新增

huliguo

2 天以前 8e0decd25b9ec86b02d58de53dee1451f83d1566

 src/main/java/com/linghu/controller/CollectController.java

@@ -1,55 +1,49 @@
package com.linghu.controller;

import java.time.Duration;
import java.time.LocalDateTime;
import java.time.format.DateTimeFormatter;
import java.util.ArrayList;
import java.util.List;
import java.util.Map;
import java.util.Optional;
import java.util.*;
import java.util.stream.Collectors;

import javax.servlet.http.HttpServletRequest;

import com.fasterxml.jackson.core.JsonProcessingException;
import com.fasterxml.jackson.databind.JsonNode;
import com.fasterxml.jackson.databind.ObjectMapper;
import com.linghu.model.dto.*;
import org.springframework.beans.BeanUtils;
import com.linghu.model.entity.*;
import com.linghu.service.*;
import lombok.extern.slf4j.Slf4j;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.beans.factory.annotation.Value;
import org.springframework.core.ParameterizedTypeReference;
import org.springframework.http.*;
import org.springframework.web.client.RestTemplate;
import org.springframework.web.reactive.function.client.WebClient;
import org.springframework.http.client.HttpComponentsClientHttpRequestFactory;

import com.baomidou.mybatisplus.core.conditions.query.LambdaQueryWrapper;
import com.baomidou.mybatisplus.core.conditions.update.LambdaUpdateWrapper;
import com.linghu.model.common.ResponseResult;
import com.linghu.model.entity.Keyword;
import com.linghu.model.entity.Question;
import com.linghu.model.entity.User;
import com.linghu.service.KeywordService;
import com.linghu.service.QuestionService;
import com.linghu.service.ReferenceService;
import com.linghu.utils.JwtUtils;

import io.jsonwebtoken.lang.Collections;
import io.swagger.annotations.Api;
import io.swagger.annotations.ApiOperation;
import reactor.core.publisher.Flux;
import reactor.core.publisher.Mono;

import org.springframework.web.bind.annotation.* ;
import org.springframework.http.HttpStatus;
import com.linghu.model.dto.TaskResultResponse.QuestionResult;
import com.linghu.model.dto.TaskResultResponse.UserResult;
import com.linghu.model.entity.Reference;
import reactor.core.scheduler.Schedulers;

import java.util.stream.Collectors;
import java.util.stream.IntStream;
import java.util.stream.Stream;

@RestController
@RequestMapping("/collect")
@Api(value = "采集接口", tags = "采集管理")
@Slf4j
public class CollectController {

    @Autowired
@@ -67,26 +61,19 @@
    private KeywordService keywordService;
    @Autowired
    private QuestionService questionService;
    @Autowired
    private KeywordTaskService keywordTaskService;
    @Autowired
    private PlatformService platformService;
    @Autowired
    private TypeService typeService;

    @PostMapping("/search")

 /*   @PostMapping("/search")
    @ApiOperation(value = "开始采集")
    public Mono<SearchTaskResponse> createSearchTask(
            @RequestBody SearchTaskRequest searchTaskRequest,
            HttpServletRequest request) throws JsonProcessingException {
//        String token = request.getHeader("Authorization");
//        User user = jwtUtils.parseToken(token);
//        // 复制到UserDto
//        UserDto userDto = new UserDto();
//        userDto.setName(user.getUser_name());
//        userDto.setEmail(user.getUser_email());
//        userDto.setPassword(user.getPassword());
//
//        List<UserDto> users = new ArrayList<>();
//        users.add(userDto);
//        searchTaskRequest.setUsers(users);
        // json格式
        ObjectMapper objectMapper = new ObjectMapper();
        System.out.println(objectMapper.writeValueAsString(searchTaskRequest));

        return webClient.post()
                .uri(baseUrl + "/api/v1/search")
@@ -104,10 +91,12 @@
                        // 保存任务ID到关键词
                        LambdaUpdateWrapper<Keyword> updateWrapper = new LambdaUpdateWrapper<>();
                        updateWrapper.eq(Keyword::getKeyword_id, searchTaskRequest.getKeyword_id());
                        updateWrapper.set(Keyword::getStatus,"Submitted");
                        updateWrapper.set(Keyword::getTask_id, taskResponse.getTask_id());
                        keywordService.update(updateWrapper);

