From a7c9ed23a07e22b525942d989eb8a159f2a4f130 Mon Sep 17 00:00:00 2001
From: guyue <1721849008@qq.com>
Date: 星期三, 30 七月 2025 14:32:48 +0800
Subject: [PATCH] 修改重复次数统计增加num
---
src/main/java/com/linghu/controller/CollectController.java | 1618 ++++++++++++++++++++++++++++++++++++++++++-----------------
1 files changed, 1,144 insertions(+), 474 deletions(-)
diff --git a/src/main/java/com/linghu/controller/CollectController.java b/src/main/java/com/linghu/controller/CollectController.java
index e0fdeeb..0fd9d17 100644
--- a/src/main/java/com/linghu/controller/CollectController.java
+++ b/src/main/java/com/linghu/controller/CollectController.java
@@ -4,6 +4,8 @@
import java.time.LocalDateTime;
import java.time.format.DateTimeFormatter;
import java.util.*;
+import java.util.concurrent.*;
+import java.util.concurrent.locks.ReentrantLock;
import java.util.stream.Collectors;
import javax.servlet.http.HttpServletRequest;
@@ -16,7 +18,10 @@
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.beans.factory.annotation.Value;
import org.springframework.core.ParameterizedTypeReference;
+import org.springframework.dao.DuplicateKeyException;
import org.springframework.http.*;
+import org.springframework.transaction.annotation.Transactional;
+import org.springframework.web.reactive.function.client.ExchangeStrategies;
import org.springframework.web.reactive.function.client.WebClient;
import com.baomidou.mybatisplus.core.conditions.query.LambdaQueryWrapper;
@@ -35,10 +40,6 @@
import com.linghu.model.dto.TaskResultResponse.QuestionResult;
import com.linghu.model.dto.TaskResultResponse.UserResult;
import reactor.core.scheduler.Schedulers;
-
-import java.util.stream.Collectors;
-import java.util.stream.IntStream;
-import java.util.stream.Stream;
@RestController
@RequestMapping("/collect")
@@ -67,129 +68,193 @@
private PlatformService platformService;
@Autowired
private TypeService typeService;
+ @Autowired
+ private UserService userService;
+ @Autowired
+ private OrderService orderService;
+ @Autowired
+ private QuestionResultService questionResultService;
+ // 替换为线程安全队列
+ private static final Queue<SearchTaskRequest> taskQueue = new ConcurrentLinkedQueue<>();
+ // 全局映射：关键词ID -> 批次队列
+ private static final ConcurrentMap<Integer, Queue<List<UserDto>>> batchQueues = new ConcurrentHashMap<>();
- /* @PostMapping("/search")
+// private static boolean isProcessing = false;
+ private static volatile boolean isProcessing = false; // 添加 volatile
+
+ @PostMapping("/search")
@ApiOperation(value = "开始采集")
- public Mono<SearchTaskResponse> createSearchTask(
+ public Mono<ResponseResult<?>> createSearchTask(
@RequestBody SearchTaskRequest searchTaskRequest,
HttpServletRequest request) throws JsonProcessingException {
- return webClient.post()
- .uri(baseUrl + "/api/v1/search")
- .contentType(MediaType.APPLICATION_JSON)
- .bodyValue(searchTaskRequest)
- .retrieve()
- .onStatus(HttpStatus::is4xxClientError, response -> response.bodyToMono(String.class)
- .flatMap(errorBody -> Mono.error(new RuntimeException(errorBody))))
- .bodyToMono(new ParameterizedTypeReference<SearchTaskResponse>() {
- })
- .flatMap(responseResult -> {
- // 提取任务ID
- SearchTaskResponse taskResponse = responseResult;
- if (taskResponse != null && taskResponse.getTask_id() != null) {
- // 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);
+ // 首先检查服务器资源
+ return getServerResource()
+ .flatMap(resourceResponse -> {
+ double cpuUsage = parseUsage(resourceResponse.getCpu_usage_percent());
+ double memoryUsage = parseUsage(resourceResponse.getMemory_usage_percent());
- // 可选：更新响应中的其他信息
+ if (cpuUsage >= 90.0 || memoryUsage >= 90.0) {
+ String errorMsg = String.format("服务器资源不足,请稍后再试");
+
+ log.warn(errorMsg);
+ return Mono.just(ResponseResult.error(503, errorMsg));
}
- return Mono.just(taskResponse);
+
+ // 将新的任务请求加入队列
+ taskQueue.add(searchTaskRequest);
+
+ // 如果当前没有任务在处理中，则启动任务队列的处理
+ if (!isProcessing) {
+ processNextTaskInQueue();
+ }
+
+ // 返回响应，通知用户任务已开始
+ return Mono.just(ResponseResult.success("任务已加入队列，正在处理..."));
})
.onErrorResume(e -> {
- // return Mono.just(ResponseResult.error("调用失败: " + e.getMessage()));
- SearchTaskResponse task = new SearchTaskResponse();
- task.setMessage("调用失败: " + e.getMessage());
- return Mono.just(task);
+ 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()));
});
- }*/
+ }
-// public SearchTaskController(WebClient.Builder webClientBuilder, KeywordService keywordService) {
-// this.webClient = webClientBuilder.build();
-// this.keywordService = keywordService;
-// }
+ private void processNextTaskInQueue() {
+ // 设置为正在处理
+ isProcessing = true;
- /* @PostMapping("/search")
- @ApiOperation(value = "开始采集")
- public Mono<SearchTaskResponse> createSearchTask(
- @RequestBody SearchTaskRequest searchTaskRequest,
- HttpServletRequest request) throws JsonProcessingException {
+ // 从队列中取出下一个任务
+ SearchTaskRequest nextTaskRequest = taskQueue.poll();
+ if (nextTaskRequest != null) {
+ Integer keywordId = nextTaskRequest.getKeyword_id();
+ log.info("开始处理任务队列，keywordId: {}", keywordId);
+
+ executeBatchTask(nextTaskRequest)
+ .doFinally(signal -> {
+ isProcessing = false;
+ if (!taskQueue.isEmpty()) {
+ processNextTaskInQueue();
+ }
+ })
+ .subscribe(
+ result -> log.info("任务处理完成，keywordId: {}", keywordId), // 成功日志
+ error -> { // 关键：添加错误处理
+ log.error("任务队列处理异常，keywordId: {}", keywordId, error);
+ }
+ );
+ } else {
+ isProcessing = false; // 无任务时重置状态
+ }
+ }
+ private Mono<ResponseResult<String>> executeBatchTask(SearchTaskRequest searchTaskRequest) {
+ Integer keywordId = searchTaskRequest.getKeyword_id();
+ //
int maxConcurrentUsers = searchTaskRequest.getConfig() != null ?
