{"id":690694,"date":"2026-03-24T04:32:49","date_gmt":"2026-03-24T03:32:49","guid":{"rendered":"https:\/\/blog.jetbrains.com\/?post_type=kotlin&p=690694"},"modified":"2026-03-24T04:32:52","modified_gmt":"2026-03-24T03:32:52","slug":"introducing-tracy-the-ai-observability-library-for-kotlin","status":"publish","type":"kotlin","link":"https:\/\/blog.jetbrains.com\/zh-hans\/kotlin\/2026\/03\/introducing-tracy-the-ai-observability-library-for-kotlin\/","title":{"rendered":"\u63a8\u51fa Tracy\uff1a\u9002\u7528\u4e8e Kotlin \u7684 AI \u53ef\u89c2\u6d4b\u6027\u5e93"},"content":{"rendered":"

Tracy<\/a> \u662f\u4e00\u6b3e\u5f00\u6e90 Kotlin \u5e93\uff0c\u53ef\u4ee5\u5728\u51e0\u5206\u949f\u5185\u4e3a AI \u8d4b\u80fd\u7684\u5e94\u7528\u7a0b\u5e8f\u6dfb\u52a0\u751f\u4ea7\u7ea7\u53ef\u89c2\u6d4b\u6027\u3002\u5b83\u53ef\u4ee5\u5e2e\u52a9\u60a8\u8c03\u8bd5\u6545\u969c\u3001\u6d4b\u7b97\u6267\u884c\u65f6\u95f4\uff0c\u5e76\u8ddf\u8e2a\u6a21\u578b\u8c03\u7528\u3001\u5de5\u5177\u8c03\u7528\u548c\u60a8\u7684\u81ea\u5b9a\u4e49\u5e94\u7528\u7a0b\u5e8f\u903b\u8f91\u4e2d\u7684 LLM \u4f7f\u7528\u60c5\u51b5\u3002\u5f52\u6839\u7ed3\u5e95\uff0c\u5168\u9762\u7684\u53ef\u89c2\u6d4b\u6027\u786e\u4fdd\u60a8\u53ef\u4ee5\u83b7\u53d6\u4e86\u89e3\u771f\u5b9e\u5e94\u7528\u7a0b\u5e8f\u884c\u4e3a\u7684\u51c6\u786e\u6570\u636e\uff0c\u4ece\u5b8f\u89c2\u8d8b\u52bf\u5230\u7ec6\u7c92\u5ea6\u8ddf\u8e2a\u5206\u6790\u6027\u80fd\uff0c\u4e3a\u5168\u9762\u7684\u5728\u7ebf\u548c\u79bb\u7ebf\u8bc4\u4f30\u63d0\u4f9b\u652f\u6301\u3002<\/p>\n

Tracy \u4e0e\u5e38\u7528\u7684 Kotlin\/LLM \u6280\u672f\u6808\uff08\u5305\u62ec OkHttp<\/a> \u548c Ktor<\/a> \u5ba2\u6237\u7aef\uff0c\u4ee5\u53ca OpenAI<\/a>\u3001Anthropic<\/a> \u548c Gemini<\/a> \u670d\u52a1\uff09\u65e0\u7f1d\u534f\u4f5c\uff0c\u5b83\u7684\u5e95\u5c42\u57fa\u4e8e OpenTelemetry \u6784\u5efa\u3002\u6b64\u67b6\u6784\u53ef\u4ee5\u786e\u4fdd\u5f00\u53d1\u8005\u5168\u9762\u7075\u6d3b\u638c\u63a7\u8ddf\u8e2a\u6570\u636e\uff0c\u652f\u6301\u4ee5\u6807\u51c6\u5f62\u5f0f\u5c06\u6570\u636e\u5bfc\u51fa\u81f3\u4efb\u4f55\u517c\u5bb9\u7684\u540e\u7aef\uff08\u5982 Jaeger<\/a>\u3001Zipkin \u6216 Grafana\uff09\uff0c\u5e76\u76f4\u63a5\u4e0e Langfuse<\/a> \u548c W&B Weave<\/a> \u8fd9\u7c7b\u4e13\u7528 LLM \u5de5\u7a0b\u5e73\u53f0\u96c6\u6210\u3002<\/p>\n

