n }\n\n /**\n * 文档总结底层都依赖这个方法\n *\n * @param url 请求的URL\n * @param appId 用户的AppId\n * @param apiSecret 用户的ApiSecret\n * @param fileId 文件ID\n * @return 请求结果\n */\n private DocumentSummaryResponse documentSummary(String url, String appId, String apiSecret, String fileId) {\n long ts = DateUtil.currentSeconds();\n return this.getSparkApiServer().documentSummary(url, appId, String.valueOf(ts), AuthUtils.getSignature(appId, apiSecret, ts), RequestBody.create(MediaType.parse(\"multipart/form-data\"), fileId)).blockingGet();\n }\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/achieve/defaults/session/DefaultEmbeddingSession.java",
"content": "package com.ai.spark.achieve.defaults.session;\n\nimport cn.hutool.http.ContentType;\nimport com.ai.common.utils.JsonUtils;\nimport com.ai.spark.achieve.ApiData;\nimport com.ai.spark.achieve.Configuration;\nimport com.ai.spark.achieve.standard.session.EmbeddingSession;\nimport com.ai.spark.common.SparkApiUrl;\nimport com.ai.spark.common.utils.AuthUtils;\nimport com.ai.spark.endPoint.embedding.req.EmbeddingRequest;\nimport com.ai.spark.endPoint.embedding.resp.EmbeddingResponse;\nimport lombok.Data;\nimport lombok.EqualsAndHashCode;\nimport lombok.SneakyThrows;\nimport lombok.ToString;\nimport okhttp3.MediaType;\nimport okhttp3.Request;\nimport okhttp3.RequestBody;\n\nimport static com.ai.common.utils.ValidationUtils.ensureNotNull;\n\n@Data\n@ToString(callSuper = true)\n@EqualsAndHashCode(callSuper = true)\npublic class DefaultEmbeddingSession extends Session implements EmbeddingSession {\n\n\n public DefaultEmbeddingSession(Configuration configuration) {\n this.setConfiguration(ensureNotNull(configuration, \"configuration\"));\n this.setSparkApiServer(ensureNotNull(configuration.getSparkApiServer(), \"sparkApiServer\"));\n }\n\n @Override\n public EmbeddingResponse embed(EmbeddingRequest embeddingRequest) {\n ApiData apiData = this.getConfiguration().getSystemApiData();\n return this.embed(apiData.getApiKey(), apiData.getApiSecret(), embeddingRequest);\n }\n\n @Override\n @SneakyThrows\n public EmbeddingResponse embed(String apiKey, String apiSecret, EmbeddingRequest embeddingRequest) {\n // 鉴权，得到请求路径\n String authUrl = AuthUtils.getAuthUrl(AuthUtils.RequestMethod.POST.getMethod(), SparkApiUrl.ApiUrl.embeddingq.getUrl(), apiKey, apiSecret);\n // 创建请求，设置请求URL和json数据\n Request request = new Request.Builder().url(authUrl).post(RequestBody.create(MediaType.parse(ContentType.JSON.getValue()), JsonUtils.toJson(embeddingRequest))).build();\n // 发起请求，获取返回的json字符串\n String response = this.getConfiguration().getOkHttpClient().newCall(request).execute().body().string();\n // 将json映射到对象上\n return JsonUtils.fromJson(response, EmbeddingResponse.class);\n }\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/achieve/defaults/session/DefaultImageSession.java",
"content": "package com.ai.spark.achieve.defaults.session;\n\nimport cn.hutool.http.ContentType;\nimport com.ai.common.utils.JsonUtils;\nimport com.ai.spark.achieve.ApiData;\nimport com.ai.spark.achieve.Configuration;\nimport com.ai.spark.achieve.defaults.listener.ImageUnderstandingListener;\nimport com.ai.spark.achieve.standard.session.ImageSession;\nimport com.ai.spark.common.SparkApiUrl;\nimport com.ai.spark.common.utils.AuthUtils;\nimport com.ai.spark.endPoint.images.req.ImageCreateRequest;\nimport com.ai.spark.endPoint.images.req.ImageUnderstandingRequest;\nimport com.ai.spark.endPoint.images.resp.ImageCreateResponse;\nimport lombok.Data;\nimport lombok.EqualsAndHashCode;\nimport lombok.SneakyThrows;\nimport lombok.ToString;\nimport okhttp3.MediaType;\nimport okhttp3.Request;\nimport okhttp3.RequestBody;\nimport okhttp3.WebSocket;\n\nimport static com.ai.common.utils.ValidationUtils.ensureNotNull;\n\n@Data\n@ToString(callSuper = true)\n@EqualsAndHashCode(callSuper = true)\npublic class DefaultImageSession extends Session implements ImageSession {\n\n public DefaultImageSession(Configuration configuration) {\n this.setConfiguration(ensureNotNull(configuration, \"configuration\"));\n this.setSparkApiServer(ensureNotNull(configuration.getSparkApiServer(), \"sparkApiServer\"));\n }\n\n @Override\n @SneakyThrows\n public ImageCreateResponse imageCreate(ImageCreateRequest imageCreateRequest) {\n ApiData apiData = this.getConfiguration().getSystemApiData();\n return this.imageCreate(apiData.getApiKey(), apiData.getApiSecret(), imageCreateRequest);\n }\n\n @Override\n @SneakyThrows\n public ImageCreateResponse imageCreate(String apiKey, String apiSecret, ImageCreateRequest imageCreateRequest) {\n // 鉴权，得到请求路径\n String authUrl = AuthUtils.getAuthUrl(AuthUtils.RequestMethod.POST.getMethod(), SparkApiUrl.ApiUrl.imageCreate.getUrl(), apiKey, apiSecret);\n // 创建请求，设置请求URL和json数据\n Request request = new Request.Builder().url(authUrl).post(RequestBody.create(MediaType.parse(ContentType.JSON.getValue()), JsonUtils.toJson(imageCreateRequest))).build();\n // 发起请求，获取返回的json字符串\n String response = this.getConfiguration().getOkHttpClient().newCall(request).execute().body().string();\n // 将json映射到对象上\n return JsonUtils.fromJson(response, ImageCreateResponse.class);\n }\n\n @Override\n public WebSocket imageUnderstanding(ImageUnderstandingRequest imageUnderstandingRequest, ImageUnderstandingListener imageUnderstandingListener) {\n ApiData apiData = this.getConfiguration().getSystemApiData();\n return imageUnderstanding(apiData.getApiKey(), apiData.getApiSecret(), imageUnderstandingRequest, imageUnderstandingListener);\n }\n\n @Override\n @SneakyThrows\n public WebSocket imageUnderstanding(String apiKey, String apiSecret, ImageUnderstandingRequest imageUnderstandingRequest, ImageUnderstandingListener imageUnderstandingListener) {\n // 生成请求的URL\n String url = AuthUtils.replaceAllHttp(\n AuthUtils.getAuthUrl(AuthUtils.RequestMethod.GET.getMethod(), SparkApiUrl.ApiUrl.imageUnderstanding.getUrl(), apiKey, apiSecret)\n );\n // 发起请求返回结果\n return this.getConfiguration().getOkHttpClient().newWebSocket(new Request.Builder().url(url).build(), imageUnderstandingListener);\n }\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/achieve/defaults/session/Session.java",
"content": "package com.ai.spark.achieve.defaults.session;\n\nimport com.ai.spark.achieve.Configuration;\nimport com.ai.spark.achieve.standard.api.SparkApiServer;\nimport lombok.Data;\n\n@Data\npublic class Session {\n\n private Configuration configuration;\n\n private SparkApiServer sparkApiServer;\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/achieve/standard/api/SparkApiServer.java",
"content": "package com.ai.spark.achieve.standard.api;\n\n\nimport com.ai.spark.endPoint.document.resp.DocumentSummaryResponse;\nimport com.ai.spark.endPoint.document.resp.FileUploadResponse;\nimport io.reactivex.Single;\nimport okhttp3.MultipartBody;\nimport okhttp3.RequestBody;\nimport retrofit2.http.*;\n\nimport java.util.Map;\n\nimport static com.ai.spark.common.Constants.*;\nimport static com.ai.spark.common.SparkApiUrl.FILE_UPLOAD_API_URL;\n\n/**\n * @Description: 讯飞星火 API接口\n */\npublic interface SparkApiServer {\n\n /**\n * 文件上传接口\n *\n * @param appId 用户appId\n * @param timestamp 当前时间戳（秒）\n * @param signature 生成的签名\n * @param file 需要上传的文件\n * @param requestBodyMap 其他字段\n * @return 文件上传返回参数\n */\n @Multipart\n @POST(FILE_UPLOAD_API_URL)\n Single fileUpload(@Header(APP_ID) String appId, @Header(TIMESTAMP) String timestamp, @Header(SIGNATURE) String signature, @Part MultipartBody.Part file, @PartMap Map requestBodyMap);\n\n /**\n * 文档总结，文档总结结果查询两个接口二合一，参数都一样，请求路径不同\n *\n * @param url 请求地址\n * @param appId 用户appId\n * @param timestamp 当前时间戳（秒）\n * @param signature 生成的签名\n * @param fileId 文件ID\n * @return 请求结果信息\n */\n @Multipart\n @POST\n Single documentSummary(@Url String url, @Header(APP_ID) String appId, @Header(TIMESTAMP) String timestamp, @Header(SIGNATURE) String signature, @Part(\"fileId\") RequestBody fileId);\n\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/achieve/standard/session/AggregationSession.java",
"content": "package com.ai.spark.achieve.standard.session;\n\n\npublic interface AggregationSession {\n\n ChatSession getChatSession();\n\n DocumentSession getDocumentSession();\n\n EmbeddingSession getEmbeddingSession();\n\n ImageSession getImageSession();\n\n AudioSession getAudioSession();\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/achieve/standard/session/AudioSession.java",
"content": "package com.ai.spark.achieve.standard.session;\n\n\npublic interface AudioSession {\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/achieve/standard/session/ChatSession.java",
"content": "package com.ai.spark.achieve.standard.session;\n\n\nimport com.ai.spark.achieve.defaults.listener.ChatListener;\nimport com.ai.spark.achieve.defaults.listener.DocumentChatListener;\nimport okhttp3.WebSocket;\n\n/**\n * 对话场景下的接口\n */\npublic interface ChatSession {\n\n /**\n * 聊天接口\n *\n * @param chatListener 对话监听器\n * @return 返回信息\n */\n WebSocket chat(T chatListener);\n\n /**\n * 聊天接口，自定义所使用的到的ApiKey和ApiSecret\n *\n * @param apiKey 用户的ApiKey\n * @param apiSecret 用户的ApiSecret\n * @param chatListener 对话监听器\n * @return 返回信息\n */\n WebSocket chat(String apiKey, String apiSecret, T chatListener);\n\n /**\n * 文档聊天接口\n *\n * @param documentChatListener 对话监听器\n * @return 返回信息\n */\n WebSocket documentChat(T documentChatListener);\n\n /**\n * 聊天接口，自定义所使用的到的AppId和ApiSecret\n *\n * @param appId 用户的AppId\n * @param apiSecret 用户的ApiSecret\n * @param documentChatListener 对话监听器\n * @return 返回信息\n */\n WebSocket documentChat(String appId, String apiSecret, T documentChatListener);\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/achieve/standard/session/DocumentSession.java",
"content": "package com.ai.spark.achieve.standard.session;\n\n\nimport com.ai.spark.endPoint.document.req.FileUploadRequest;\nimport com.ai.spark.endPoint.document.resp.DocumentSummaryResponse;\nimport com.ai.spark.endPoint.document.resp.FileUploadResponse;\n\n/**\n * 文档场景下的接口\n */\npublic interface DocumentSession {\n\n /**\n * 文件上传\n *\n * @param fileUploadRequest 上传的信息\n * @return 返回信息\n */\n FileUploadResponse fileUpload(FileUploadRequest fileUploadRequest);\n\n /**\n * 文件上传\n *\n * @param appId 用户的AppId\n * @param apiSecret 用户的ApiSecret\n * @param fileUploadRequest 上传的信息\n * @return 返回信息\n */\n FileUploadResponse fileUpload(String appId, String apiSecret, FileUploadRequest fileUploadRequest);\n\n /**\n * 发起文档总结接口\n *\n * @param fileId 文件ID\n * @return 返回信息\n */\n DocumentSummaryResponse documentSummaryStart(String fileId);\n\n /**\n * 发起文档总结接口\n *\n * @param appId 用户的AppId\n * @param apiSecret 用户的ApiSecret\n * @param fileId 文件ID\n * @return 返回信息\n */\n DocumentSummaryResponse documentSummaryStart(String appId, String apiSecret, String fileId);\n\n /**\n * 查询文档总结结果\n *\n * @param fileId 文件ID\n * @return 返回信息\n */\n DocumentSummaryResponse documentSummaryQuery(String fileId);\n\n /**\n * 查询文档总结结果\n *\n * @param appId 用户的AppId\n * @param apiSecret 用户的ApiSecret\n * @param fileId 文件ID\n * @return 返回信息\n */\n DocumentSummaryResponse documentSummaryQuery(String appId, String apiSecret, String fileId);\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/achieve/standard/session/EmbeddingSession.java",
