feat(ai-chat): 新增豆包API + AI模型配置项支持动态切换
- 后端新增豆包（火山引擎Ark）API集成：DoubaoController、ToolDoubaoServiceImpl， 使用OkHttp3 SSE流式对话，兼容OpenAI Chat Completions格式 - 新增DoubaoConfig配置类，读取doubao.api.*配置 - 在eb_system_config表新增ai_chat_model配置项，支持doubao/coze/gemini三种模型切换 - 新增GET /api/front/doubao/ai-model-config接口供前端读取当前模型配置 - 前端ai-nutritionist.vue的sendToAI按系统配置分发到_sendViaDoubao/_sendViaCoze/_sendViaGemini - 前端models-api.js新增doubaoChatStream/doubaoChat/getAiModelConfig函数 - 附带豆包API测试脚本和数据库初始化SQL Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
This commit is contained in:
msh-agent
@@ -0,0 +1,43 @@
|
||||
package com.zbkj.common.config;
|
||||
|
||||
import lombok.Data;
|
||||
import org.springframework.boot.context.properties.ConfigurationProperties;
|
||||
import org.springframework.context.annotation.Configuration;
|
||||
|
||||
/**
|
||||
* 豆包（火山引擎 Ark）API 配置类
|
||||
* 使用 OpenAI 兼容的 Chat Completions 接口
|
||||
* +----------------------------------------------------------------------
|
||||
* | Author:ScottPan
|
||||
* +----------------------------------------------------------------------
|
||||
*/
|
||||
@Data
|
||||
@Configuration
|
||||
@ConfigurationProperties(prefix = "doubao.api")
|
||||
public class DoubaoConfig {
|
||||
|
||||
/**
|
||||
* API 基础 URL
|
||||
*/
|
||||
private String baseUrl = "https://ark.cn-beijing.volces.com/api/v3";
|
||||
|
||||
/**
|
||||
* API Key（火山引擎控制台获取）
|
||||
*/
|
||||
private String apiKey;
|
||||
|
||||
/**
|
||||
* 默认模型名称
|
||||
*/
|
||||
private String model = "doubao-seed-2-0-pro-260215";
|
||||
|
||||
/**
|
||||
* 连接超时时间（毫秒）
|
||||
*/
|
||||
private Integer connectTimeout = 30000;
|
||||
|
||||
/**
|
||||
* 读取超时时间（毫秒）
|
||||
*/
|
||||
private Integer readTimeout = 120000;
|
||||
}
|
||||
@@ -202,4 +202,7 @@ public class SysConfigConstants {
|
||||
/** 京东云存储端点 */
|
||||
public static final String CONFIG_JD_CLOUD_ENDPOINT = "jdEndpoint";
|
||||
|
||||
// ====== AI 模型配置 ======
|
||||
/** AI 营养师对话模型选择: doubao / coze / gemini */
|
||||
public static final String CONFIG_KEY_AI_CHAT_MODEL = "ai_chat_model";
|
||||
}
|
||||
|
||||
@@ -0,0 +1,58 @@
|
||||
package com.zbkj.common.request.doubao;
|
||||
|
||||
import io.swagger.annotations.ApiModel;
|
||||
import io.swagger.annotations.ApiModelProperty;
|
||||
import lombok.Data;
|
||||
import lombok.NoArgsConstructor;
|
||||
|
||||
import javax.validation.Valid;
|
||||
import javax.validation.constraints.NotEmpty;
|
||||
import javax.validation.constraints.NotNull;
|
||||
import java.io.Serializable;
|
||||
import java.util.List;
|
||||
|
||||
/**
|
||||
* 豆包（火山引擎 Ark）Chat Completions 请求 DTO
|
||||
* 兼容 OpenAI Chat Completions 格式
|
||||
* +----------------------------------------------------------------------
|
||||
* | Author:ScottPan
|
||||
* +----------------------------------------------------------------------
|
||||
*/
|
||||
@Data
|
||||
@NoArgsConstructor
|
||||
@ApiModel(value = "DoubaoChatRequest", description = "豆包 Chat Completions 请求")
|
||||
public class DoubaoChatRequest implements Serializable {
|
||||
|
||||
private static final long serialVersionUID = 1L;
|
||||
|
||||
@Valid
|
||||
@NotEmpty(message = "messages不能为空")
|
||||
@ApiModelProperty(value = "消息列表", required = true)
|
||||
private List<Message> messages;
|
||||
|
||||
@ApiModelProperty(value = "是否流式返回", example = "false")
|
||||
private Boolean stream;
|
||||
|
||||
@ApiModelProperty(value = "模型名称（可选，不传则使用服务端默认配置）")
|
||||
private String model;
|
||||
|
||||
@ApiModelProperty(value = "温度参数 (0-2)", example = "0.7")
|
||||
private Double temperature;
|
||||
|
||||
@ApiModelProperty(value = "最大输出 token 数")
|
||||
private Integer maxTokens;
|
||||
|
||||
@Data
|
||||
@NoArgsConstructor
|
||||
@ApiModel(value = "Message", description = "单条消息")
|
||||
public static class Message implements Serializable {
|
||||
private static final long serialVersionUID = 1L;
|
||||
|
||||
@ApiModelProperty(value = "角色: system, user, assistant")
|
||||
private String role;
|
||||
|
||||
@NotNull(message = "content不能为空")
|
||||
@ApiModelProperty(value = "内容：字符串或多模态数组", required = true)
|
||||
private Object content;
|
||||
}
|
||||
}
|
||||
@@ -0,0 +1,74 @@
|
||||
package com.zbkj.front.controller;
|
||||
|
||||
import com.zbkj.common.constants.SysConfigConstants;
|
||||
import com.zbkj.common.request.doubao.DoubaoChatRequest;
|
||||
import com.zbkj.common.result.CommonResult;
|
||||
import com.zbkj.service.service.SystemConfigService;
|
||||
import com.zbkj.service.service.tool.ToolDoubaoService;
|
||||
import io.swagger.annotations.Api;
|
||||
import io.swagger.annotations.ApiOperation;
|
||||
import org.springframework.beans.factory.annotation.Autowired;
|
||||
import org.springframework.validation.annotation.Validated;
|
||||
import org.springframework.web.bind.annotation.*;
|
||||
import org.springframework.web.servlet.mvc.method.annotation.SseEmitter;
|
||||
|
||||
import javax.servlet.http.HttpServletResponse;
|
||||
import java.util.HashMap;
|
||||
import java.util.Map;
|
||||
|
||||
/**
|
||||
* 豆包（火山引擎 Ark）对话接口 + AI 模型配置
|
||||
* +----------------------------------------------------------------------
|
||||
* | Author:ScottPan
|
||||
* +----------------------------------------------------------------------
|
||||
*/
|
||||
@RestController
|
||||
@RequestMapping("/api/front/doubao")
|
||||
@Api(tags = "豆包 AI 对话")
|
||||
public class DoubaoController {
|
||||
|
||||
@Autowired
|
||||
private ToolDoubaoService toolDoubaoService;
|
||||
|
||||
@Autowired
|
||||
private SystemConfigService systemConfigService;
|
||||
|
||||
/**
|
||||
* 获取当前 AI 对话模型配置
|
||||
* 返回 { model: "doubao" | "coze" | "gemini" }
|
||||
*/
|
||||
@ApiOperation(value = "获取AI模型配置", notes = "获取当前系统配置的AI对话模型")
|
||||
@GetMapping("/ai-model-config")
|
||||
public CommonResult<Map<String, String>> getAiModelConfig() {
|
||||
String model = systemConfigService.getValueByKey(SysConfigConstants.CONFIG_KEY_AI_CHAT_MODEL);
|
||||
if (model == null || model.isEmpty()) {
|
||||
model = "doubao"; // 默认使用豆包
|
||||
}
|
||||
Map<String, String> result = new HashMap<>();
|
||||
result.put("model", model.trim().toLowerCase());
|
||||
return CommonResult.success(result);
|
||||
}
|
||||
|
||||
/**
|
||||
* 非流式对话
|
||||
*/
|
||||
@ApiOperation(value = "对话", notes = "与豆包模型进行对话")
|
||||
@PostMapping("/chat")
|
||||
public Map<String, Object> chat(@RequestBody @Validated DoubaoChatRequest request) {
|
||||
request.setStream(false);
|
||||
return toolDoubaoService.chat(request);
|
||||
}
|
||||
|
||||
/**
|
||||
* 流式对话（SSE）
|
||||
*/
|
||||
@ApiOperation(value = "流式对话", notes = "与豆包模型进行流式对话，使用 SSE 实时推送响应")
|
||||
@PostMapping(value = "/chat/stream", produces = "text/event-stream")
|
||||
public SseEmitter chatStream(@RequestBody @Validated DoubaoChatRequest request,
|
||||
HttpServletResponse response) {
|
||||
response.setHeader("X-Accel-Buffering", "no");
|
||||
response.setHeader("Cache-Control", "no-cache");
|
||||