\u5c3d\u7ba1 Spring AI \u6216 Koog \u7b49\u6210\u719f\u7684 AI \u6846\u67b6\u63d0\u4f9b\u5185\u7f6e\u53ef\u89c2\u6d4b\u6027\uff0c\u4f46 LLM \u8c03\u7528\u5fc5\u987b\u5b8c\u5168\u901a\u8fc7\u5176\u6846\u67b6\u53d1\u8d77\u624d\u80fd\u5b9e\u73b0\u8ddf\u8e2a\uff0c\u5e76\u4e14\u8fd9\u7c7b\u6846\u67b6\u65e0\u6cd5\u901a\u8fc7\u7b80\u6613\u7684\u65b9\u5f0f\u8ddf\u8e2a\u5185\u90e8\u5e94\u7528\u7a0b\u5e8f\u6d41\u3002\u76f8\u6bd4\u4e4b\u4e0b\uff0cTracy \u53ef\u4ee5\u901a\u8fc7 API \u6216 HTTP \u5ba2\u6237\u7aef\u63d2\u6869\u7684\u65b9\u5f0f\u6765\u76d1\u6d4b LLM \u4f7f\u7528\u60c5\u51b5\u3002\u5b83\u8fd8\u901a\u8fc7\u4e3a Kotlin \u51fd\u6570\u6216\u4ee3\u7801\u5757\u6dfb\u52a0\u6ce8\u89e3\u7684\u65b9\u5f0f\u5e2e\u52a9\u60a8\u7406\u6e05 AI \u7ec4\u4ef6\u6216\u5185\u90e8 AI \u667a\u80fd\u4f53\u72b6\u6001\u7684\u65f6\u5e8f\u4e0e\u56e0\u679c\u5173\u7cfb\u3002<\/p>\n

\u6211\u4eec\u5df2\u5f00\u653e Tracy \u7684\u6e90\u4ee3\u7801\uff0c\u8bda\u9080\u60a8\u5e2e\u52a9\u6211\u4eec\u6269\u5c55\u5176\u529f\u80fd \u2013 \u65e0\u8bba\u662f\u8bf7\u6c42\u65b0\u589e AI \u540e\u7aef\u6216 API \u5ba2\u6237\u7aef\u96c6\u6210\uff0c\u8fd8\u662f\u63d0\u4ea4\u62c9\u53d6\u8bf7\u6c42\u5b9e\u73b0\u76f8\u5e94\u529f\u80fd\u3002<\/p>\n

AI \u53ef\u89c2\u6d4b\u6027\u7684\u7ec4\u4ef6\u548c Tracy \u7684\u89e3\u51b3\u65b9\u6848<\/h2>\n

\u4f5c\u4e3a\u5de5\u7a0b\u5e08\uff0c\u65e0\u8bba\u662f\u4e3a\u73b0\u6709\u5e94\u7528\u7a0b\u5e8f\u6dfb\u52a0\u53ef\u89c2\u6d4b\u6027\uff0c\u8fd8\u662f\u4ece\u5934\u5f00\u59cb\u6784\u5efa\u65b0\u5e94\u7528\u7a0b\u5e8f\uff0c\u6211\u4eec\u90fd\u60f3\u8981\u8ddf\u8e2a\u3001\u5b58\u50a8\u548c\u5206\u6790\u4ee5\u4e0b\u5185\u5bb9\uff1a<\/p>\n

    \n
  1. LLM \u8c03\u7528\u5143\u6570\u636e\uff0c\u5305\u62ec\u88ab\u8c03\u7528\u7684 API\u3001\u6a21\u578b\u53ca\u5176\u53c2\u6570\u3002\u6211\u4eec\u53ef\u4ee5\u9009\u62e9\u5728\u5f00\u53d1\u8fc7\u7a0b\u4e2d\u8ddf\u8e2a LLM \u8f93\u5165\u548c\u8f93\u51fa\uff0c\u4ee5\u8fdb\u884c\u8c03\u8bd5\uff0c\u540c\u65f6\u786e\u4fdd\u751f\u4ea7\u73af\u5883\u4e2d\u4e0d\u4f1a\u8ddf\u8e2a\u8fd9\u4e9b\u6570\u636e\u3002<\/li>\n