"content": "package com.ai.spark.achieve.standard.session;\n\n\nimport com.ai.spark.endPoint.embedding.req.EmbeddingRequest;\nimport com.ai.spark.endPoint.embedding.resp.EmbeddingResponse;\n\n/**\n * 文本嵌入场景下的接口\n */\npublic interface EmbeddingSession {\n\n /**\n * 文本嵌入\n *\n * @param embeddingRequest 请求参数\n * @return 请求结果\n */\n EmbeddingResponse embed(EmbeddingRequest embeddingRequest);\n\n /**\n * 文本嵌入\n *\n * @param apiKey 用户的ApiKey\n * @param apiSecret 用户的ApiSecret\n * @param embeddingRequest 请求参数\n * @return 请求结果\n */\n EmbeddingResponse embed(String apiKey, String apiSecret, EmbeddingRequest embeddingRequest);\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/achieve/standard/session/ImageSession.java",
"content": "package com.ai.spark.achieve.standard.session;\n\n\nimport com.ai.spark.achieve.defaults.listener.ImageUnderstandingListener;\nimport com.ai.spark.endPoint.images.req.ImageCreateRequest;\nimport com.ai.spark.endPoint.images.req.ImageUnderstandingRequest;\nimport com.ai.spark.endPoint.images.resp.ImageCreateResponse;\nimport okhttp3.WebSocket;\n\n/**\n * 图片生成场景下的接口\n */\npublic interface ImageSession {\n\n /**\n * 图片创作接口，使用系统默认的ApiData\n *\n * @param imageCreateRequest 请求参数\n * @return 请求结果\n */\n ImageCreateResponse imageCreate(ImageCreateRequest imageCreateRequest);\n\n /**\n * 图片创作接口，使用自定义的ApiData\n *\n * @param apiKey 用户的ApiKey\n * @param apiSecret 用户的ApiSecret\n * @param imageCreateRequest 请求参数\n * @return 请求结果\n */\n ImageCreateResponse imageCreate(String apiKey, String apiSecret, ImageCreateRequest imageCreateRequest);\n\n /**\n * 图片理解\n *\n * @param imageUnderstandingRequest 请求参数\n * @return 请求结果\n */\n public WebSocket imageUnderstanding(ImageUnderstandingRequest imageUnderstandingRequest, ImageUnderstandingListener imageUnderstandingListener);\n\n /**\n * 图片理解\n *\n * @param apiKey 用户的ApiKey\n * @param apiSecret 用户的ApiSecret\n * @param imageUnderstandingRequest 请求参数\n * @return 请求结果\n */\n public WebSocket imageUnderstanding(String apiKey, String apiSecret, ImageUnderstandingRequest imageUnderstandingRequest, ImageUnderstandingListener imageUnderstandingListener);\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/common/Constants.java",
"content": "package com.ai.spark.common;\n\n\npublic class Constants {\n\n public static final String APP_ID = \"appId\";\n public static final String TIMESTAMP = \"timestamp\";\n public static final String SIGNATURE = \"signature\";\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/common/SparkApiUrl.java",
"content": "package com.ai.spark.common;\n\nimport lombok.AllArgsConstructor;\nimport lombok.Getter;\nimport lombok.extern.slf4j.Slf4j;\n\nimport java.util.HashMap;\nimport java.util.Map;\n\n/**\n * 记录各个API的请求URL\n */\n@Slf4j\npublic class SparkApiUrl {\n // 星火模型对话链接\n public final static String GENERAL_V1 = \"general\";\n public final static String GENERAL_V2 = \"generalv2\";\n public final static String GENERAL_V3 = \"generalv3\";\n public final static String SPARK_API_HOST_WS_V1_1_URL = \"http://spark-api.xf-yun.com/v1.1/chat\";\n public final static String SPARK_API_HOST_WSS_V1_1_URL = \"https://spark-api.xf-yun.com/v1.1/chat\";\n public final static String SPARK_API_HOST_WS_V2_1_URL = \"http://spark-api.xf-yun.com/v2.1/chat\";\n public final static String SPARK_API_HOST_WSS_V2_1_URL = \"https://spark-api.xf-yun.com/v2.1/chat\";\n public final static String SPARK_API_HOST_WS_V3_1_URL = \"http://spark-api.xf-yun.com/v3.1/chat\";\n public final static String SPARK_API_HOST_WSS_V3_1_URL = \"https://spark-api.xf-yun.com/v3.1/chat\";\n\n // 文档对话文件上传链接\n public final static String FILE_UPLOAD = \"fileUpload\";\n public final static String FILE_UPLOAD_API_URL = \"https://chatdoc.xfyun.cn/openapi/fileUpload\";\n\n // 文档对话链接\n public final static String DOCUMENT_CHAT = \"documentChat\";\n public final static String DOCUMENT_CHAT_API_URL = \"wss://chatdoc.xfyun.cn/openapi/chat\";\n\n // 文档总结链接\n public final static String DOCUMENT_SUMMARY_START = \"documentSummaryStart\";\n public final static String DOCUMENT_SUMMARY_START_API_URL = \"https://chatdoc.xfyun.cn/openapi/startSummary\";\n\n // 查询文档总结结果链接\n public final static String DOCUMENT_SUMMARY_QUERY = \"documentSummaryStart\";\n public final static String DOCUMENT_SUMMARY_QUERY_API_URL = \"https://chatdoc.xfyun.cn/openapi/fileSummary\";\n\n // 文本嵌入接口\n public final static String EMBEDDING_P = \"Embeddingp\";\n public final static String EMBEDDING_P_API_URL = \"https://cn-huabei-1.xf-yun.com/v1/private/sa8a05c27\";\n public final static String EMBEDDING_Q = \"Embeddingq\";\n public final static String EMBEDDING_Q_API_URL = \"https://cn-huabei-1.xf-yun.com/v1/private/s50d55a16\";\n\n // 图片生成接口\n public final static String IMAGE_CREATE = \"imageCreate\";\n public final static String IMAGE_CREATE_API_URL = \"https://spark-api.cn-huabei-1.xf-yun.com/v2.1/tti\";\n\n // 图片理解接口\n public final static String IMAGE_UNDERSANDING = \"imageUnderstanding\";\n public final static String IMAGE_UNDERSANDING_API_URL = \"https://spark-api.cn-huabei-1.xf-yun.com/v2.1/image\";\n\n // 超拟人合成协议接口\n public final static String HYPERMIMETIC_SYNTHESIS = \"hypermimeticSynthesis\";\n public final static String HYPERMIMETIC_SYNTHESIS_API_URL = \"wss://cbm01.cn-huabei-1.xf-yun.com/v1/private/medd90fec\";\n\n public final static Map urlMap = new HashMap<>();\n\n static {\n urlMap.put(GENERAL_V1, SPARK_API_HOST_WSS_V1_1_URL);\n urlMap.put(GENERAL_V2, SPARK_API_HOST_WSS_V2_1_URL);\n urlMap.put(GENERAL_V3, SPARK_API_HOST_WSS_V3_1_URL);\n urlMap.put(FILE_UPLOAD, FILE_UPLOAD_API_URL);\n urlMap.put(DOCUMENT_CHAT, DOCUMENT_CHAT_API_URL);\n urlMap.put(DOCUMENT_SUMMARY_START, DOCUMENT_SUMMARY_START_API_URL);\n urlMap.put(DOCUMENT_SUMMARY_QUERY, DOCUMENT_SUMMARY_QUERY_API_URL);\n urlMap.put(EMBEDDING_P, EMBEDDING_P_API_URL);\n urlMap.put(EMBEDDING_Q, EMBEDDING_Q_API_URL);\n urlMap.put(IMAGE_CREATE, IMAGE_CREATE_API_URL);\n urlMap.put(IMAGE_UNDERSANDING, IMAGE_UNDERSANDING_API_URL);\n urlMap.put(HYPERMIMETIC_SYNTHESIS, HYPERMIMETIC_SYNTHESIS_API_URL);\n }\n\n public static String getUrl(String key) {\n if (!urlMap.containsKey(key)) {\n log.error(\"No corresponding URL path found for {}\", key);\n return null;\n }\n return urlMap.get(key);\n }\n\n @Getter\n @AllArgsConstructor\n public enum ApiUrl {\n general(SPARK_API_HOST_WSS_V1_1_URL),\n generalV2(SPARK_API_HOST_WSS_V2_1_URL),\n generalV3(SPARK_API_HOST_WSS_V3_1_URL),\n fileUpload(FILE_UPLOAD_API_URL),\n documentChat(DOCUMENT_CHAT_API_URL),\n documentSummaryStart(DOCUMENT_SUMMARY_START_API_URL),\n documentSummaryQuery(DOCUMENT_SUMMARY_QUERY_API_URL),\n embeddingp(EMBEDDING_P_API_URL),\n embeddingq(EMBEDDING_Q_API_URL),\n imageCreate(IMAGE_CREATE_API_URL),\n imageUnderstanding(IMAGE_UNDERSANDING_API_URL),\n hypermimeticSynthesis(HYPERMIMETIC_SYNTHESIS_API_URL);\n private String url;\n }\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/common/Usage.java",
"content": "package com.ai.spark.common;\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport com.fasterxml.jackson.annotation.JsonProperty;\nimport lombok.AllArgsConstructor;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\n\n@Data\n@NoArgsConstructor\n@AllArgsConstructor\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class Usage {\n\n @JsonProperty(\"text\")\n private UsageText usageText;\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/common/UsageText.java",
"content": "package com.ai.spark.common;\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport com.fasterxml.jackson.annotation.JsonProperty;\nimport lombok.AllArgsConstructor;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\n\n@Data\n@NoArgsConstructor\n@AllArgsConstructor\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class UsageText {\n\n /**\n * 保留字段，可忽略\n */\n @JsonProperty(\"question_tokens\")\n private Integer questionTokens;\n\n /**\n * 包含历史问题的总tokens大小\n */\n @JsonProperty(\"prompt_tokens\")\n private Integer promptTokens;\n\n /**\n * 回答的tokens大小\n */\n @JsonProperty(\"completion_tokens\")\n private Integer completionTokens;\n\n /**\n * prompt_tokens和completion_tokens的和\n */\n @JsonProperty(\"total_tokens\")\n private Integer totalTokens;\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/common/utils/AuthUtils.java",
"content": "package com.ai.spark.common.utils;\n\n\nimport lombok.AllArgsConstructor;\nimport lombok.Getter;\n\nimport javax.crypto.Mac;\nimport javax.crypto.spec.SecretKeySpec;\nimport java.io.UnsupportedEncodingException;\nimport java.net.URI;\nimport java.net.URISyntaxException;\nimport java.net.URLEncoder;\nimport java.nio.charset.StandardCharsets;\nimport java.security.InvalidKeyException;\nimport java.security.MessageDigest;\nimport java.security.NoSuchAlgorithmException;\nimport java.security.SignatureException;\nimport java.time.ZoneId;\nimport java.time.ZonedDateTime;\nimport java.time.format.DateTimeFormatter;\nimport java.util.Base64;\nimport java.util.Locale;\n\n/**\n * 鉴权工具类\n */\npublic class AuthUtils {\n\n /**\n * 日期格式化\n */\n public final static DateTimeFormatter dateTimeFormatter = DateTimeFormatter.ofPattern(\"EEE, dd MMM yyyy HH:mm:ss z\", Locale.US);\n public final static String preStr = \"host: %s\\n\" +\n \"date: %s\\n\" +\n \"%s %s HTTP/1.1\";\n private static final char[] MD5_TABLE = {'0', '1', '2', '3', '4', '5', '6', '7', '8', '9', 'a', 'b', 'c', 'd', 'e', 'f'};\n\n public static String replaceAllHttp(String authUrl) {\n return authUrl.replaceAll(\"http://\", \"ws://\").replaceAll(\"https://\", \"wss://\");\n }\n\n /**\n * 鉴权方法，适用于对话接口\n *\n * @param requestMethod 请求方式\n * @param hostUrl 地址\n * @param apiKey apikey\n * @param apiSecret apiSecret\n * @return 鉴权信息\n */\n public static String getAuthUrl(String requestMethod, String hostUrl, String apiKey, String apiSecret) throws InvalidKeyException, NoSuchAlgorithmException, URISyntaxException, UnsupportedEncodingException {\n URI uri = new URI(hostUrl);\n String date = ZonedDateTime.now(ZoneId.of(\"GMT\")).format(dateTimeFormatter);\n // SHA256加密\n Mac mac = Mac.getInstance(\"hmacsha256\");\n mac.init(new SecretKeySpec(apiSecret.getBytes(StandardCharsets.UTF_8), \"hmacsha256\"));\n byte[] hexDigits = mac.doFinal(String.format(preStr, uri.getHost(), date, requestMethod, uri.getPath()).getBytes(StandardCharsets.UTF_8));\n String authorization = String.format(\"api_key=\\\"%s\\\", algorithm=\\\"%s\\\", headers=\\\"%s\\\", signature=\\\"%s\\\"\", apiKey, \"hmac-sha256\", \"host date request-line\", Base64.getEncoder().encodeToString(hexDigits));\n // 拼接地址\n return new StringBuilder(hostUrl)\n .append(\"?authorization=\").append(Base64.getEncoder().encodeToString(authorization.getBytes(StandardCharsets.UTF_8)))\n .append(\"&date=\").append(URLEncoder.encode(date, StandardCharsets.UTF_8.name()))\n .append(\"&host=\").append(uri.getHost())\n .toString();\n }\n\n /**\n * 获取签名，适用于文档问答\n *\n * @param appId 签名的key\n * @param secret 签名秘钥\n * @return 返回签名\n */\n public static String getSignature(String appId, String secret, long ts) {\n try {\n String auth = md5(appId + ts);\n return hmacSHA1Encrypt(auth, secret);\n } catch (SignatureException e) {\n return null;\n }\n }\n\n /**\n * sha1加密\n *\n * @param encryptText 加密文本\n * @param encryptKey 加密键\n * @return 加密\n */\n private static String hmacSHA1Encrypt(String encryptText, String encryptKey)\n throws SignatureException {\n byte[] rawHmac;\n try {\n byte[] data = encryptKey.getBytes(StandardCharsets.UTF_8);\n SecretKeySpec secretKey = new SecretKeySpec(data, \"HmacSHA1\");\n Mac mac = Mac.getInstance(\"HmacSHA1\");\n mac.init(secretKey);\n byte[] text = encryptText.getBytes(StandardCharsets.UTF_8);\n rawHmac = mac.doFinal(text);\n } catch (InvalidKeyException e) {\n throw new SignatureException(\"InvalidKeyException:\" + e.getMessage());\n } catch (NoSuchAlgorithmException e) {\n throw new SignatureException(\n \"NoSuchAlgorithmException:\" + e.getMessage()\n );\n }\n return Base64.getEncoder().encodeToString(rawHmac);\n }\n\n private static String md5(String cipherText) {\n try {\n byte[] data = cipherText.getBytes();\n // 信息摘要是安全的单向哈希函数，它接收任意大小的数据，并输出固定长度的哈希值。\n MessageDigest mdInst = MessageDigest.getInstance(\"MD5\");\n\n // MessageDigest对象通过使用 update方法处理数据， 使用指定的byte数组更新摘要\n mdInst.update(data);\n\n // 摘要更新之后，通过调用digest（）执行哈希计算，获得密文\n byte[] md = mdInst.digest();\n\n // 把密文转换成十六进制的字符串形式\n int j = md.length;\n char[] str = new char[j * 2];\n int k = 0;\n for (byte byte0 : md) { // i = 0\n str[k++] = MD5_TABLE[byte0 >>> 4 & 0xf]; // 5\n str[k++] = MD5_TABLE[byte0 & 0xf]; // F\n }\n // 返回经过加密后的字符串\n return new String(str);\n } catch (Exception e) {\n return null;\n }\n }\n\n @Getter\n @AllArgsConstructor\n public enum RequestMethod {\n GET(\"GET\"),\n POST(\"POST\");\n private String method;\n }\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/audio/Audio.java",
"content": "package com.ai.spark.endPoint.audio;\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport com.fasterxml.jackson.annotation.JsonProperty;\nimport lombok.AllArgsConstructor;\nimport lombok.Builder;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class Audio {\n\n /**\n * ⾳频编码\n */\n private String encoding;\n\n /**\n * ⾳频编码\n */\n @JsonProperty(\"sample_rate\")\n private Integer sampleRate;\n\n /**\n * 声道数\n */\n private Integer channels;\n\n /**\n * 位深\n */\n @JsonProperty(\"bit_depth\")\n private Integer bitDepth;\n\n /**\n * 帧⼤⼩\n */\n @JsonProperty(\"frame_size\")\n private Integer frameSize;\n\n /**\n * 数据状态\n * 0:开始, 1:开始, 2:结束\n */\n private Integer status;\n\n /**\n * 数据序号\n * 最⼩值:0, 最⼤值:9999999\n */\n private Integer seq;\n\n /**\n * ⾳频数据\n * 最⼩尺⼨:0B, 最⼤尺⼨:10485760B\n */\n private String audio;\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/audio/AudioHeader.java",
"content": "package com.ai.spark.endPoint.audio;\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport com.fasterxml.jackson.annotation.JsonProperty;\nimport lombok.AllArgsConstructor;\nimport lombok.Builder;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class AudioHeader {\n\n /**\n * 在平台申请的appid信息，必传\n */\n @JsonProperty(\"app_id\")\n private String appId;\n\n /**\n * 请求⽤户服务返回的uid，⽤户及设备级别个性化功能依赖此参数\n */\n private String uid;\n\n /**\n * 请求⽅确保唯⼀的设备标志，设备级别个性化功能依赖此参数\n */\n private String did;\n\n /**\n * 设备imei信息\n */\n private String imei;\n\n /**\n * 设备imsi信息\n */\n private String imsi;\n\n /**\n * 设备mac信息\n */\n private String mac;\n\n /**\n * ⽹络类型，可选值为wifi、2G、3G、4G、5G\n */\n @JsonProperty(\"net_type\")\n private String netType;\n\n /**\n * 运营商信息，可选值为CMCC、CUCC、CTCC、other\n */\n @JsonProperty(\"net_isp\")\n private String netIsp;\n\n /**\n * 客户端请求的会话唯⼀标识\n */\n @JsonProperty(\"request_id\")\n private String requestId;\n\n /**\n * 个性化资源ID\n */\n @JsonProperty(\"res_id\")\n private String resId;\n\n /**\n * 请求状态，可选值为：0-开始、1-继续、2-结束\n */\n private Integer status;\n\n // 以下是请求返回时所需参数\n /**\n * 错误码，0表示正常，非0表示出错；\n */\n private Integer code;\n\n /**\n * 会话是否成功的描述信息\n */\n private String message;\n\n /**\n * 会话的唯一id，用于讯飞技术人员查询服务端会话日志使用,出现调用错误时建议留存该字段\n */\n private String sid;\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/audio/AudioParameter.java",
"content": "package com.ai.spark.endPoint.audio;\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport lombok.AllArgsConstructor;\nimport lombok.Builder;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class AudioParameter {\n\n private Oral oral;\n\n private Tts tts;\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/audio/AudioPayload.java",
"content": "package com.ai.spark.endPoint.audio;\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport com.fasterxml.jackson.annotation.JsonProperty;\nimport lombok.AllArgsConstructor;\nimport lombok.Builder;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class AudioPayload {\n\n /**\n * 待合成⽂本\n */\n private AudioText text;\n\n /**\n * ⽤户原始输⼊，场景化合成开启时必传，不开启为⾮必传\n */\n @JsonProperty(\"user_text\")\n private AudioText userText;\n\n /**\n * 响应数据块\n */\n private Audio audio;\n\n /**\n * 响应数据块\n */\n private Pybuf pybuf;\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/audio/AudioText.java",
"content": "package com.ai.spark.endPoint.audio;\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport lombok.AllArgsConstructor;\nimport lombok.Builder;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class AudioText {\n\n /**\n * ⽂本编码\n */\n private String encoding;\n\n /**\n * ⽂本压缩格式\n */\n private String compress;\n\n /**\n * ⽂本格式\n */\n private String format;\n\n /**\n * 数据状态\n */\n private Integer status;\n\n /**\n * 数据序号\n */\n private Integer seq;\n\n /**\n * ⽂本数据\n */\n private String text;\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/audio/Oral.java",
"content": "package com.ai.spark.endPoint.audio;\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport com.fasterxml.jackson.annotation.JsonProperty;\nimport lombok.AllArgsConstructor;\nimport lombok.Builder;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class Oral {\n\n @JsonProperty(\"spark_assist\")\n private Integer sparkAssist;\n\n @JsonProperty(\"oral_level\")\n private String oralLevel;\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/audio/Pybuf.java",
"content": "package com.ai.spark.endPoint.audio;\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport lombok.AllArgsConstructor;\nimport lombok.Builder;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class Pybuf {\n\n /**\n * ⽂本编码\n */\n private String encoding;\n\n /**\n * ⽂本压缩格式\n */\n private String compress;\n\n /**\n * ⽂本格式\n */\n private String format;\n\n /**\n * 数据状态\n * 0:开始, 1:开始, 2:结束\n */\n private Integer status;\n\n /**\n * 数据序号\n * 最⼩值:0, 最⼤值:9999999\n */\n private Integer seq;\n\n /**\n * ⽂本数据\n * 最⼩尺⼨:0B, 最⼤尺⼨:1048576B\n */\n private String text;\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/audio/Tts.java",
"content": "package com.ai.spark.endPoint.audio;\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport lombok.*;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class Tts {\n\n /**\n * 发⾳⼈，必传\n */\n private String vcn;\n\n /**\n * 语速：0对应默认语速的1/2，100 对应默认语速的2倍\n * 最⼩值:0, 最⼤值:100\n */\n private Integer speed;\n\n /**\n * ⾳量：0是 静⾳，1对 应默认⾳量 1/2，100对应默认⾳量的2倍\n * 最⼩值:0, 最⼤值:100\n */\n private Integer volume;\n\n /**\n * 语调：0对应默认语速的1/2，100对应默认语速的2倍\n * 最⼩值:0, 最⼤值:100\n */\n private Integer pitch;\n\n /**\n * 背景⾳\n * 0:⽆背景⾳, 1:内置背景⾳1, 2:内置背景⾳2\n */\n private Integer bgs;\n\n /**\n * 英⽂发⾳⽅式\n * 0:⾃动判断处理，如果不确定将按照英⽂词语拼写处理（缺省）,\n * 1:所有英⽂按字⺟发⾳,\n * 2:⾃动判断处理，如果不确定将按照字⺟朗读\n */\n private Integer reg;\n\n /**\n * 合成⾳频数字发⾳⽅式\n * 0:⾃动判断, 1:完全数值, 2:完全字符串, 3:字符串优先\n */\n private Integer rdn;\n\n /**\n * 是否返回拼⾳标注\n * 0:不返回拼⾳, 1:返回拼⾳（纯⽂本格式，utf8编码）\n */\n private Integer rhy;\n\n /**\n * 场景\n * 0:⽆, 1:散⽂阅读, 2:⼩说阅读, 3:新闻, 4:⼴告, 5:交互\n */\n private Integer scn;\n\n /**\n * 引擎初始化，是否返回版本信息+时间戳信息\n * 0:不返回, 1:返回版本信息+时间戳信息。如XXX.18928127 XXX表示版本号，后接秒为单位的时间戳\n */\n private Integer version;\n\n /**\n * 控制L5静⾳时⻓，取值范围为 0~10000ms\n * 最⼩值:0, 最⼤值:10000\n */\n private Integer L5SilLen;\n\n /**\n * 段落静⾳时⻓，取值范围为0~10000ms\n * 最⼩值:0, 最⼤值:10000\n */\n private Integer ParagraphSilLen;\n\n private Audio audio;\n\n private Pybuf pybuf;\n\n @Getter\n @AllArgsConstructor\n public enum Vcn {\n lxx(\"x4_lingxiaoxuan_oral\"),\n lfz(\"x4_lingfeizhe_oral\"),\n lyz(\"x4_lingyuzhao_oral\");\n private String name;\n }\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/audio/req/AudioRequest.java",
"content": "package com.ai.spark.endPoint.audio.req;\n\nimport com.ai.spark.endPoint.audio.AudioHeader;\nimport com.ai.spark.endPoint.audio.AudioParameter;\nimport com.ai.spark.endPoint.audio.AudioPayload;\nimport com.ai.spark.endPoint.audio.Tts;\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport lombok.AllArgsConstructor;\nimport lombok.Builder;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class AudioRequest {\n\n /**\n * 协议头部，⽤于描述平台特性的参数\n */\n private AudioHeader header;\n\n /**\n * AI 能⼒功能参数，⽤于控制 AI 引擎特性的开关。\n */\n private AudioParameter parameter;\n\n private AudioPayload payload;\n\n public static AudioRequest baseBuild(Tts.Vcn vcn, String appId) {\n AudioRequest request = AudioRequest.builder()\n .header(AudioHeader.builder().appId(appId).status(0).build())\n .parameter(AudioParameter.builder().tts(Tts.builder().vcn(vcn.getName()).build()).build())\n .build();\n return request;\n }\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/audio/resp/AudioResponse.java",