searchTaskRequest.getConfig().getMax_concurrent_users() : 3;
- List<List<UserDto>> userBatches = splitUsersIntoBatches(searchTaskRequest.getUsers(), maxConcurrentUsers);
+ List<List<UserDto>> userBatches = splitUsersIntoBatches(searchTaskRequest.getUsers(), maxConcurrentUsers, keywordId,searchTaskRequest.getIs_first());
- return processBatchesSequentially(userBatches, searchTaskRequest)
- .onErrorResume(e -> {
- SearchTaskResponse task = new SearchTaskResponse();
- task.setMessage("调用失败: " + e.getMessage());
- return Mono.just(task);
+
+ // 创建批次队列并存入全局映射
+ Queue<List<UserDto>> batchQueue = new ConcurrentLinkedQueue<>(userBatches);
+ batchQueues.put(keywordId, batchQueue); // 存储到全局映射
+
+ return Mono.just(ResponseResult.success("第一个批次已开始"))
+ .doOnTerminate(() -> {
+ executeBatchTask(batchQueue, searchTaskRequest, keywordId)
+ .subscribe(
+ result -> log.info("批次任务启动成功，keywordId: {}", keywordId),
+ error -> { // 处理批次执行异常
+ log.error("批次任务执行异常，keywordId: {}", keywordId, error);
+ // 可选：异常时清理资源
+ batchQueues.remove(keywordId);
+ }
+ );
});
}
+ private Mono<ResponseResult<?>> executeBatchTask(Queue<List<UserDto>> batchQueue, SearchTaskRequest searchTaskRequest, Integer keywordId) {
+ // 如果队列为空，说明所有批次已经完成
- 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())));
+ if (batchQueue == null || batchQueue.isEmpty()) {
+ // 清理资源
+ batchQueues.remove(keywordId);
+ return Mono.just(ResponseResult.success("所有批次已完成"));
}
- 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());
+ List<UserDto> currentBatch = batchQueue.poll(); // 从队列中获取当前批次
+ SearchTaskRequest batchRequest = new SearchTaskRequest();
+ batchRequest.setUsers(currentBatch);
+ 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);
- 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;
+ return createSingleBatchTask(batchRequest)
+ .flatMap(taskResponse -> {
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);
+ // 直接等待任务完成，不再保存任务关联信息
+ return waitForTaskCompletion(taskResponse.getTask_id(), batchQueue, searchTaskRequest, keywordId);
+ } else {
+ return Mono.just(ResponseResult.error("创建批次任务失败"));
}
- return waitForTaskCompletion(taskResponse.getTask_id())
- .then(Mono.just(taskResponse));
+ })
+ .onErrorResume(e -> {
+ log.error("调用第三方接口失败: {}", e.getMessage(), e); // 关键日志
+ return Mono.error(new RuntimeException("调用第三方接口失败: " + e.getMessage()));
+ })
+ .doFinally(signal -> {
+ // 任务完成时清理资源
+ if (batchQueue.isEmpty()) {
+ batchQueues.remove(keywordId);
+ }
});
}
- 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>
- }*/
+ private Mono<ResponseResult<?>> waitForTaskCompletion(String taskId, Queue<List<UserDto>> batchQueue, SearchTaskRequest searchTaskRequest, Integer keywordId) {
+ // 查询任务状态
+ return getTaskStatus(taskId)
+ .flatMap(statusResponse -> {
+ // 检查任务是否被取消
+ if ("cancelled".equalsIgnoreCase(statusResponse.getStatus())) {
+ batchQueues.remove(keywordId); // 清理资源
+ return Mono.just(ResponseResult.success("任务已被取消"));
+ }
+ // 如果任务状态是"submitted"或"running"，继续轮询
+ if (!"completed".equalsIgnoreCase(statusResponse.getStatus()) && !"failed".equalsIgnoreCase(statusResponse.getStatus()) && !"cancelled".equalsIgnoreCase(statusResponse.getStatus()) && !("ERROR".equalsIgnoreCase(statusResponse.getStatus()) && statusResponse.getMessage().contains("Task not found")) ) {
+ return Mono.delay(Duration.ofSeconds(5)) // 延迟 5 秒后再次查询
+ .flatMap(aLong -> waitForTaskCompletion(taskId, batchQueue, searchTaskRequest, keywordId)); // 递归调用继续等待
+ } else {
+ // 如果状态为其他状态，则继续处理下一个批次
+ return executeBatchTask(batchQueue, searchTaskRequest, keywordId);
+ }
+ })
+ .onErrorResume(e -> {
+ // 处理查询任务状态时的错误
+ return Mono.just(ResponseResult.error("查询任务状态失败: " + e.getMessage()));
+ });
+ }
+ @ApiOperation(value = "查询任务状态")
+ @GetMapping("/status")
+ public Mono<TaskStatusResponse> getTaskStatus(String taskId) {
+ return webClient.get()
+ .uri(baseUrl + "/api/v1/tasks/" + taskId)
+ .accept(MediaType.APPLICATION_JSON)
+ .retrieve()
+ .onStatus(HttpStatus::isError, response -> response.bodyToMono(TaskStatusResponse.class)
+ .flatMap(errorBody -> Mono.error(new RuntimeException(errorBody.getDetail()))))
+ .bodyToMono(TaskStatusResponse.class)
+ .onErrorResume(e -> {
+ // 处理错误，创建一个自定义的错误响应对象
+ TaskStatusResponse errorResponse = new TaskStatusResponse();
+ errorResponse.setStatus("ERROR");
+ errorResponse.setMessage(e.getMessage());
+ errorResponse.setDetail(e.getMessage());
+ return Mono.just(errorResponse);
+ });
+ }
+
// 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 {
@@ -204,105 +269,42 @@
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));
- }
-
- int maxConcurrentUsers = searchTaskRequest.getConfig() != null ?
- searchTaskRequest.getConfig().getMax_concurrent_users() : 3;
- List<List<UserDto>> userBatches = splitUsersIntoBatches(searchTaskRequest.getUsers(), maxConcurrentUsers);
-
- // 获取 keywordId
- Integer keywordId = searchTaskRequest.getKeyword_id();
- //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) {
+ private List<List<UserDto>> splitUsersIntoBatches(List<UserDto> users, int batchSize,Integer keywordId,Boolean isFirst) {
Keyword keyword = keywordService.getById(keywordId);
- if (null==keyword.getNum()){
- keyword.setNum(0);
+ if (isFirst){
+ keyword.setNum(1);
+ }else {
+ keyword.setNum(keyword.getNum()+1);
}
- 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())));
+
}
+ for (int i = 0; i < batches.size(); i++){
+ // 创建 KeywordTask 关联，task_id 设置为 null，表示任务尚未开始
+ KeywordTask keywordTask = new KeywordTask();
+ keywordTask.setKeyword_id(keywordId);
+ keywordTask.setTask_id(null); // 任务ID为空
+
+ keywordTask.setNum(keyword.getNum());
+ keywordTaskService.save(keywordTask); // 保存 KeywordTask
+ }
+
+
return batches;
}
private Mono<SearchTaskResponse> createSingleBatchTask(SearchTaskRequest batchRequest) {
+ // 记录请求第三方的基本信息（便于排查）
+ String thirdPartyUrl = baseUrl + "/api/v1/search";
+ Integer keywordId = batchRequest.getKeyword_id();
+ log.info("开始向第三方提交任务，keywordId: {}, URL: {}, 请求参数: {}",
+ keywordId, thirdPartyUrl, batchRequest.toString()); // 打印请求参数（建议用工具类转JSON）
return webClient.post()
.uri(baseUrl + "/api/v1/search")
.contentType(MediaType.APPLICATION_JSON)
@@ -310,17 +312,61 @@
.retrieve()
.onStatus(HttpStatus::is4xxClientError, response -> response.bodyToMono(String.class)
.flatMap(errorBody -> Mono.error(new RuntimeException(errorBody))))
+ // 处理第三方返回的5xx服务器错误（如第三方服务异常）
+ .onStatus(HttpStatus::is5xxServerError, response ->
+ response.bodyToMono(String.class)
+ .flatMap(errorBody -> {
+ String errorMsg = String.format("第三方接口5xx错误，keywordId: %d, URL: %s, 状态码: %d, 错误详情: %s",
+ keywordId, thirdPartyUrl, response.statusCode().value(), errorBody);
+ log.error(errorMsg);