                        // 可选：更新响应中的其他信息
                        // taskResponse.setMessage("任务已提交并保存，ID: " + taskResponse.getTaskId());

                    }
                    return Mono.just(taskResponse);
                })
@@ -117,8 +106,230 @@
                    task.setMessage("调用失败: " + e.getMessage());
                    return Mono.just(task);
                });
    }*/

//    public SearchTaskController(WebClient.Builder webClientBuilder, KeywordService keywordService) {
//        this.webClient = webClientBuilder.build();
//        this.keywordService = keywordService;
//    }

   /* @PostMapping("/search")
    @ApiOperation(value = "开始采集")
    public Mono<SearchTaskResponse> createSearchTask(
            @RequestBody SearchTaskRequest searchTaskRequest,
            HttpServletRequest request) throws JsonProcessingException {

        int maxConcurrentUsers = searchTaskRequest.getConfig() != null ?
                searchTaskRequest.getConfig().getMax_concurrent_users() : 3;
        List<List<UserDto>> userBatches = splitUsersIntoBatches(searchTaskRequest.getUsers(), maxConcurrentUsers);

        return processBatchesSequentially(userBatches, searchTaskRequest)
                .onErrorResume(e -> {
                    SearchTaskResponse task = new SearchTaskResponse();
                    task.setMessage("调用失败: " + e.getMessage());
                    return Mono.just(task);
                });
    }

    private List<List<UserDto>> splitUsersIntoBatches(List<UserDto> users, int batchSize) {
        List<List<UserDto>> batches = new ArrayList<>();
        for (int i = 0; i < users.size(); i += batchSize) {
            batches.add(users.subList(i, Math.min(i + batchSize, users.size())));
        }
        return batches;
    }

    private Mono<SearchTaskResponse> processBatchesSequentially(List<List<UserDto>> userBatches, SearchTaskRequest originalRequest) {
        Mono<SearchTaskResponse> resultMono = Mono.empty();
        for (List<UserDto> batch : userBatches) {
            SearchTaskRequest batchRequest = new SearchTaskRequest();
            batchRequest.setUsers(batch);
            batchRequest.setQuestions(originalRequest.getQuestions());
            batchRequest.setConfig(originalRequest.getConfig());
            batchRequest.setSave_to_database(originalRequest.getSave_to_database());
            batchRequest.setWebhook_url(originalRequest.getWebhook_url());
            batchRequest.setKeyword_id(originalRequest.getKeyword_id());

            resultMono = resultMono.then(createSingleBatchTask(batchRequest));
        }
        return resultMono;
    }

    private Mono<SearchTaskResponse> createSingleBatchTask(SearchTaskRequest batchRequest) {
        return webClient.post()
                .uri(baseUrl + "/api/v1/search")
                .contentType(MediaType.APPLICATION_JSON)
                .bodyValue(batchRequest)
                .retrieve()
                .onStatus(HttpStatus::is4xxClientError, response -> response.bodyToMono(String.class)
                        .flatMap(errorBody -> Mono.error(new RuntimeException(errorBody))))
                .bodyToMono(new ParameterizedTypeReference<SearchTaskResponse>() {
                })
                .flatMap(responseResult -> {
                    SearchTaskResponse taskResponse = responseResult;
                    if (taskResponse != null && taskResponse.getTask_id() != null) {
                        LambdaUpdateWrapper<Keyword> updateWrapper = new LambdaUpdateWrapper<>();
                        updateWrapper.eq(Keyword::getKeyword_id, batchRequest.getKeyword_id());
                        updateWrapper.set(Keyword::getStatus, "Submitted");
                        updateWrapper.set(Keyword::getTask_id, taskResponse.getTask_id());
                        keywordService.update(updateWrapper);
                    }
                    return waitForTaskCompletion(taskResponse.getTask_id())
                            .then(Mono.just(taskResponse));
                });
    }