"content": "package com.ai.spark.endPoint.audio.resp;\n\nimport com.ai.spark.endPoint.audio.AudioHeader;\nimport com.ai.spark.endPoint.audio.AudioPayload;\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport lombok.AllArgsConstructor;\nimport lombok.Builder;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class AudioResponse {\n\n private AudioHeader header;\n\n private AudioPayload payload;\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/chat/Chat.java",
"content": "package com.ai.spark.endPoint.chat;\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport com.fasterxml.jackson.annotation.JsonProperty;\nimport lombok.*;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class Chat {\n\n /**\n * 必传\n * 指定访问的领域,general指向V1.5版本,generalv2指向V2版本,generalv3指向V3版本 。注意：不同的取值对应的url也不一样！\n * 取值为[general,generalv2,generalv3]\n */\n private String domain;\n\n /**\n * 非必传\n * 核采样阈值。用于决定结果随机性，取值越高随机性越强即相同的问题得到的不同答案的可能性越高\n * 取值范围 (0，1] ，默认值0.5\n */\n private Double temperature;\n\n /**\n * 非必传\n * 模型回答的tokens的最大长度\n * V1.5取值为[1,4096]\n * V2.0取值为[1,8192]，默认为2048。\n * V3.0取值为[1,8192]，默认为2048。\n */\n @JsonProperty(\"max_tokens\")\n private Integer maxTokens;\n\n /**\n * 非必传\n * 从k个候选中随机选择⼀个（⾮等概率）\n * 取值为[1，6],默认为4\n */\n @JsonProperty(\"top_k\")\n private Integer topK;\n\n /**\n * 非必传\n * 用于关联用户会话，需要保障用户下的唯一性\n */\n @JsonProperty(\"chat_id\")\n private String chatId;\n\n @Getter\n @AllArgsConstructor\n public enum General {\n general(\"general\"),\n generalV2(\"generalv2\"),\n generalV3(\"generalv3\");\n private String msg;\n }\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/chat/ChatHeader.java",
"content": "package com.ai.spark.endPoint.chat;\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport com.fasterxml.jackson.annotation.JsonProperty;\nimport lombok.*;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class ChatHeader {\n\n /**\n * 必传\n * 应用appid，从开放平台控制台创建的应用中获取\n */\n @JsonProperty(\"app_id\")\n private String appId;\n\n /**\n * 非必传\n * 每个用户的id，用于区分不同用户\n */\n private String uid;\n\n // 以下是请求返回时所需参数\n /**\n * 错误码，0表示正常，非0表示出错；\n */\n private Integer code;\n\n /**\n * 会话是否成功的描述信息\n */\n private String message;\n\n /**\n * 会话的唯一id，用于讯飞技术人员查询服务端会话日志使用,出现调用错误时建议留存该字段\n */\n private String sid;\n\n /**\n * 会话状态，取值为[0,1,2]；0代表首次结果；1代表中间结果；2代表最后一个结果\n */\n private Integer status;\n\n @Getter\n @AllArgsConstructor\n public enum Code {\n SUCCESS(0),\n ;\n\n private final int value;\n }\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/chat/ChatParameter.java",
"content": "package com.ai.spark.endPoint.chat;\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport lombok.AllArgsConstructor;\nimport lombok.Builder;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class ChatParameter {\n\n private Chat chat;\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/chat/ChatPayload.java",
"content": "package com.ai.spark.endPoint.chat;\n\nimport com.ai.spark.common.Usage;\nimport com.ai.spark.endPoint.chat.function.Function;\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport com.fasterxml.jackson.annotation.JsonProperty;\nimport lombok.AllArgsConstructor;\nimport lombok.Builder;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class ChatPayload {\n\n private Message message;\n\n @JsonProperty(\"functions\")\n private Function function;\n\n // 以下是请求返回时所需参数\n @JsonProperty(\"choices\")\n private Choice choice;\n\n private Usage usage;\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/chat/ChatText.java",
"content": "package com.ai.spark.endPoint.chat;\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport com.fasterxml.jackson.annotation.JsonProperty;\nimport lombok.*;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class ChatText {\n\n /**\n * user表示是用户的问题，assistant表示AI的回复\n * 取值为[user,assistant]\n */\n private String role;\n\n /**\n * 用户和AI的对话内容\n * 所有content的累计tokens需控制8192以内\n */\n private String content;\n\n /**\n * 结果序号，取值为[0,10];\n */\n private Integer index;\n\n /**\n * 数据的类型\n */\n @JsonProperty(\"content_type\")\n private String contentType;\n\n\n public static ChatText baseBuild(Role role, String content) {\n return ChatText.builder().role(role.getRoleName()).content(content).build();\n }\n\n @Getter\n @AllArgsConstructor\n public enum Role {\n\n USER(\"user\"),\n ASSISTANT(\"assistant\"),\n ;\n\n private String RoleName;\n }\n\n @Getter\n @AllArgsConstructor\n public enum ContentType {\n\n TEXT(\"text\"),\n IMAGE(\"image\"),\n ;\n\n private String type;\n }\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/chat/Choice.java",
"content": "package com.ai.spark.endPoint.chat;\n\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport com.fasterxml.jackson.annotation.JsonProperty;\nimport lombok.*;\n\nimport java.util.List;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class Choice {\n\n /**\n * 文本响应状态，取值为[0,1,2]; 0代表首个文本结果；1代表中间文本结果；2代表最后一个文本结果\n */\n private Integer status;\n\n /**\n * 返回的数据序号，取值为[0,9999999]\n */\n private Integer seq;\n\n @JsonProperty(\"text\")\n private List texts;\n\n @Getter\n public enum Status {\n START(0),\n ING(1),\n END(2),\n ;\n\n private final int value;\n\n Status(int value) {\n this.value = value;\n }\n }\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/chat/Message.java",
"content": "package com.ai.spark.endPoint.chat;\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport com.fasterxml.jackson.annotation.JsonProperty;\nimport lombok.AllArgsConstructor;\nimport lombok.Builder;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\n\nimport java.util.List;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class Message {\n\n @JsonProperty(\"text\")\n private List chatTexts;\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/chat/document/ChatExtends.java",
"content": "package com.ai.spark.endPoint.chat.document;\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport lombok.AllArgsConstructor;\nimport lombok.Builder;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class ChatExtends {\n\n /**\n * wiki 大模型问答模板，在某些场景服务默认的 prompt 回答效果不好时，业务可以考虑通过自定义 prompt 来改善。替换的问题标识，替换的文本内容标识\n */\n private String wikiPromptTpl;\n\n /**\n * wiki 结果分数阈值，低于这个阈值的结果丢弃。取值范围为(0,1] 参考值为：0.80非常宽松 0.82宽松 0.83标准0.84严格 0.86非常严格。服务会根据问题检索文件列表中内容相关的文段，该值设置的越高，可能丢弃的内容越多，但保留下来的内容越准确；但过高也可能导致无匹配内容\n */\n private Float wikiFilterScore;\n\n /**\n * 用户问题未匹配到文档内容时，是否使用大模型兜底回答问题\n */\n private Boolean sparkWhenWithoutEmbedding;\n\n /**\n * 大模型问答时的温度，取值范围 (0，1] ，temperature 越大，大模型回答随机度越高\n */\n private Float temperature;\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/chat/function/Function.java",
"content": "package com.ai.spark.endPoint.chat.function;\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport com.fasterxml.jackson.annotation.JsonProperty;\nimport lombok.AllArgsConstructor;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\n\nimport java.util.List;\n\n@Data\n@NoArgsConstructor\n@AllArgsConstructor\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class Function {\n\n @JsonProperty(\"functions\")\n private List functionTexts;\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/chat/function/FunctionParameter.java",
"content": "package com.ai.spark.endPoint.chat.function;\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport lombok.AllArgsConstructor;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\n\nimport java.util.List;\n\n@Data\n@NoArgsConstructor\n@AllArgsConstructor\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class FunctionParameter {\n\n /**\n * 参数类型\n */\n private String type;\n\n /**\n * 该内容由用户定义，命中该方法时需要返回哪些参数\n */\n private Object properties;\n\n /**\n * 该内容由用户定义，命中方法时必须返回的字段\n */\n private List required;\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/chat/function/FunctionText.java",
"content": "package com.ai.spark.endPoint.chat.function;\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport com.fasterxml.jackson.annotation.JsonProperty;\nimport lombok.AllArgsConstructor;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\n\n@Data\n@NoArgsConstructor\n@AllArgsConstructor\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class FunctionText {\n\n /**\n * 用户输入命中后，会返回该名称\n */\n private String name;\n\n /**\n * 描述function功能即可，越详细越有助于大模型理解该function\n */\n private String description;\n\n @JsonProperty(\"parameters\")\n private FunctionParameter functionParameter;\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/chat/req/ChatRequest.java",
"content": "package com.ai.spark.endPoint.chat.req;\n\nimport com.ai.spark.endPoint.chat.*;\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport com.fasterxml.jackson.annotation.JsonProperty;\nimport lombok.AllArgsConstructor;\nimport lombok.Builder;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\n\nimport java.util.ArrayList;\nimport java.util.Arrays;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class ChatRequest {\n\n @JsonProperty(\"header\")\n private ChatHeader chatHeader;\n @JsonProperty(\"parameter\")\n private ChatParameter chatParameter;\n @JsonProperty(\"payload\")\n private ChatPayload chatPayload;\n\n public static ChatRequest baseBuild(String question, String appId) {\n ChatHeader chatHeader = ChatHeader.builder().appId(appId).build();\n Chat chat = Chat.builder().domain(Chat.General.generalV3.getMsg()).build();\n ChatParameter chatParameter = ChatParameter.builder().chat(chat).build();\n ChatText chatText = ChatText.builder().role(ChatText.Role.USER.getRoleName()).content(question).build();\n Message message = Message.builder().chatTexts(new ArrayList<>(Arrays.asList(chatText))).build();\n ChatPayload chatPayload = ChatPayload.builder().message(message).build();\n return ChatRequest.builder().chatHeader(chatHeader).chatParameter(chatParameter).chatPayload(chatPayload).build();\n }\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/chat/req/DocumentChatRequest.java",
"content": "package com.ai.spark.endPoint.chat.req;\n\nimport com.ai.spark.endPoint.chat.ChatText;\nimport com.ai.spark.endPoint.chat.document.ChatExtends;\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport com.fasterxml.jackson.annotation.JsonProperty;\nimport lombok.AllArgsConstructor;\nimport lombok.Builder;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\n\nimport java.util.ArrayList;\nimport java.util.Arrays;\nimport java.util.List;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class DocumentChatRequest {\n\n private ChatExtends chatExtends;\n\n /**\n * 必传\n * 提问问题检索的文件 id 列表\n */\n private List fileIds;\n\n /**\n * 问答内容列表，按时间正序，最后一条为最新提问\n */\n @JsonProperty(\"messages\")\n private List chatTexts;\n\n public static DocumentChatRequest baseBuild(String question, List fileIds) {\n return DocumentChatRequest.builder()\n .fileIds(fileIds)\n .chatTexts(new ArrayList<>(Arrays.asList(ChatText.baseBuild(ChatText.Role.USER, question))))\n .build();\n }\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/chat/resp/ChatResponse.java",