request.setStream(true);
|
||||
return toolDoubaoService.chatStream(request);
|
||||
}
|
||||
}
|
||||
@@ -41,7 +41,7 @@ spring:
|
||||
min-idle: 0 # 连接池中的最小空闲连接
|
||||
time-between-eviction-runs: -1 #逐出扫描的时间间隔(毫秒) 如果为负数,则不运行逐出线程, 默认-1
|
||||
second:
|
||||
database: 23 # 微信accessToken存储库
|
||||
database: 3 # 微信accessToken存储库
|
||||
|
||||
|
||||
debug: true
|
||||
@@ -87,6 +87,15 @@ coze:
|
||||
nutrition-analysis-workflow-id: 1180790412263 #饮食打卡记录ai分析
|
||||
diet-analysis-id: 1180790412263
|
||||
|
||||
# 豆包（火山引擎 Ark）API 配置
|
||||
doubao:
|
||||
api:
|
||||
base-url: https://ark.cn-beijing.volces.com/api/v3
|
||||
api-key: 18480c26-ebcd-4263-8a6f-48359b8bd65d
|
||||
model: doubao-seed-2-0-pro-260215
|
||||
connect-timeout: 30000
|
||||
read-timeout: 120000
|
||||
|
||||
# 腾讯云语音识别配置
|
||||
tencent-asr:
|
||||
# API密钥ID
|
||||
|
||||
@@ -21,6 +21,7 @@ crmeb:
|
||||
- api/front/upload/imageOuter
|
||||
- api/front/coze/**
|
||||
- api/front/kieai/**
|
||||
- api/front/doubao/**
|
||||
|
||||
# 配置端口
|
||||
server:
|
||||
|
||||
@@ -0,0 +1,171 @@
|
||||
package com.zbkj.service.service.impl.tool;
|
||||
|
||||
import com.alibaba.fastjson.JSON;
|
||||
import com.alibaba.fastjson.JSONObject;
|
||||
import com.zbkj.common.config.DoubaoConfig;
|
||||
import com.zbkj.common.request.doubao.DoubaoChatRequest;
|
||||
import com.zbkj.common.utils.SseEmitterUtil;
|
||||
import com.zbkj.service.service.tool.ToolDoubaoService;
|
||||
import okhttp3.*;
|
||||
import org.slf4j.Logger;
|
||||
import org.slf4j.LoggerFactory;
|
||||
import org.springframework.beans.factory.annotation.Autowired;
|
||||
import org.springframework.http.client.SimpleClientHttpRequestFactory;
|
||||
import org.springframework.stereotype.Service;
|
||||
import org.springframework.web.client.RestTemplate;
|
||||
import org.springframework.web.servlet.mvc.method.annotation.SseEmitter;
|
||||
|
||||
import java.io.BufferedReader;
|
||||
import java.io.IOException;
|
||||
import java.io.InputStreamReader;
|
||||
import java.nio.charset.StandardCharsets;
|
||||
import java.util.*;
|
||||
import java.util.concurrent.TimeUnit;
|
||||
|
||||
/**
|
||||
* 豆包（火山引擎 Ark）服务实现类
|
||||
* 使用 OpenAI 兼容的 Chat Completions 接口 + OkHttp SSE 流式
|
||||
* +----------------------------------------------------------------------
|
||||
* | Author:ScottPan
|
||||
* +----------------------------------------------------------------------
|
||||
*/
|
||||
@Service
|
||||
public class ToolDoubaoServiceImpl implements ToolDoubaoService {
|
||||
|
||||
private static final Logger logger = LoggerFactory.getLogger(ToolDoubaoServiceImpl.class);
|
||||
|
||||
private static final String CHAT_COMPLETIONS_PATH = "/chat/completions";
|
||||
|
||||
@Autowired
|
||||
private DoubaoConfig config;
|
||||
|
||||
/**
|
||||
* 构建发送给豆包 API 的请求体
|
||||
*/
|
||||
private Map<String, Object> buildRequestBody(DoubaoChatRequest request, boolean stream) {
|
||||
Map<String, Object> body = new LinkedHashMap<>();
|
||||
// 模型：优先使用请求指定的，否则使用配置默认的
|
||||
String model = (request.getModel() != null && !request.getModel().isEmpty())
|
||||
? request.getModel() : config.getModel();
|
||||
body.put("model", model);
|
||||
|
||||
// 消息列表 — 用 FastJSON 解析保留原始结构（与 KieAI 实现一致）
|
||||
List<Object> messages = JSON.parseArray(JSON.toJSONString(request.getMessages()));
|
||||
body.put("messages", messages);
|
||||
|
||||
body.put("stream", stream);
|
||||
|
||||
if (request.getTemperature() != null) {
|
||||
body.put("temperature", request.getTemperature());
|
||||
}
|
||||
if (request.getMaxTokens() != null) {
|
||||
body.put("max_tokens", request.getMaxTokens());
|
||||
}
|
||||
return body;
|
||||
}
|
||||
|
||||
@Override
|
||||
public Map<String, Object> chat(DoubaoChatRequest request) {
|
||||
if (config.getApiKey() == null || config.getApiKey().isEmpty()) {
|
||||
throw new RuntimeException("豆包 API Key 未配置");
|
||||
}
|
||||
Map<String, Object> body = buildRequestBody(request, false);
|
||||
String url = config.getBaseUrl() + CHAT_COMPLETIONS_PATH;
|
||||
|
||||
SimpleClientHttpRequestFactory factory = new SimpleClientHttpRequestFactory();
|
||||
factory.setReadTimeout(config.getReadTimeout());
|
||||
factory.setConnectTimeout(config.getConnectTimeout());
|
||||
RestTemplate rt = new RestTemplate(factory);
|
||||
|
||||
org.springframework.http.HttpHeaders headers = new org.springframework.http.HttpHeaders();
|
||||
headers.set("Authorization", "Bearer " + config.getApiKey());
|
||||
headers.set("Content-Type", "application/json");
|
||||
|
||||
org.springframework.http.HttpEntity<String> entity =
|
||||
new org.springframework.http.HttpEntity<>(JSON.toJSONString(body), headers);
|
||||
|
||||
try {
|
||||
org.springframework.http.ResponseEntity<String> resp =
|
||||
rt.postForEntity(url, entity, String.class);
|
||||
if (resp.getBody() != null) {
|
||||
return JSON.parseObject(resp.getBody(), Map.class);
|
||||
}
|
||||
return Collections.emptyMap();
|
||||
} catch (Exception e) {
|
||||
logger.error("Doubao chat error", e);
|
||||
throw new RuntimeException("豆包对话失败: " + e.getMessage());
|
||||
}
|
||||
}
|
||||
|
||||
@Override
|
||||
public SseEmitter chatStream(DoubaoChatRequest request) {
|
||||
if (config.getApiKey() == null || config.getApiKey().isEmpty()) {
|
||||
throw new RuntimeException("豆包 API Key 未配置");
|
||||
}
|
||||
SseEmitter emitter = new SseEmitter(120000L);
|
||||
|
||||
Map<String, Object> body = buildRequestBody(request, true);
|
||||
String url = config.getBaseUrl() + CHAT_COMPLETIONS_PATH;
|
||||
|
||||
OkHttpClient client = new OkHttpClient.Builder()
|
||||
.connectTimeout(config.getConnectTimeout(), TimeUnit.MILLISECONDS)
|
||||
.readTimeout(config.getReadTimeout(), TimeUnit.MILLISECONDS)
|
||||
.writeTimeout(30, TimeUnit.SECONDS)
|
||||
.build();
|
||||
|
||||
RequestBody requestBody = RequestBody.create(
|
||||
MediaType.parse("application/json; charset=utf-8"),
|
||||
JSON.toJSONString(body));
|
||||
|
||||
Request okRequest = new Request.Builder()
|
||||
.url(url)
|
||||
.post(requestBody)
|
||||
.addHeader("Authorization", "Bearer " + config.getApiKey())
|
||||
.addHeader("Content-Type", "application/json")
|
||||
.addHeader("Accept", "text/event-stream")
|
||||
.build();
|
||||
|
||||
try {
|
||||
client.newCall(okRequest).enqueue(new Callback() {
|
||||
@Override
|
||||
public void onFailure(Call call, IOException e) {
|
||||
logger.error("Doubao chat stream request failed", e);
|
||||
SseEmitterUtil.completeWithError(emitter, e);
|
||||
}
|
||||
|
||||
@Override
|
||||
public void onResponse(Call call, Response response) throws IOException {
|
||||
if (!response.isSuccessful() || response.body() == null) {
|
||||
String errBody = response.body() != null ? response.body().string() : "";
|
||||
logger.error("Doubao stream failed: {} {}", response.code(), errBody);
|
||||
SseEmitterUtil.completeWithError(emitter,
|
||||
new RuntimeException("豆包流式请求失败: " + response.code()));
