历史题库导入

src/main/java/com/project/classicpaper/domain/service/ImportClassicPaperDomainService.java

@@ -0,0 +1,23 @@
+package com.project.classicpaper.domain.service;
+
+import com.project.classicpaper.domain.dto.ClassicPaperDTO;
+import org.springframework.web.multipart.MultipartFile;
+
+/**
+ * 历史套卷导入服务
+ * 支持导入旧系统的 Word 历史套卷直接入库
+ */
+public interface ImportClassicPaperDomainService {
+
+    /**
+     * 从 Word 文件导入经典套卷
+     * 解析 Word → 提取题目（含图片） → 创建套卷关联
+     * 导入的题目 sourceType = Historical_Import(2)，无知识点关联
+     *
+     * @param file      Word 文件
+     * @param subLineId 子产品线ID
+     * @param paperName 套卷名称
+     * @return 套卷详情
+     */
+    ClassicPaperDTO importFromWord(MultipartFile file, Long subLineId, String paperName) throws Exception;
+}

src/main/java/com/project/classicpaper/domain/service/impl/ImportClassicPaperDomainServiceImpl.java

@@ -0,0 +1,123 @@
+package com.project.classicpaper.domain.service.impl;
+
+import com.project.classicpaper.domain.dto.ClassicPaperDTO;
+import com.project.classicpaper.domain.entity.ClassicPaperEntity;
+import com.project.classicpaper.domain.entity.ClassicPaperQuestionEntity;
+import com.project.classicpaper.domain.enums.ClassicPaperSourceTypeEnum;
+import com.project.classicpaper.domain.enums.ClassicPaperStatusEnum;
+import com.project.classicpaper.domain.service.ClassicPaperBaseService;
+import com.project.classicpaper.domain.service.ClassicPaperQuestionBaseService;
+import com.project.classicpaper.domain.service.ImportClassicPaperDomainService;
+import com.project.classicpaper.domain.service.parser.WordQuestionParser;
+import com.project.information.utils.MinIoUtils;
+import com.project.question.domain.dto.QuestionDTO;
+import com.project.question.domain.enums.QuestionSourceTypeEnum;
+import com.project.question.domain.service.SaveQuestionDomainService;
+import lombok.extern.slf4j.Slf4j;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.stereotype.Service;
+import org.springframework.transaction.annotation.Transactional;
+import org.springframework.web.multipart.MultipartFile;
+
+import java.util.ArrayList;
+import java.util.List;
+import java.util.Map;
+import java.util.TreeMap;
+
+/**
+ * 历史套卷导入服务实现
+ * 解析 Word → 提取题目（含图片上传 MinIO） → 创建套卷关联
+ * 导入的题目 sourceType = Historical_Import(2)，无知识点关联
+ */
+@Service
+@Slf4j
+public class ImportClassicPaperDomainServiceImpl implements ImportClassicPaperDomainService {
+
+    @Autowired
+    private ClassicPaperBaseService classicPaperBaseService;
+
+    @Autowired
+    private ClassicPaperQuestionBaseService classicPaperQuestionBaseService;
+
+    @Autowired
+    private SaveQuestionDomainService saveQuestionDomainService;
+
+    @Autowired
+    private MinIoUtils minIoUtils;
+
+    @Override
+    @Transactional(rollbackFor = Exception.class)
+    public ClassicPaperDTO importFromWord(MultipartFile file, Long subLineId, String paperName) throws Exception {
+        log.info(">>> [历史导入] 开始导入套卷, fileName={}, subLineId={}", file.getOriginalFilename(), subLineId);
+
+        // 1. 解析 Word 文件，提取题目列表（含图片上传 MinIO）
+        WordQuestionParser parser = new WordQuestionParser(minIoUtils);
+        WordQuestionParser.ParseResult parseResult = parser.parse(file.getInputStream());
+        List<WordQuestionParser.ParsedQuestion> parsedQuestions = parseResult.getQuestions();
+
+        if (parsedQuestions.isEmpty()) {
+            log.warn(">>> [历史导入] Word 文件中未提取到题目");
+        }
+
+        // 2. 创建套卷
+        ClassicPaperEntity paper = new ClassicPaperEntity();
+        paper.setName(paperName != null ? paperName : parseResult.getTitle());
+        paper.setSubLineId(subLineId);
+        paper.setSourceType(ClassicPaperSourceTypeEnum.WORD_IMPORT.getValue());
+        paper.setStatus(ClassicPaperStatusEnum.DRAFT.getValue());
+        paper.setQuestionCount(parsedQuestions.size());
+        classicPaperBaseService.save(paper);
+
+        // 3. 保存题目并创建关联
+        List<ClassicPaperQuestionEntity> paperQuestions = new ArrayList<>();
+        for (int i = 0; i < parsedQuestions.size(); i++) {
+            WordQuestionParser.ParsedQuestion parsed = parsedQuestions.get(i);
+
+            // 构建选项 Map
+            Map<String, String> optionsMap = new TreeMap<>();
+            if (parsed.getOptions() != null) {
+                for (WordQuestionParser.ParsedOption opt : parsed.getOptions()) {
+                    optionsMap.put(opt.getLabel(), opt.getContent());
+                }
+            }
+
+            // 创建题目（sourceType = Historical_Import，无知识点关联）
+            QuestionDTO questionDTO = new QuestionDTO();
+            questionDTO.setSourceType(QuestionSourceTypeEnum.Historical_Import.getValue());
+            questionDTO.setQuestionType(parsed.getQuestionType());
+            questionDTO.setKpIdList(new ArrayList<>()); // 无知识点关联
+
+            QuestionDTO.QuestionDetailDTO detailDTO = new QuestionDTO.QuestionDetailDTO();
+            detailDTO.setQuestionContent(parsed.getStem());
+            detailDTO.setType(parsed.getQuestionType());
+            detailDTO.setOptions(optionsMap.isEmpty() ? null : optionsMap);
+            detailDTO.setRightAnswer(parsed.getAnswer());
+            detailDTO.setAnalysis(parsed.getAnalysis());
+            questionDTO.setQuestionDetailDTO(detailDTO);
+
+            // 保存题目
+            saveQuestionDomainService.save(questionDTO);
+
+            // 创建套卷-题目关联
+            ClassicPaperQuestionEntity paperQuestion = new ClassicPaperQuestionEntity();
+            paperQuestion.setPaperId(paper.getId());
+            paperQuestion.setQuestionId(questionDTO.getId());
+            paperQuestion.setSortOrder(i + 1);
+            paperQuestion.setQuestionType(parsed.getQuestionType());
+            paperQuestions.add(paperQuestion);
+        }
+
+        classicPaperQuestionBaseService.saveBatch(paperQuestions);
+
+        log.info(">>> [历史导入] 导入完成, paperId={}, 题目数={}", paper.getId(), paperQuestions.size());
+
+        // 4. 返回套卷详情
+        ClassicPaperDTO dto = new ClassicPaperDTO();
+        dto.setId(paper.getId());
+        dto.setName(paper.getName());
+        dto.setSourceType(paper.getSourceType());
+        dto.setStatus(paper.getStatus());
+        dto.setQuestionCount(paper.getQuestionCount());
+        return dto;
+    }
+}