"content": "package com.ai.spark.endPoint.chat.resp;\n\nimport com.ai.spark.endPoint.chat.ChatHeader;\nimport com.ai.spark.endPoint.chat.ChatPayload;\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport com.fasterxml.jackson.annotation.JsonProperty;\nimport lombok.AllArgsConstructor;\nimport lombok.Builder;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class ChatResponse {\n\n @JsonProperty(\"header\")\n private ChatHeader chatHeader;\n @JsonProperty(\"payload\")\n private ChatPayload chatPayload;\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/chat/resp/DocumentChatResponse.java",
"content": "package com.ai.spark.endPoint.chat.resp;\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport lombok.*;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class DocumentChatResponse {\n\n /**\n * 错误码 ，0 标识成功\n */\n private Integer code;\n\n /**\n * 错误描述\n */\n private String content;\n\n /**\n * 文档引用，status=99 的时候返回；结构是个 Map，key=文件 id，value=引用的文段列表（对应 fileTrunks 的 index）\n */\n private String fileRefer;\n\n /**\n * 会话唯一标识\n */\n private String sid;\n\n /**\n * 会话状态，取值为[0,1,2,99]；0 代表首次结果；1 代表中间结果；2 代表最后一个结果；99 代表引用的文档及文段\n */\n private Integer status;\n\n @Getter\n @AllArgsConstructor\n public enum Code {\n SUCCESS(0),\n ;\n\n private final int value;\n }\n\n @Getter\n public enum Status {\n START(0),\n ING(1),\n END(2),\n DOCUMENT(99),\n ;\n\n private final int value;\n\n Status(int value) {\n this.value = value;\n }\n }\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/document/Data.java",
"content": "package com.ai.spark.endPoint.document;\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport lombok.AllArgsConstructor;\nimport lombok.Builder;\nimport lombok.NoArgsConstructor;\n\n@lombok.Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class Data {\n /**\n * 返回上传的 fileId\n */\n private String fileId;\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/document/req/FileUploadRequest.java",
"content": "package com.ai.spark.endPoint.document.req;\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport lombok.*;\nimport lombok.experimental.FieldNameConstants;\n\nimport java.io.File;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@FieldNameConstants\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class FileUploadRequest {\n\n /**\n * 要上传的文件\n */\n private File file;\n /**\n * 文件 url （文件和文件 url 必须有一个）\n */\n private String url;\n /**\n * 文件名称，带后缀。文件用 url 的方式，该字段必传；传 file 的话，该字段可不传\n */\n private String fileName;\n /**\n * 文件类型，目前传固定值\"wiki\"\n */\n @Builder.Default\n private String fileType = FileType.wiki.getType();\n /**\n * 文件状态回调地址，文件状态有变动时服务会调用该 url。调用的时候会带上鉴权头，鉴权方式同【接口鉴权】，业务可根据需要是否做鉴权校验\n */\n private String callbackUrl;\n\n @Getter\n @AllArgsConstructor\n public enum FileType {\n wiki(\"wiki\");\n private String type;\n }\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/document/resp/DocumentSummaryResponse.java",
"content": "package com.ai.spark.endPoint.document.resp;\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport lombok.AllArgsConstructor;\nimport lombok.Builder;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class DocumentSummaryResponse {\n\n private Boolean flag;\n private Integer code;\n private String sid;\n private String desc;\n private Object data;\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/document/resp/FileUploadResponse.java",
"content": "package com.ai.spark.endPoint.document.resp;\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport lombok.AllArgsConstructor;\nimport lombok.Builder;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class FileUploadResponse {\n /**\n * 返回上传的 fileId\n */\n private Integer code;\n /**\n * 请求唯一 id，用于问题定位\n */\n private String sid;\n /**\n * 结果描述\n */\n private String desc;\n /**\n * 返回结果\n */\n private com.ai.spark.endPoint.document.Data data;\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/embedding/Emb.java",
"content": "package com.ai.spark.endPoint.embedding;\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport lombok.AllArgsConstructor;\nimport lombok.Builder;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\nimport lombok.experimental.FieldNameConstants;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@FieldNameConstants\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class Emb {\n private Feature feature;\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/embedding/EmbeddingHeader.java",
"content": "package com.ai.spark.endPoint.embedding;\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport com.fasterxml.jackson.annotation.JsonProperty;\nimport lombok.AllArgsConstructor;\nimport lombok.Builder;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\nimport lombok.experimental.FieldNameConstants;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@FieldNameConstants\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class EmbeddingHeader {\n\n /**\n * 必传\n * 在平台申请的app id信息\n */\n @JsonProperty(\"app_id\")\n private String appId;\n\n /**\n * 非必传\n * 每个用户的id，用于区分不同用户\n */\n private String uid;\n\n /**\n * 发送状态标识，3为一次性发完\n */\n @Builder.Default\n private Integer status = 3;\n\n // 以下是请求返回时所需参数\n /**\n * 错误码，0表示正常，非0表示出错；\n */\n private Integer code;\n\n /**\n * 会话是否成功的描述信息\n */\n private String message;\n\n /**\n * 会话的唯一id，用于讯飞技术人员查询服务端会话日志使用,出现调用错误时建议留存该字段\n */\n private String sid;\n\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/embedding/EmbeddingMessage.java",
"content": "package com.ai.spark.endPoint.embedding;\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport lombok.AllArgsConstructor;\nimport lombok.Builder;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\nimport lombok.experimental.FieldNameConstants;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@FieldNameConstants\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class EmbeddingMessage {\n private String text;\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/embedding/EmbeddingParameter.java",
"content": "package com.ai.spark.endPoint.embedding;\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport lombok.AllArgsConstructor;\nimport lombok.Builder;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\nimport lombok.experimental.FieldNameConstants;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@FieldNameConstants\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class EmbeddingParameter {\n @Builder.Default\n private Emb emb = Emb.builder().feature(Feature.builder().build()).build();\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/embedding/EmbeddingPayload.java",
"content": "package com.ai.spark.endPoint.embedding;\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport com.fasterxml.jackson.annotation.JsonProperty;\nimport lombok.AllArgsConstructor;\nimport lombok.Builder;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\nimport lombok.experimental.FieldNameConstants;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@FieldNameConstants\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class EmbeddingPayload {\n @JsonProperty(\"messages\")\n private EmbeddingMessage embeddingMessage;\n\n private Feature feature;\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/embedding/Feature.java",
"content": "package com.ai.spark.endPoint.embedding;\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport lombok.AllArgsConstructor;\nimport lombok.Builder;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\nimport lombok.experimental.FieldNameConstants;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@FieldNameConstants\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class Feature {\n @Builder.Default\n private String encoding = \"utf8\";\n private String seq;\n private String status;\n private String text;\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/embedding/req/EmbeddingRequest.java",
"content": "package com.ai.spark.endPoint.embedding.req;\n\nimport com.ai.common.utils.JsonUtils;\nimport com.ai.spark.endPoint.chat.ChatText;\nimport com.ai.spark.endPoint.embedding.EmbeddingHeader;\nimport com.ai.spark.endPoint.embedding.EmbeddingMessage;\nimport com.ai.spark.endPoint.embedding.EmbeddingParameter;\nimport com.ai.spark.endPoint.embedding.EmbeddingPayload;\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport com.fasterxml.jackson.annotation.JsonProperty;\nimport lombok.AllArgsConstructor;\nimport lombok.Builder;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\nimport lombok.experimental.FieldNameConstants;\n\nimport java.nio.charset.StandardCharsets;\nimport java.util.Base64;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@FieldNameConstants\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class EmbeddingRequest {\n\n @JsonProperty(\"header\")\n private EmbeddingHeader embeddingHeader;\n\n @JsonProperty(\"parameter\")\n private EmbeddingParameter embeddingParameter;\n\n @JsonProperty(\"payload\")\n private EmbeddingPayload embeddingPayload;\n\n\n public static EmbeddingRequest baseBuild(ChatText text, String appId) {\n return EmbeddingRequest.builder()\n .embeddingHeader(EmbeddingHeader.builder().appId(appId).build())\n .embeddingParameter(EmbeddingParameter.builder().build())\n .embeddingPayload(EmbeddingPayload.builder().embeddingMessage(EmbeddingMessage\n .builder()\n .text(Base64.getEncoder().encodeToString(JsonUtils.toJson(text).getBytes(StandardCharsets.UTF_8)))\n .build()).build()\n )\n .build();\n }\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/embedding/resp/EmbeddingResponse.java",
"content": "package com.ai.spark.endPoint.embedding.resp;\n\nimport com.ai.spark.endPoint.embedding.EmbeddingHeader;\nimport com.ai.spark.endPoint.embedding.EmbeddingPayload;\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport com.fasterxml.jackson.annotation.JsonProperty;\nimport lombok.AllArgsConstructor;\nimport lombok.Builder;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\nimport lombok.experimental.FieldNameConstants;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@FieldNameConstants\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class EmbeddingResponse {\n\n @JsonProperty(\"header\")\n private EmbeddingHeader embeddingHeader;\n\n @JsonProperty(\"payload\")\n private EmbeddingPayload embeddingPayload;\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/images/ImageChat.java",