  2. \u5f15\u53d1\u548c\u54cd\u5e94 LLM \u8c03\u7528\u7684\u5e94\u7528\u7a0b\u5e8f\u903b\u8f91\u6d41 \u2013 \u5373\u67d0\u6b21\u8c03\u7528\u7684\u53d1\u8d77\u4f4d\u7f6e\u53ca\u6d89\u53ca\u7684\u5de5\u5177\u3002<\/li>\n<\/ol>\n

    \u60f3\u8c61\u4e00\u4e2a\u975e\u5e38\u7b80\u5355\u7684\u95ee\u5019\u7528\u6237\u7684 LLM \u804a\u5929\u5e94\u7528\u7a0b\u5e8f\uff0c\u901a\u8fc7\u90e8\u7f72\u5de5\u5177\u8ba9\u95ee\u5019\u66f4\u5177\u4e2a\u6027\u5316\u3002\u5229\u7528 OpenAI \u5ba2\u6237\u7aef\uff0c\u5e94\u7528\u7a0b\u5e8f\u4ee3\u7801\u53ef\u80fd\u662f\u8fd9\u6837\u7684\uff1a<\/p>\n

    \/** Interface for LLM tool *\/\ninterface Tool {\n   \/** Tool call *\/\n   fun execute(): T\n}\n\n\/** Gets the current user's name from the system *\/\nclass GetUserName() : Tool { ... }\n\n\/** Gets the current date and time *\/\nclass GetCurrentDateTime() : Tool { ... }\n\nfun main() {\n   \/\/ Create OpenAI-client using environment variables\n   val client: OpenAIClient = OpenAIOkHttpClient.fromEnv()\n   ...\n   val params = ResponseCreateParams.builder()\n       .model(ChatModel.GPT_4O_MINI)\n       .maxOutputTokens(2048)\n       .addTool(GetUserName::class.java)\n       .addTool(GetCurrentDateTime::class.java)\n       .input(ResponseCreateParams.Input.ofResponse(inputs))\n       .build()\n\n   \/\/ Get the response. \n   \/\/ In a real application, it would use a loop to process tool calls.\n   val response: Response = client.responses().create(params)\n   ...\n   println(finalGreeting)\n}<\/pre>\n
    \u6b64\u65f6\uff0c\u9700\u8981\u8ddf\u8e2a\u7684\u91cd\u8981\u5185\u5bb9\u4e3a\uff1a<\/p>\n
      \n
    1. \u95ee\u5019\u667a\u80fd\u4f53\u88ab\u8c03\u7528\u7684\u60c5\u51b5\u3002<\/li>\n
    2. LLM \u8c03\u7528\u3002<\/li>\n
    3. \u5de5\u5177\u6267\u884c\u3002<\/li>\n<\/ol>\n
      \u6211\u4eec\u53ef\u4ee5\u4f7f\u7528\u57fa\u672c\u7684 OpenTelemetry SDK\uff0c\u4f46\u8fd9\u6837\u4fbf\u9700\u8981\u624b\u52a8\u6dfb\u52a0\u63d2\u6869\u4ee3\u7801\uff0c\u4e14\u8fd9\u4f1a\u5bfc\u81f4\u5de5\u5177\u8c03\u7528\u8ddf\u8e2a\u7684\u4ee3\u7801\u91cd\u590d\u3002\u00a0<\/p>\n