|
||||
return;
|
||||
}
|
||||
try (BufferedReader reader = new BufferedReader(
|
||||
new InputStreamReader(response.body().byteStream(), StandardCharsets.UTF_8))) {
|
||||
String line;
|
||||
while ((line = reader.readLine()) != null) {
|
||||
if (line.startsWith("data: ")) {
|
||||
String json = line.substring(6).trim();
|
||||
if ("[DONE]".equals(json) || json.isEmpty()) {
|
||||
break;
|
||||
}
|
||||
SseEmitterUtil.send(emitter, json);
|
||||
}
|
||||
}
|
||||
SseEmitterUtil.complete(emitter);
|
||||
} catch (Exception e) {
|
||||
logger.error("Doubao chat stream read error", e);
|
||||
SseEmitterUtil.completeWithError(emitter, e);
|
||||
}
|
||||
}
|
||||
});
|
||||
} catch (Exception e) {
|
||||
logger.error("Doubao chat stream error", e);
|
||||
SseEmitterUtil.completeWithError(emitter, e);
|
||||
}
|
||||
|
||||
return emitter;
|
||||
}
|
||||
}
|
||||
@@ -0,0 +1,31 @@
|
||||
package com.zbkj.service.service.tool;
|
||||
|
||||
import com.zbkj.common.request.doubao.DoubaoChatRequest;
|
||||
import org.springframework.web.servlet.mvc.method.annotation.SseEmitter;
|
||||
|
||||
import java.util.Map;
|
||||
|
||||
/**
|
||||
* 豆包（火山引擎 Ark）服务接口
|
||||
* +----------------------------------------------------------------------
|
||||
* | Author:ScottPan
|
||||
* +----------------------------------------------------------------------
|
||||
*/
|
||||
public interface ToolDoubaoService {
|
||||
|
||||
/**
|
||||
* 豆包对话（非流式）
|
||||
*
|
||||
* @param request 对话请求参数
|
||||
* @return 对话响应
|
||||
*/
|
||||
Map<String, Object> chat(DoubaoChatRequest request);
|
||||
|
||||
/**
|
||||
* 豆包对话（流式 SSE）
|
||||
*
|
||||
* @param request 对话请求参数
|
||||
* @return SSE 事件流
|
||||
*/
|
||||
SseEmitter chatStream(DoubaoChatRequest request);
|
||||
}
|
||||
@@ -812,6 +812,163 @@ function cozeTextToSpeech(data) {
|
||||
})
|
||||
}
|
||||
|
||||
/**
|
||||
* 获取 AI 对话模型配置
|
||||
* @returns {Promise} { data: { model: "doubao" | "coze" | "gemini" } }
|
||||
*/
|
||||
function getAiModelConfig() {
|
||||
return request('/api/front/doubao/ai-model-config')
|
||||
}
|
||||
|
||||
/**
|
||||
* 豆包（火山引擎 Ark）- 非流式对话
|
||||
* @param {object} data 请求参数 { messages: [{role, content}], model?, temperature?, maxTokens? }
|
||||
* @returns {Promise} 对话响应（OpenAI Chat Completions 格式）
|
||||
*/
|
||||
function doubaoChat(data) {
|
||||
return request('/api/front/doubao/chat', {
|
||||
method: 'POST',
|
||||
data: data
|
||||
})
|
||||
}
|
||||
|
||||
/**
|
||||
* 豆包（火山引擎 Ark）- 流式对话 (SSE + enableChunked)
|
||||
* 返回 OpenAI 兼容的 SSE 事件流：choices[0].delta.content
|
||||
* @param {object} data 请求参数 { messages: [{role, content}] }
|
||||
* @returns {object} 控制器 { onMessage(deltaText), onError, onComplete, abort }
|
||||
*/
|
||||
function doubaoChatStream(data) {
|
||||
let _onMessage = () => {}
|
||||
let _onError = () => {}
|
||||
let _onComplete = () => {}
|
||||
let _buffer = ''
|
||||
let _task = null
|
||||
let _gotChunks = false
|
||||
|
||||
const controller = {
|
||||
onMessage(fn) { _onMessage = fn; return controller },
|
||||
onError(fn) { _onError = fn; return controller },
|
||||
onComplete(fn) { _onComplete = fn; return controller },
|
||||
abort() { if (_task) _task.abort() },
|
||||
getTask() { return _task }
|
||||
}
|
||||
|
||||
/** 从 SSE data JSON 中提取增量文本（OpenAI 兼容格式） */
|
||||
const extractDeltaText = (evt) => {
|
||||
if (evt && Array.isArray(evt.choices) && evt.choices[0]) {
|
||||
const delta = evt.choices[0].delta
|
||||
if (delta && typeof delta.content === 'string') return delta.content
|
||||
const msg = evt.choices[0].message
|
||||
if (msg && typeof msg.content === 'string') return msg.content
|
||||
}
|
||||
return ''
|
||||
}
|
||||
|
||||
const parseSseLines = (text) => {
|
||||
_buffer += text
|
||||
const lines = _buffer.split('\n')
|
||||
_buffer = lines.pop() || ''
|
||||
for (const line of lines) {
|
||||
const trimmed = line.trim()
|
||||
if (!trimmed || trimmed.startsWith(':')) continue
|
||||
if (trimmed === 'data: [DONE]') continue
|
||||
if (trimmed.startsWith('data:')) {
|
||||
const jsonStr = trimmed.slice(5).trim()
|
||||
if (!jsonStr) continue
|
||||
try {
|
||||
const evt = JSON.parse(jsonStr)
|
||||
const delta = extractDeltaText(evt)
|
||||
if (delta) _onMessage(delta)
|
||||
} catch (e) {
|
||||
console.warn('[doubaoChatStream] JSON parse failed:', jsonStr.slice(0, 200))
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
const parseSseResponseBody = (body) => {
|
||||
if (!body || typeof body !== 'string') return
|
||||
const lines = body.split('\n')
|
||||
for (const line of lines) {
|
||||
const trimmed = line.trim()
|
||||
if (!trimmed || trimmed.startsWith(':')) continue
|
||||
if (trimmed === 'data: [DONE]') continue
|
||||
if (trimmed.startsWith('data:')) {
|
||||
const jsonStr = trimmed.slice(5).trim()
|
||||
if (!jsonStr) continue
|
||||
try {
|
||||
const evt = JSON.parse(jsonStr)
|
||||
const delta = extractDeltaText(evt)
|
||||
if (delta) _onMessage(delta)
|
||||
} catch (e) {
|
||||
console.warn('[doubaoChatStream] JSON parse failed in body fallback:', jsonStr.slice(0, 200))
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
const token = store.state && store.state.app && store.state.app.token
|
||||
_task = uni.request({
|
||||
url: `${API_BASE_URL}/api/front/doubao/chat/stream`,
|
||||
method: 'POST',
|
||||
data: data,
|
||||
header: {
|
||||
'Content-Type': 'application/json',
|
||||
'Accept': 'text/event-stream',
|
||||
...(token ? { [TOKENNAME]: token } : {})
|
||||
},
|
||||
enableChunked: true,
|
||||
responseType: 'text',
|
||||
success: (res) => {
|
||||
console.log('[doubaoChatStream] success: statusCode=', res.statusCode, '_gotChunks=', _gotChunks)
|
||||
if (res.statusCode !== 200) {
|
||||
let errMsg = '请求失败: ' + res.statusCode
|
||||
try {
|
||||
const body = typeof res.data === 'string' ? JSON.parse(res.data) : res.data
|
||||
if (body && body.message) errMsg = body.message
|
||||
else if (body && body.error && body.error.message) errMsg = body.error.message
|
||||
} catch (e) { /* keep default message */ }
|
||||
_onError(new Error(errMsg))
|
||||
return
|
||||
}
|
||||
// 处理 buffer 中残余内容
|
||||
if (_buffer.trim()) {
|
||||
parseSseLines('\n')
|
||||
}
|
||||
// 不支持 chunked 的环境降级：从完整 response body 解析
|
||||
if (!_gotChunks && res && res.data) {
|
||||
const body = typeof res.data === 'string' ? res.data : JSON.stringify(res.data)
|
||||
parseSseResponseBody(body)
|
||||
}
|
||||
_onComplete()
|
||||
},
|
||||
fail: (err) => {
|
||||
console.error('[doubaoChatStream] request fail:', err)
|
||||
_onError(err)
|
||||
}
|
||||
})
|
||||
|
||||
if (_task && _task.onChunkReceived) {
|
||||
_task.onChunkReceived((res) => {
|
||||
_gotChunks = true
|
||||
try {
|
||||
const bytes = new Uint8Array(res.data)
|
||||
let text = ''
|
||||
for (let i = 0; i < bytes.length; i++) {
|
||||
text += String.fromCharCode(bytes[i])
|
||||
}
|
||||
text = decodeURIComponent(escape(text))