    private Mono<Void> waitForTaskCompletion(String taskId) {
        return Flux.interval(Duration.ofSeconds(5)) // 每5秒执行一次
                .flatMap(tick -> webClient.get()
                        .uri(baseUrl + "/api/v1/tasks/" + taskId)
                        .retrieve()
                        .bodyToMono(TaskStatusResponse.class)
                )
                .filter(response -> "completed".equals(response.getStatus()))
                .next() // 找到第一个完成的响应后结束流
                .then(); // 转换为Mono<Void>
    }*/
   // 添加一个辅助方法来安全地将字符串转换为double
   private double parseUsage(String usageStr) {
       try {
           if (usageStr != null) {
               // 移除可能存在的百分号
               usageStr = usageStr.replace("%", "").trim();
               return Double.parseDouble(usageStr);
           }
           return 0.0;
       } catch (NumberFormatException e) {
           log.error("解析资源使用率失败: {}", e.getMessage());
           return 0.0;
       }
   }
   @PostMapping("/search")
   @ApiOperation(value = "开始采集")
   public Mono<ResponseResult<?>> createSearchTask(
           @RequestBody SearchTaskRequest searchTaskRequest,
           HttpServletRequest request) throws JsonProcessingException {

       // 首先检查服务器资源
       return getServerResource()
               .flatMap(resourceResponse -> {
                   // 将字符串类型的使用率转换为double类型
                   double cpuUsage = parseUsage(resourceResponse.getCpu_usage_percent());
                   double memoryUsage = parseUsage(resourceResponse.getMemory_usage_percent());
                   // 检查CPU和内存使用率
                   if (cpuUsage >= 90.0 || memoryUsage >= 90.0) {
                       String errorMsg = String.format("服务器资源不足：CPU使用率 %.1f%%，内存使用率 %.1f%%",
                               resourceResponse.getCpu_usage_percent(), resourceResponse.getMemory_usage_percent());
                       log.warn(errorMsg);
                       return Mono.just(ResponseResult.error(503, errorMsg));
                   }
                   Integer keywordId = searchTaskRequest.getKeyword_id();

                   int maxConcurrentUsers = searchTaskRequest.getConfig() != null ?
                           searchTaskRequest.getConfig().getMax_concurrent_users() : 3;
                   List<List<UserDto>> userBatches = splitUsersIntoBatches(searchTaskRequest.getUsers(), maxConcurrentUsers,keywordId);

                   // 获取 keywordId

                   //分割

     
                   return Flux.fromIterable(userBatches)
                           .flatMap(batch -> {
                               SearchTaskRequest batchRequest = new SearchTaskRequest();
                               batchRequest.setUsers(batch);
                               batchRequest.setQuestions(searchTaskRequest.getQuestions());
                               batchRequest.setConfig(searchTaskRequest.getConfig());
                               batchRequest.setSave_to_database(searchTaskRequest.getSave_to_database());
                               batchRequest.setWebhook_url(searchTaskRequest.getWebhook_url());
                               batchRequest.setKeyword_id(keywordId);

                               return createSingleBatchTask(batchRequest)
                                       .delaySubscription(Duration.ofSeconds(2)); // 批次之间添加延迟
                           }, 1) // 限制并发数为1，确保顺序执行
                           .collectList() // 收集所有批次的响应
                           .flatMap(responses ->
                                   saveKeywordTasks(keywordId, responses) // 保存关联关系
                                           .thenReturn(responses) // 返回原始响应
                           )
                           .map(responses -> ResponseResult.success(responses)) // 使用ResponseResult包装结果
                           .onErrorResume(e -> {
                               log.error("创建搜索任务失败: {}", e.getMessage(), e);
                               return Mono.just(ResponseResult.error("创建搜索任务失败: " + e.getMessage()));
                           });
               })
               .onErrorResume(e -> {
                   log.error("检查服务器资源失败: {}", e.getMessage(), e);
                   return Mono.just(ResponseResult.error("检查服务器资源失败: " + e.getMessage()));
               });
   }