      \u5728\u7406\u60f3\u573a\u666f\u4e0b\uff0c\u6211\u4eec\u5c06\u80fd\u591f\u901a\u8fc7\u914d\u7f6e\u4e00\u6b21\u5de5\u5177\u8ddf\u8e2a\u5bf9\u6240\u6709\u5b9e\u73b0\u8fdb\u884c\u81ea\u52a8\u8ddf\u8e2a\uff0c\u4ece\u800c\u786e\u4fdd\u7edd\u4e0d\u4f1a\u51fa\u73b0\u65b0\u589e\u5de5\u5177\u672a\u88ab\u8ddf\u8e2a\u7684\u95ee\u9898\u3002Tracy \u8ba9\u7406\u60f3\u573a\u666f\u53d8\u4e3a\u73b0\u5b9e\u3002<\/p>\n
      \u901a\u8fc7 Tracy \u6dfb\u52a0\u53ef\u89c2\u6d4b\u6027<\/h2>\n
      Tracy \u63d0\u4f9b\u7684\u4e09\u4e2a\u9ad8\u7ea7 API \u53ef\u4ee5\u5e2e\u52a9\u6211\u4eec\u5168\u9762\u8ddf\u8e2a\u804a\u5929\u5e94\u7528\u7a0b\u5e8f\u3002<\/p>\n
      \u9650\u5b9a\u4f5c\u7528\u57df\u7684 span<\/h3>\n
      withSpan<\/code> API \u53ef\u7528\u4e8e\u521b\u5efa\u9650\u5b9a\u4f5c\u7528\u57df\u7684 span\u3002\u8fd9\u4e9b span \u4f1a\u81ea\u52a8\u5728\u5757\u5f00\u59cb\u65f6\u6fc0\u6d3b\uff0c\u5728\u5757\u7ed3\u675f\u65f6\u7ec8\u6b62\uff0c\u4ece\u800c\u786e\u4fdd\u6b63\u786e\u7684\u5d4c\u5957\u548c\u65f6\u5e8f\u3002\u00a0<\/p>\n
      fun main() {\n\u00a0\u00a0\u00a0\/\/ Encapsulation into withSpan ensures that all nested events will be\n\u00a0\u00a0\u00a0\/\/ traced as part of the greeting agent\u2019s work.\u00a0\u00a0\n\u00a0\u00a0\u00a0withSpan(\"Greeting agent\") {\n\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0...\n\u00a0\u00a0\u00a0}\u00a0\u00a0\n}<\/pre>\n
      LLM \u5ba2\u6237\u7aef\u63d2\u6869\u00a0<\/h3>\n
      LLM \u8c03\u7528\u662f\u4efb\u610f AI \u667a\u80fd\u4f53\u7684\u91cd\u8981\u7ec4\u6210\u90e8\u5206\u3002\u5b83\u4eec\u51b3\u5b9a\u4e86\u5e94\u7528\u7a0b\u5e8f\u7684\u6210\u672c\u3001\u5ef6\u8fdf\u548c\u6548\u7387\uff0c\u662f\u51fa\u73b0\u95ee\u9898\u65f6\u9996\u5148\u8981\u6392\u67e5\u7684\u5bf9\u8c61\u3002\u56e0\u6b64\uff0c\u4e3a LLM \u5ba2\u6237\u7aef\u589e\u52a0\u53ef\u89c2\u6d4b\u6027\u5e94\u7b80\u5355\u6613\u884c\uff0c\u5e76\u5c3d\u91cf\u51cf\u5c11\u5bf9\u4ee3\u7801\u5e93\u7684\u66f4\u6539\u3002\u4f8b\u5982\uff0c\u4e3a OpenAI \u5ba2\u6237\u7aef\u6dfb\u52a0\u63d2\u6869\u5c31\u975e\u5e38\u7b80\u5355\uff0c\u53ea\u9700\u5b8c\u6210\u4ee5\u4e0b\u6b65\u9aa4\uff1a<\/p>\n
      val client = OpenAIOkHttpClient.fromEnv()\n\/\/ All calls made with the instrumented client are traced.\ninstrument(client)<\/pre>\n