+ return Mono.error(new RuntimeException(errorMsg));
+ })
+ )
.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(() -> {
+
+ //更新关键词状态
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);
+ //设置轮数
+ Keyword keyword = keywordService.getById(batchRequest.getKeyword_id());
+ // 更新关键词任务与任务ID的关联
+ // 获取与关键词相关的任务，task_id 为 null，确保只取一个任务
+ List<KeywordTask> keywordTasks = keywordTaskService.list(new LambdaQueryWrapper<KeywordTask>()
+ .eq(KeywordTask::getKeyword_id, keyword.getKeyword_id())
+ .eq(KeywordTask::getNum, keyword.getNum())
+ .isNull(KeywordTask::getTask_id));
+ if (keywordTasks.size() > 0) {
+ KeywordTask keywordTask = keywordTasks.get(0);
+ keywordTask.setTask_id(taskResponse.getTask_id());
+ keywordTask.setStatus("pending");
+ keywordTaskService.updateById(keywordTask);
+ }
+ //将提问词列表的状态转为pending
+ for (String questionName : batchRequest.getQuestions()) {
+ questionService.update(new LambdaUpdateWrapper<Question>().eq(Question::getKeyword_id, keyword.getKeyword_id()).eq(Question::getQuestion,questionName).set(Question::getStatus, "pending"));
+
+ }
+ //所有关键词都在采集中或者已完成或者错误设置订单进入采集状态
+ List<Keyword> orderKeywords = keywordService.list(new LambdaQueryWrapper<Keyword>()
+ .eq(Keyword::getOrder_id, keyword.getOrder_id()));
+ if (!orderKeywords.isEmpty() && orderKeywords.stream().allMatch(k ->
+ "completed".equals(k.getStatus()) || "submitted".equals(k.getStatus())
+ )) {
+ Orders orders = orderService.getById(keyword.getOrder_id());
+ if (orders != null) {
+ orders.setStatus(2);
+ orderService.updateById(orders);
+
+ }
+ }
}).subscribeOn(Schedulers.boundedElastic()) // 在弹性线程池执行
.thenReturn(taskResponse);
}
@@ -328,74 +374,193 @@
});
}
- // 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) {
- return webClient.get()
- .uri(baseUrl + "/api/v1/tasks/" + taskId)
- .accept(MediaType.APPLICATION_JSON)
- .retrieve()
- .onStatus(HttpStatus::isError, response -> response.bodyToMono(TaskStatusResponse.class)
- .flatMap(errorBody -> Mono.error(new RuntimeException(errorBody.getDetail()))))
- .bodyToMono(TaskStatusResponse.class)
- .flatMap(result -> {
- TaskStatusResponse taskStatusResponse = result;
- if (taskStatusResponse != null && taskStatusResponse.getStatus() != null) {
- List<Question> updateQuestions = taskStatusResponse.getQuestions_status().stream()
- .map(qs -> {
- Question question = new Question();
- question.setQuestion_id(qs.getQuestion_id());
- question.setStatus(qs.getStatus());
- return question;
- }).collect(Collectors.toList());
+ @PostMapping("/cancel/{keywordId}")
+ @ApiOperation(value = "取消任务")
+ public Mono<ResponseResult<TaskCancelResponse>> cancelTask(@PathVariable Integer keywordId) {
+ // 1. 从主队列移除任务
+ List<SearchTaskRequest> removedMainQueueTasks = removeTasksFromQueueByKeywordId(keywordId);
+ int removedMainQueueCount = removedMainQueueTasks.size(); // 获取移除的任务数量
- // 包装成响应式操作
- return Mono.fromCallable(() -> {
- questionService.updateBatchById(updateQuestions);
- return result;
- });
+ // 2. 从批次队列移除任务 (新增逻辑)
+ int removedBatchQueue = removeBatchTasksByKeywordId(keywordId);
- }
- return Mono.just(result);
- })
- .onErrorResume(e -> {
- // 创建一个自定义的错误响应对象
- TaskStatusResponse errorResponse = new TaskStatusResponse();
- errorResponse.setStatus("ERROR");
- errorResponse.setMessage(e.getMessage());
- errorResponse.setDetail(e.getMessage());
+ // 3. 查询所有与关键词相关的任务
+ List<KeywordTask> tasks = keywordTaskService.list(
+ new LambdaQueryWrapper<KeywordTask>().eq(KeywordTask::getKeyword_id, keywordId)
+ );
- return Mono.just(errorResponse);
+ // 4. 筛选出需要远程取消的任务
+ List<KeywordTask> tasksToCancelRemotely = tasks.stream()
+ .filter(task -> task.getTask_id() != null && "pending".equalsIgnoreCase(task.getStatus()))
+ .collect(Collectors.toList());
+
+ return Flux.fromIterable(tasksToCancelRemotely)
+ .flatMap(task -> {
+ // 创建状态更新和远程取消的组合操作
+ Mono<Void> updateStatus = updateTaskStatus(task.getTask_id(), "cancelled");
+ Mono<ResponseResult<?>> cancelOp = cancelRemoteTask(task.getTask_id())
+ .onErrorResume(e -> {
+ log.error("取消任务 {} 失败: {}", task.getTask_id(), e.getMessage());
+ return Mono.just(ResponseResult.error("取消任务失败: " + e.getMessage()));
+ });
+
+ return Mono.zip(cancelOp, updateStatus)
+ .thenReturn(true);
+ }, 10)
+ .collectList()
+ .flatMap(canceledTasks -> {
+ return updateKeywordAndOrderStatus(keywordId)
+ .thenReturn(ResponseResult.success(
+ new TaskCancelResponse(
+ String.format("任务已取消: 主队列移除%d, 批次队列移除%d, 远程取消%d",
+ removedMainQueueCount ,
+ removedBatchQueue,
+ tasksToCancelRemotely.size())
+ )
+ ));
});
}
- @PostMapping("/cancel/{taskId}")
- @ApiOperation(value = "取消任务")
- public Mono<ResponseResult<TaskCancelResponse>> cancelTask(@PathVariable String taskId) {
+ // 新增方法：移除批次队列
+ private int removeBatchTasksByKeywordId(Integer keywordId) {
+ Queue<List<UserDto>> batchQueue = batchQueues.remove(keywordId);
+ if (batchQueue != null) {
+ int count = batchQueue.size();
+ batchQueue.clear();
+ log.info("从批次队列中移除关键词 {} 的 {} 个批次任务", keywordId, count);
+ return count;
+ }
+ return 0;
+ }
+ // 辅助方法：获取待取消任务
+ private List<KeywordTask> getTasksToCancel(Integer keywordId) {
+ return keywordTaskService.list(
+ new LambdaQueryWrapper<KeywordTask>()
+ .eq(KeywordTask::getKeyword_id, keywordId)
+ .isNotNull(KeywordTask::getTask_id)
+ .eq(KeywordTask::getStatus, "pending")
+ );
+ }
+
+ // 提取关键词和订单状态更新的逻辑为单独方法
+ private Mono<Void> updateKeywordAndOrderStatus(Integer keywordId) {
+ return Mono.fromRunnable(() -> {
+ try {
+ // 查询关键词
+ Keyword keyword = keywordService.getById(keywordId);
+ if (keyword == null) {
+ log.warn("未找到关键词，keywordId: {}", keywordId);
+ return;
+ }
+
+ //把任务id为空的删除
+ LambdaUpdateWrapper<KeywordTask> updateWrapper = new LambdaUpdateWrapper<>();
+ updateWrapper.eq(KeywordTask::getKeyword_id, keywordId);
+ updateWrapper.isNull(KeywordTask::getTask_id);
+ keywordTaskService.remove(updateWrapper);
+ // 查询该关键词下的所有任务
+ LambdaQueryWrapper<KeywordTask> keywordTaskWrapper = new LambdaQueryWrapper<>();
+ keywordTaskWrapper.eq(KeywordTask::getKeyword_id, keywordId);
+
+ List<KeywordTask> keywordTasks = keywordTaskService.list(keywordTaskWrapper);
+
+ // 更新关键词状态
+ keyword.setStatus("completed");
+ keywordService.updateById(keyword);
+
+ //更新提问词状态为取消
+ questionService.update(new LambdaUpdateWrapper<Question>().eq(Question::getKeyword_id, keywordId).isNull(Question::getResponse).set(Question::getStatus, "cancelled").set(Question::getError, "任务已取消"));
+// 将所有提问词设置eroor为已取消
+// questionService.update(new LambdaUpdateWrapper<Question>().eq(Question::getKeyword_id, keywordId).set(Question::getStatus, "failed").set(Question::getError, "任务已取消"));
+
+ // 更新订单状态
+ String orderId = keyword.getOrder_id();