|
||||
parseSseLines(text)
|
||||
} catch (e) {
|
||||
console.warn('[doubaoChatStream] chunk decode error:', e)
|
||||
}
|
||||
})
|
||||
}
|
||||
|
||||
return controller
|
||||
}
|
||||
|
||||
export default {
|
||||
request,
|
||||
getArticleById,
|
||||
@@ -835,5 +992,10 @@ export default {
|
||||
cozeWorkflowStream,
|
||||
cozeWorkflowResume,
|
||||
cozeUploadFile,
|
||||
cozeTextToSpeech
|
||||
cozeTextToSpeech,
|
||||
// 豆包 API
|
||||
doubaoChat,
|
||||
doubaoChatStream,
|
||||
// AI 模型配置
|
||||
getAiModelConfig
|
||||
}
|
||||
|
||||
@@ -224,6 +224,7 @@ export default {
|
||||
},
|
||||
data() {
|
||||
return {
|
||||
aiModel: 'doubao', // 当前AI模型: doubao / coze / gemini（从系统配置读取）
|
||||
botId: '7591133240535449654',
|
||||
conversationId: '',
|
||||
scrollTop: 0,
|
||||
@@ -261,6 +262,7 @@ export default {
|
||||
});
|
||||
this.initRecorder();
|
||||
this.initAudioContext();
|
||||
this.loadAiModelConfig();
|
||||
},
|
||||
onUnload() {
|
||||
this.stopRecordTimer();
|
||||
@@ -277,6 +279,19 @@ export default {
|
||||
}
|
||||
},
|
||||
methods: {
|
||||
/** 加载 AI 模型配置 */
|
||||
loadAiModelConfig() {
|
||||
api.getAiModelConfig().then(res => {
|
||||
const data = res && res.data
|
||||
if (data && data.model) {
|
||||
this.aiModel = data.model.trim().toLowerCase()
|
||||
console.log('[ai-nutritionist] AI模型配置:', this.aiModel)
|
||||
}
|
||||
}).catch(err => {
|
||||
console.warn('[ai-nutritionist] 获取AI模型配置失败，使用默认 doubao:', err)
|
||||
})
|
||||
},
|
||||
|
||||
// 初始化录音管理器
|
||||
initRecorder() {
|
||||
// #ifdef MP-WEIXIN || APP-PLUS
|
||||
@@ -744,87 +759,65 @@ export default {
|
||||
return '';
|
||||
},
|
||||
|
||||
/** 从 Gemini(KieAI) 非流式响应中提取回复文本 */
|
||||
getGeminiReplyFromResponse(response) {
|
||||
if (!response || typeof response !== 'object') return ''
|
||||
let data = response.data
|
||||
if (typeof data === 'string') {
|
||||
try { data = JSON.parse(data) } catch (e) { return '' }
|
||||
}
|
||||
const payload = this.getGeminiPayload(data || response)
|
||||
if (!payload) return ''
|
||||
// OpenAI 形态: choices[0].message.content
|
||||
if (Array.isArray(payload.choices) && payload.choices[0]) {
|
||||
const msg = payload.choices[0].message
|
||||
if (msg) return this.extractReplyContent(msg.content)
|
||||
const delta = payload.choices[0].delta
|
||||
if (delta) return this.extractReplyContent(delta.content)
|
||||
}
|
||||
// Gemini 形态: candidates[0].content
|
||||
if (Array.isArray(payload.candidates) && payload.candidates[0]) {
|
||||
return this.extractReplyContent(payload.candidates[0].content)
|
||||
}
|
||||
return ''
|
||||
},
|
||||
|
||||
/** 工具方法：sleep ms 毫秒 */
|
||||
sleep(ms) {
|
||||
return new Promise(resolve => setTimeout(resolve, ms));
|
||||
},
|
||||
|
||||
buildGeminiMessages(content, type) {
|
||||
if (type === 'text') {
|
||||
return [{ role: 'user', content: typeof content === 'string' ? content : String(content) }];
|
||||
}
|
||||
const parts = Array.isArray(content) ? content : [{ type: 'text', text: String(content) }];
|
||||
return [{ role: 'user', content: parts }];
|
||||
},
|
||||
|
||||
/**
|
||||
* BUG-005：主路径为 response.data.choices[0].message.content（models-api 已尽量规范 data 形态）。
|
||||
* 兼容上游 candidates、或根级即 completion 等变体；仅使用接口返回字段，
|
||||
* 成功路径不使用 api/tool.js 的 getAIResponse 或本地关键词话术。
|
||||
* 构建 OpenAI 兼容格式的消息列表（用于豆包 API）
|
||||
* @param {string|Array} content 消息内容
|
||||
* @param {string} type 消息类型 text / image / multimodal
|
||||
* @returns {Array} messages 数组 [{role, content}]
|
||||
*/
|
||||
getGeminiReplyFromResponse(response) {
|
||||
if (!response || typeof response !== 'object') return '';
|
||||
const looksLikeCompletion = (o) =>
|
||||
o && typeof o === 'object' &&
|
||||
(Array.isArray(o.choices) || Array.isArray(o.candidates));
|
||||
let data = response.data;
|
||||
if (typeof data === 'string') {
|
||||
try {
|
||||
data = JSON.parse(data);
|
||||
} catch (e) {
|
||||
return '';
|
||||
buildChatMessages(content, type) {
|
||||
if (type === 'text') {
|
||||
return [{ role: 'user', content: typeof content === 'string' ? content : String(content) }]
|
||||
}
|
||||
if (type === 'image') {
|
||||
// 图片：尝试构建多模态消息
|
||||
let fileInfo = content
|
||||
if (typeof fileInfo === 'string') {
|
||||
try { fileInfo = JSON.parse(fileInfo) } catch (e) { /* 非JSON */ }
|
||||
}
|
||||
// 主路径：OpenAI 形态 choices[0].message.content（与需求 data.choices[0].message.content 一致）
|
||||
if (data && typeof data === 'object' && Array.isArray(data.choices) && data.choices.length > 0) {
|
||||
const choice0 = data.choices[0];
|
||||
const msg = choice0 && choice0.message;
|
||||
if (msg && typeof msg === 'object') {
|
||||
const fromMsg = this.extractReplyContent(msg.content);
|
||||
if (fromMsg.trim()) return fromMsg;
|
||||
const imageUrl = (fileInfo && fileInfo.url) || (fileInfo && fileInfo.path) || ''
|
||||
if (imageUrl) {
|
||||
return [{
|
||||
role: 'user',
|
||||
content: [
|
||||
{ type: 'image_url', image_url: { url: imageUrl } },
|
||||
{ type: 'text', text: '请分析这张图片' }
|
||||
]
|
||||
}]
|
||||
}
|
||||
const delta = choice0 && choice0.delta;
|
||||
if (delta && typeof delta === 'object') {
|
||||
const fromDelta = this.extractReplyContent(delta.content);
|
||||
if (fromDelta.trim()) return fromDelta;
|
||||
return [{ role: 'user', content: '我发送了一张图片，请帮我分析' }]
|
||||
}
|
||||
if (choice0 && typeof choice0.text === 'string' && choice0.text.trim()) {
|
||||
return choice0.text;
|
||||
}
|
||||
}
|
||||
if (!looksLikeCompletion(data) && looksLikeCompletion(response)) {
|
||||
data = response;
|
||||
}
|
||||
if (!data || typeof data !== 'object') return '';
|
||||
const payload = this.getGeminiPayload(data);
|
||||
if (!payload || typeof payload !== 'object') return '';
|
||||
const choices = payload.choices;
|
||||
if (Array.isArray(choices) && choices.length > 0) {
|
||||
const choice0 = choices[0];
|
||||
const msg = choice0 && choice0.message;
|
||||
if (msg && typeof msg === 'object') {
|
||||
const fromMsg = this.extractReplyContent(msg.content);
|
||||
if (fromMsg.trim()) return fromMsg;
|
||||
}
|
||||
const delta = choice0 && choice0.delta;
|
||||
if (delta && typeof delta === 'object') {
|
||||
const fromDelta = this.extractReplyContent(delta.content);
|
||||
if (fromDelta.trim()) return fromDelta;
|
||||
}
|
||||
if (choice0 && typeof choice0.text === 'string' && choice0.text.trim()) {
|
||||
return choice0.text;
|
||||
}
|
||||
}
|
||||
const cands = payload.candidates;
|
||||
if (Array.isArray(cands) && cands.length > 0) {
|
||||
const c0 = cands[0];
|
||||
if (c0 && typeof c0 === 'object') {
|
||||
const fromCand = this.extractReplyContent(c0.content);
|
||||
if (fromCand.trim()) return fromCand;
|
||||
if (typeof c0.text === 'string' && c0.text.trim()) return c0.text;
|
||||
}
|
||||
}
|
||||
return '';
|
||||
// multimodal：直接传多模态 parts
|
||||
const parts = Array.isArray(content) ? content : [{ type: 'text', text: String(content) }]
