From d8143b9121bbe941f116230eaa5524ab2cc12a66 Mon Sep 17 00:00:00 2001
From: huliguo <2023611923@qq.com>
Date: 星期四, 10 七月 2025 21:20:21 +0800
Subject: [PATCH] 新增
---
src/main/java/com/linghu/controller/CollectController.java | 140 +++++++++++++++++++++++++++++++++-------------
1 files changed, 100 insertions(+), 40 deletions(-)
diff --git a/src/main/java/com/linghu/controller/CollectController.java b/src/main/java/com/linghu/controller/CollectController.java
index 09acdce..bbd9f8e 100644
--- a/src/main/java/com/linghu/controller/CollectController.java
+++ b/src/main/java/com/linghu/controller/CollectController.java
@@ -47,6 +47,7 @@
import java.util.stream.Collectors;
import java.util.stream.IntStream;
+import java.util.stream.Stream;
@RestController
@RequestMapping("/collect")
@@ -206,10 +207,11 @@
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
+ List<List<UserDto>> userBatches = splitUsersIntoBatches(searchTaskRequest.getUsers(), maxConcurrentUsers,keywordId);
return Flux.fromIterable(userBatches)
.flatMap(batch -> {
@@ -226,6 +228,7 @@
}, 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
);
@@ -238,6 +241,7 @@
KeywordTask keywordTask = new KeywordTask();
keywordTask.setKeyword_id(keywordId);
keywordTask.setTask_id(response.getTask_id());
+ keywordTask.setStatus("pending");
return keywordTask;
})
.collect(Collectors.toList());
@@ -253,7 +257,15 @@
.then();
}
- private List<List<UserDto>> splitUsersIntoBatches(List<UserDto> users, int batchSize) {
+ 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())));
@@ -272,11 +284,12 @@
.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::getStatus, "submitted");
updateWrapper.set(Keyword::getTask_id, taskResponse.getTask_id());
keywordService.update(updateWrapper);
}).subscribeOn(Schedulers.boundedElastic()) // 在弹性线程池执行
@@ -342,9 +355,9 @@
.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, "任务不存在"));
} 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, "任务已完成，无法取消"));
}
return Mono.just(ResponseResult.error(500, "取消任务失败: " + e.getMessage()));
});
@@ -453,20 +466,12 @@
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<KeywordTask> keywordTaskWrapper = new LambdaQueryWrapper<>();
keywordTaskWrapper.eq(KeywordTask::getTask_id, result.getTask_id());
KeywordTask keywordTask = keywordTaskService.getOne(keywordTaskWrapper);
-// LambdaQueryWrapper<Keyword> keywordQuery = new LambdaQueryWrapper<>();
-// keywordQuery.eq(Keyword::getTask_id, keywordTask.getTask_id());
Keyword keyword = keywordService.getById(keywordTask.getKeyword_id());
if (keyword == null) {
@@ -475,6 +480,8 @@
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<>();
@@ -488,7 +495,7 @@
// 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()) {
@@ -514,17 +521,6 @@
questionsToUpdate.add(question);
- //如果查询结果不为空查询num
- Integer maxNumByKeywordId = referenceService.getMaxNumByKeywordId(keyword.getKeyword_id());
- if (maxNumByKeywordId != null){
- maxNumByKeywordId++;
- }else {
- maxNumByKeywordId = 1;
- }
-
-
- // 收集引用数据，处理空集合情况
- Integer finalMaxNumByKeywordId = maxNumByKeywordId;
List<Reference> references =
Optional.ofNullable(questionResult.getReferences())
.orElse(Collections.emptyList())
@@ -535,28 +531,38 @@
reference.setTitle(ref.getTitle());
reference.setUrl(ref.getUrl());
reference.setDomain(ref.getDomain());
- reference.setNum(finalMaxNumByKeywordId);
+ 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) {
-// throw new RuntimeException("未找到对应的平台: " + reference.getDomain());
-// }
- if (platform != null){
+ 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) {
-// throw new RuntimeException("未找到对应的类型: " + reference.getDomain());
-// }
if (type != null){
reference.setType_id(type.getType_id());
}
}
-
-
- // 根据 domain 查询类型
-
-
return reference;
})
.collect(Collectors.toList());
@@ -565,6 +571,53 @@
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);
@@ -579,7 +632,7 @@
questionService.updateBatchById(questionsToUpdate);
System.out.println("成功批量更新 " + questionsToUpdate.size() + " 个问题");
}
- referenceService.saveBatch(allReferences);
+
// 5. 批量插入引用，使用流式分批处理
// if (!allReferences.isEmpty()) {
// int batchSize = 1000;
@@ -621,6 +674,7 @@
}
@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
@@ -634,6 +688,7 @@
* 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")
@@ -642,4 +697,9 @@
.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