    private Mono<Void> saveKeywordTasks(Integer keywordId, List<SearchTaskResponse> taskResponses) {
        List<KeywordTask> keywordTasks = taskResponses.stream()
                .filter(response -> response.getTask_id() != null)
                .map(response -> {
                    KeywordTask keywordTask = new KeywordTask();
                    keywordTask.setKeyword_id(keywordId);
                    keywordTask.setTask_id(response.getTask_id());
                    keywordTask.setStatus("pending");
                    return keywordTask;
                })
                .collect(Collectors.toList());

        // 将 MyBatis-Plus 的同步方法包装为 Mono<Void>
        return Mono.fromRunnable(() -> {
                    boolean success = keywordTaskService.saveOrUpdateBatch(keywordTasks);
                    if (!success) {
                        throw new RuntimeException("保存关键词任务关联失败");
                    }
                })
                .doFinally(signalType -> log.info("成功保存 {} 个关键词任务关联", keywordTasks.size()))
                .then();
    }

    private List<List<UserDto>> splitUsersIntoBatches(List<UserDto> users, int batchSize,Integer keywordId) {

        Keyword keyword = keywordService.getById(keywordId);
        if (null==keyword.getNum()){
            keyword.setNum(0);
        }
        keyword.setNum(keyword.getNum()+1);
        keywordService.updateById(keyword);

        List<List<UserDto>> batches = new ArrayList<>();
        for (int i = 0; i < users.size(); i += batchSize) {
            batches.add(users.subList(i, Math.min(i + batchSize, users.size())));
        }
        return batches;
    }

    private Mono<SearchTaskResponse> createSingleBatchTask(SearchTaskRequest batchRequest) {
        return webClient.post()
                .uri(baseUrl + "/api/v1/search")
                .contentType(MediaType.APPLICATION_JSON)
                .bodyValue(batchRequest)
                .retrieve()
                .onStatus(HttpStatus::is4xxClientError, response -> response.bodyToMono(String.class)
                        .flatMap(errorBody -> Mono.error(new RuntimeException(errorBody))))
                .bodyToMono(new ParameterizedTypeReference<SearchTaskResponse>() {})
                .flatMap(taskResponse -> {
                    if (taskResponse != null && taskResponse.getTask_id() != null) {

                        // 使用 Reactor 的方式更新数据库
                        return Mono.fromRunnable(() -> {
                                    LambdaUpdateWrapper<Keyword> updateWrapper = new LambdaUpdateWrapper<>();
                                    updateWrapper.eq(Keyword::getKeyword_id, batchRequest.getKeyword_id());
                                    updateWrapper.set(Keyword::getStatus, "submitted");
                                    updateWrapper.set(Keyword::getTask_id, taskResponse.getTask_id());
                                    keywordService.update(updateWrapper);
                                }).subscribeOn(Schedulers.boundedElastic()) // 在弹性线程池执行
                                .thenReturn(taskResponse);
                    }
                    return Mono.just(taskResponse);
                });
    }

    // 移除原来的waitForTaskCompletion方法，不再需要同步等待
    @ApiOperation(value = "查询任务状态")
    @GetMapping("/status")
    public Mono<TaskStatusResponse> getTaskStatus(String taskId) {
@@ -126,7 +337,7 @@
                .uri(baseUrl + "/api/v1/tasks/" + taskId)
                .accept(MediaType.APPLICATION_JSON)
                .retrieve()
                .onStatus(HttpStatus::is4xxClientError, response -> response.bodyToMono(TaskStatusResponse.class)
                .onStatus(HttpStatus::isError, response -> response.bodyToMono(TaskStatusResponse.class)
                        .flatMap(errorBody -> Mono.error(new RuntimeException(errorBody.getDetail()))))
                .bodyToMono(TaskStatusResponse.class)
                .flatMap(result -> {
@@ -140,7 +351,12 @@
                                    return question;
                                }).collect(Collectors.toList());