|
||||
return [{ role: 'user', content: parts }]
|
||||
},
|
||||
|
||||
async sendToAI(content, type) {
|
||||
@@ -833,54 +826,15 @@ export default {
|
||||
this.messageList.push(aiMsg);
|
||||
this.scrollToBottom();
|
||||
|
||||
const messages = this.buildGeminiMessages(content, type);
|
||||
|
||||
// 优先尝试流式输出以改善响应速度感知
|
||||
try {
|
||||
await new Promise((resolve, reject) => {
|
||||
const ctrl = api.kieaiGeminiChatStream({ messages })
|
||||
this._streamCtrl = ctrl
|
||||
|
||||
// 收到第一个 chunk 后切换为 streaming 状态
|
||||
ctrl.onMessage((deltaText) => {
|
||||
if (aiMsg.loading) {
|
||||
aiMsg.loading = false
|
||||
aiMsg.streaming = true
|
||||
}
|
||||
aiMsg.content += deltaText
|
||||
this.messageList = [...this.messageList]
|
||||
this.scrollToBottom()
|
||||
})
|
||||
|
||||
ctrl.onError((err) => {
|
||||
console.warn('[sendToAI] 流式请求失败，降级为非流式:', err)
|
||||
this._streamCtrl = null
|
||||
reject(err)
|
||||
})
|
||||
|
||||
ctrl.onComplete(() => {
|
||||
this._streamCtrl = null
|
||||
resolve()
|
||||
})
|
||||
})
|
||||
|
||||
// 流式完成，检查内容
|
||||
if (!aiMsg.content.trim()) {
|
||||
aiMsg.content = '模型未返回有效内容，请稍后重试。'
|
||||
}
|
||||
} catch (streamError) {
|
||||
// 流式失败，降级为非流式请求
|
||||
try {
|
||||
const response = await api.kieaiGeminiChat({
|
||||
messages,
|
||||
stream: false
|
||||
});
|
||||
const reply = this.getGeminiReplyFromResponse(response);
|
||||
aiMsg.content = reply.trim() ? reply : '模型未返回有效内容，请稍后重试。';
|
||||
} catch (error) {
|
||||
console.error('KieAI Gemini 对话失败:', error);
|
||||
const errText = error && error.message ? String(error.message) : '';
|
||||
aiMsg.content = errText || '抱歉，处理您的请求时出现错误，请稍后再试。';
|
||||
// 根据系统配置的 aiModel 分发到不同的 API
|
||||
if (this.aiModel === 'coze') {
|
||||
await this._sendViaCoze(content, type, aiMsg)
|
||||
} else if (this.aiModel === 'gemini') {
|
||||
await this._sendViaGemini(content, type, aiMsg)
|
||||
} else {
|
||||
// 默认：豆包
|
||||
await this._sendViaDoubao(content, type, aiMsg)
|
||||
}
|
||||
} finally {
|
||||
aiMsg.loading = false;
|
||||
@@ -897,6 +851,159 @@ export default {
|
||||
}
|
||||
},
|
||||
|
||||
// ========== 豆包 API（OpenAI 兼容，默认） ==========
|
||||
async _sendViaDoubao(content, type, aiMsg) {
|
||||
const messages = this.buildChatMessages(content, type)
|
||||
try {
|
||||
await new Promise((resolve, reject) => {
|
||||
const ctrl = api.doubaoChatStream({ messages })
|
||||
this._streamCtrl = ctrl
|
||||
ctrl.onMessage((deltaText) => {
|
||||
if (aiMsg.loading) { aiMsg.loading = false; aiMsg.streaming = true }
|
||||
aiMsg.content += deltaText
|
||||
this.messageList = [...this.messageList]
|
||||
this.scrollToBottom()
|
||||
})
|
||||
ctrl.onError((err) => { this._streamCtrl = null; reject(err) })
|
||||
ctrl.onComplete(() => { this._streamCtrl = null; resolve() })
|
||||
})
|
||||
if (!aiMsg.content.trim()) aiMsg.content = '模型未返回有效内容，请稍后重试。'
|
||||
} catch (streamError) {
|
||||
console.warn('[sendToAI:doubao] 流式失败，降级非流式:', streamError)
|
||||
try {
|
||||
const response = await api.doubaoChat({ messages, stream: false })
|
||||
let reply = ''
|
||||
if (response && response.choices && response.choices[0]) {
|
||||
const msg = response.choices[0].message
|
||||
if (msg && msg.content) reply = msg.content
|
||||
}
|
||||
aiMsg.content = reply.trim() ? reply : '模型未返回有效内容，请稍后重试。'
|
||||
} catch (error) {
|
||||
console.error('豆包对话失败:', error)
|
||||
aiMsg.content = (error && error.message) || '抱歉，处理您的请求时出现错误，请稍后再试。'
|
||||
}
|
||||
}
|
||||
},
|
||||
|
||||
// ========== Coze API ==========
|
||||
async _sendViaCoze(content, type, aiMsg) {
|
||||
const userId = this.uid || (uni.getStorageSync('userInfo') || {}).id || 'default_user'
|
||||
const cozeMessages = this.buildChatMessages(content, type).map(m => ({
|
||||
...m, content_type: 'text'
|
||||
}))
|
||||
const reqData = {
|
||||
botId: this.botId,
|
||||
userId: String(userId),
|
||||
additionalMessages: cozeMessages
|
||||
}
|
||||
if (this.conversationId) reqData.conversationId = this.conversationId
|
||||
|
||||
try {
|
||||
await new Promise((resolve, reject) => {
|
||||
const ctrl = api.cozeChatStream(reqData)
|
||||
this._streamCtrl = ctrl
|
||||
ctrl.onMessage((evt) => {
|
||||
if (evt.conversation_id && !this.conversationId) {
|
||||
this.conversationId = evt.conversation_id
|
||||
}
|
||||
if (evt.event === 'conversation.message.delta' && evt.type === 'answer' && evt.content) {
|
||||
if (aiMsg.loading) { aiMsg.loading = false; aiMsg.streaming = true }
|
||||
aiMsg.content += evt.content
|
||||
this.messageList = [...this.messageList]
|
||||
this.scrollToBottom()
|
||||
}
|
||||
if (evt.event === 'conversation.chat.completed' && evt.conversation_id) {
|
||||
this.conversationId = evt.conversation_id
|
||||
}
|
||||
})
|
||||
ctrl.onError((err) => { this._streamCtrl = null; reject(err) })
|
||||
ctrl.onComplete(() => { this._streamCtrl = null; resolve() })
|
||||
})
|
||||
if (!aiMsg.content.trim()) aiMsg.content = '模型未返回有效内容，请稍后重试。'
|
||||
} catch (streamError) {
|
||||
console.warn('[sendToAI:coze] 流式失败，降级非流式:', streamError)
|
||||
try {
|
||||
reqData.stream = false
|
||||
const response = await api.cozeChat(reqData)
|
||||
if (response && response.data && response.data.chat) {
|
||||
const { conversation_id, id: chat_id } = response.data.chat
|
||||
if (conversation_id) this.conversationId = conversation_id
|
||||
await this.pollCozeResult(conversation_id, chat_id, aiMsg)
|
||||
} else {
|
||||
aiMsg.content = '模型未返回有效内容，请稍后重试。'
|
||||
}
|
||||
} catch (error) {
|
||||
console.error('Coze 对话失败:', error)
|
||||
aiMsg.content = (error && error.message) || '抱歉，处理您的请求时出现错误，请稍后再试。'
|
||||
}
|
||||
}
|
||||
},
|
||||
|
||||
// ========== Gemini API (KieAI) ==========
|
||||
async _sendViaGemini(content, type, aiMsg) {
|
||||
const messages = this.buildChatMessages(content, type)
|
||||
try {
|
||||
await new Promise((resolve, reject) => {
|
||||
const ctrl = api.kieaiGeminiChatStream({ messages })
|
||||
this._streamCtrl = ctrl
|
||||
ctrl.onMessage((deltaText) => {
|
||||
if (aiMsg.loading) { aiMsg.loading = false; aiMsg.streaming = true }
|
||||
aiMsg.content += deltaText
|
||||
this.messageList = [...this.messageList]
|
||||
this.scrollToBottom()
|
||||
})
|
||||
ctrl.onError((err) => { this._streamCtrl = null; reject(err) })
|
||||
ctrl.onComplete(() => { this._streamCtrl = null; resolve() })
|
||||
})
|
||||
if (!aiMsg.content.trim()) aiMsg.content = '模型未返回有效内容，请稍后重试。'
|
||||
} catch (streamError) {
|
||||
console.warn('[sendToAI:gemini] 流式失败，降级非流式:', streamError)
|
||||
try {
|
||||
const response = await api.kieaiGeminiChat({ messages, stream: false })
|
||||
const reply = this.getGeminiReplyFromResponse(response)
|
||||
aiMsg.content = reply.trim() ? reply : '模型未返回有效内容，请稍后重试。'
|
||||
} catch (error) {
|
||||
console.error('Gemini 对话失败:', error)
|
||||
aiMsg.content = (error && error.message) || '抱歉，处理您的请求时出现错误，请稍后再试。'
|
||||
}
|
||||
}