src/main/java/com/project/classicpaper/domain/service/parser/WordQuestionParser.java

@@ -0,0 +1,236 @@
+package com.project.classicpaper.domain.service.parser;
+
+import com.project.information.utils.MinIoUtils;
+import cn.hutool.core.date.DateUtil;
+import cn.hutool.core.util.IdUtil;
+import lombok.Data;
+import lombok.extern.slf4j.Slf4j;
+import org.apache.poi.xwpf.usermodel.*;
+
+import java.io.InputStream;
+import java.util.*;
+import java.util.regex.Matcher;
+import java.util.regex.Pattern;
+
+/**
+ * Word 文档题目解析器
+ * 逐段落状态机解析，提取题目、选项、答案、解析、图片
+ * 图片上传 MinIO，占位符替换为 HTML img 标签
+ */
+@Slf4j
+public class WordQuestionParser {
+
+    private static final Pattern QUESTION_PATTERN = Pattern.compile("^\\d+[.．\\u3001]");
+    private static final Pattern OPTION_PATTERN = Pattern.compile("^[A-Z][.．\\u3001:：、]\\s*");
+    private static final Pattern ANSWER_PATTERN = Pattern.compile("^答案[：:\\u3001]\\s*(.*)");
+    private static final Pattern ANALYSIS_PATTERN = Pattern.compile("^解析[：:\\u3001]\\s*(.*)");
+
+    private final MinIoUtils minIoUtils;
+
+    private String title;
+    private final List<ParsedQuestion> questions = new ArrayList<>();
+    private ParsedQuestion currentQuestion;
+    private int imageCounter = 0;
+
+    public WordQuestionParser(MinIoUtils minIoUtils) {
+        this.minIoUtils = minIoUtils;
+    }
+
+    /**
+     * 解析 Word 文档
+     * @param inputStream Word 文件流
+     * @return 解析结果
+     */
+    public ParseResult parse(InputStream inputStream) throws Exception {
+        try (XWPFDocument document = new XWPFDocument(inputStream)) {
+            for (XWPFParagraph paragraph : document.getParagraphs()) {
+                processParagraph(paragraph);
+            }
+            // 保存最后一道题
+            if (currentQuestion != null) {
+                validateAndSave(currentQuestion);
+            }
+        }
+
+        ParseResult result = new ParseResult();
+        result.setTitle(title);
+        result.setQuestions(questions);
+        return result;
+    }
+
+    private void processParagraph(XWPFParagraph paragraph) {
+        // 提取图片并上传 MinIO，返回替换后的文本
+        String text = extractImagesAndUpload(paragraph);
+
+        if (text == null || text.isBlank()) {
+            return;
+        }
+
+        text = text.trim();
+
+        // 第一个非空段落作为标题
+        if (title == null) {
+            title = text;
+            return;
+        }
+
+        // 题号匹配
+        if (QUESTION_PATTERN.matcher(text).find()) {
+            if (currentQuestion != null) {
+                validateAndSave(currentQuestion);
+            }
+            currentQuestion = new ParsedQuestion();
+            // 去掉题号
+            currentQuestion.stem = QUESTION_PATTERN.matcher(text).replaceFirst("").trim();
+            currentQuestion.questionType = detectQuestionType(currentQuestion.stem);
+            return;
+        }
+
+        if (currentQuestion == null) {
+            return;
+        }
+
+        // 答案行
+        Matcher answerMatcher = ANSWER_PATTERN.matcher(text);
+        if (answerMatcher.matches()) {
+            currentQuestion.answer = answerMatcher.group(1).trim();
+            if (currentQuestion.answer.length() > 1 && currentQuestion.questionType == 0) {
+                currentQuestion.questionType = 1; // 升级为多选
+            }
+            return;
+        }
+
+        // 解析行
+        Matcher analysisMatcher = ANALYSIS_PATTERN.matcher(text);
+        if (analysisMatcher.matches()) {
+            currentQuestion.analysis = analysisMatcher.group(1).trim();
+            return;
+        }
+
+        // 选项行
+        Matcher optionMatcher = OPTION_PATTERN.matcher(text);
+        if (optionMatcher.find()) {
+            String label = text.substring(0, 1);
+            String content = optionMatcher.replaceFirst("").trim();
+            currentQuestion.options.add(new ParsedOption(label, content));
+            return;
+        }
+
+        // 独立图片段落：合并到题干或最后一个选项