                        questionService.updateBatchById(updateQuestions);
                        // 包装成响应式操作
                        return Mono.fromCallable(() -> {
                            questionService.updateBatchById(updateQuestions);
                            return result;
                        });

                    }
                    return Mono.just(result);
                })
@@ -166,14 +382,39 @@
                .onStatus(HttpStatus::isError, response -> response.bodyToMono(TaskCancelResponse.class)
                        .flatMap(errorBody -> Mono.error(new RuntimeException(errorBody.getDetail()))))
                .bodyToMono(TaskCancelResponse.class)
                .flatMap(cancelResponse -> {
                    // 更新关键词状态
                    Mono<Void> updateKeyword = Mono.fromRunnable(() -> {
                                LambdaUpdateWrapper<Keyword> updateWrapper = new LambdaUpdateWrapper<>();
                                updateWrapper.eq(Keyword::getTask_id, taskId);
                                updateWrapper.set(Keyword::getStatus, "canceled"); // 统一使用"canceled"
                                keywordService.update(updateWrapper);
                            })
                            .subscribeOn(Schedulers.boundedElastic())
                            .then();

                    // 更新关键词任务状态
                    Mono<Void> updateKeywordTask = Mono.fromRunnable(() -> {
                                LambdaUpdateWrapper<KeywordTask> updateWrapper = new LambdaUpdateWrapper<>();
                                updateWrapper.eq(KeywordTask::getTask_id, taskId);
                                updateWrapper.set(KeywordTask::getStatus, "canceled"); // 统一使用"canceled"
                                keywordTaskService.update(updateWrapper);
                            })
                            .subscribeOn(Schedulers.boundedElastic())
                            .then();

                    // 并行执行两个更新操作，并在完成后返回cancelResponse
                    return Mono.when(updateKeyword, updateKeywordTask)
                            .thenReturn(cancelResponse);
                })
                .map(data -> ResponseResult.success(data))
                .onErrorResume(e -> {
                    if (e.getMessage().contains("任务不存在")) {
                        return Mono.just(ResponseResult.error(404, "任务不存在"));
                        return Mono.just(ResponseResult.error(200, e.getMessage()));
                    } else if (e.getMessage().contains("无法取消")) {
                        return Mono.just(ResponseResult.error(400, "任务已完成，无法取消"));
                        return Mono.just(ResponseResult.error(200, e.getMessage()));
                    }
                    return Mono.just(ResponseResult.error(500, "取消任务失败: " + e.getMessage()));
                    return Mono.just(ResponseResult.error(500,  e.getMessage()));
                });
    }

@@ -280,21 +521,22 @@
    private Mono<Void> updateQuestionAndReference(TaskResultResponse result) {
        return Mono.fromRunnable(() -> {
            try {
                // 1. 更新关键词状态
                LambdaUpdateWrapper<Keyword> keywordUpdate = new LambdaUpdateWrapper<>();
                keywordUpdate.eq(Keyword::getTask_id, result.getTask_id())
                        .set(Keyword::getStatus, "completed");
                keywordService.update(keywordUpdate);

                // 1. 根据KeywordTask更新关键词状态
                // 查询关键词ID
                LambdaQueryWrapper<Keyword> keywordQuery = new LambdaQueryWrapper<>();
                keywordQuery.eq(Keyword::getTask_id, result.getTask_id());
                Keyword keyword = keywordService.getOne(keywordQuery);
                LambdaQueryWrapper<KeywordTask> keywordTaskWrapper = new LambdaQueryWrapper<>();
                keywordTaskWrapper.eq(KeywordTask::getTask_id, result.getTask_id());
                KeywordTask keywordTask = keywordTaskService.getOne(keywordTaskWrapper);