      \u9ed8\u8ba4\u60c5\u51b5\u4e0b\uff0c\u5ba2\u6237\u7aef\u63d2\u6869\u4ec5\u4f1a\u8ddf\u8e2a\u5143\u6570\u636e\u3002\u8981\u8ddf\u8e2a\u53ef\u80fd\u5305\u542b\u654f\u611f\u6570\u636e\u7684 LLM \u8f93\u5165\u548c\u8f93\u51fa\uff0c\u5fc5\u987b\u4ee5\u7f16\u7a0b\u65b9\u5f0f\u660e\u786e\u542f\u7528\u6b64\u529f\u80fd\uff0c\u4ee3\u7801\u4e3a\uff1a<\/p>\n
      TracingManager.traceSensitiveContent()<\/pre>\n
      \u4e5f\u53ef\u4ee5\u5728\u8fd0\u884c\u65f6\u542f\u7528\u8ddf\u8e2a\uff0c\u5c06\u73af\u5883\u53d8\u91cf TRACY_CAPTURE_INPUT<\/code> \u548c TRACY_CAPTURE_OUTPUT<\/code> \u8bbe\u4e3a true<\/code> \u5373\u53ef\u3002<\/p>\n
      \u5de5\u5177\u8c03\u7528\u548c\u51fd\u6570\u8ddf\u8e2a<\/h3>\n
      LLM \u9ad8\u5ea6\u4f9d\u8d56\u5de5\u5177\uff1a\u5404\u7c7b\u5de5\u5177\u53ef\u4ee5\u5e2e\u52a9 LLM \u9ad8\u6548\u5b8c\u6210\u786e\u5b9a\u6027\u4efb\u52a1\u3001\u8282\u7701 token\uff0c\u8fd8\u80fd\u4e0e\u5176\u8fd0\u884c\u73af\u5883\u8fdb\u884c\u4ea4\u4e92\u3002\u4f5c\u4e3a\u5f00\u53d1\u8005\uff0c\u6211\u4eec\u540c\u6837\u91cd\u89c6\u5de5\u5177\uff0c\u4f46\u4e3a\u4ee3\u7801\u5e93\u4e2d\u7684 LLM \u5de5\u5177\u9010\u4e00\u6dfb\u52a0\u53ef\u89c2\u6d4b\u6027\u662f\u4e00\u9879\u7e41\u7410\u4e14\u6781\u6613\u9057\u6f0f\u7684\u4efb\u52a1\u3002<\/p>\n
      \u867d\u7136 Python \u6846\u67b6\u4e2d\u901a\u8fc7\u88c5\u9970\u5668\u5f88\u597d\u5730\u89e3\u51b3\u4e86\u8fd9\u7c7b\u95ee\u9898\uff0c\u4f46 Kotlin \u5f00\u53d1\u8005\u6b64\u524d\u53ea\u80fd\u6295\u4ee5\u7fa1\u6155\u7684\u76ee\u5149\u3002Tracy \u5f7b\u5e95\u6539\u53d8\u4e86\u8fd9\u4e00\u73b0\u72b6\u3002\u501f\u52a9\u57fa\u4e8e\u6ce8\u89e3\u7684\u8ddf\u8e2a\uff0c\u5f00\u53d1\u8005\u53ea\u9700\u4e3a\u63a5\u53e3\u65b9\u6cd5\u6dfb\u52a0 @Trace<\/code> \u6ce8\u89e3\uff0c\u4fbf\u53ef\u5728\u6240\u6709\u5b9e\u73b0\u7c7b\u4e2d\u5b9e\u73b0\u8ddf\u8e2a\u3002\u5982\u679c\u53ea\u662f\u60f3\u8ddf\u8e2a\u67d0\u4e2a\u9694\u79bb\u7684\u65b9\u6cd5\uff0c\u64cd\u4f5c\u4e5f\u540c\u6837\u7b80\u5355\u3002@Trace<\/code> \u6ce8\u89e3\u540c\u6837\u9002\u7528\u4e8e\u4e2a\u522b\u65b9\u6cd5\u6216\u51fd\u6570\u3002<\/p>\n
      \/** Interface for LLM tool *\/\ninterface Tool {\n   \/\/ All tool calls are now traced\n   @Trace(name = \"Tool Call\")\n   fun execute(): T\n}<\/pre>\n
      \u603b\u7ed3<\/h3>\n