+        if (text.startsWith("[IMAGE:") && currentQuestion.answer == null) {
+            if (currentQuestion.options.isEmpty()) {
+                currentQuestion.stem += text;
+            } else {
+                ParsedOption lastOpt = currentQuestion.options.get(currentQuestion.options.size() - 1);
+                lastOpt.content += text;
+            }
+            return;
+        }
+
+        // 其他文本：追加到题干
+        if (currentQuestion.answer == null) {
+            currentQuestion.stem += "\n" + text;
+        }
+    }
+
+    /**
+     * 提取段落中的图片，上传 MinIO，替换为 HTML img 标签
+     */
+    private String extractImagesAndUpload(XWPFParagraph paragraph) {
+        StringBuilder sb = new StringBuilder();
+
+        for (XWPFRun run : paragraph.getRuns()) {
+            String runText = run.text();
+            if (runText != null) {
+                sb.append(sanitizeText(runText));
+            }
+
+            // 提取嵌入图片
+            List<XWPFPicture> pictures = run.getEmbeddedPictures();
+            for (XWPFPicture picture : pictures) {
+                try {
+                    imageCounter++;
+                    // MinIO 路径规范：classicpaper/{yyyyMMdd}/{uuid}.png（不依赖 subLineId）
+                    String fileName = String.format("classicpaper/%s/%s.png",
+                            DateUtil.format(new java.util.Date(), "yyyyMMdd"),
+                            IdUtil.fastSimpleUUID());
+                    byte[] imageData = picture.getPictureData().getData();
+
+                    // 上传到 MinIO
+                    minIoUtils.uploadFile(new java.io.ByteArrayInputStream(imageData), fileName);
+                    String imageUrl = minIoUtils.getPreviewUrl(fileName);
+
+                    // 替换为 HTML img 标签
+                    sb.append(String.format("<img src='%s'/>", imageUrl));
+                    log.debug(">>> [Word解析] 图片已上传 MinIO: {}", fileName);
+                } catch (Exception e) {
+                    log.error(">>> [Word解析] 图片上传失败", e);
+                    sb.append("[IMAGE:上传失败]");
+                }
+            }
+        }
+
+        return sb.toString().trim();
+    }
+
+    private String sanitizeText(String text) {
+        if (text == null) return "";
+        return text
+                .replace('、', ' ')  // 顿号
+                .replace(' ', ' ')  // 不间断空格
+                .replace('​', ' ')  // 零宽空格
+                .replace('', ' ')  // BOM
+                .replaceAll("\\s+", " ")
+                .trim();
+    }
+
+    /**
+     * 检测题型
+     */
+    private int detectQuestionType(String stem) {
+        if (stem.contains("[判断题]")) return 2;  // TRUE_FALSE
+        if (stem.contains("[简答题]")) return 4;  // SHORT_ANSWER
+        if (stem.contains("[多选题]")) return 1;  // MULTIPLE_CHOICE
+        if (stem.contains("[单选题]")) return 0;  // SINGLE_CHOICE
+        return 0; // 默认单选
+    }
+
+    private void validateAndSave(ParsedQuestion q) {
+        if (q.stem == null || q.stem.isBlank()) {
+            log.warn(">>> [Word解析] 题目内容为空，跳过");
+            return;
+        }
+        // 清理题型标签
+        q.stem = q.stem.replaceAll("\\[.*?题\\]", "").trim();
+        questions.add(q);
+    }
+
+    // =============== 内部数据结构 ===============
+
+    @Data
+    public static class ParseResult {
+        private String title;
+        private List<ParsedQuestion> questions = new ArrayList<>();
+    }
+
+    @Data
+    public static class ParsedQuestion {
+        /** 题型：0-单选，1-多选，2-判断，4-简答 */
+        int questionType = 0;
+        String stem = "";
+        List<ParsedOption> options = new ArrayList<>();
+        String answer;
+        String analysis;
+    }
+
+    @Data
+    public static class ParsedOption {
+        String label;
+        String content;
+
+        public ParsedOption(String label, String content) {
+            this.label = label;
+            this.content = content;
+        }
+    }
+}

src/main/java/com/project/information/application/impl/GenerateFromFilesApplicationServiceImpl.java