+ if (orderId != null && !orderId.isEmpty()) {
+ // 查询订单下所有关键词
+ LambdaQueryWrapper<Keyword> orderKeywordsWrapper = new LambdaQueryWrapper<>();
+ orderKeywordsWrapper.eq(Keyword::getOrder_id, orderId);
+ List<Keyword> orderKeywords = keywordService.list(orderKeywordsWrapper);
+
+ // 所有关键词均已完成，则更新订单状态为3
+ if (!orderKeywords.isEmpty() && orderKeywords.stream().allMatch(k ->
+ "completed".equals(k.getStatus()) || "false".equals(k.getStatus())
+ )) {
+ Orders orders = orderService.getById(orderId);
+ if (orders != null) {
+ orders.setStatus(3);
+ orderService.updateById(orders);
+ log.info("订单 {} 所有关键词已完成，更新状态为3", orderId);
+ }
+ }
+ if (!orderKeywords.isEmpty() && orderKeywords.stream().allMatch(k ->
+ !"completed".equals(k.getStatus()) || "false".equals(k.getStatus()) || "cancelled".equals(k.getStatus())
+ )) {
+ Orders orders = orderService.getById(orderId);
+ if (orders != null) {
+ orders.setStatus(1);
+ orderService.updateById(orders);
+ log.info("订单 {} 所有关键词已完成或者取消，更新状态为1", orderId);
+ }
+ }
+ }
+ } catch (Exception e) {
+ log.error("更新关键词和订单状态失败: {}", e.getMessage(), e);
+ }
+ });
+ }
+private List<SearchTaskRequest> removeTasksFromQueueByKeywordId(Integer keywordId) {
+ List<SearchTaskRequest> removedTasks = new ArrayList<>();
+
+ Iterator<SearchTaskRequest> iterator = taskQueue.iterator();
+ while (iterator.hasNext()) {
+ SearchTaskRequest task = iterator.next();
+ if (task.getKeyword_id() != null && task.getKeyword_id().equals(keywordId)) {
+ removedTasks.add(task);
+ iterator.remove();
+ }
+ }
+
+
+ log.info("从队列中移除了 {} 个与关键词ID {} 相关的任务", removedTasks.size(), keywordId);
+ return removedTasks;
+}
+
+// 发送远程取消请求
+ private Mono<ResponseResult<?>> cancelRemoteTask(String taskId) {
+ // 使用Collections.singletonMap或手动创建Map
+ Map<String, Object> requestBody = new HashMap<>();
+ requestBody.put("status", "pending");
+
return webClient.post()
.uri(baseUrl + "/api/v1/tasks/" + taskId + "/cancel")
.contentType(MediaType.APPLICATION_JSON)
- .bodyValue(Collections.emptyMap()) // 添加空请求体
+ .bodyValue(requestBody)
.retrieve()
- .onStatus(HttpStatus::isError, response -> response.bodyToMono(TaskCancelResponse.class)
- .flatMap(errorBody -> Mono.error(new RuntimeException(errorBody.getDetail()))))
- .bodyToMono(TaskCancelResponse.class)
- .map(data -> ResponseResult.success(data))
- .onErrorResume(e -> {
- if (e.getMessage().contains("任务不存在")) {
- return Mono.just(ResponseResult.error(200, "任务不存在"));
- } else if (e.getMessage().contains("无法取消")) {
- return Mono.just(ResponseResult.error(200, "任务已完成，无法取消"));
- }
- return Mono.just(ResponseResult.error(500, "取消任务失败: " + e.getMessage()));
- });
+ .onStatus(HttpStatus::isError, response -> response.bodyToMono(String.class)
+ .flatMap(errorBody -> Mono.error(new RuntimeException("取消失败: " + errorBody))))
+ .bodyToMono(Void.class)
+ .thenReturn(ResponseResult.success("任务已取消"));
}
+ // 更新单个任务状态
+ private Mono<Void> updateTaskStatus(String taskId, String status) {
+ return Mono.fromRunnable(() -> {
+ LambdaUpdateWrapper<KeywordTask> updateWrapper = new LambdaUpdateWrapper<>();
+ updateWrapper.eq(KeywordTask::getTask_id, taskId);
+ updateWrapper.set(KeywordTask::getStatus, status);
+ keywordTaskService.update(updateWrapper);
+ }).subscribeOn(Schedulers.boundedElastic()).then();
+ }
@ApiOperation(value = "获取任务结果")
@GetMapping("/tasks/{taskId}")
public Mono<TaskResultResponse> getTaskResult(@PathVariable String taskId) {
- return webClient.get()
+ WebClient webClient2 = WebClient.builder()
+ .exchangeStrategies(ExchangeStrategies.builder()
+ .codecs(configurer -> configurer.defaultCodecs()
+ .maxInMemorySize(10 * 1024 * 1024)) // 10MB
+ .build())
+ .build();
+ return webClient2.get()
.uri(baseUrl + "/api/v1/tasks/" + taskId + "/result")
.accept(MediaType.APPLICATION_JSON)
.retrieve()
@@ -412,6 +577,7 @@
.bodyToMono(TaskResultResponse.class)
.flatMap(responseResult -> {
TaskResultResponse result = responseResult;
+
if (result != null && result.getResults() != null) {
return updateQuestionAndReference(result)
.thenReturn(responseResult);
@@ -420,268 +586,776 @@
})
.onErrorResume(e -> {
System.out.println("获取任务结果失败");
- TaskResultResponse result = new TaskResultResponse();
+ log.error("获取任务结果失败: {}", e.getMessage(), e);
+ TaskResultResponse result = new TaskResultResponse();
result.setDetail("获取任务结果失败: " + e.getMessage());
return Mono.just(result);
});
}
+ /**
+ * 获取或创建平台（确保同一domain只创建一次）
+ * @param domain 平台域名
+ * @return 已存在或新创建的Platform
+ */
+ private Platform getOrCreatePlatform(String domain,String platformName) {
+ // 1. 先尝试查询已存在的平台
+
+ Platform platform = platformService.getPlatformByDomain(domain);
+ if (platform != null) {
+ return platform;
+ }
+
+ // 2. 若不存在，尝试创建（处理并发场景）
+ try {
+ // 2.1 获取或创建“默认”类型（Type也需避免重复，建议Type表的type_name也加唯一约束）
+ Type defaultType = typeService.getOne(new LambdaQueryWrapper<Type>()
+ .eq(Type::getType_name, "默认"));
+ if (defaultType == null) {
+ defaultType = new Type();
+ defaultType.setType_name("默认");
+ typeService.save(defaultType); // 若Type可能重复，此处也需处理DuplicateKeyException
+ }
+
+ // 2.2 构建新平台对象
+ Platform newPlatform = new Platform();
+ newPlatform.setDomain(domain);
+ if (platformName != null) {
+ newPlatform.setPlatform_name(platformName);
+ }else {
+ newPlatform.setPlatform_name(domain);
+ }
+ // 平台名称默认使用域名，可根据实际需求调整
+ newPlatform.setType_id(defaultType.getType_id());
+ newPlatform.setCreate_time(LocalDateTime.now()); // 补充创建时间
+
+ // 2.3 尝试保存，若因唯一约束冲突失败，则捕获异常
+ platformService.save(newPlatform);
+ return newPlatform; // 保存成功，返回新创建的平台
+
+ } catch (DuplicateKeyException e) {
+ // 3. 若捕获到重复键异常，说明并发创建了，重新查询即可（此时数据库中已存在该平台）
+ log.warn("平台domain={}已存在，无需重复创建", domain, e);
+ return platformService.getPlatformByDomain(domain); // 重新查询，一定能获取到
+ } catch (Exception e) {
+ // 处理其他异常（如数据库连接失败等）
+ log.error("创建平台失败，domain={}", domain, e);
+ throw new RuntimeException("创建平台失败", e);
+ }
+ }
+ //更新提问词和引用数据
// private Mono<Void> updateQuestionAndReference(TaskResultResponse result) {
// return Mono.fromRunnable(() -> {
-// // 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);
+// try {
+// //查看每个账号信息的status是否正常
//
-// // 查询关键词ID
-// LambdaQueryWrapper<Keyword> keywordQuery = new LambdaQueryWrapper<>();
-// keywordQuery.eq(Keyword::getTask_id, result.getTask_id());
-// Keyword keyword = keywordService.getOne(keywordQuery);
+// // 1. 根据KeywordTask更新关键词状态
+// // 查询关键词ID
+// LambdaQueryWrapper<KeywordTask> keywordTaskWrapper = new LambdaQueryWrapper<>();
+// keywordTaskWrapper.eq(KeywordTask::getTask_id, result.getTask_id());
+// KeywordTask keywordTask = keywordTaskService.getOne(keywordTaskWrapper);
+// keywordTask.setStatus("completed");
+// keywordTaskService.updateById(keywordTask);