"content": "package com.ai.spark.endPoint.images;\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport lombok.AllArgsConstructor;\nimport lombok.Builder;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\nimport lombok.experimental.FieldNameConstants;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@FieldNameConstants\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class ImageChat {\n\n @Builder.Default\n private String domain = \"general\";\n\n @Builder.Default\n private Integer width = 512;\n\n @Builder.Default\n private Integer height = 512;\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/images/ImageHeader.java",
"content": "package com.ai.spark.endPoint.images;\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport com.fasterxml.jackson.annotation.JsonProperty;\nimport lombok.AllArgsConstructor;\nimport lombok.Builder;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\nimport lombok.experimental.FieldNameConstants;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@FieldNameConstants\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class ImageHeader {\n\n @JsonProperty(\"app_id\")\n private String appId;\n\n private String uid;\n\n // 下面是返回时用到的属性\n\n private Integer code;\n\n private String message;\n\n private String sid;\n\n private Integer status;\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/images/ImageParameter.java",
"content": "package com.ai.spark.endPoint.images;\n\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport com.fasterxml.jackson.annotation.JsonProperty;\nimport lombok.AllArgsConstructor;\nimport lombok.Builder;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\nimport lombok.experimental.FieldNameConstants;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@FieldNameConstants\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class ImageParameter {\n\n @Builder.Default\n @JsonProperty(\"chat\")\n private ImageChat imageChat = ImageChat.builder().build();\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/images/ImagePayload.java",
"content": "package com.ai.spark.endPoint.images;\n\nimport com.ai.spark.endPoint.chat.Choice;\nimport com.ai.spark.endPoint.chat.Message;\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport com.fasterxml.jackson.annotation.JsonProperty;\nimport lombok.AllArgsConstructor;\nimport lombok.Builder;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\nimport lombok.experimental.FieldNameConstants;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@FieldNameConstants\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class ImagePayload {\n\n private Message message;\n\n @JsonProperty(\"choices\")\n private Choice choice;\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/images/ImageUnderstandingChat.java",
"content": "package com.ai.spark.endPoint.images;\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport com.fasterxml.jackson.annotation.JsonProperty;\nimport lombok.AllArgsConstructor;\nimport lombok.Builder;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\nimport lombok.experimental.FieldNameConstants;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@FieldNameConstants\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class ImageUnderstandingChat {\n\n @Builder.Default\n private String domain = \"general\";\n\n private String auditing;\n\n private Double temperature;\n\n @JsonProperty(\"top_k\")\n private Integer topK;\n\n @JsonProperty(\"max_tokens\")\n private Integer maxTokens;\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/images/ImageUnderstandingParameter.java",
"content": "package com.ai.spark.endPoint.images;\n\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport com.fasterxml.jackson.annotation.JsonProperty;\nimport lombok.AllArgsConstructor;\nimport lombok.Builder;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\nimport lombok.experimental.FieldNameConstants;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@FieldNameConstants\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class ImageUnderstandingParameter {\n\n @Builder.Default\n @JsonProperty(\"chat\")\n private ImageUnderstandingChat imageUnderstandingChat = ImageUnderstandingChat.builder().build();\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/images/ImageUnderstandingPayload.java",
"content": "package com.ai.spark.endPoint.images;\n\n\nimport com.ai.spark.common.Usage;\nimport com.ai.spark.endPoint.chat.Choice;\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport com.fasterxml.jackson.annotation.JsonProperty;\nimport lombok.AllArgsConstructor;\nimport lombok.Builder;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\nimport lombok.experimental.FieldNameConstants;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@FieldNameConstants\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class ImageUnderstandingPayload {\n\n @JsonProperty(\"choices\")\n private Choice choice;\n\n private Usage usage;\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/images/req/ImageCreateRequest.java",
"content": "package com.ai.spark.endPoint.images.req;\n\nimport com.ai.spark.endPoint.chat.ChatText;\nimport com.ai.spark.endPoint.chat.Message;\nimport com.ai.spark.endPoint.images.ImageHeader;\nimport com.ai.spark.endPoint.images.ImageParameter;\nimport com.ai.spark.endPoint.images.ImagePayload;\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport com.fasterxml.jackson.annotation.JsonProperty;\nimport lombok.AllArgsConstructor;\nimport lombok.Builder;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\nimport lombok.experimental.FieldNameConstants;\n\nimport java.util.Arrays;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@FieldNameConstants\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class ImageCreateRequest {\n\n @JsonProperty(\"header\")\n private ImageHeader imageHeader;\n\n @JsonProperty(\"parameter\")\n private ImageParameter imageParameter;\n\n @JsonProperty(\"payload\")\n private ImagePayload imagePayload;\n\n public static ImageCreateRequest baseBuild(String content, String appId) {\n return ImageCreateRequest\n .builder()\n .imageHeader(ImageHeader.builder().appId(appId).build())\n .imageParameter(ImageParameter.builder().build())\n .imagePayload(ImagePayload.builder().message(Message.builder().chatTexts(Arrays.asList(ChatText.baseBuild(ChatText.Role.USER, content))).build()).build())\n .build();\n }\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/images/req/ImageUnderstandingRequest.java",
"content": "package com.ai.spark.endPoint.images.req;\n\nimport com.ai.spark.endPoint.chat.ChatText;\nimport com.ai.spark.endPoint.chat.Message;\nimport com.ai.spark.endPoint.images.ImageHeader;\nimport com.ai.spark.endPoint.images.ImagePayload;\nimport com.ai.spark.endPoint.images.ImageUnderstandingParameter;\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport com.fasterxml.jackson.annotation.JsonProperty;\nimport lombok.AllArgsConstructor;\nimport lombok.Builder;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\nimport lombok.experimental.FieldNameConstants;\n\nimport java.io.File;\nimport java.io.FileInputStream;\nimport java.io.FileNotFoundException;\nimport java.io.IOException;\nimport java.util.ArrayList;\nimport java.util.Arrays;\nimport java.util.Base64;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@FieldNameConstants\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class ImageUnderstandingRequest {\n\n @JsonProperty(\"header\")\n private ImageHeader imageHeader;\n\n @JsonProperty(\"parameter\")\n private ImageUnderstandingParameter imageUnderstandingParameter;\n\n @JsonProperty(\"payload\")\n private ImagePayload imagePayload;\n\n public static ImageUnderstandingRequest baseBuild(String content, String appId, File image) {\n byte[] data = null;\n try (FileInputStream fileInputStream = new FileInputStream(image)) {\n data = new byte[fileInputStream.available()];\n fileInputStream.read(data); // 读取文件数据到数组中\n } catch (FileNotFoundException e) {\n e.printStackTrace();\n } catch (IOException e) {\n e.printStackTrace();\n }\n String imageStr = Base64.getEncoder().encodeToString(data); // 将byte数组进行base64编码\n ImageHeader imageHeader = ImageHeader.builder().appId(appId).build();\n ImageUnderstandingParameter imageUnderstandingParameter = ImageUnderstandingParameter.builder().build();\n ChatText chatText = ChatText.baseBuild(ChatText.Role.USER, content);\n chatText.setContentType(ChatText.ContentType.TEXT.getType());\n ChatText imgText = ChatText.baseBuild(ChatText.Role.USER, imageStr);\n imgText.setContentType(ChatText.ContentType.IMAGE.getType());\n Message message = Message.builder().chatTexts(new ArrayList<>(Arrays.asList(imgText, chatText))).build();\n ImagePayload imagePayload = ImagePayload.builder().message(message).build();\n return ImageUnderstandingRequest\n .builder()\n .imageHeader(imageHeader)\n .imagePayload(imagePayload)\n .imageUnderstandingParameter(imageUnderstandingParameter).build();\n\n }\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/images/resp/ImageCreateResponse.java",
"content": "package com.ai.spark.endPoint.images.resp;\n\nimport com.ai.spark.endPoint.images.ImageHeader;\nimport com.ai.spark.endPoint.images.ImagePayload;\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport com.fasterxml.jackson.annotation.JsonProperty;\nimport lombok.AllArgsConstructor;\nimport lombok.Builder;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\nimport lombok.experimental.FieldNameConstants;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@FieldNameConstants\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class ImageCreateResponse {\n\n @JsonProperty(\"header\")\n private ImageHeader imageHeader;\n\n @JsonProperty(\"payload\")\n private ImagePayload imagePayload;\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/endPoint/images/resp/ImageUnderstandingResponse.java",
"content": "package com.ai.spark.endPoint.images.resp;\n\nimport com.ai.spark.endPoint.images.ImageHeader;\nimport com.ai.spark.endPoint.images.ImageUnderstandingPayload;\nimport com.fasterxml.jackson.annotation.JsonIgnoreProperties;\nimport com.fasterxml.jackson.annotation.JsonInclude;\nimport com.fasterxml.jackson.annotation.JsonProperty;\nimport lombok.AllArgsConstructor;\nimport lombok.Builder;\nimport lombok.Data;\nimport lombok.NoArgsConstructor;\nimport lombok.experimental.FieldNameConstants;\n\n@Data\n@Builder\n@NoArgsConstructor\n@AllArgsConstructor\n@FieldNameConstants\n@JsonIgnoreProperties(ignoreUnknown = true)\n@JsonInclude(JsonInclude.Include.NON_NULL)\npublic class ImageUnderstandingResponse {\n\n @JsonProperty(\"header\")\n private ImageHeader imageHeader;\n\n @JsonProperty(\"payload\")\n private ImageUnderstandingPayload imageUnderstandingPayload;\n\n}\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/interceptor/BaseUrlInterceptor.java",