                Keyword keyword = keywordService.getById(keywordTask.getKeyword_id());

                if (keyword == null) {
                    System.out.println("未找到关联的关键词，task_id: " + result.getTask_id());
                    return;
                    //报错
                    throw new Exception("未找到关联的关键词，task_id: " + result.getTask_id());
//                    return;
                }
                keyword.setStatus("completed");
                keywordService.updateById(keyword);

                // 2. 批量查询所有问题
                LambdaQueryWrapper<Question> queryWrapper = new LambdaQueryWrapper<>();
@@ -308,7 +550,7 @@
                // 3. 收集所有需要更新的问题和引用
                List<Question> questionsToUpdate = new ArrayList<>();
                List<Reference> allReferences = new ArrayList<>();

                List<Reference> resultList = new ArrayList<>();
                // 遍历结果
                for (UserResult userResult : result.getResults()) {
                    for (QuestionResult questionResult : userResult.getQuestions_results()) {
@@ -320,6 +562,7 @@
                                question.setResponse(questionResult.getResponse());
                                question.setExtracted_count(questionResult.getExtracted_count());
                                question.setError(questionResult.getError());
                                question.setKeyword_id(keyword.getKeyword_id());

                                // 解析时间戳
                                if (questionResult.getTimestamp() != null) {
@@ -333,7 +576,6 @@

                                questionsToUpdate.add(question);

                                // 收集引用数据，处理空集合情况
                                List<Reference> references =
                                        Optional.ofNullable(questionResult.getReferences())
                                                .orElse(Collections.emptyList())
@@ -344,7 +586,38 @@
                                                    reference.setTitle(ref.getTitle());
                                                    reference.setUrl(ref.getUrl());
                                                    reference.setDomain(ref.getDomain());
                                                    reference.setNum(keyword.getNum());
                                                    reference.setTask_id(result.getTask_id());
                                                    reference.setKeyword_id(keyword.getKeyword_id());
                                                    //域名和平台id映射
                                                    reference.setCreate_time(LocalDateTime.now());
                                                    Platform platform = platformService.getPlatformByDomain(reference.getDomain());
                                                    if (platform == null) {
                                                        //平台为空 创建平台 类型为“默认”
                                                        Type type = typeService.getOne(new LambdaQueryWrapper<Type>().eq(Type::getType_name,"默认"));
                                                        if (type == null) {
                                                            Type newType = new Type();
                                                            newType.setType_name("默认");
                                                            typeService.save(newType);
                                                            type = newType;
                                                        }
                                                        Platform platform1 = new Platform();
                                                        platform1.setDomain(reference.getDomain());
                                                        platform1.setPlatform_name(reference.getDomain());
                                                        platform1.setType_id(type.getType_id());
                                                        platformService.save(platform1);

                                                        reference.setType_id(type.getType_id());
                                                        reference.setPlatform_id(platform1.getPlatform_id());

                                                    }
                                                    else {
                                                        reference.setPlatform_id(platform.getPlatform_id());
                                                        Type type = typeService.getById(platform.getType_id());
                                                        if (type != null){
                                                            reference.setType_id(type.getType_id());
                                                        }
                                                    }
                                                    return reference;
                                                })
                                                .collect(Collectors.toList());
@@ -353,8 +626,56 @@
                                if (!references.isEmpty()) {
                                    allReferences.addAll(references);
                                }

                                //取数据库中当前关键词的当前轮次的当前问题id结果拿出来
                                List<Reference> dbList = referenceService.list(new LambdaQueryWrapper<Reference>().eq(Reference::getKeyword_id, keyword.getKeyword_id())
                                        .eq(Reference::getNum, keyword.getNum())
                                        .eq(Reference::getQuestion_id, question.getQuestion_id())
                                );

                                // 1. 合并两个列表
                                List<Reference> combinedList = new ArrayList<>();
                                combinedList.addAll(allReferences);
                                combinedList.addAll(dbList);

                                // 2. 创建复合键的Map，用于统计完全匹配的记录
                                Map<String, List<Reference>> compositeKeyMap = combinedList.stream()
                                        .collect(Collectors.groupingBy(
                                                ref -> ref.getTitle() + "|" + ref.getUrl() + "|" + ref.getDomain()
                                        ));