+// 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;
-// }
+// if (keyword == null) {
+// System.out.println("未找到关联的关键词，task_id: " + result.getTask_id());
+// //报错
+// throw new Exception("未找到关联的关键词，task_id: " + result.getTask_id());
//
-// // 2. 处理每个用户的问题结果
-// for (UserResult userResult : result.getResults()) {
-// for (QuestionResult questionResult : userResult.getQuestions_results()) {
-// // 2.1 查询问题ID
-// LambdaQueryWrapper<Question> queryWrapper = new LambdaQueryWrapper<>();
-// queryWrapper.eq(Question::getQuestion, questionResult.getQuestion())
-// .eq(Question::getKeyword_id, keyword.getKeyword_id());
-// Question question = questionService.getOne(queryWrapper);
+// }
+// LambdaQueryWrapper<KeywordTask> keywordTaskWrapper2 = new LambdaQueryWrapper<>();
+// keywordTaskWrapper2.eq(KeywordTask::getKeyword_id, keyword.getKeyword_id());
+// List<KeywordTask> keywordTasks = keywordTaskService.list(keywordTaskWrapper2);
//
-// if (question != null) {
-// // 更新问题状态
-// LambdaUpdateWrapper<Question> updateWrapper = new LambdaUpdateWrapper<>();
-// updateWrapper.eq(Question::getQuestion_id, question.getQuestion_id())
-// .set(Question::getStatus, questionResult.getStatus())
-// .set(Question::getResponse, questionResult.getResponse())
-// .set(Question::getExtracted_count, questionResult.getExtracted_count())
-// .set(Question::getError, questionResult.getError())
-// .set(Question::getTimestamp, LocalDateTime.parse(
-// questionResult.getTimestamp(),
-// DateTimeFormatter.ofPattern("yyyy-MM-dd'T'HH:mm:ss.SSSSSS")
-// ));
-// questionService.update(updateWrapper);
+// //如果全部为completed 或者错误、取消、任务不存在 关键词也为completed ，如果关联关系没有任务id，或者状态为running ，关键词为submitted，
+// if (keywordTasks.stream().allMatch(task -> "completed".equals(task.getStatus()) || "false".equals(task.getStatus()) || "cancelled".equals(task.getStatus()) ||"canceled".equals(task.getStatus()) || "nonentity".equals(task.getStatus())) ) {
+// keyword.setStatus("completed");
+// keywordService.updateById(keyword);
//
-// // 2.2 保存引用数据
-// List<Reference> references = questionResult.getReferences().stream()
-// .map(ref -> {
+// }
+//
+// String orderId = keyword.getOrder_id();
+// if (orderId == null || orderId.isEmpty()) {
+// System.out.println("关键词[" + keyword.getKeyword_id() + "]未关联订单，跳过订单状态更新");
+// return;
+// }
+//
+// // 2.更新订单状态为待处理 查询该订单下的所有关键词，更新订单状态（有取消）
+// LambdaQueryWrapper<Keyword> orderKeywordsWrapper = new LambdaQueryWrapper<>();
+// orderKeywordsWrapper.eq(Keyword::getOrder_id, orderId);
+// List<Keyword> orderKeywords = keywordService.list(orderKeywordsWrapper);
+//
+// if (orderKeywords.isEmpty()) {
+// System.out.println("订单[" + orderId + "]下无关键词，跳过状态更新");
+// return;
+// }
+// boolean allValid2 = orderKeywords.stream()
+// .allMatch(k -> "completed".equals(k.getStatus()) || "false".equals(k.getStatus()) || "cancelled".equals(k.getStatus()));
+// if (allValid2) {
+// Orders orders = orderService.getById(orderId);
+// if (orders != null) {
+// orders.setStatus(1); // 假设Orders有Integer类型的status字段
+// orderService.updateById(orders);
+// System.out.println("订单[" + orderId + "]所有关键词采集完成或者取消，已更新状态为1");
+// } else {
+// System.out.println("未找到订单[" + orderId + "]，无法更新状态");
+// }
+// }
+// // 3.更新订单状态为完成 检查所有关键词的状态是否均为 completed 或 false
+// boolean allValid = orderKeywords.stream()
+// .allMatch(k -> "completed".equals(k.getStatus()) || "false".equals(k.getStatus()));
+//
+// // 4. 若所有关键词状态均有效，更新订单状态为3
+// if (allValid) {
+// Orders orders = orderService.getById(orderId);
+// if (orders != null) {
+// orders.setStatus(3); // 假设Orders有Integer类型的status字段
+// orderService.updateById(orders);
+// System.out.println("订单[" + orderId + "]所有关键词状态符合条件，已更新状态为3");
+// } else {
+// System.out.println("未找到订单[" + orderId + "]，无法更新状态");
+// }
+// }
+//
+//
+// Orders orders = orderService.getById(keyword.getOrder_id());
+//
+// // 2. 批量查询所有问题
+// LambdaQueryWrapper<Question> queryWrapper = new LambdaQueryWrapper<>();
+// queryWrapper.eq(Question::getKeyword_id, keyword.getKeyword_id());
+// List<Question> questions = questionService.list(queryWrapper);
+//
+// // 构建问题映射表，用于快速查找
+// Map<String, Question> questionMap = questions.stream()
+// .collect(Collectors.toMap(Question::getQuestion, q -> q));
+//
+// // 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<>();
+//
+// // 遍历账号
+// for (UserResult userResult : result.getResults()) {
+// //更新账号状态
+// if ( "failed".equals(userResult.getStatus())){
+// if (userResult.getError().contains("登录失败")){
+// LambdaUpdateWrapper<User> userWrapper = new LambdaUpdateWrapper<>();
+// userWrapper.eq(User::getUser_email, userResult.getUser_email());
+// userWrapper.set(User::getStatus, "无法登录");
+// userService.update(userWrapper);
+// //更新所有提问词的状态
+// questionService.update(new LambdaUpdateWrapper<Question>().eq(Question::getKeyword_id, keyword.getKeyword_id())
+// .set(Question::getStatus, "failed")
+// .set(Question::getError, "账户登录失败"));
+//
+// }else if (userResult.getError().contains("信息错误")){
+// LambdaUpdateWrapper<User> userWrapper = new LambdaUpdateWrapper<>();
+// userWrapper.eq(User::getUser_email, userResult.getUser_email());
+// userWrapper.set(User::getStatus, "信息错误");
+// userService.update(userWrapper);
+// }
+// }
+// for (QuestionResult questionResult : userResult.getQuestions_results()) {
+// try {
+// Question question = questionMap.get(questionResult.getQuestion());
+// if (question != null) {
+//
+//
+// //保存问题结果
+// QuestionResultList questionResultList = new QuestionResultList();
+// questionResultList.setKeyword_id(keyword.getKeyword_id());
+// questionResultList.setQuestion(questionResult.getQuestion());
+// questionResultList.setResponse(questionResult.getResponse());
+// questionResultList.setStatus(questionResult.getStatus());
+// questionResultList.setExtracted_count(questionResult.getExtracted_count());
+// questionResultList.setKeyword_task_id(result.getTask_id());
+// questionResultList.setError(questionResult.getError());
+// questionResultList.setNum(keyword.getNum());
+// if (questionResult.getTimestamp() != null) {
+// DateTimeFormatter formatter = DateTimeFormatter
+// .ofPattern("yyyy-MM-dd'T'HH:mm:ss.SSSSSS");
+// questionResultList.setTimestamp(
+// LocalDateTime.parse(questionResult.getTimestamp(), formatter));
+// }
+// // 保存问题结果列表（新增保存逻辑）
+// questionResultService.save(questionResultList);
+// // 查询当前轮次下该提问词的所有结果
+// List<QuestionResultList> allResults = questionResultService.list(
+// new LambdaQueryWrapper<QuestionResultList>()
+// .eq(QuestionResultList::getKeyword_id, keyword.getKeyword_id())
+// .eq(QuestionResultList::getQuestion, question.getQuestion())
+// .eq(QuestionResultList::getNum, keyword.getNum())