|
||||
},
|
||||
|
||||
/**
|
||||
* 非流式 Coze 降级：轮询对话状态并获取回复
|
||||
*/
|
||||
async pollCozeResult(conversationId, chatId, aiMsg) {
|
||||
const maxAttempts = 30
|
||||
const interval = 2000
|
||||
for (let i = 0; i < maxAttempts; i++) {
|
||||
await this.sleep(interval)
|
||||
try {
|
||||
const res = await api.cozeRetrieveChat({ conversationId, chatId })
|
||||
const status = res && res.data && res.data.status
|
||||
if (status === 'completed') {
|
||||
const msgRes = await api.cozeMessageList({ conversationId, chatId })
|
||||
if (msgRes && msgRes.data && Array.isArray(msgRes.data)) {
|
||||
const answer = msgRes.data.find(m => m.role === 'assistant' && m.type === 'answer')
|
||||
if (answer && answer.content) {
|
||||
aiMsg.content = answer.content
|
||||
this.messageList = [...this.messageList]
|
||||
this.scrollToBottom()
|
||||
return
|
||||
}
|
||||
}
|
||||
aiMsg.content = '模型未返回有效内容，请稍后重试。'
|
||||
return
|
||||
} else if (status === 'failed') {
|
||||
aiMsg.content = '对话处理失败，请稍后重试。'
|
||||
return
|
||||
}
|
||||
} catch (e) {
|
||||
console.warn('[pollCozeResult] 轮询出错:', e)
|
||||
}
|
||||
}
|
||||
aiMsg.content = '等待超时，请稍后重试。'
|
||||
},
|
||||
|
||||
// ---------- TTS 方法 ----------
|
||||
|
||||
initAudioContext() {
|
||||
|
||||
21
sql/add_ai_chat_model_config.sql
Normal file
21
sql/add_ai_chat_model_config.sql
Normal file
@@ -0,0 +1,21 @@
|
||||
-- ============================================================
|
||||
-- AI 营养师对话模型配置项
|
||||
-- 在 eb_system_config 表中新增模型选择配置
|
||||
-- 可选值: doubao / coze / gemini
|
||||
-- ============================================================
|
||||
|
||||
-- 插入配置项（如果不存在）
|
||||
INSERT INTO eb_system_config (`name`, `title`, `value`, `form_id`, `status`, `create_time`, `update_time`)
|
||||
SELECT 'ai_chat_model', 'AI对话模型选择(doubao/coze/gemini)', 'doubao', 0, 0, NOW(), NOW()
|
||||
FROM DUAL
|
||||
WHERE NOT EXISTS (
|
||||
SELECT 1 FROM eb_system_config WHERE `name` = 'ai_chat_model'
|
||||
);
|
||||
|
||||
-- 如果需要修改模型，可以直接更新：
|
||||
-- UPDATE eb_system_config SET `value` = 'coze', `update_time` = NOW() WHERE `name` = 'ai_chat_model';
|
||||
-- UPDATE eb_system_config SET `value` = 'gemini', `update_time` = NOW() WHERE `name` = 'ai_chat_model';
|
||||
-- UPDATE eb_system_config SET `value` = 'doubao', `update_time` = NOW() WHERE `name` = 'ai_chat_model';
|
||||
|
||||
-- 验证：
|
||||
-- SELECT * FROM eb_system_config WHERE `name` = 'ai_chat_model';
|
||||
104
test-doubao-api.py
Normal file
104
test-doubao-api.py
Normal file
@@ -0,0 +1,104 @@
|
||||
#!/usr/bin/env python3
|
||||
"""
|
||||
豆包（火山引擎 Ark）API 快速测试
|
||||
用法: python3 test-doubao-api.py
|
||||
"""
|
||||
import json, time, sys
|
||||
|
||||
try:
|
||||
import requests
|
||||
except ImportError:
|
||||
print("需要安装 requests: pip install requests")
|
||||
sys.exit(1)
|
||||
|
||||
API_KEY = "18480c26-ebcd-4263-8a6f-48359b8bd65d"
|
||||
BASE_URL = "https://ark.cn-beijing.volces.com/api/v3"
|
||||
MODEL = "doubao-seed-2-0-pro-260215"
|
||||
|
||||
headers = {
|
||||
"Content-Type": "application/json",
|
||||
"Authorization": f"Bearer {API_KEY}"
|
||||
}
|
||||
|
||||
print("=" * 50)
|
||||
print(" 豆包 API 测试")
|
||||
print("=" * 50)
|
||||
|
||||
# ---- 测试 1: 非流式 ----
|
||||
print("\n【测试 1】非流式调用")
|
||||
payload = {
|
||||
"model": MODEL,
|
||||
"messages": [{"role": "user", "content": "你好，请用一句话介绍你自己"}],
|
||||
"stream": False
|
||||
}
|
||||
|
||||
start = time.time()
|
||||
try:
|
||||
resp = requests.post(f"{BASE_URL}/chat/completions", headers=headers, json=payload, timeout=30)
|
||||
elapsed = (time.time() - start) * 1000
|
||||
print(f"HTTP 状态码: {resp.status_code}")
|
||||
print(f"耗时: {elapsed:.0f} ms")
|
||||
if resp.status_code == 200:
|
||||
data = resp.json()
|
||||
choices = data.get("choices", [])
|
||||
if choices:
|
||||
print(f"回复: {choices[0]['message']['content']}")
|
||||
usage = data.get("usage", {})
|
||||
print(f"Token: prompt={usage.get('prompt_tokens',0)}, completion={usage.get('completion_tokens',0)}, total={usage.get('total_tokens',0)}")
|
||||
print(f"模型: {data.get('model', '?')}")
|
||||
else:
|
||||
print(f"错误: {resp.text[:500]}")
|
||||
except Exception as e:
|
||||
elapsed = (time.time() - start) * 1000
|
||||
print(f"请求失败 ({elapsed:.0f}ms): {e}")
|
||||
|
||||
# ---- 测试 2: 流式 ----
|
||||
print("\n【测试 2】流式调用 (SSE)")
|
||||
payload_stream = {
|
||||
"model": MODEL,
|
||||
"messages": [{"role": "user", "content": "简单说一下健康饮食的三个要点，每个要点一句话"}],
|
||||
"stream": True
|
||||
}
|
||||
|
||||
start2 = time.time()
|
||||
first_token_time = None
|
||||
full_content = ""
|
||||
|
||||
try:
|
||||
resp2 = requests.post(f"{BASE_URL}/chat/completions", headers=headers, json=payload_stream, stream=True, timeout=30)
|
||||
print(f"HTTP 状态码: {resp2.status_code}")
|
||||
if resp2.status_code == 200:
|
||||
print("流式输出: ", end="", flush=True)
|
||||
for line in resp2.iter_lines(decode_unicode=True):
|
||||
if not line:
|
||||
continue
|
||||
if line.startswith("data: "):
|
||||
json_str = line[6:].strip()
|
||||
if json_str == "[DONE]":
|
||||
break
|
||||
try:
|
||||
evt = json.loads(json_str)
|
||||
delta = evt.get("choices", [{}])[0].get("delta", {}).get("content", "")
|
||||
if delta:
|
||||
if first_token_time is None:
|
||||
first_token_time = time.time()
|
||||
full_content += delta
|
||||
print(delta, end="", flush=True)
|
||||
except json.JSONDecodeError:
|
||||
pass
|
||||
|
||||
total_time = (time.time() - start2) * 1000
|
||||
ttft = (first_token_time - start2) * 1000 if first_token_time else 0
|
||||
print(f"\n---")
|
||||
print(f"首字时间 (TTFT): {ttft:.0f} ms")
|
||||
print(f"总耗时: {total_time:.0f} ms")
|
||||
print(f"回复长度: {len(full_content)} 字符")
|
||||
else:
|
||||
print(f"错误: {resp2.text[:500]}")
|
||||
except Exception as e:
|
||||
elapsed2 = (time.time() - start2) * 1000
|
||||
print(f"\n请求失败 ({elapsed2:.0f}ms): {e}")
|
||||
|
||||
print("\n" + "=" * 50)
|
||||
print(" 测试完成")
|
||||
print("=" * 50)
|
||||
144
test-doubao-api.sh
Normal file
144
test-doubao-api.sh
Normal file
@@ -0,0 +1,144 @@
|
||||
#!/bin/bash
|
||||
# 豆包 API 连通性测试脚本
|
||||
# 用法: bash test-doubao-api.sh
|
||||
|
||||
API_KEY="18480c26-ebcd-4263-8a6f-48359b8bd65d"
|
||||
BASE_URL="https://ark.cn-beijing.volces.com/api/v3"
|
||||
MODEL="doubao-seed-2-0-pro-260215"
|
||||
|
||||
echo "============================================"
|
||||
echo " 豆包（火山引擎 Ark）API 测试"
|
||||
echo "============================================"
|
||||
echo ""
|
||||
|
||||
# ---- 测试 1: 非流式调用 ----
|
||||
echo "【测试 1】非流式调用"
|
||||
echo "请求: 你好，请用一句话介绍你自己"
|
||||
echo "---"
|
||||
|
||||
START=$(date +%s%3N 2>/dev/null || python3 -c "import time; print(int(time.time()*1000))")
|
||||
|
||||
RESPONSE=$(curl -s -w "\n---HTTP_CODE:%{http_code}---" \
|
||||
-X POST "${BASE_URL}/chat/completions" \
|
||||
-H "Content-Type: application/json" \
|
||||
-H "Authorization: Bearer ${API_KEY}" \
|
||||
-d "{
|
||||
\"model\": \"${MODEL}\",
|
||||