@@ -7,6 +7,7 @@ import com.project.information.application.GenerateFromFilesApplicationService;
 import com.project.information.domain.dto.UploadedFileDTO;
 import com.project.information.domain.entity.InformationEntity;
 import com.project.information.domain.entity.InformationFileEntity;
+import com.project.information.domain.enums.ContentCategoryEnum;
 import com.project.information.domain.enums.DraftStatusEnum;
 import com.project.information.domain.enums.FileParseStatusEnum;
 import com.project.information.domain.enums.MaterialTypeEnum;
@@ -18,6 +19,7 @@ import lombok.extern.slf4j.Slf4j;
 import org.springframework.beans.factory.annotation.Autowired;
 import org.springframework.stereotype.Service;
 import org.springframework.transaction.annotation.Transactional;
+import org.springframework.util.StringUtils;
 
 import java.util.ArrayList;
 import java.util.List;
@@ -49,10 +51,32 @@ public class GenerateFromFilesApplicationServiceImpl implements GenerateFromFile
         if (param.getSubLineId() == null) {
             throw new BusinessErrorException("子产品线不能为空");
         }
-        if (param.getFiles() == null || param.getFiles().isEmpty()) {
+        if (!StringUtils.hasText(param.getFileName())) {
+            throw new BusinessErrorException("文件名不能为空");
+        }
+
+        // 解析 files JSON 字符串
+        List<UploadedFileDTO> files = param.parseFiles();
+        if (files.isEmpty()) {
             throw new BusinessErrorException("请先上传文件");
         }
 
+        // 校验内容分类是否合法
+        for (int i = 0; i < files.size(); i++) {
+            if (!ContentCategoryEnum.isValid(files.get(i).getContentCategory())) {
+                throw new BusinessErrorException("第" + (i + 1) + "个文件的内容分类不合法");
+            }
+        }
+
+        // 检查同子产品线下是否有重复文件名
+        boolean exists = informationBaseService.lambdaQuery()
+                .eq(InformationEntity::getSubLineId, param.getSubLineId())
+                .eq(InformationEntity::getName, param.getFileName())
+                .exists();
+        if (exists) {
+            throw new BusinessErrorException("该子产品线下已存在同名资料「" + param.getFileName() + "」，请修改文件名");
+        }
+
         // 1. 创建虚拟 InformationEntity（materialType=原始资料, draftStatus=PENDING）
         InformationEntity virtualInfo = new InformationEntity();
         virtualInfo.setSubLineId(param.getSubLineId());
@@ -68,7 +92,7 @@ public class GenerateFromFilesApplicationServiceImpl implements GenerateFromFile
 
         // 2. 创建子文件记录并关联虚拟资料
         List<InformationFileEntity> fileEntities = new ArrayList<>();
-        for (UploadedFileDTO file : param.getFiles()) {
+        for (UploadedFileDTO file : files) {
             InformationFileEntity fileEntity = new InformationFileEntity();
             fileEntity.setInformationId(informationId);
             fileEntity.setFileName(file.getFileName());

src/main/java/com/project/information/application/impl/UploadFilesApplicationServiceImpl.java

@@ -1,6 +1,7 @@
 package com.project.information.application.impl;
 
 import cn.hutool.core.io.FileUtil;
+import cn.hutool.core.util.IdUtil;
 import com.project.base.config.CustomIdGenerator;
 import com.project.base.domain.exception.BusinessErrorException;
 import com.project.base.domain.result.Result;
@@ -73,7 +74,7 @@ public class UploadFilesApplicationServiceImpl implements UploadFilesApplication
             // MinIO 路径规范：classicpaper/{yyyyMMdd}/{uuid}.{ext}
             String filePath = String.format("classicpaper/%s/%s.%s",
                     new java.text.SimpleDateFormat("yyyyMMdd").format(new Date()),
-                    customIdGenerator.nextId(null),
+                    IdUtil.fastSimpleUUID(),
                     suffix);
 
             // 上传到 MinIO

src/main/java/com/project/information/controller/InformationController.java

@@ -77,6 +77,7 @@ public class InformationController {
 
     /**
      * Step2：提交表单（创建虚拟资料 + 关联子文件 + 触发解析）
+     * form-data 参数：subLineId, fileName, fileNames[], filePaths[], fileSuffixes[], contentCategories[]
      */
     @PostMapping("/generateFromFiles")
     public Result<String> generateFromFiles(GenerateFromFilesParam param) throws Exception {

src/main/java/com/project/information/domain/param/GenerateFromFilesParam.java

@@ -1,23 +1,47 @@
 package com.project.information.domain.param;
 