+// );
+//
+// // 判断最终状态
+// String finalStatus = determineFinalStatus(allResults);
+// if ("success".equals(finalStatus)){
+// question.setStatus("success");
+// question.setError("");
+// }else if ("no_results".equals(finalStatus)){
+// question.setStatus("success");
+// question.setError("采集结果无引用数据");
+// }else if ("busyness".equals(finalStatus)){
+// question.setStatus("failed");
+// question.setError("DeepSeek繁忙，请稍后尝试");
+// }
+//
+// // 更新问题对象
+// question.setResponse(questionResult.getResponse());
+// question.setExtracted_count(questionResult.getExtracted_count());
+//// question.setError(questionResult.getError());
+// question.setKeyword_id(keyword.getKeyword_id());
+//
+// if (questionResult.getTimestamp() != null) {
+// DateTimeFormatter formatter = DateTimeFormatter
+// .ofPattern("yyyy-MM-dd'T'HH:mm:ss.SSSSSS");
+// question.setTimestamp(
+// LocalDateTime.parse(questionResult.getTimestamp(), formatter));
+// }
+//
+// questionsToUpdate.add(question);
+// // 初始化引用列表（避免null）
+// List<Reference> references = new ArrayList<>();
+// List<TaskResultResponse.Reference> originalReferences = questionResult.getReferences();
+// if (originalReferences == null) {
+// originalReferences = Collections.emptyList();
+// }
+//
+// // 遍历原始引用列表，转换为Reference对象
+// for (TaskResultResponse.Reference ref : originalReferences) {
// Reference reference = new Reference();
+// // 设置基本字段
// reference.setQuestion_id(question.getQuestion_id());
// reference.setTitle(ref.getTitle());
// reference.setUrl(ref.getUrl());
// reference.setDomain(ref.getDomain());
-// reference.setCreate_time(LocalDateTime.now());
-// return reference;
-// })
-// .collect(Collectors.toList());
+// reference.setNum(keyword.getNum());
+// reference.setTask_id(result.getTask_id());
+// reference.setKeyword_id(keyword.getKeyword_id());
+// if (null!=ref.getPublish_time()) {
+// reference.setCreate_time(ref.getPublish_time().atStartOfDay());
+// }
//
-// if (!references.isEmpty()) {
-// referenceService.saveBatch(references);
+// // 关键：使用优化后的方法获取平台，避免重复创建
+// Platform platform = getOrCreatePlatform(ref.getDomain(),ref.getPlatform_name());
+// reference.setPlatform_id(platform.getPlatform_id());
+// reference.setType_id(platform.getType_id()); // 直接从平台获取类型ID，更可靠
+// // 添加到结果列表
+// references.add(reference);
+// }
+// // 添加到总引用列表
+// 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 = new Reference();
+// 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 ? 1 : 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(1+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());
// }
-// } else {
-// System.out.println("未找到匹配的问题，question " + question.getQuestion());
-//
// }
+//
+//
// }
+//
+// // 4. 批量更新问题
+// System.out.println(questionsToUpdate);
+// if (!questionsToUpdate.isEmpty()) {
+// questionService.updateBatchById(questionsToUpdate);
+// System.out.println("成功批量更新 " + questionsToUpdate.size() + " 个问题");
+// }
+//
+// } catch (Exception e) {
+// log.error("更新问题和引用数据失败: " ,e.getMessage(), e);
+// throw new RuntimeException("更新问题和引用数据失败", e);
// }
// });
// }
+// // 根据所有批次的结果判断最终状态
+// private String determineFinalStatus(List<QuestionResultList> results) {
+// if (results.isEmpty()) {
+// return "no_results"; // 无结果
+// }
+//
+// // 统计关键指标
+// int totalCount = results.size();
+// int emptyResponseCount = 0;
+// int systemBusyCount = 0;
+//
+// for (QuestionResultList result : results) {
+// // 判断回答是否为空
+// if (result.getExtracted_count() == 0 ) {
+// emptyResponseCount++;
+// }
+//
+// // 判断是否为系统繁忙
+// if ("success".equals(result.getStatus()) && (result.getResponse().isEmpty()|| result.getResponse().contains("WebDriver连接中断") || result.getResponse().contains("响应超时"))) {
+// systemBusyCount++;
+// }
+//
+// }
+//
+// // 全返回系统繁忙
+// if (systemBusyCount == totalCount) {
+// return "busyness";
+// }
+// // 全返回信息为空
+// if (emptyResponseCount == totalCount) {
+// return "no_results";
+// }
+//
+//
+// // 系统繁忙比例超过阈值（可配置，这里设为70%）
+//// double busyRate = (double) systemBusyCount / totalCount;
+//// if (busyRate >= 0.7) {
+//// return "系统繁忙，请稍后尝试";
+//// }
+//
+// // 其他情况返回成功
+// return "success";
+// }
private Mono<Void> updateQuestionAndReference(TaskResultResponse result) {
- return Mono.fromRunnable(() -> {
- try {
- // 1. 根据KeywordTask更新关键词状态
- // 查询关键词ID
- LambdaQueryWrapper<KeywordTask> keywordTaskWrapper = new LambdaQueryWrapper<>();
- keywordTaskWrapper.eq(KeywordTask::getTask_id, result.getTask_id());
- KeywordTask keywordTask = keywordTaskService.getOne(keywordTaskWrapper);
+ return Mono.fromRunnable(() -> doUpdateQuestionAndReference(result))
+ .onErrorResume(e -> {
+ log.error("处理任务结果失败", e);
+ return Mono.error(e); // 传播异常，触发事务回滚
+ }).then();
+ }
- 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());
- //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<>();
- queryWrapper.eq(Question::getKeyword_id, keyword.getKeyword_id());
- List<Question> questions = questionService.list(queryWrapper);
-
- // 构建问题映射表，用于快速查找
- Map<String, Question> questionMap = questions.stream()
- .collect(Collectors.toMap(Question::getQuestion, q -> q));
-
- // 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()) {
- try {
- Question question = questionMap.get(questionResult.getQuestion());
- if (question != null) {
- // 更新问题对象
- question.setStatus(questionResult.getStatus());
- 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) {
- DateTimeFormatter formatter = DateTimeFormatter
- .ofPattern("yyyy-MM-dd'T'HH:mm:ss.SSSSSS");
- question.setTimestamp(
- LocalDateTime.parse(questionResult.getTimestamp(), formatter));
- }
- //更新
-// questionService.updateById(question);
-
- questionsToUpdate.add(question);
-
- List<Reference> references =
- Optional.ofNullable(questionResult.getReferences())
- .orElse(Collections.emptyList())
- .stream()
- .map(ref -> {
- Reference reference = new Reference();
- reference.setQuestion_id(question.getQuestion_id());
- 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());
-
- // 添加到总引用列表
- 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());
- }
- }
- }
-
- // 4. 批量更新问题
- System.out.println(questionsToUpdate);
- if (!questionsToUpdate.isEmpty()) {
- questionService.updateBatchById(questionsToUpdate);
- System.out.println("成功批量更新 " + questionsToUpdate.size() + " 个问题");
- }
-
- // 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");
-// }
-
-
- } catch (Exception e) {
- 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);
+ // 核心业务逻辑，添加事务注解保证原子性
+ @Transactional(rollbackFor = Exception.class)
+ public void doUpdateQuestionAndReference(TaskResultResponse result) {
+ try {
+ // 1. 查询关键词任务并更新状态
+ LambdaQueryWrapper<KeywordTask> keywordTaskWrapper = new LambdaQueryWrapper<>();