                                // 3. 处理每组重复记录

                                compositeKeyMap.forEach((key, refGroup) -> {
                                    // 3.1 找出组内有ID的记录（优先从dbList中获取）
                                    Optional<Reference> existingRecord = refGroup.stream()
                                            .filter(ref -> ref.getReference_id() != null)
                                            .findFirst();

                                    // 3.2 统计该组的重复次数（总数-1）
                                    int repetitionCount = refGroup.size() - 1;

                                    // 3.3 决定最终保留的记录
                                    Reference recordToSave;
                                    if (existingRecord.isPresent()) {
                                        // 使用已有ID的记录并更新重复次数
                                        recordToSave = existingRecord.get();
                                        recordToSave.setRepetition_num(
                                                (recordToSave.getRepetition_num() == null ? 0 : recordToSave.getRepetition_num())
                                                        + repetitionCount
                                        );
                                    } else {
                                        // 没有ID记录则取第一条并设置重复次数
                                        recordToSave = refGroup.get(0);
                                        recordToSave.setRepetition_num(repetitionCount);
                                    }

                                    resultList.add(recordToSave);
                                });
                                referenceService.saveOrUpdateBatch(resultList);
                            }
                        } catch (Exception e) {
                            log.error(e.getMessage(), e);
                            System.out.println("处理问题结果失败: " + e.getMessage());
                        }
                    }
@@ -368,20 +689,22 @@
                }

                // 5. 批量插入引用，使用流式分批处理
                if (!allReferences.isEmpty()) {
                    int batchSize = 1000;
                    IntStream.iterate(0, i -> i + batchSize)
                            .limit((allReferences.size() + batchSize - 1) / batchSize)
                            .forEach(i -> {
                                List<Reference> batch = allReferences.subList(
                                        i, Math.min(i + batchSize, allReferences.size()));
                                referenceService.saveBatch(batch);
                            });
                    System.out.println("成功批量插入 " + allReferences.size() + " 条引用数据");
                }
//                if (!allReferences.isEmpty()) {
//                    int batchSize = 1000;
//                    IntStream.iterate(0, i -> i + batchSize)
//                            .limit((allReferences.size() + batchSize - 1) / batchSize)
//                            .forEach(i -> {
//                                List<Reference> batch = allReferences.subList(
//                                        i, Math.min(i + batchSize, allReferences.size()));
//                                referenceService.saveBatch(batch);
//                            });
//                    System.out.println("成功批量插入 " + allReferences.size() + " 条引用数据");
//                }


            } catch (Exception e) {
                System.out.println("更新问题和引用数据失败: " + e.getMessage());
                log.error("更新问题和引用数据失败: " ,e.getMessage(), e);
//                System.out.println("更新问题和引用数据失败: " + e.getMessage());
                throw new RuntimeException("更新问题和引用数据失败", e);
            }
        });
@@ -406,6 +729,7 @@
    }

    @GetMapping("/health")
    @ApiOperation("健康检查")
    public Mono<HealthResponse> checkThirdPartyHealth() {
        return webClient.get()
                .uri(baseUrl + "/health") // 假设第三方健康检查接口路径为/health
@@ -414,4 +738,23 @@
                .onErrorResume(e -> Mono.just(
                        new HealthResponse("unhealthy", null, "", e.getMessage())));
    }

    /**
     * 查询服务器资源
     */
    @GetMapping("/server/resource")
    @ApiOperation(value = "查询服务器资源")
    public Mono<ServerResourceResponse> getServerResource() {
        return webClient.get()
                .uri(baseUrl + "/api/v1/system/resources")
                .retrieve()
                .bodyToMono(ServerResourceResponse.class)
                .onErrorResume(e -> Mono.just(
                        new ServerResourceResponse( e.getMessage())));
    }
    /**
     * 传入orderid查所有关键词id以及关键词下面的所有任务id，轮询所有任务状态，如果状态为completed，则循环调用获取结果接口，处理结果
     */


}