+import com.fasterxml.jackson.core.JsonProcessingException;
+import com.fasterxml.jackson.core.type.TypeReference;
+import com.fasterxml.jackson.databind.ObjectMapper;
+import com.project.base.domain.exception.BusinessErrorException;
 import com.project.information.domain.dto.UploadedFileDTO;
 import lombok.Data;
+import org.springframework.util.StringUtils;
 
+import java.util.Collections;
 import java.util.List;
 
 /**
  * 原始资料表单提交请求参数
- * Step2：创建虚拟资料 + 子文件记录 + 触发解析
+ * Step2：创建虚拟资料 + 子文件记录 + 触发数据服务解析
+ * subLineId、fileName 用普通 form-data 传参
+ * files 用 JSON 字符串传参（仿照 ClusterCallbackDTO 模式）
  */
 @Data
 public class GenerateFromFilesParam {
 
+    private static final ObjectMapper MAPPER = new ObjectMapper();
+
     /** 子产品线ID */
     private Long subLineId;
 
     /** 虚拟资料文件名（用户填写的占位名称） */
     private String fileName;
 
-    /** Step1 上传的文件信息列表 */
-    private List<UploadedFileDTO> files;
+    /** Step1 上传的文件信息（JSON 字符串） */
+    private String files;
+
+    /**
+     * 解析 files JSON 字符串为列表（Lombok 不会覆盖此方法）
+     */
+    public List<UploadedFileDTO> parseFiles() {
+        if (!StringUtils.hasText(files)) {
+            return Collections.emptyList();
+        }
+        try {
+            return MAPPER.readValue(files, new TypeReference<List<UploadedFileDTO>>() {});
+        } catch (JsonProcessingException e) {
+            throw new BusinessErrorException("文件信息解析失败");
+        }
+    }
 }

src/main/java/com/project/information/domain/service/ConfirmKnowledgePointDraftDomainService.java

@@ -0,0 +1,21 @@
+package com.project.information.domain.service;
+
+import com.project.base.domain.result.Result;
+
+/**
+ * 确认固化知识点草稿域服务
+ * 将草稿复制到正式表，触发聚类，不可逆
+ */
+public interface ConfirmKnowledgePointDraftDomainService {
+
+    /**
+     * 确认提交（固化，不可逆）
+     * 1. 幂等校验
+     * 2. 草稿 content → evaluator_knowledge_point
+     * 3. 更新 draftStatus = APPROVED
+     * 4. 触发聚类
+     *
+     * @param informationId 虚拟资料ID
+     */
+    Result<String> confirm(Long informationId);
+}

src/main/java/com/project/information/domain/service/impl/ConfirmKnowledgePointDraftDomainServiceImpl.java

@@ -0,0 +1,103 @@
+package com.project.information.domain.service.impl;
+
+import com.baomidou.mybatisplus.core.conditions.query.LambdaQueryWrapper;
+import com.project.base.domain.exception.BusinessErrorException;
+import com.project.base.domain.result.Result;
+import com.project.information.domain.dto.KnowledgePointDTO;
+import com.project.information.domain.entity.InformationEntity;
+import com.project.information.domain.entity.KnowledgePointDraftEntity;
+import com.project.information.domain.entity.KnowledgePointEntity;
+import com.project.information.domain.enums.AuditStatusEnum;
+import com.project.information.domain.enums.DraftStatusEnum;
+import com.project.information.domain.service.*;
+import com.project.interaction.application.AlgorithmApplicationService;
+import lombok.extern.slf4j.Slf4j;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.stereotype.Service;
+import org.springframework.transaction.annotation.Transactional;
+
+import java.util.ArrayList;
+import java.util.List;
+
+/**
+ * 确认固化知识点草稿域服务实现
+ */
+@Service
+@Slf4j
+public class ConfirmKnowledgePointDraftDomainServiceImpl implements ConfirmKnowledgePointDraftDomainService {
+
+    @Autowired
+    private KnowledgePointDraftBaseService knowledgePointDraftBaseService;
+
+    @Autowired
+    private InformationBaseService informationBaseService;
+
+    @Autowired
+    private KnowledgePointBaseService knowledgePointBaseService;
+
+    @Autowired
+    private AlgorithmApplicationService algorithmApplicationService;
+
+    @Override
+    @Transactional(rollbackFor = Exception.class)
+    public Result<String> confirm(Long informationId) {
+        if (informationId == null) {
+            throw new BusinessErrorException("资料ID不能为空");
+        }
+
+        // 1. 查询资料，幂等校验
+        InformationEntity information = informationBaseService.getById(informationId);
+        if (information == null) {
+            throw new BusinessErrorException("资料不存在");
+        }
+        if (DraftStatusEnum.APPROVED.getValue().equals(information.getDraftStatus())) {
+            return Result.success("已固化，不可重复操作");
+        }
+        if (!DraftStatusEnum.PENDING.getValue().equals(information.getDraftStatus())) {
+            throw new BusinessErrorException("该资料不在待审核状态");
+        }
+
+        // 2. 查询所有待审核草稿
+        List<KnowledgePointDraftEntity> drafts = knowledgePointDraftBaseService.list(
+                new LambdaQueryWrapper<KnowledgePointDraftEntity>()
+                        .eq(KnowledgePointDraftEntity::getInformationId, informationId)
+                        .eq(KnowledgePointDraftEntity::getAuditStatus, AuditStatusEnum.PENDING.getValue()));
+        if (drafts.isEmpty()) {
+            throw new BusinessErrorException("没有待审核的草稿");
+        }
+
+        // 3. 将草稿 content 复制到正式知识点表
+        List<KnowledgePointEntity> knowledgePoints = new ArrayList<>();
+        for (KnowledgePointDraftEntity draft : drafts) {
+            KnowledgePointEntity kp = new KnowledgePointEntity();
+            kp.setContent(draft.getContent());
+            kp.setKnowledgeType(draft.getKnowledgeType());
+            kp.setInformationId(informationId);
+            kp.setParseName(draft.getParseName());
+            knowledgePoints.add(kp);
+        }
+        knowledgePointBaseService.saveBatch(knowledgePoints);
+
+        // 4. 更新草稿 auditStatus 为已通过
+        List<Long> draftIds = drafts.stream().map(KnowledgePointDraftEntity::getId).toList();
+        knowledgePointDraftBaseService.lambdaUpdate()
+                .in(KnowledgePointDraftEntity::getId, draftIds)
+                .set(KnowledgePointDraftEntity::getAuditStatus, AuditStatusEnum.APPROVED.getValue())
+                .update();
+
+        // 5. 更新虚拟资料 draftStatus = APPROVED(2)
+        informationBaseService.lambdaUpdate()
+                .eq(InformationEntity::getId, informationId)
+                .set(InformationEntity::getDraftStatus, DraftStatusEnum.APPROVED.getValue())
+                .update();
+
+        // 6. 触发聚类
+        List<KnowledgePointDTO> kpList = knowledgePoints.stream()
+                .map(entity -> entity.toDTO(KnowledgePointDTO::new))
+                .toList();
+        algorithmApplicationService.postToClusteringByInformationId(informationId, kpList);
+
+        log.info(">>> [草稿审核] 确认固化完成, informationId={}, 草稿数={}", informationId, drafts.size());
+        return Result.success("确认成功");
+    }
+}