\"messages\": [{\"role\": \"user\", \"content\": \"你好，请用一句话介绍你自己\"}],
|
||||
\"stream\": false
|
||||
}" 2>&1)
|
||||
|
||||
END=$(date +%s%3N 2>/dev/null || python3 -c "import time; print(int(time.time()*1000))")
|
||||
ELAPSED=$((END - START))
|
||||
|
||||
HTTP_CODE=$(echo "$RESPONSE" | grep -oP '(?<=---HTTP_CODE:)\d+(?=---)')
|
||||
BODY=$(echo "$RESPONSE" | sed 's/---HTTP_CODE:[0-9]*---//')
|
||||
|
||||
echo "HTTP 状态码: ${HTTP_CODE}"
|
||||
echo "耗时: ${ELAPSED} ms"
|
||||
|
||||
if [ "$HTTP_CODE" = "200" ]; then
|
||||
# 提取回复内容
|
||||
CONTENT=$(echo "$BODY" | python3 -c "
|
||||
import sys, json
|
||||
data = json.load(sys.stdin)
|
||||
choices = data.get('choices', [])
|
||||
if choices:
|
||||
print('回复:', choices[0].get('message', {}).get('content', '(empty)'))
|
||||
usage = data.get('usage', {})
|
||||
print(f'Token: prompt={usage.get(\"prompt_tokens\",0)}, completion={usage.get(\"completion_tokens\",0)}, total={usage.get(\"total_tokens\",0)}')
|
||||
print(f'模型: {data.get(\"model\", \"?\")}')
|
||||
" 2>&1)
|
||||
echo "$CONTENT"
|
||||
else
|
||||
echo "错误响应: $BODY"
|
||||
fi
|
||||
|
||||
echo ""
|
||||
echo ""
|
||||
|
||||
# ---- 测试 2: 流式调用 ----
|
||||
echo "【测试 2】流式调用 (SSE)"
|
||||
echo "请求: 简单说一下健康饮食的三个要点"
|
||||
echo "---"
|
||||
|
||||
START2=$(date +%s%3N 2>/dev/null || python3 -c "import time; print(int(time.time()*1000))")
|
||||
|
||||
# 流式请求，逐行输出
|
||||
STREAM_OUTPUT=""
|
||||
FIRST_CHUNK_TIME=""
|
||||
|
||||
curl -s -N \
|
||||
-X POST "${BASE_URL}/chat/completions" \
|
||||
-H "Content-Type: application/json" \
|
||||
-H "Authorization: Bearer ${API_KEY}" \
|
||||
-d "{
|
||||
\"model\": \"${MODEL}\",
|
||||
\"messages\": [{\"role\": \"user\", \"content\": \"简单说一下健康饮食的三个要点，每个要点一句话\"}],
|
||||
\"stream\": true
|
||||
}" 2>/dev/null | while IFS= read -r line; do
|
||||
if [[ "$line" == data:* ]]; then
|
||||
JSON_DATA="${line#data: }"
|
||||
if [ "$JSON_DATA" = "[DONE]" ]; then
|
||||
break
|
||||
fi
|
||||
# 提取 delta content
|
||||
DELTA=$(echo "$JSON_DATA" | python3 -c "
|
||||
import sys, json
|
||||
try:
|
||||
d = json.load(sys.stdin)
|
||||
c = d.get('choices',[{}])[0].get('delta',{}).get('content','')
|
||||
if c: print(c, end='', flush=True)
|
||||
except: pass
|
||||
" 2>/dev/null)
|
||||
if [ -z "$FIRST_CHUNK_TIME" ] && [ -n "$DELTA" ]; then
|
||||
FIRST_CHUNK_TIME=$(date +%s%3N 2>/dev/null || python3 -c "import time; print(int(time.time()*1000))")
|
||||
fi
|
||||
fi
|
||||
done
|
||||
|
||||
END2=$(date +%s%3N 2>/dev/null || python3 -c "import time; print(int(time.time()*1000))")
|
||||
ELAPSED2=$((END2 - START2))
|
||||
|
||||
echo ""
|
||||
echo "---"
|
||||
echo "流式总耗时: ${ELAPSED2} ms"
|
||||
|
||||
echo ""
|
||||
echo ""
|
||||
|
||||
# ---- 测试 3: 通过后端接口调用（需先启动后端服务） ----
|
||||
echo "【测试 3】通过后端接口调用（需后端服务运行中）"
|
||||
echo "请确认后端已启动，默认地址: http://localhost:8081"
|
||||
echo "---"
|
||||
|
||||
BACKEND_URL="http://localhost:8081/api/front/doubao/chat"
|
||||
|
||||
RESP3=$(curl -s -w "\n---HTTP_CODE:%{http_code}---" \
|
||||
-X POST "${BACKEND_URL}" \
|
||||
-H "Content-Type: application/json" \
|
||||
-d "{
|
||||
\"messages\": [{\"role\": \"user\", \"content\": \"你好\"}]
|
||||
}" 2>&1)
|
||||
|
||||
HTTP3=$(echo "$RESP3" | grep -oP '(?<=---HTTP_CODE:)\d+(?=---)')
|
||||
BODY3=$(echo "$RESP3" | sed 's/---HTTP_CODE:[0-9]*---//')
|
||||
|
||||
if [ "$HTTP3" = "200" ]; then
|
||||
echo "后端接口调用成功! HTTP $HTTP3"
|
||||
echo "$BODY3" | python3 -c "
|
||||
import sys, json
|
||||
data = json.load(sys.stdin)
|
||||
choices = data.get('choices', [])
|
||||
if choices:
|
||||
print('回复:', choices[0].get('message', {}).get('content', '(empty)')[:100])
|
||||
" 2>&1
|
||||
else
|
||||
echo "后端接口调用失败: HTTP ${HTTP3:-连接失败}"
|
||||
echo "（如果后端未启动，这是正常的）"
|
||||
fi
|
||||
|
||||
echo ""
|
||||
echo "============================================"
|
||||
echo " 测试完成"
|
||||
echo "============================================"
|
||||
322
客户反馈分析报告.md
Normal file
322
客户反馈分析报告.md
Normal file
@@ -0,0 +1,322 @@
|
||||
# 测试问题与优化建议 — 修改解决方案
|
||||
|
||||
**分析日期：** 2026年4月11日
|
||||
**项目：** msh_single_uniapp（UniApp 小程序）
|
||||
**涉及页面：** calculator-result / ai-nutritionist / food-encyclopedia
|
||||
|
||||
---
|
||||
|
||||
## 一、总览：需处理项 vs 暂不处理项
|
||||
|
||||
| # | 问题 | 决策 | 涉及文件 |
|
||||
|---|------|------|----------|
|
||||
| 1 | 食物份数 → 克数 | **采纳** | `pages/tool/calculator-result.vue` |
|
||||
| 2 | AI对话交互按钮 | **部分采纳** | `pages/tool/ai-nutritionist.vue` |
|
||||
| 3 | AI回复速度慢 | **已优化（切换Coze流式API）** | `ai-nutritionist.vue` + `models-api.js` |
|
||||
| 4 | 食物百科图片缺失 | **不处理**（后台可改） | — |
|
||||
| 5 | 百科卡片误导点击 | **采纳**（跳转详情页） | `pages/tool/food-encyclopedia.vue` |
|
||||
| 6 | 角色权限与分销 | **暂不处理**（后台人工配置） | — |
|
||||
| 7 | 百科点击报错 Bug | **必须修复** | `pages/tool/food-encyclopedia.vue` |
|
||||
|
||||
---
|
||||
|
||||
## 二、需修改的问题及代码级解决方案
|
||||
|
||||
---
|
||||
|
||||
### 问题1：食物份数建议改为克数
|
||||
|
||||
**页面：** `pages/tool/calculator-result.vue`
|
||||
|
||||
**现状分析：**
|
||||
当前"食物份数建议"卡片（约第94-113行）中，食物列表渲染逻辑为：
|
||||
|
||||
```html
|
||||
<text class="food-portion">{{ item.portion }} 份</text>
|
||||
```
|
||||
|
||||
数据结构 `foodList` 中每个 item 包含 `{ number, name, portion }`，`portion` 为份数值。
|
||||
|
||||
**修改方案：**
|
||||
|
||||
1. **后端接口改造**（推荐）：`getCalculatorResult` 接口返回数据中增加 `gram` 字段，或将 `portion` 的含义改为克数。
|
||||
|
||||
2. **前端模板修改**（`calculator-result.vue` 约第94-113行）：
|
||||
|
||||
```html
|
||||
<!-- 修改前 -->
|
||||
<text class="card-title">食物份数建议</text>
|
||||
...
|
||||
<text class="food-portion">{{ item.portion }} 份</text>
|
||||
|
||||
<!-- 修改后 -->
|
||||
<text class="card-title">每日食物建议</text>
|
||||
...
|
||||
<text class="food-portion">{{ item.gram || item.portion }} 克</text>
|
||||
```
|
||||
|
||||
3. **数据层适配**（`applyResult` 方法，约第320-350行）：解析接口返回时，确保 `foodList` 中的 item 携带 `gram` 字段。如后端暂未改造，可前端用换算公式临时过渡：
|
||||
|
||||
```javascript
|
||||
// 在 applyResult 方法中
|
||||
this.foodList = (res.data.foodList || []).map(item => ({
|
||||
...item,
|
||||
gram: item.gram || Math.round(item.portion * item.gramPerServing) || item.portion
|
||||
}))
|
||||
```
|
||||
|
||||
**工作量估计：** 前端 0.5天，后端（如需改接口）0.5天
|
||||
|
||||
---
|
||||
|
||||
### 问题2：AI营养师对话页面新增交互按钮
|
||||
|
||||
**页面：** `pages/tool/ai-nutritionist.vue`
|
||||
|
||||
**现状分析：**
|
||||
当前每条 AI 消息仅有 **1个操作按钮**：TTS 语音朗读（约第84-90行）。消息数据结构为 `{ role, content, type, loading, streaming }`。AI 回复通过 `api.kieaiGeminiChat()` 调用（实际走 `/api/front/kieai/gemini/chat`，`stream: false` 非流式），整体响应一次性返回。
|
||||
|
||||
**部分采纳后的需求：** 新增"复制"、"重新生成"、"删除"按钮（语音朗读已有）。
|
||||
|
||||
**修改方案：**
|
||||
|
||||
在消息气泡下方的操作区域（约第84-90行 TTS 按钮位置），扩展为按钮组：
|
||||
|
||||
```html
|
||||
<!-- 修改前：仅 TTS 按钮 -->
|
||||
<view class="tts-play-btn" v-if="msg.role === 'ai' && !msg.loading && !msg.streaming && msg.content && msg.type !== 'image'" @click="...">
|
||||
<text>{{ ttsPlayingIndex === index ? '⏹' : '▶' }}</text>
|
||||
</view>
|
||||
|
||||
<!-- 修改后：操作按钮组 -->
|
||||