      \u6355\u83b7\u5e94\u7528\u7a0b\u5e8f\u9065\u6d4b\u6570\u636e\u53ea\u662f\u7b2c\u4e00\u6b65\u3002\u540e\u7eed\u8fd8\u9700\u8981\u5c06\u8fd9\u4e9b\u6570\u636e\u8def\u7531\u81f3\u5408\u9002\u7684\u540e\u7aef\u8fdb\u884c\u5b58\u50a8\u548c\u5206\u6790\u3002\u5c3d\u7ba1\u6211\u4eec\u5f3a\u70c8\u5efa\u8bae\u5f00\u53d1\u8005\u4f7f\u7528\u4e13\u4e3a LLM \u8ddf\u8e2a\u8bbe\u8ba1\u7684\u53ef\u89c2\u6d4b\u6027\u89e3\u51b3\u65b9\u6848\uff0c\u5e76\u76f4\u63a5\u63d0\u4f9b\u5bf9 Langfuse \u548c W&B Weave \u7684\u652f\u6301\uff0c\u4f46 Tracy \u4e5f\u63d0\u4f9b\u4fbf\u6377\u7684\u65b9\u5f0f\uff0c\u53ef\u5c06\u8ddf\u8e2a\u53d1\u9001\u81f3\u4efb\u4f55\u517c\u5bb9 OpenTelemetry \u7684\u540e\u7aef\u3001\u6587\u4ef6\u6216\u63a7\u5236\u53f0\u3002\u4ed3\u5e93\u4e2d\u5305\u542b\u5927\u91cf\u793a\u4f8b<\/a>\uff0c\u5982\u9700\u83b7\u53d6\u672c\u6587\u793a\u4f8b\u7684\u5b8c\u6574\u4ee3\u7801\uff0c\u8bf7\u70b9\u51fb\u6b64\u5904<\/a>\u3002<\/p>\n
      \"\"<\/figure>\n
      \u501f\u52a9 Tracy\uff0c\u53ea\u9700\u51e0\u79d2\u5373\u53ef\u914d\u7f6e\u5411 Langfuse \u7684\u9065\u6d4b\u6570\u636e\u5bfc\u51fa\u3002\u6700\u540e\uff0c\u60a8\u4f1a\u5f97\u5230\u6355\u83b7\u4e86 LLM \u548c\u5de5\u5177\u8c03\u7528\u7684\u5c42\u7ea7\u5316\u5e94\u7528\u7a0b\u5e8f\u8ddf\u8e2a\u3002<\/em><\/p>\n
      \u672a\u6765\u8ba1\u5212<\/h2>\n
      \u6211\u4eec\u575a\u4fe1\uff0c\u65e0\u8bba\u672a\u6765\u51e0\u5e74 LLM \u5982\u4f55\u53d1\u5c55\uff0c\u53ef\u89c2\u6d4b\u6027\u59cb\u7ec8\u662f\u9ad8\u6548\u3001\u53ef\u9760 AI \u5de5\u7a0b\u7684\u6838\u5fc3\u3002\u65e0\u8bba\u5e95\u5c42 LLM \u7684\u6027\u80fd\u53d8\u5f97\u591a\u4e48\u5f3a\u5927\uff0c\u5f00\u53d1\u9636\u6bb5\u548c\u751f\u4ea7\u73af\u5883\u4e2d\u4ecd\u9700\u5bf9\u4f7f\u7528 LLM \u7684\u5e94\u7528\u7a0b\u5e8f\u8fdb\u884c\u8c03\u8bd5\u548c\u8bc4\u4f30\u3002Tracy \u6b63\u662f\u4e3a\u4e86\u6ee1\u8db3\u8fd9\u4e00\u9700\u6c42\u800c\u521b\u5efa\u7684\uff0c\u5373\u4e3a Kotlin \u751f\u6001\u7cfb\u7edf\u5f15\u5165\u751f\u4ea7\u7ea7 AI \u53ef\u89c2\u6d4b\u6027\u3002<\/p>\n
      \u8fd9\u53ea\u662f\u4e00\u4e2a\u5f00\u7aef\uff01 \u60a8\u53ef\u4ee5\u901a\u8fc7\u63d0\u4ea4\u95ee\u9898<\/a>\u3001\u63d0\u4ea4\u62c9\u53d6\u8bf7\u6c42\uff0c\u6216\u5728\u9879\u76ee\u4e2d\u8bd5\u7528 Tracy \u5e76\u5206\u4eab\u53cd\u9988<\/a>\uff0c\u4e3aKotlin AI \u751f\u6001\u7cfb\u7edf\u7684\u53d1\u5c55\u505a\u51fa\u8d21\u732e\u3002<\/a>\u8ba9\u6211\u4eec\u5171\u540c\u5f00\u542f\u8ddf\u8e2a\u4e4b\u65c5\uff01<\/a>\u00a0\u00a0<\/a><\/p>\n\n\n