src/main/java/com/project/interaction/application/AiExtractCallbackApplicationService.java

@@ -0,0 +1,10 @@
+package com.project.interaction.application;
+
+import com.project.interaction.domain.dto.AiExtractCallbackDTO;
+
+/**
+ * 算法服务知识点提取回调应用服务
+ */
+public interface AiExtractCallbackApplicationService {
+    void handleCallback(AiExtractCallbackDTO callback);
+}

src/main/java/com/project/interaction/application/DataExtractCallbackApplicationService.java

@@ -0,0 +1,10 @@
+package com.project.interaction.application;
+
+import com.project.interaction.domain.dto.DataExtractCallbackDTO;
+
+/**
+ * 数据服务解析回调应用服务
+ */
+public interface DataExtractCallbackApplicationService {
+    void handleCallback(DataExtractCallbackDTO callback);
+}

src/main/java/com/project/interaction/application/impl/AiExtractCallbackApplicationServiceImpl.java

@@ -0,0 +1,22 @@
+package com.project.interaction.application.impl;
+
+import com.project.interaction.application.AiExtractCallbackApplicationService;
+import com.project.interaction.domain.dto.AiExtractCallbackDTO;
+import com.project.interaction.domain.service.SaveAiExtractDraftDomainService;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.stereotype.Service;
+
+/**
+ * 算法服务知识点提取回调应用服务实现
+ */
+@Service
+public class AiExtractCallbackApplicationServiceImpl implements AiExtractCallbackApplicationService {
+
+    @Autowired
+    private SaveAiExtractDraftDomainService saveAiExtractDraftDomainService;
+
+    @Override
+    public void handleCallback(AiExtractCallbackDTO callback) {
+        saveAiExtractDraftDomainService.handleCallback(callback);
+    }
+}

src/main/java/com/project/interaction/application/impl/DataExtractCallbackApplicationServiceImpl.java

@@ -0,0 +1,22 @@
+package com.project.interaction.application.impl;
+
+import com.project.interaction.application.DataExtractCallbackApplicationService;
+import com.project.interaction.domain.dto.DataExtractCallbackDTO;
+import com.project.interaction.domain.service.SaveDataExtractResultDomainService;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.stereotype.Service;
+
+/**
+ * 数据服务解析回调应用服务实现
+ */
+@Service
+public class DataExtractCallbackApplicationServiceImpl implements DataExtractCallbackApplicationService {
+
+    @Autowired
+    private SaveDataExtractResultDomainService saveDataExtractResultDomainService;
+
+    @Override
+    public void handleCallback(DataExtractCallbackDTO callback) {
+        saveDataExtractResultDomainService.handleCallback(callback);
+    }
+}

src/main/java/com/project/interaction/domain/service/SaveAiExtractDraftDomainService.java

@@ -0,0 +1,11 @@
+package com.project.interaction.domain.service;
+
+import com.project.interaction.domain.dto.AiExtractCallbackDTO;
+
+/**
+ * 保存AI提取知识点草稿域服务
+ * 将算法返回的草稿知识点写入草稿表
+ */
+public interface SaveAiExtractDraftDomainService {
+    void handleCallback(AiExtractCallbackDTO callback);
+}

src/main/java/com/project/interaction/domain/service/SaveDataExtractResultDomainService.java