"content": "package com.ai.spark.interceptor;\n\nimport lombok.extern.slf4j.Slf4j;\nimport okhttp3.Interceptor;\nimport okhttp3.Request;\nimport okhttp3.Response;\n\nimport java.io.IOException;\n\n\n@Slf4j\npublic class BaseUrlInterceptor implements Interceptor {\n // TODO 对路径进行拦截\n @Override\n public Response intercept(Chain chain) throws IOException {\n //获取request\n Request request = chain.request();\n return chain.proceed(request);\n }\n}\n\n"
},
{
"path": "ai-spark/src/main/java/com/ai/spark/interceptor/ResponseInterceptor.java",
"content": "package com.ai.spark.interceptor;\n\nimport com.ai.core.exception.BaseException;\nimport com.ai.core.exception.Constants;\nimport lombok.extern.slf4j.Slf4j;\nimport okhttp3.Interceptor;\nimport okhttp3.Request;\nimport okhttp3.Response;\n\nimport java.io.IOException;\n\n/**\n * 返回信息拦截器\n */\n@Slf4j\npublic class ResponseInterceptor implements Interceptor {\n @Override\n public Response intercept(Chain chain) throws IOException {\n // 1. 获取 req 和 resp\n Request original = chain.request();\n Response response = chain.proceed(original);\n // 2. 排除webSocket连接，判断返回状态\n if (!\"websocket\".equalsIgnoreCase(response.header(\"Upgrade\"))\n && !\"Upgrade\".equalsIgnoreCase(response.header(\"Connection\"))\n && !response.isSuccessful()\n && response.body() != null) {\n // 2.1 获取返回的错误信息\n log.error(response.body().string());\n throw new BaseException(Constants.ErrorMsg.RETRY_ERROR);\n }\n return response;\n }\n}\n"
},
{
"path": "ai-spark/src/test/java/com/ai/spark/AudioApiTest.java",
"content": "package com.ai.spark;\n\nimport com.ai.core.strategy.impl.FirstKeyStrategy;\nimport com.ai.spark.achieve.ApiData;\nimport com.ai.spark.achieve.Configuration;\nimport com.ai.spark.achieve.defaults.DefaultSparkSessionFactory;\nimport com.ai.spark.achieve.standard.session.AggregationSession;\nimport org.junit.Before;\n\nimport java.util.Arrays;\n\n\npublic class AudioApiTest {\n\n private AggregationSession aggregationSession;\n\n @Before\n public void before() {\n // 1. 创建配置类\n Configuration configuration = new Configuration();\n configuration.setApiHost(\"https://spark-api.xf-yun.com\");\n // 3. 设置鉴权所需的API Key,可设置多个。\n ApiData apiData = ApiData.builder()\n .apiKey(\"***********************\")\n .apiSecret(\"***********************\")\n .appId(\"***********************\")\n .build();\n configuration.setKeyList(Arrays.asList(apiData));\n // 4. 设置请求时 key 的使用策略，默认实现了：随机获取 和 固定第一个Key 两种方式。\n configuration.setKeyStrategy(new FirstKeyStrategy());\n// configuration.setKeyStrategy(new RandomKeyStrategy());\n // 5. 设置代理，若不需要可不设置\n// configuration.setProxy(new Proxy(Proxy.Type.HTTP, new InetSocketAddress(\"127.0.0.1\", 7890)));\n // 6. 创建 session 工厂，制造不同场景的 session\n DefaultSparkSessionFactory factory = new DefaultSparkSessionFactory(configuration);\n this.aggregationSession = factory.openAggregationSession();\n }\n\n}\n"
},
{
"path": "ai-spark/src/test/java/com/ai/spark/ChatApiTest.java",
"content": "package com.ai.spark;\n\n\nimport com.ai.core.strategy.impl.FirstKeyStrategy;\nimport com.ai.spark.achieve.ApiData;\nimport com.ai.spark.achieve.Configuration;\nimport com.ai.spark.achieve.defaults.DefaultSparkSessionFactory;\nimport com.ai.spark.achieve.defaults.listener.ChatListener;\nimport com.ai.spark.achieve.defaults.listener.DocumentChatListener;\nimport com.ai.spark.achieve.standard.session.AggregationSession;\nimport com.ai.spark.endPoint.chat.ChatText;\nimport com.ai.spark.endPoint.chat.req.ChatRequest;\nimport com.ai.spark.endPoint.chat.req.DocumentChatRequest;\nimport com.ai.spark.endPoint.chat.resp.ChatResponse;\nimport com.ai.spark.endPoint.chat.resp.DocumentChatResponse;\nimport lombok.SneakyThrows;\nimport org.junit.Before;\nimport org.junit.Test;\n\nimport java.util.Arrays;\nimport java.util.concurrent.CountDownLatch;\n\npublic class ChatApiTest {\n\n private AggregationSession aggregationSession;\n\n @Before\n public void before() {\n // 1. 创建配置类\n Configuration configuration = new Configuration();\n configuration.setApiHost(\"https://spark-api.xf-yun.com\");\n // 3. 设置鉴权所需的API Key,可设置多个。\n ApiData apiData = ApiData.builder()\n .apiKey(\"**********************\")\n .apiSecret(\"**********************\")\n .appId(\"**********************\")\n .build();\n configuration.setKeyList(Arrays.asList(apiData));\n // 4. 设置请求时 key 的使用策略，默认实现了：随机获取 和 固定第一个Key 两种方式。\n configuration.setKeyStrategy(new FirstKeyStrategy());\n// configuration.setKeyStrategy(new RandomKeyStrategy());\n // 5. 设置代理，若不需要可不设置\n// configuration.setProxy(new Proxy(Proxy.Type.HTTP, new InetSocketAddress(\"127.0.0.1\", 7890)));\n // 6. 创建 session 工厂，制造不同场景的 session\n DefaultSparkSessionFactory factory = new DefaultSparkSessionFactory(configuration);\n this.aggregationSession = factory.openAggregationSession();\n }\n\n /**\n * 测试聊天功能\n */\n @Test\n public void test_chat() {\n // 创建参数\n ChatRequest request = ChatRequest.baseBuild(\"讲一个笑话\", \"c8f362b8\");\n // 设置参数并发起请求，监听事件信息\n aggregationSession.getChatSession().chat(new ChatListener(request) {\n // 异常处理\n @SneakyThrows\n @Override\n public void onChatError(ChatResponse chatResponse) {\n System.out.println(chatResponse);\n }\n\n // 获取正常返回的数据\n @Override\n public void onChatOutput(ChatResponse chatResponse) {\n System.out.println(chatResponse);\n System.out.print(chatResponse.getChatPayload().getChoice().getTexts().get(0).getContent());\n }\n\n // 结束处理\n @Override\n public void onChatEnd() {\n System.out.println(\"当前会话结束了\");\n }\n\n });\n // 等待会话结束\n CountDownLatch countDownLatch = new CountDownLatch(1);\n try {\n countDownLatch.await();\n } catch (InterruptedException e) {\n e.printStackTrace();\n }\n }\n\n /**\n * 测试多轮聊天功能\n */\n @Test\n public void test_chat_multiple() {\n // 创建参数\n ChatRequest request = ChatRequest.baseBuild(\"1+1=\", \"c8f362b8\");\n // 设置第一轮对话的结果\n ChatText chatText1 = ChatText.baseBuild(ChatText.Role.ASSISTANT, \"2\");\n // 设置第二轮对话的问题\n ChatText chatText2 = ChatText.baseBuild(ChatText.Role.USER, \"2+2=\");\n // 将对话过程注入到参数当中\n request.getChatPayload().getMessage().getChatTexts().add(chatText1);\n request.getChatPayload().getMessage().getChatTexts().add(chatText2);\n // 设置参数并发起请求，监听事件信息\n aggregationSession.getChatSession().chat(new ChatListener(request) {\n @Override\n public void onChatError(ChatResponse chatResponse) {\n System.out.println(chatResponse);\n }\n\n @Override\n public void onChatOutput(ChatResponse chatResponse) {\n System.out.print(chatResponse.getChatPayload().getChoice().getTexts().get(0).getContent());\n }\n\n @Override\n public void onChatEnd() {\n System.out.println(\"当前会话结束了\");\n }\n\n });\n // 等待会话结束\n CountDownLatch countDownLatch = new CountDownLatch(1);\n try {\n countDownLatch.await();\n } catch (InterruptedException e) {\n e.printStackTrace();\n }\n }\n\n /**\n * 测试文档对话功能\n */\n @Test\n public void test_document_chat() {\n // 构建参数\n DocumentChatRequest documentChatRequest = DocumentChatRequest.baseBuild(\"总结一下故事一说了什么？\", Arrays.asList(\"c42a68fd31964d43b4f57eab11e9a833\"));\n // 设置阐述并发起请求\n aggregationSession.getChatSession().documentChat(new DocumentChatListener(documentChatRequest) {\n @Override\n public void onChatError(DocumentChatResponse documentChatResponse) {\n System.err.println(documentChatResponse);\n }\n\n @Override\n public void onChatOutput(DocumentChatResponse documentChatResponse) {\n System.out.println(documentChatResponse);\n }\n\n @Override\n public void onChatEnd() {\n System.out.println(\"当前会话结束了\");\n }\n });\n // 等待会话结束\n CountDownLatch countDownLatch = new CountDownLatch(1);\n try {\n countDownLatch.await();\n } catch (InterruptedException e) {\n e.printStackTrace();\n }\n }\n}\n"
},
{
"path": "ai-spark/src/test/java/com/ai/spark/DocumentApiTest.java",
"content": "package com.ai.spark;\n\nimport com.ai.core.strategy.impl.FirstKeyStrategy;\nimport com.ai.spark.achieve.ApiData;\nimport com.ai.spark.achieve.Configuration;\nimport com.ai.spark.achieve.defaults.DefaultSparkSessionFactory;\nimport com.ai.spark.achieve.standard.session.AggregationSession;\nimport com.ai.spark.endPoint.document.req.FileUploadRequest;\nimport com.ai.spark.endPoint.document.resp.DocumentSummaryResponse;\nimport com.ai.spark.endPoint.document.resp.FileUploadResponse;\nimport org.junit.Before;\nimport org.junit.Test;\n\nimport java.io.File;\nimport java.util.Arrays;\n\npublic class DocumentApiTest {\n\n private AggregationSession aggregationSession;\n\n @Before\n public void before() {\n // 1. 创建配置类\n Configuration configuration = new Configuration();\n configuration.setApiHost(\"https://spark-api.xf-yun.com\");\n // 3. 设置鉴权所需的API Key,可设置多个。\n ApiData apiData = ApiData.builder()\n .apiKey(\"***********************\")\n .apiSecret(\"***********************\")\n .appId(\"***********************\")\n .build();\n configuration.setKeyList(Arrays.asList(apiData));\n // 4. 设置请求时 key 的使用策略，默认实现了：随机获取 和 固定第一个Key 两种方式。\n configuration.setKeyStrategy(new FirstKeyStrategy());\n// configuration.setKeyStrategy(new RandomKeyStrategy());\n // 5. 设置代理，若不需要可不设置\n// configuration.setProxy(new Proxy(Proxy.Type.HTTP, new InetSocketAddress(\"127.0.0.1\", 7890)));\n // 6. 创建 session 工厂，制造不同场景的 session\n DefaultSparkSessionFactory factory = new DefaultSparkSessionFactory(configuration);\n this.aggregationSession = factory.openAggregationSession();\n }\n\n /**\n * 测试文件上传功能\n */\n @Test\n public void test_file_upload() {\n // 读取文件\n File file = new File(\"D:\\\\chatGPT-api\\\\AI-java\\\\doc\\\\test\\\\test_file_upload.pdf\");\n // 构建参数\n FileUploadRequest request = FileUploadRequest.builder().file(file).build();\n // 发起请求获取结果\n FileUploadResponse fileUploadResponse = this.aggregationSession.getDocumentSession().fileUpload(request);\n System.out.println(fileUploadResponse);\n// FileUploadResponse(code=0, sid=28db14303e054046aabd2e96e7e65c51, desc=null, data=Data(fileId=1a477e7e9cb44e23ad4abd98076e3f70))\n// FileUploadResponse(code=0, sid=8e4f267415d84827a6ec7a1580e1ce64, desc=null, data=Data(fileId=004c3c6e79bc4d738a7e94a12697ea75))\n }\n\n // 文档总结和文档总结查询这两个接口其实只有请求路径不同，类似于异步的效果。\n // 调用文档总结接口，并不会直接返回结果，而是通知模型开始进行总结。\n // 然后调用文档总结查询接口查询结果，如果结果已经存在的情况下，不管是调用文档总结接口还是文档总结查询接口，返回的数据都是一样的。\n\n /**\n * 测试文档总结功能\n */\n @Test\n public void test_document_summary_start() {\n // 传入文档ID，发起请求\n DocumentSummaryResponse documentSummaryResponse = this.aggregationSession.getDocumentSession()\n .documentSummaryStart(\"004c3c6e79bc4d738a7e94a12697ea75\");\n System.out.println(documentSummaryResponse);\n }\n\n /**\n * 测试文档总结结果查询功能\n */\n @Test\n public void test_document_summary_query() {\n // 传入文档ID，发起请求\n DocumentSummaryResponse documentSummaryResponse = this.aggregationSession.getDocumentSession()\n .documentSummaryQuery(\"004c3c6e79bc4d738a7e94a12697ea75\");\n System.out.println(documentSummaryResponse);\n }\n}\n"