<view class="msg-actions" v-if="msg.role === 'ai' && !msg.loading && !msg.streaming && msg.content && msg.type !== 'image'">
|
||||
<!-- 复制 -->
|
||||
<view class="action-btn" @click="copyMessage(index)">
|
||||
<text class="action-icon">📋</text>
|
||||
</view>
|
||||
<!-- 重新生成（仅最后一条AI消息显示） -->
|
||||
<view class="action-btn" v-if="isLastAiMessage(index)" @click="regenerateMessage(index)">
|
||||
<text class="action-icon">🔄</text>
|
||||
</view>
|
||||
<!-- 语音朗读（已有） -->
|
||||
<view class="action-btn" @click="ttsPlayingIndex === index ? stopTTS() : playTTS(index)">
|
||||
<text class="action-icon">{{ ttsPlayingIndex === index ? '⏹' : '▶' }}</text>
|
||||
</view>
|
||||
<!-- 删除 -->
|
||||
<view class="action-btn" @click="deleteMessage(index)">
|
||||
<text class="action-icon">🗑️</text>
|
||||
</view>
|
||||
</view>
|
||||
```
|
||||
|
||||
**新增 methods：**
|
||||
|
||||
```javascript
|
||||
// 复制消息
|
||||
copyMessage(index) {
|
||||
const msg = this.messageList[index]
|
||||
uni.setClipboardData({
|
||||
data: msg.content,
|
||||
success: () => uni.showToast({ title: '已复制', icon: 'success' })
|
||||
})
|
||||
},
|
||||
|
||||
// 删除单条消息
|
||||
deleteMessage(index) {
|
||||
uni.showModal({
|
||||
title: '提示',
|
||||
content: '确定删除这条消息吗？',
|
||||
success: (res) => {
|
||||
if (res.confirm) {
|
||||
this.messageList.splice(index, 1)
|
||||
}
|
||||
}
|
||||
})
|
||||
},
|
||||
|
||||
// 重新生成（重发上一条用户消息）
|
||||
regenerateMessage(index) {
|
||||
// 找到该AI消息对应的上一条用户消息
|
||||
let userMsgIndex = index - 1
|
||||
while (userMsgIndex >= 0 && this.messageList[userMsgIndex].role !== 'user') {
|
||||
userMsgIndex--
|
||||
}
|
||||
if (userMsgIndex < 0) return
|
||||
// 移除当前AI回复
|
||||
this.messageList.splice(index, 1)
|
||||
// 重新发送
|
||||
this.sendToAI()
|
||||
},
|
||||
|
||||
// 判断是否为最后一条AI消息
|
||||
isLastAiMessage(index) {
|
||||
for (let i = this.messageList.length - 1; i >= 0; i--) {
|
||||
if (this.messageList[i].role === 'ai') return i === index
|
||||
}
|
||||
return false
|
||||
}
|
||||
```
|
||||
|
||||
**新增样式：**
|
||||
|
||||
```scss
|
||||
.msg-actions {
|
||||
display: flex;
|
||||
gap: 16rpx;
|
||||
margin-top: 8rpx;
|
||||
justify-content: flex-start;
|
||||
}
|
||||
.action-btn {
|
||||
width: 56rpx;
|
||||
height: 56rpx;
|
||||
display: flex;
|
||||
align-items: center;
|
||||
justify-content: center;
|
||||
border-radius: 50%;
|
||||
background: #f4f5f7;
|
||||
}
|
||||
.action-icon {
|
||||
font-size: 24rpx;
|
||||
}
|
||||
```
|
||||
|
||||
**工作量估计：** 1天
|
||||
|
||||
---
|
||||
|
||||
### 问题3：AI回复速度慢（已优化 — 切换至 Coze API）
|
||||
|
||||
**优化前调用链路：**
|
||||
|
||||
```
|
||||
小程序 → POST /api/front/kieai/gemini/chat/stream
|
||||
→ Spring KieAIController (SseEmitter)
|
||||
→ OkHttp3 异步 POST https://api.kie.ai/gemini-2.5-flash/v1/chat/completions
|
||||
→ KieAI 代理网关 → Gemini 2.5 Flash 模型
|
||||
```
|
||||
|
||||
**优化后调用链路：**
|
||||
|
||||
```
|
||||
小程序 → POST /api/front/coze/chat/stream
|
||||
→ Spring CozeController (SseEmitter, 120s timeout, 15s heartbeat)
|
||||
→ Coze 官方 SDK (client.chat().stream())
|
||||
→ https://api.coze.cn → Coze Bot (7591133240535449654)
|
||||
```
|
||||
|
||||
**已完成的优化：**
|
||||
|
||||
1. **前端切换至 Coze 流式 API：** `sendToAI()` 改为调用 `api.cozeChatStream()`，直接使用 Coze 的流式事件（`conversation.message.delta`），逐字渲染 AI 回复。首字可见时间大幅缩短。
|
||||
|
||||
2. **Coze 事件解析：** 正确解析 Coze SSE 事件格式 `{ event, conversation_id, chat_id, content, role, type }`，仅累积 `event=conversation.message.delta` 且 `type=answer` 的文本增量。
|
||||
|
||||
3. **多轮对话支持：** 自动保存 `conversation_id`，后续消息在同一会话中进行，Coze Bot 可获取完整上下文。
|
||||
|
||||
4. **消息格式适配：** 新增 `buildCozeMessages()` / `buildCozeRequest()` 方法，将用户输入（文本/图片/多模态）转换为 Coze 格式的 `additionalMessages`。
|
||||
|
||||
5. **降级策略：** 流式失败时自动降级为非流式 `api.cozeChat()`，并通过 `pollCozeResult()` 轮询获取最终回复。
|
||||
|
||||
**Coze 相比 KieAI 代理的优势：**
|
||||
|
||||
- 去掉一层 KieAI 代理网关，减少一跳网络延迟
|
||||
- Coze 后端已配置 `X-Accel-Buffering: no` + `Cache-Control: no-cache`，SSE 不会被 nginx 缓冲
|
||||
- Coze SDK 内置连接池管理，不再每次请求 new OkHttpClient
|
||||
- 内置 15s heartbeat 防止连接超时断开
|
||||
- 支持通过 `conversation_id` 进行多轮上下文对话
|
||||
|
||||
---
|
||||
|
||||
### 问题4：食物百科图片缺失
|
||||
|
||||
**决策：不处理。** 图片数据通过商城 PC 管理后台维护，属运营侧工作。
|
||||
|
||||
---
|
||||
|
||||
### 问题5：百科卡片点击 → 跳转食物详情页
|
||||
|
||||
**页面：** `pages/tool/food-encyclopedia.vue`
|
||||
|
||||
**现状分析：**
|
||||
当前卡片已绑定了 `@click="goToFoodDetail(item)"`（约第103行），`goToFoodDetail` 方法（约第1012-1030行）会尝试提取 item 的 id 并跳转到 `/pages/tool/food-detail`。项目中已存在 `pages/tool/food-detail.vue` 页面。
|
||||
|
||||
**问题：** 跳转逻辑本身已实现，但存在 Bug（见问题7），导致点击报错而非正常跳转，用户看到的效果就是"点击无反应"或"点击后闪退"。
|
||||
|
||||
**修改方案：** 修复问题7的 Bug 后，卡片点击跳转详情页的功能即可正常工作。需确认：
|
||||
|
||||
1. `food-detail.vue` 详情页内容是否完整可用
|
||||
2. `pages.json` 中是否已注册 `/pages/tool/food-detail` 路由
|
||||
|
||||
**工作量估计：** 与问题7合并处理，0.5天
|
||||
|
||||
---
|
||||
|
||||
### 问题6：角色权限与分销系统
|
||||
|
||||
**决策：暂不处理。** 临时方案为运营人员在后台手动为指定用户账号配置角色标签（医生/护士/普通用户）。
|
||||
|
||||
---
|
||||
|
||||
### 问题7（Bug）：食物百科点击报错 TypeError
|
||||
|
||||
**页面：** `pages/tool/food-encyclopedia.vue`
|
||||
**错误信息：**
|
||||
```
|
||||
TypeError: Cannot read property 'id' of undefined
|
||||
at pickNumericIdStr (food-encyclopedia.vue:1015)
|
||||
at VueComponent.goToFoodDetail (food-encyclopedia.vue:1022)
|
||||
```
|
||||
|
||||
**根因分析：**
|
||||
|
||||
`goToFoodDetail(item)` 方法（第1012行）中，内部闭包 `pickNumericIdStr` 直接访问 `item.id`（第1015行），但未做空值校验。当 `filteredFoodList` 中的某个 item 为 `undefined` 或 `null` 时（可能因 API 返回脏数据、`normalizeFoodItem` 边界情况），点击即触发 TypeError。
|
||||
|
||||
**修复方案：**
|
||||
|
||||
在 `goToFoodDetail` 方法开头增加防御性校验（约第1012行）：
|
||||
|
||||
```javascript
|
||||
goToFoodDetail(item) {
|
||||
// 防御性校验：避免 item 为空时报错
|
||||
if (!item || typeof item !== 'object') {
|
||||
console.warn('[food-encyclopedia] goToFoodDetail: item 为空，跳过跳转')
|
||||
return
|
||||
}
|
||||
|
||||
const pickNumericIdStr = () => {
|
||||
const cands = [item.id, item.foodId, item.food_id, item.v2FoodId, item.v2_food_id, item.foodID]
|
||||
// ... 原有逻辑不变
|
||||
}
|
||||
// ... 后续逻辑不变
|
||||
}
|
||||
```
|
||||
|
||||
同时建议加固 `filteredFoodList` 计算属性（约第261行），确保过滤掉无效项：
|
||||
|
||||
```javascript
|
||||
filteredFoodList() {
|
||||
const list = (this.foodList || []).filter(item => item != null && typeof item === 'object' && item.name)
|
||||
// ... 后续过滤逻辑不变
|
||||
}
|
||||
```
|
||||
|
||||
**工作量估计：** 0.5天（含测试验证）
|
||||
|
||||
---
|
||||
|
||||
## 三、执行优先级排序
|
||||
|
||||
| 优先级 | 问题 | 预估工时 | 负责方 |
|
||||
|--------|------|----------|--------|
|
||||
| **P0** | #7 百科点击报错 Bug | 0.5天 | 前端 |
|
||||
| **P1** | #1 份数→克数 | 0.5-1天 | 前端+后端 |
|
||||
| **P1** | #5 百科卡片跳转详情页 | 与#7合并 | 前端 |
|
||||
| **P2** | #3 AI响应速度（已切换Coze流式API） | ✅ 已完成 | 前端 |
|
||||
| **P2** | #2 AI对话新增操作按钮 | 1天 | 前端 |
|
||||
| **—** | #4 百科图片 | 不处理 | 运营 |
|
||||
| **—** | #6 角色权限 | 暂不处理 | 运营（后台手动） |
|
||||
|
||||
**总预估：** 前端约 3-4天，后端约 2-3天
|
||||
|
||||
---
|
||||
|
||||
*基于 msh_single_uniapp 项目代码分析生成，所有代码引用均已核对源文件。*
|
||||
Reference in New Issue
Block a user