@@ -0,0 +1,11 @@
+package com.project.interaction.domain.service;
+
+import com.project.interaction.domain.dto.DataExtractCallbackDTO;
+
+/**
+ * 保存数据服务解析结果域服务
+ * 更新子文件解析状态和文本，检查批次完成，触发算法服务
+ */
+public interface SaveDataExtractResultDomainService {
+    void handleCallback(DataExtractCallbackDTO callback);
+}

src/main/java/com/project/interaction/domain/service/impl/SaveAiExtractDraftDomainServiceImpl.java

@@ -0,0 +1,68 @@
+package com.project.interaction.domain.service.impl;
+
+import com.project.information.domain.entity.InformationEntity;
+import com.project.information.domain.entity.KnowledgePointDraftEntity;
+import com.project.information.domain.enums.AuditStatusEnum;
+import com.project.information.domain.enums.DraftStatusEnum;
+import com.project.information.domain.service.InformationBaseService;
+import com.project.information.domain.service.KnowledgePointDraftBaseService;
+import com.project.interaction.domain.dto.AiExtractCallbackDTO;
+import com.project.interaction.domain.service.SaveAiExtractDraftDomainService;
+import lombok.extern.slf4j.Slf4j;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.stereotype.Service;
+import org.springframework.transaction.annotation.Transactional;
+
+import java.util.ArrayList;
+import java.util.List;
+
+/**
+ * 保存AI提取知识点草稿域服务实现
+ */
+@Service
+@Slf4j
+public class SaveAiExtractDraftDomainServiceImpl implements SaveAiExtractDraftDomainService {
+
+    @Autowired
+    private KnowledgePointDraftBaseService knowledgePointDraftBaseService;
+
+    @Autowired
+    private InformationBaseService informationBaseService;
+
+    @Override
+    @Transactional(rollbackFor = Exception.class)
+    public void handleCallback(AiExtractCallbackDTO callback) {
+        Long informationId = callback.getInformationId();
+        List<AiExtractCallbackDTO.DraftItem> drafts = callback.getDrafts();
+
+        log.info(">>> [算法服务回调] 收到知识点提取回调, informationId={}, 草稿数={}",
+                informationId, drafts != null ? drafts.size() : 0);
+
+        if (drafts == null || drafts.isEmpty()) {
+            log.warn(">>> [算法服务回调] 算法返回的草稿为空, informationId={}", informationId);
+            return;
+        }
+
+        // 1. 保存草稿（aiContent = content 的副本）
+        List<KnowledgePointDraftEntity> draftEntities = new ArrayList<>();
+        for (AiExtractCallbackDTO.DraftItem item : drafts) {
+            KnowledgePointDraftEntity draft = new KnowledgePointDraftEntity();
+            draft.setInformationId(informationId);
+            draft.setContent(item.getContent());
+            draft.setAiContent(item.getContent());
+            draft.setKnowledgeType(item.getKnowledgeType());
+            draft.setParseName(item.getParseName());
+            draft.setAuditStatus(AuditStatusEnum.PENDING.getValue());
+            draftEntities.add(draft);
+        }
+        knowledgePointDraftBaseService.saveBatch(draftEntities);
+
+        // 2. 更新虚拟资料 draftStatus = PENDING(1)
+        informationBaseService.lambdaUpdate()
+                .eq(InformationEntity::getId, informationId)
+                .set(InformationEntity::getDraftStatus, DraftStatusEnum.PENDING.getValue())
+                .update();
+
+        log.info(">>> [算法服务回调] 草稿保存完成, informationId={}, 草稿数={}", informationId, draftEntities.size());
+    }
+}

src/main/java/com/project/interaction/domain/service/impl/SaveDataExtractResultDomainServiceImpl.java