},
{
"path": "ai-spark/src/test/java/com/ai/spark/EmbeddingApiTest.java",
"content": "package com.ai.spark;\n\n\nimport com.ai.core.strategy.impl.FirstKeyStrategy;\nimport com.ai.spark.achieve.ApiData;\nimport com.ai.spark.achieve.Configuration;\nimport com.ai.spark.achieve.defaults.DefaultSparkSessionFactory;\nimport com.ai.spark.achieve.standard.session.AggregationSession;\nimport com.ai.spark.endPoint.chat.ChatText;\nimport com.ai.spark.endPoint.embedding.req.EmbeddingRequest;\nimport com.ai.spark.endPoint.embedding.resp.EmbeddingResponse;\nimport org.junit.Before;\nimport org.junit.Test;\n\nimport java.util.Arrays;\n\npublic class EmbeddingApiTest {\n\n\n private AggregationSession aggregationSession;\n\n @Before\n public void before() {\n // 1. 创建配置类\n Configuration configuration = new Configuration();\n configuration.setApiHost(\"https://spark-api.xf-yun.com\");\n // 3. 设置鉴权所需的API Key,可设置多个。\n ApiData apiData = ApiData.builder()\n .apiKey(\"**********************\")\n .apiSecret(\"**********************\")\n .appId(\"**********************\")\n .build();\n configuration.setKeyList(Arrays.asList(apiData));\n // 4. 设置请求时 key 的使用策略，默认实现了：随机获取 和 固定第一个Key 两种方式。\n configuration.setKeyStrategy(new FirstKeyStrategy());\n// configuration.setKeyStrategy(new RandomKeyStrategy());\n // 5. 设置代理，若不需要可不设置\n// configuration.setProxy(new Proxy(Proxy.Type.HTTP, new InetSocketAddress(\"127.0.0.1\", 7890)));\n // 6. 创建 session 工厂，制造不同场景的 session\n DefaultSparkSessionFactory factory = new DefaultSparkSessionFactory(configuration);\n this.aggregationSession = factory.openAggregationSession();\n }\n\n /**\n * 测试文本嵌入功能\n */\n @Test\n public void test_embedding() {\n // 构造参数信息\n ChatText chatText = ChatText.baseBuild(ChatText.Role.USER, \"这是一段文字\");\n EmbeddingRequest request = EmbeddingRequest.baseBuild(chatText, \"c8f362b8\");\n // 发起请求\n EmbeddingResponse response = aggregationSession.getEmbeddingSession().embed(request);\n System.out.println(response);\n }\n\n}\n"
},
{
"path": "ai-spark/src/test/java/com/ai/spark/ImageApiTest.java",
"content": "package com.ai.spark;\n\nimport com.ai.core.strategy.impl.FirstKeyStrategy;\nimport com.ai.spark.achieve.ApiData;\nimport com.ai.spark.achieve.Configuration;\nimport com.ai.spark.achieve.defaults.DefaultSparkSessionFactory;\nimport com.ai.spark.achieve.defaults.listener.ImageUnderstandingListener;\nimport com.ai.spark.achieve.standard.session.AggregationSession;\nimport com.ai.spark.endPoint.images.req.ImageCreateRequest;\nimport com.ai.spark.endPoint.images.req.ImageUnderstandingRequest;\nimport com.ai.spark.endPoint.images.resp.ImageCreateResponse;\nimport com.ai.spark.endPoint.images.resp.ImageUnderstandingResponse;\nimport org.junit.Before;\nimport org.junit.Test;\n\nimport javax.imageio.ImageIO;\nimport javax.xml.bind.DatatypeConverter;\nimport java.awt.image.BufferedImage;\nimport java.io.ByteArrayInputStream;\nimport java.io.File;\nimport java.io.IOException;\nimport java.util.Arrays;\nimport java.util.concurrent.CountDownLatch;\n\npublic class ImageApiTest {\n\n private AggregationSession aggregationSession;\n\n @Before\n public void before() {\n // 1. 创建配置类\n Configuration configuration = new Configuration();\n configuration.setApiHost(\"https://spark-api.xf-yun.com\");\n // 3. 设置鉴权所需的API Key,可设置多个。\n ApiData apiData = ApiData.builder()\n .apiKey(\"**********************\")\n .apiSecret(\"**********************\")\n .appId(\"**********************\")\n .build();\n configuration.setKeyList(Arrays.asList(apiData));\n // 4. 设置请求时 key 的使用策略，默认实现了：随机获取 和 固定第一个Key 两种方式。\n configuration.setKeyStrategy(new FirstKeyStrategy());\n// configuration.setKeyStrategy(new RandomKeyStrategy());\n // 5. 设置代理，若不需要可不设置\n// configuration.setProxy(new Proxy(Proxy.Type.HTTP, new InetSocketAddress(\"127.0.0.1\", 7890)));\n // 6. 创建 session 工厂，制造不同场景的 session\n DefaultSparkSessionFactory factory = new DefaultSparkSessionFactory(configuration);\n this.aggregationSession = factory.openAggregationSession();\n }\n\n /**\n * 测试图片生成功能\n */\n @Test\n public void test_image_create() throws IOException {\n // 创建请求参数\n ImageCreateRequest request = ImageCreateRequest.baseBuild(\"画一座大山\", \"c8f362b8\");\n // 发起请求获取结果\n ImageCreateResponse imageCreateResponse = aggregationSession.getImageSession().imageCreate(request);\n // 得到结果当中的base64 图片字符串\n String content = imageCreateResponse.getImagePayload().getChoice().getTexts().get(0).getContent();\n // 转换为byte数组\n byte[] imageBytes = DatatypeConverter.parseBase64Binary(content.substring(content.indexOf(\",\") + 1));\n // 读取byte数组，存放到指定文件路径\n BufferedImage bufferedImage = ImageIO.read(new ByteArrayInputStream(imageBytes));\n File outputFile = new File(\"D:\\\\chatGPT-api\\\\AI-java\\\\doc\\\\test\\\\test_create_image.png\");\n ImageIO.write(bufferedImage, \"png\", outputFile);\n }\n\n @Test\n public void test_image_understanding() {\n String filePath = \"D:\\\\chatGPT-api\\\\AI-java\\\\doc\\\\test\\\\test_create_image.png\";\n File file = new File(filePath);\n ImageUnderstandingRequest request = ImageUnderstandingRequest.baseBuild(\"这张图片的内容是什么？\", \"c8f362b8\", file);\n aggregationSession.getImageSession().imageUnderstanding(request, new ImageUnderstandingListener(request) {\n @Override\n public void onChatError(ImageUnderstandingResponse imageUnderstandingResponse) {\n System.err.println(imageUnderstandingResponse);\n }\n\n @Override\n public void onChatOutput(ImageUnderstandingResponse imageUnderstandingResponse) {\n System.out.println(imageUnderstandingResponse.getImageUnderstandingPayload().getChoice().getTexts().get(0).getContent());\n }\n\n @Override\n public void onChatEnd() {\n }\n });\n // 等待会话结束\n CountDownLatch countDownLatch = new CountDownLatch(1);\n try {\n countDownLatch.await();\n } catch (InterruptedException e) {\n e.printStackTrace();\n }\n }\n\n\n}\n"
},
{
"path": "doc/test/test_file_upload.txt",
"content": "故事1：小老虎问路\n\n一头骄傲的小老虎在大森林里迷了路。\n\n他走啊走，看到了一头正在蒙头大睡的野猪。小老虎对着野猪的耳朵，大声喊道：“喂，蠢猪，别打呼噜了，告诉我回家的路怎么走吧!”野猪生气地眨了眨眼睛，一言不发，把屁股转向了小老虎，继续睡大觉。小老虎讨了个没趣，无奈地走了。\n\n小老虎问路\n\n路上，他看到一只正在忙碌的小松鼠，于是他用自己的大嗓门儿喊道：“喂，如果你告诉我回家的路怎么走，我就让妈妈给你最好的礼物!”小松鼠就像没听见一样，不搭理小老虎，照样干自己的活儿。小老虎勃然大怒，冲向一只戴眼镜的老灰兔：“嘿，花眼的老兔头，快给我指一条回家的路!”老灰兔慢慢地抬起头，和蔼地说：“森林里的路大家都熟悉，可你这样没礼貌，哪怕你问遍所有的动物，你还是找不到回家的路。”听了老灰兔的话，小老虎猛然醒悟过来——对人说话，要有礼貌才行!\n\n这时，前面过来一只梅花鹿。小老虎走过去，礼貌地说：“梅花鹿你好，请你告诉我回家的路怎么走，好吗?”梅花鹿热情地告诉了小老虎，小老虎高兴的连声说：“谢谢你，梅花鹿，谢谢你!”\n\n小老虎终于安全地回到了自己的小屋。\n\n\n\n故事2：鸡和猫调工作\n\n张三家里养了一只猫和一只公鸡。\n\n一天早晨，“喔，喔，喔……”公鸡的长鸣把沉睡的猫叫醒了，猫揉了揉眼说：“死公鸡，没娘教的孩子，吵死啊!”公鸡大声地回骂道：“死懒猫，你才没娘教，太阳晒屁股了，还不起来干活!”猫又回敬了一句：“哼，我晚上辛辛苦苦晚上捉老鼠，白天还不让我多睡会儿?”公鸡无语了。\n\n过了一会儿，猫想出一个好办法，对公鸡说：“不如，我们调换一下工作，你去捉老鼠，我来打鸣叫时钟，你敢不敢?”公鸡自信地说：“啊有不敢?我就怕你不敢?”猫说：“好，那我们就调工作吧!”\n\n第二天早晨，猫早早地起来，站在房顶上，“喵，喵，喵……”地叫着，可是人们都听习惯了公鸡叫天亮，猫的叫声只把附近的人叫醒了，远处的人根本听不见。后来，猫的嗓子都喊破了，被送到宠物医院治疗。\n\n晚上，公鸡来到老鼠洞前，对老鼠说：“鸡大爷来了，快给我出来!”鼠王以为是猫的诡计，就叫了一只小老鼠去侦察，小老鼠侦察了一番回来说：“站在洞门口的是一只公鸡。”于是，鼠们就往老鼠洞里钻。公鸡见一只老鼠都不出来，就把锋利的爪子伸进老鼠洞里，一只老鼠胆大包天串到洞门口对准鸡爪狠狠地咬了一口，鸡痛得哇哇大叫，也被送进了宠物医院。\n\n猫和公鸡在宠物医院相遇了，当他们见彼此都受了伤后，明白了一个道理：尺有所短，寸有所长。\n\n\n\n故事3：固执的鱼\n\n在一个池塘里住着一条小鱼。他有一个好朋友，是一只蝌蚪。这条小鱼和蝌蚪总是在一起游泳，一起找食物，一起玩。\n\n一天早上，小鱼吃惊地看见在蝌蚪尾巴的两边长出了一对腿。小鱼问蝌蚪为什么他会有腿。\n\n“我不是鱼，我是一只蝌蚪，一只年幼的青蛙。以后，等我长大了，我就不再呆在这儿了。”蝌蚪回答说。\n\n“你说谎!”小鱼说。\n\n“如果你不相信，就等着瞧吧!”蝌蚪说。\n\n小鱼已经三天没有看见蝌蚪了，他很担心。为了寻找朋友，他搜索了每一个地方，蝌蚪能去哪儿呢?\n\n几天以后，蝌蚪又出现了。小鱼非常高兴，可是他又吃了一惊，蝌蚪又长出了一对腿，哦，还有，尾巴也变短了。\n\n“这几天你到什么地方去了?”小鱼问。\n\n“我去陆地上了，我不是告诉你我在这儿呆不长吗?看，现在我有四条腿了。不久我就要长期在地上生活了。”蝌蚪说，“现在请不要再叫我蝌蚪，就叫我青蛙吧。再见了，鱼!”\n\n小鱼眨眨眼睛，被他朋友的话搞糊涂了。他不能相信听到和看到的，因为从前他的朋友能像鱼一样游泳，从前他没有腿，而现在却不是这样了。\n\n小鱼独自留在池塘里，最后，他变成了一条大鱼。\n\n一天，当这条鱼在池塘里游水寻找食物的时候，一只青蛙跳进水里。他就是鱼的老朋友。看到朋友，鱼非常高兴。\n\n“你去哪儿啦?”鱼问。\n\n“我一直在陆地上。”青蛙说，并且把他在陆地上遇到的事情告诉鱼。\n\n鱼听了青蛙的故事后，问：“在那边谁是你的朋友呀?”\n\n“我有好多朋友，像牛啦，鸟啦，猫啦，还有其它许多动物。”青蛙说。\n\n“我能跟你去陆地吗?我想见见他们。”鱼说。\n\n“那怎么行!你在陆地上不能呼吸，你会死的。”青蛙解释说。\n\n“可是我想去看看牛啦，鸟啦，还有别的你刚才告诉我的朋友们。”鱼请求道。\n\n“你不必亲自去，我给你说说不就行了。”青蛙说。于是青蛙向鱼讲了许多他的陆地朋友的事。鱼试着想象那些动物的样子，但是他总是不满意。\n\n“你在陆地上还看见什么别的吗?”鱼又问。\n\n“还有人，有孩子，有玩具和别的许多东西。”青蛙继续说。\n\n他们一直聊到晚上。鱼很不开心，因为他不能去陆地上看这些奇怪的事，这天晚上他失眠了。他满脑子都是白天听到的各种各样的事情。\n\n第二天早上，鱼去寻找食物。忽然，他看见水面上天空飞鸟的倒影。他太想看看鸟了，就鼓起勇气试着跳到河岸上。鱼一纵身跳上了岸。但是在他睁开眼睛之前，他已经喘不过气了。他开始呻吟起来。算他走运，青蛙正好在附近找吃的。\n\n青蛙马上跳到奄奄一息的鱼的身边。他一点也没耽搁，把大鱼拉进池塘里。鱼一进水，立刻苏醒了。他很惊讶，问青蛙发生了什么事。\n\n青蛙微笑着说：“我跟你说过，你不能到陆地上去。不管你是在陆地上还是在水里都没关系，一切都是美好的，美丽的，为什么你不愿意听我的话!”\n\n“可是只呆在这儿我觉得不满意。”鱼继续说。\n\n“你应该满足了，”青蛙劝鱼说：“没有多少生命能像你一样呆在水里。”\n\n鱼笑了，高兴地在水草间游来游去，他认识到他的朋友说的是真的。\n\n\n\n故事4：橘子老虎\n\n秋天，橘子熟了，那黄澄澄的蜜橘挂满枝头，远远望去，就像一个个的小灯笼。一天，一个最大、最沉的橘子，看到同伴被人摘走，伤心地对橘子树说：“妈妈，难道我们橘子生来就应该被人吃掉吗?”\n\n“是啊，孩子，我们的最大愿望就是丰富人类的美好生活。我们身上的果核将会落入泥土，然后新的生命又会破土而出。”\n\n“不，妈妈，我可不愿为他人活着，更不愿被人随意摘取，我要变成一个人见人怕的老虎。”\n\n“孩子，那可不是我们橘子的风格。”\n\n“不嘛!”大橘子纵身一跳，从树上跳到了地面。奇怪的事情发生了：大橘子的身体不断地胀大、胀大，最后圆圆的橘子肚子拉长了，橘皮上竟然浮现出色彩斑斓的花纹来，前面拱出一个脑袋，脑门正中有一个醒目的“王”字，后面露出一条长尾巴来。哈，大橘子竟然变成一只威风凛凛的橘子老虎啦!\n\n橘子老虎非常高兴，她告别妈妈，决定到各地去旅行，让大家见识见识橘子老虎的威风。他翻过一座山冈，看见一只小羊正趴在大树下低声哭泣。他决定吓唬吓唬小羊。他蹑手蹑脚地走上前去，却发现一只大灰狼也在偷偷地逼近小羊。橘子老虎见状忙喊道：“小羊，当心大灰狼!”\n\n小羊一惊，扭头要跑，大灰狼扑上来抓住了小羊。在这千钧一发的时刻，只见橘子老虎一个箭步蹿上前去，对着大灰狼喊道：“大灰狼，快放开他，要不我就撕碎你!”\n\n大灰狼一愣，见一只猛虎向自己扑来，不由两腿直打颤。虽然舍不得到嘴的肥嫩小羊，但这只老虎可不是等闲之辈，只得丢下小羊逃走了。面对橘子老虎，小羊流着眼泪说道：“虎大王，我妈妈得了重病，想吃橘子，您能否等我找到橘子后再吃我呢?”\n\n看看楚楚可怜的小羊，橘子老虎被他爱妈妈的孝心所感动了。他安慰小羊：“别怕，让我帮助你完成心愿。”说着，他便撕开自己的橘子肚皮，掰下一瓣橘子，递给小羊说，“给你妈妈送回去吧。”\n\n小羊感激地说：“您真是天下心眼最好的老虎呀!”\n\n橘子老虎笑了笑，继续往前赶路了。\n\n\n\n故事5：狐狸假扮兽王\n\n很早以前，森林中的百兽过着闲逸、安乐的生活。因没有兽王，便商议决定寻找一个有资格作兽王的动物来领导群兽，于是四处寻觅。一天，有只狐狸跑到一家染衣坊寻找食物，不慎掉进了染缸。它惊恐万分，拼命挣扎，等到爬出染缸时，已是精疲力尽。狐狸再也没有心思寻找食物，落荒而逃。它在河边喝水时，见到水中的倒影，忽然发现自己身上的颜色变得美丽异常，与众不同。狐狸自己知道那是在染缸里染上的。正在这时，寻找兽王的动物们发现了它，惊奇地问它是什么动物?是从什么地方来的?狐狸灵机一动，诈称自己是天帝派来作兽王的。群兽从来没有见到过它这样的动物，又听说是天帝派来的，便生起信心，拥立狐狸为王。\n\n当上兽王的狐狸，得意忘形，作威作福。它不但役使所有的野兽为自己做事，还忘乎所以地让狮子当坐骑，四处巡视游玩。照理说狐狸当了兽王，应该对自己的同类特别关照才是，但这兽王并没有这样做，反而痛恨狐群，百般加以折磨。动物们本以为有兽王领导，生活会更加幸福、快乐，没想到却落得如此痛苦。众狐狸更觉得是飞来的横祸，大惑不解，暗地里对兽王进行观察，它们怀疑这天帝所赐的兽王可能是狐狸装扮的。众狐狸找了个机会，偷偷地询问狮子：“每月十五，月圆之日，兽王是否仍要骑着你去游玩?”狮子说：“不，兽王每月十五都给我放假，它总是单独离去。”群狐说：“我们狐狸因为业力的关系，每到十五日就会昏迷一阵，好一会儿才能恢复。你可以在十五日那天跟踪兽王，看它是不是狐狸所扮?”\n\n等到十五日，兽王照常向远处跑去。狮子便悄悄地跟在后面，到了一个山洞里，果然看见兽王象死尸一样倒在地上，昏迷不醒。狮子这才知道动物们都上当受骗了，尤其是自己，居然被狐狸当坐骑戏弄了这么长的时间，狮子羞怒难当，一跃而上将这只狐狸吞食了……\n\n群兽因为没有好好观察，让一只卑劣的狐狸当了兽王。最后的结果是让群兽都受到了莫大的痛苦，那自作聪明的狐狸也自取灭亡。"
},
{
"path": "pom.xml",
"content": "\n\n 4.0.0\n\n com.ai\n AI-java\n pom\n 1.0\n \n ai-common\n ai-openai\n ai-spark\n ai-baidu\n ai-core\n \n AI-java\n ai-sdk-java\n\n \n AI-java\n \n \n org.apache.maven.plugins\n maven-surefire-plugin\n 2.12.4\n \n true\n \n \n \n org.apache.maven.plugins\n maven-compiler-plugin\n \n 8\n 8\n \n \n \n org.apache.maven.plugins\n maven-jar-plugin\n 2.3.1\n \n \n \n true\n \n \n \n \n \n \n\n\n"
}
]