+ keywordTaskWrapper.eq(KeywordTask::getTask_id, result.getTask_id());
+ KeywordTask keywordTask = keywordTaskService.getOne(keywordTaskWrapper);
+ if (keywordTask == null) {
+ throw new Exception("未找到关键词任务，task_id: " + result.getTask_id());
}
+ keywordTask.setStatus("completed");
+ keywordTaskService.updateById(keywordTask);
+
+ // 2. 查询关键词信息
+ Keyword keyword = keywordService.getById(keywordTask.getKeyword_id());
+ if (keyword == null) {
+ throw new Exception("未找到关联的关键词，task_id: " + result.getTask_id());
+ }
+
+ // 3. 更新关键词状态（基于关联任务状态）
+ updateKeywordStatus(keyword);
+
+ // 4. 更新订单状态（基于关键词状态）
+ updateOrderStatus(keyword);
+
+ // 5. 预查询问题列表（一次查询，内存映射）
+ LambdaQueryWrapper<Question> questionWrapper = new LambdaQueryWrapper<>();
+ questionWrapper.eq(Question::getKeyword_id, keyword.getKeyword_id());
+ List<Question> questions = questionService.list(questionWrapper);
+ Map<String, Question> questionMap = questions.stream()
+ .collect(Collectors.toMap(Question::getQuestion, q -> q));
+
+ // 6. 收集批量保存的数据（避免循环内保存）
+ List<QuestionResultList> questionResultsToSave = new ArrayList<>(); // 批量保存问题结果
+ List<Reference> allReferences = new ArrayList<>(); // 收集所有引用，后续统一处理
+
+ // 7. 遍历结果处理问题和引用
+ for (UserResult userResult : result.getResults()) {
+ // 7.1 更新用户状态（失败处理）
+ handleUserStatus(userResult, keyword);
+
+ // 7.2 处理问题结果
+ for (QuestionResult questionResult : userResult.getQuestions_results()) {
+ Question question = questionMap.get(questionResult.getQuestion());
+ if (question == null) {
+ log.warn("未找到问题记录: {}", questionResult.getQuestion());
+ continue;
+ }
+
+ // 7.2.1 构建问题结果并加入批量列表
+ QuestionResultList questionResultList = buildQuestionResultList(questionResult, keyword, result);
+ questionResultsToSave.add(questionResultList);
+
+ // 7.2.2 处理引用数据（仅收集，不立即保存）
+ List<Reference> references = buildReferences(questionResult, question, keyword, result);
+ allReferences.addAll(references);
+ }
+ }
+
+ // 8. 批量保存问题结果（一次数据库交互）
+ if (!questionResultsToSave.isEmpty()) {
+ questionResultService.saveBatch(questionResultsToSave);
+ log.info("批量保存问题结果 {} 条", questionResultsToSave.size());
+ }
+
+ // 9. 批量处理引用数据（去重+更新重复次数+批量保存）
+ if (!allReferences.isEmpty()) {
+ handleReferencesInBatch(allReferences, keyword);
+ }
+
+ // 10. 批量更新问题状态（基于最终结果）
+ updateQuestionsStatusInBatch(questions, keyword);
+
+ } catch (Exception e) {
+ log.error("更新数据失败", e);
+ throw new RuntimeException("更新数据失败", e); // 触发事务回滚
+ }
+ }
+
+ // 更新关键词状态
+ private void updateKeywordStatus(Keyword keyword) {
+ LambdaQueryWrapper<KeywordTask> taskWrapper = new LambdaQueryWrapper<>();
+ taskWrapper.eq(KeywordTask::getKeyword_id, keyword.getKeyword_id());
+ List<KeywordTask> keywordTasks = keywordTaskService.list(taskWrapper);
+
+ boolean allCompletedOrFailed = keywordTasks.stream()
+ .allMatch(task -> "completed".equals(task.getStatus())
+ || "false".equals(task.getStatus())
+ || "cancelled".equals(task.getStatus())
+ || "canceled".equals(task.getStatus())
+ || "nonentity".equals(task.getStatus()));
+
+ if (allCompletedOrFailed) {
+ keyword.setStatus("completed");
+ keywordService.updateById(keyword);
+ }
+ }
+
+ // 更新订单状态
+ private void updateOrderStatus(Keyword keyword) {
+ String orderId = keyword.getOrder_id();
+ if (orderId == null || orderId.isEmpty()) {
+ log.info("关键词[{}]未关联订单，跳过订单更新", keyword.getKeyword_id());
+ return;
+ }
+
+ LambdaQueryWrapper<Keyword> orderKeywordsWrapper = new LambdaQueryWrapper<>();
+ orderKeywordsWrapper.eq(Keyword::getOrder_id, orderId);
+ List<Keyword> orderKeywords = keywordService.list(orderKeywordsWrapper);
+ if (orderKeywords.isEmpty()) {
+ log.info("订单[{}]无关键词，跳过状态更新", orderId);
+ return;
+ }
+
+ boolean allValid2 = orderKeywords.stream()
+ .allMatch(k -> "completed".equals(k.getStatus())
+ || "false".equals(k.getStatus())
+ || "cancelled".equals(k.getStatus()));
+ if (allValid2) {
+ updateOrderStatus(orderId, 1, "所有关键词采集完成或取消");
+ }
+
+ boolean allValid = orderKeywords.stream()
+ .allMatch(k -> "completed".equals(k.getStatus()) || "false".equals(k.getStatus()));
+ if (allValid) {
+ updateOrderStatus(orderId, 3, "所有关键词状态符合条件");
+ }
+ }
+
+ // 封装订单状态更新
+ private void updateOrderStatus(String orderId, Integer status, String logMsg) {
+ Orders orders = orderService.getById(orderId);
+ if (orders != null) {
+ orders.setStatus(status);
+ orderService.updateById(orders);
+ log.info("订单[{}]{}，已更新状态为{}", orderId, logMsg, status);
+ } else {
+ log.warn("未找到订单[{}]，无法更新状态", orderId);
+ }
+ }
+
+ // 构建问题结果对象
+ private QuestionResultList buildQuestionResultList(QuestionResult questionResult, Keyword keyword, TaskResultResponse result) {
+ QuestionResultList questionResultList = new QuestionResultList();
+ questionResultList.setKeyword_id(keyword.getKeyword_id());
+ questionResultList.setQuestion(questionResult.getQuestion());
+ questionResultList.setResponse(questionResult.getResponse());
+ questionResultList.setStatus(questionResult.getStatus());
+ questionResultList.setExtracted_count(questionResult.getExtracted_count());
+ questionResultList.setKeyword_task_id(result.getTask_id());
+ questionResultList.setError(questionResult.getError());
+ questionResultList.setNum(keyword.getNum());
+
+ if (questionResult.getTimestamp() != null) {
+ DateTimeFormatter formatter = DateTimeFormatter.ofPattern("yyyy-MM-dd'T'HH:mm:ss.SSSSSS");
+ questionResultList.setTimestamp(LocalDateTime.parse(questionResult.getTimestamp(), formatter));
+ }
+ return questionResultList;
+ }
+
+ // 构建引用数据列表
+ private List<Reference> buildReferences(QuestionResult questionResult, Question question, Keyword keyword, TaskResultResponse result) {
+ List<Reference> references = new ArrayList<>();
+ List<TaskResultResponse.Reference> originalReferences = questionResult.getReferences();
+ if (originalReferences == null) {
+ return references;
+ }
+
+ for (TaskResultResponse.Reference ref : originalReferences) {
+ // 过滤无效引用（标题、URL、域名不能为空）
+ if (ref.getTitle() == null || ref.getUrl() == null || ref.getDomain() == null) {
+ log.warn("引用数据字段缺失，跳过：title={}, url={}, domain={}",
+ ref.getTitle(), ref.getUrl(), ref.getDomain());
+ continue;
+ }
+
+ Reference reference = new Reference();
+ reference.setQuestion_id(question.getQuestion_id());
+ 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());
+
+ if (ref.getPublish_time() != null) {
+ reference.setCreate_time(ref.getPublish_time().atStartOfDay());
+ }
+
+ Platform platform = getOrCreatePlatform(ref.getDomain(), ref.getPlatform_name());
+ // 校验平台信息非空
+ if (platform == null || platform.getPlatform_id() == null) {
+ log.warn("平台信息无效，跳过引用：domain={}", ref.getDomain());
+ continue;
+ }
+ reference.setPlatform_id(platform.getPlatform_id());