      <\/p>\n\n\n\n
      \u672c\u535a\u6587\u82f1\u6587\u539f\u4f5c\u8005\uff1a<\/p>\n\n\n    
      \n        
      \n            
      \n                
      \n                    \"Anton\n                <\/div>\n                
      \n                                            
      Anton Bragin<\/h4>\n                                                        <\/div>\n            <\/div>\n        <\/div>\n    <\/div>\n","protected":false},"author":1297,"featured_media":692123,"comment_status":"closed","ping_status":"closed","template":"","categories":[8899,909,4113],"tags":[],"cross-post-tag":[8396],"acf":[],"_links":{"self":[{"href":"https:\/\/blog.jetbrains.com\/zh-hans\/wp-json\/wp\/v2\/kotlin\/690694"}],"collection":[{"href":"https:\/\/blog.jetbrains.com\/zh-hans\/wp-json\/wp\/v2\/kotlin"}],"about":[{"href":"https:\/\/blog.jetbrains.com\/zh-hans\/wp-json\/wp\/v2\/types\/kotlin"}],"author":[{"embeddable":true,"href":"https:\/\/blog.jetbrains.com\/zh-hans\/wp-json\/wp\/v2\/users\/1297"}],"replies":[{"embeddable":true,"href":"https:\/\/blog.jetbrains.com\/zh-hans\/wp-json\/wp\/v2\/comments?post=690694"}],"version-history":[{"count":3,"href":"https:\/\/blog.jetbrains.com\/zh-hans\/wp-json\/wp\/v2\/kotlin\/690694\/revisions"}],"predecessor-version":[{"id":692134,"href":"https:\/\/blog.jetbrains.com\/zh-hans\/wp-json\/wp\/v2\/kotlin\/690694\/revisions\/692134"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/blog.jetbrains.com\/zh-hans\/wp-json\/wp\/v2\/media\/692123"}],"wp:attachment":[{"href":"https:\/\/blog.jetbrains.com\/zh-hans\/wp-json\/wp\/v2\/media?parent=690694"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/blog.jetbrains.com\/zh-hans\/wp-json\/wp\/v2\/categories?post=690694"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/blog.jetbrains.com\/zh-hans\/wp-json\/wp\/v2\/tags?post=690694"},{"taxonomy":"cross-post-tag","embeddable":true,"href":"https:\/\/blog.jetbrains.com\/zh-hans\/wp-json\/wp\/v2\/cross-post-tag?post=690694"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}