+ reference.setType_id(platform.getType_id());
+ references.add(reference);
+ }
+ return references;
+ }
+
+ // 批量处理引用数据（去重+更新重复次数）
+ private void handleReferencesInBatch(List<Reference> allReferences, Keyword keyword) {
+ // 1. 过滤原始列表中的 null 元素
+ List<Reference> validReferences = allReferences.stream()
+ .filter(Objects::nonNull)
+ .collect(Collectors.toList());
+
+ if (validReferences.isEmpty()) {
+ log.info("无有效引用数据，跳过批量保存");
+ return;
+ }
+
+ // 2. 查询数据库中已存在的引用并过滤 null
+ LambdaQueryWrapper<Reference> dbRefWrapper = new LambdaQueryWrapper<>();
+ dbRefWrapper.eq(Reference::getKeyword_id, keyword.getKeyword_id())
+ .eq(Reference::getNum, keyword.getNum());
+ List<Reference> dbReferences = referenceService.list(dbRefWrapper);
+ List<Reference> validDbReferences = dbReferences.stream()
+ .filter(Objects::nonNull)
+ .collect(Collectors.toList());
+
+ // 3. 合并并去重
+ Map<String, List<Reference>> compositeKeyMap = new HashMap<>();
+ validReferences.forEach(ref -> addToCompositeMap(compositeKeyMap, ref));
+ validDbReferences.forEach(ref -> addToCompositeMap(compositeKeyMap, ref));
+
+ // 4. 处理重复次数
+ List<Reference> referencesToSave = new ArrayList<>();
+ compositeKeyMap.forEach((key, refGroup) -> {
+ List<Reference> validRefGroup = refGroup.stream()
+ .filter(Objects::nonNull)
+ .collect(Collectors.toList());
+ if (validRefGroup.isEmpty()) return;
+
+ Optional<Reference> existingRef = validRefGroup.stream()
+ .filter(ref -> ref.getReference_id() != null)
+ .findFirst();
+
+ Reference finalRef = existingRef.orElse(validRefGroup.get(0));
+// 重复次数 = 老的 + （这一组里一共抓到多少－1）
+ int repetitionNum = (finalRef.getRepetition_num() == null ? 1 : finalRef.getRepetition_num())
+ + (validRefGroup.size() - 1);
+ finalRef.setRepetition_num(repetitionNum);
+ referencesToSave.add(finalRef);
});
+
+ // 5. 最终校验并保存
+ List<Reference> finalSaveList = referencesToSave.stream()
+ .filter(Objects::nonNull)
+ .collect(Collectors.toList());
+
+ // 关键校验：列表非空且元素有效
+ if (finalSaveList.isEmpty()) {
+ log.info("处理后无有效引用数据可保存");
+ return;
+ }
+ if (finalSaveList.stream().anyMatch(ref -> !(ref instanceof Reference))) {
+ log.error("引用数据类型异常，无法保存");
+ return;
+ }
+
+ // 执行保存
+ try {
+ referenceService.saveOrUpdateBatch(finalSaveList);
+ log.info("批量保存引用数据成功，数量：{}", finalSaveList.size());
+ } catch (Exception e) {
+ log.error("批量保存引用数据失败", e);
+ throw new RuntimeException("保存引用数据失败", e);
+ }
+ }
+
+ // 辅助方法：将引用添加到复合键Map
+ private void addToCompositeMap(Map<String, List<Reference>> map, Reference ref) {
+ // 再次校验引用的核心字段非空
+ if (ref.getTitle() == null || ref.getUrl() == null || ref.getDomain() == null) {
+ log.warn("引用核心字段为空，跳过映射：{}", ref);
+ return;
+ }
+ String key = ref.getQuestion_id() + "|" + ref.getTitle() + "|" + ref.getUrl() + "|" + ref.getDomain() + "|" + ref.getNum();
+ map.computeIfAbsent(key, k -> new ArrayList<>()).add(ref);
+ }
+
+ // 批量更新问题状态
+ private void updateQuestionsStatusInBatch(List<Question> questions, Keyword keyword) {
+ if (questions.isEmpty()) {
+ return;
+ }
+
+ // 一次性查询所有问题结果（基于关键词+轮次）
+ LambdaQueryWrapper<QuestionResultList> resultWrapper = new LambdaQueryWrapper<>();
+ resultWrapper.eq(QuestionResultList::getKeyword_id, keyword.getKeyword_id())
+ .eq(QuestionResultList::getNum, keyword.getNum());
+ List<QuestionResultList> allQuestionResults = questionResultService.list(resultWrapper);
+
+ // 按问题分组，便于查询
+ Map<String, List<QuestionResultList>> questionResultsMap = allQuestionResults.stream()
+ .collect(Collectors.groupingBy(QuestionResultList::getQuestion));
+
+ // 批量更新问题状态
+ List<Question> questionsToUpdate = new ArrayList<>();
+ questions.forEach(question -> {
+ List<QuestionResultList> results = questionResultsMap.getOrDefault(question.getQuestion(), Collections.emptyList());
+ String finalStatus = determineFinalStatus(results);
+
+ if ("success".equals(finalStatus)) {
+ question.setStatus("success");
+ question.setError("");
+ } else if ("no_results".equals(finalStatus)) {
+ question.setStatus("success");
+ question.setError("采集结果无引用数据");
+ } else if ("busyness".equals(finalStatus)) {
+ question.setStatus("failed");
+ question.setError("DeepSeek繁忙，请稍后尝试");
+ }
+ // 更新其他字段（响应、时间戳等）
+ results.stream().findFirst().ifPresent(result -> {
+ question.setResponse(result.getResponse());
+ question.setExtracted_count(result.getExtracted_count());
+ question.setTimestamp(result.getTimestamp());
+ });
+ questionsToUpdate.add(question);
+ });
+
+ if (!questionsToUpdate.isEmpty()) {
+ questionService.updateBatchById(questionsToUpdate);
+ log.info("批量更新问题状态 {} 条", questionsToUpdate.size());
+ }
+ }
+
+ // 处理用户状态异常
+ private void handleUserStatus(UserResult userResult, Keyword keyword) {
+ if ("failed".equals(userResult.getStatus())) {
+ LambdaUpdateWrapper<User> userWrapper = new LambdaUpdateWrapper<>();
+ userWrapper.eq(User::getUser_email, userResult.getUser_email());
+
+ if (userResult.getError().contains("登录失败")) {
+ userWrapper.set(User::getStatus, "无法登录");
+ userService.update(userWrapper);
+ // 批量更新问题状态为失败
+ questionService.update(new LambdaUpdateWrapper<Question>()
+ .eq(Question::getKeyword_id, keyword.getKeyword_id())
+ .set(Question::getStatus, "failed")
+ .set(Question::getError, "账户登录失败"));
+ } else if (userResult.getError().contains("信息错误")) {
+ userWrapper.set(User::getStatus, "信息错误");
+ userService.update(userWrapper);
+ // 批量更新问题状态为失败
+ questionService.update(new LambdaUpdateWrapper<Question>()
+ .eq(Question::getKeyword_id, keyword.getKeyword_id())
+ .set(Question::getStatus, "failed")
+ .set(Question::getError, "账户信息错误"));
+ }
+ }
+ }
+
+ // 原方法：判断最终状态（复用）
+ private String determineFinalStatus(List<QuestionResultList> results) {
+ if (results.isEmpty()) {
+ return "no_results";
+ }
+
+ int totalCount = results.size();
+ int emptyResponseCount = 0;
+ int systemBusyCount = 0;
+
+ for (QuestionResultList result : results) {
+ if (result.getExtracted_count() == 0) {
+ emptyResponseCount++;
+ }
+ if ("success".equals(result.getStatus()) &&
+ (result.getResponse() == null || result.getResponse().isEmpty()
+ || result.getResponse().contains("WebDriver连接中断")
+ || result.getResponse().contains("响应超时"))) {
+ systemBusyCount++;
+ }
+ }
+
+ if (systemBusyCount == totalCount) {
+ return "busyness";
+ }
+ if (emptyResponseCount == totalCount) {
+ return "no_results";
+ }
+ return "success";
}
@GetMapping("/tasks/all")
@@ -697,8 +1371,6 @@
TaskListResponse response = new TaskListResponse();
response.setDetail("获取任务列表失败: " + e.getMessage());
return Mono.just(response);
-
- // return Mono.just(ResponseResult.error("获取任务列表失败: " + e.getMessage()));
});
}
@@ -726,9 +1398,7 @@
.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
- */
+
}
--
Gitblit v1.7.1