@@ -0,0 +1,96 @@
+package com.project.interaction.domain.service.impl;
+
+import com.baomidou.mybatisplus.core.conditions.query.LambdaQueryWrapper;
+import com.project.information.domain.entity.InformationFileEntity;
+import com.project.information.domain.enums.FileParseStatusEnum;
+import com.project.information.domain.service.InformationFileBaseService;
+import com.project.interaction.domain.dto.AiExtractRequestDTO;
+import com.project.interaction.domain.dto.DataExtractCallbackDTO;
+import com.project.interaction.domain.service.PostToAiExtractDomainService;
+import com.project.interaction.domain.service.SaveDataExtractResultDomainService;
+import lombok.extern.slf4j.Slf4j;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.stereotype.Service;
+
+import java.util.ArrayList;
+import java.util.List;
+
+/**
+ * 保存数据服务解析结果域服务实现
+ */
+@Service
+@Slf4j
+public class SaveDataExtractResultDomainServiceImpl implements SaveDataExtractResultDomainService {
+
+    @Autowired
+    private InformationFileBaseService informationFileBaseService;
+
+    @Autowired
+    private PostToAiExtractDomainService postToAiExtractDomainService;
+
+    @Override
+    public void handleCallback(DataExtractCallbackDTO callback) {
+        Long fileId = callback.getFileId();
+        Long informationId = callback.getInformationId();
+
+        log.info(">>> [数据服务回调] 收到子文件解析回调, fileId={}, informationId={}, status={}",
+                fileId, informationId, callback.getParseStatus());
+
+        // 1. 更新子文件解析状态
+        InformationFileEntity file = informationFileBaseService.getById(fileId);
+        if (file == null) {
+            log.error(">>> [数据服务回调] 子文件不存在, fileId={}", fileId);
+            return;
+        }
+        file.setParseStatus(callback.getParseStatus());
+        if (FileParseStatusEnum.Success.getValue().equals(callback.getParseStatus())) {
+            file.setParsedText(callback.getParsedText());
+        }
+        informationFileBaseService.updateById(file);
+
+        // 2. 解析失败则记录日志
+        if (!FileParseStatusEnum.Success.getValue().equals(callback.getParseStatus())) {
+            log.warn(">>> [数据服务回调] 子文件解析失败, fileId={}, error={}", fileId, callback.getErrorMsg());
+            return;
+        }
+
+        // 3. 检查该虚拟资料下所有子文件是否全部完成
+        List<InformationFileEntity> allFiles = informationFileBaseService.list(
+                new LambdaQueryWrapper<InformationFileEntity>()
+                        .eq(InformationFileEntity::getInformationId, informationId));
+
+        boolean allDone = allFiles.stream()
+                .allMatch(f -> FileParseStatusEnum.Success.getValue().equals(f.getParseStatus())
+                        || FileParseStatusEnum.Failed.getValue().equals(f.getParseStatus()));
+
+        if (!allDone) {
+            log.info(">>> [数据服务回调] 虚拟资料 {} 下还有子文件未完成解析，等待中", informationId);
+            return;
+        }
+
+        // 4. 全部完成 → 调用算法服务知识点提取
+        log.info(">>> [数据服务回调] 虚拟资料 {} 下所有子文件解析完成，开始调用算法服务", informationId);
+
+        List<AiExtractRequestDTO.FileContent> fileContents = new ArrayList<>();
+        for (InformationFileEntity f : allFiles) {
+            if (FileParseStatusEnum.Success.getValue().equals(f.getParseStatus()) && f.getParsedText() != null) {
+                AiExtractRequestDTO.FileContent fc = new AiExtractRequestDTO.FileContent();
+                fc.setFileName(f.getFileName());
+                fc.setContentCategory(String.valueOf(f.getContentCategory()));
+                fc.setParsedText(f.getParsedText());
+                fileContents.add(fc);
+            }
+        }
+
+        if (fileContents.isEmpty()) {
+            log.warn(">>> [数据服务回调] 虚拟资料 {} 下没有成功解析的子文件", informationId);
+            return;
+        }
+
+        AiExtractRequestDTO request = new AiExtractRequestDTO();
+        request.setInformationId(informationId);
+        request.setFiles(fileContents);
+
+        postToAiExtractDomainService.postToAiExtract(request);
+    }
+}

src/main/resources/application-prod.yml

@@ -0,0 +1,97 @@
+server:
+  port: 7088
+
+spring:
+  datasource:
+    dynamic:
+      primary: master
+      datasource:
+        master:
+          driverClassName: com.mysql.cj.jdbc.Driver
+          password: Itc@123456
+          url: jdbc:mysql://8.129.84.155:3306/ai_evaluator?serverTimezone=GMT%2B8&useUnicode=true&characterEncoding=utf-8&allowMultiQueries=true
+          username: root
+  # Redis
+  data:
+    redis:
+      host: 8.129.84.155
+      port: 6379
+      password: Itc@123456
+      database: 3
+      timeout: 5000ms
+      lettuce:
+        pool:
+          max-active: 8
+          max-idle: 30
+          max-wait: 10000
+          min-idle: 10
+  jpa:
+    hibernate:
+      ddl-auto: update
+      naming:
+        physical-strategy: org.hibernate.boot.model.naming.PhysicalNamingStrategyStandardImpl
+    properties:
+      hibernate:
+        dialect: org.hibernate.dialect.MySQL8Dialect
+    show-sql: true
+  # 上传下载限制
+  servlet:
+    multipart:
+      max-file-size: 100MB
+      max-request-size: 100MB
+  jackson:
+    date-format: yyyy-MM-dd HH:mm:ss
+    time-zone: GMT+8
+    serialization:
+      write-dates-as-timestamps: false
+    generator:
+      write-numbers-as-strings: true
+minio:
+  endpoint: ${MINIO_ENDPOINT:http://localhost:9000}
+  accessKey: ${MINIO_ASSESSKEY:DTKYZDZM1i31XOvd24SP}
+  secretKey: ${MINIO_SECRETKEY:PnfLPcJbvaUboZIwYZAADPB0pDtPZgbi0QiLSs3C}
+  bucket: ${MINIO_BUCKET:ai-evaluator}
+  tempAccessFileUrl: http://8.129.84.155/minio-api
+mybatis-plus:
+  configuration:
+    map-underscore-to-camel-case: true
+  mapper-locations: classpath*:mapper/**/*.xml
+  type-aliases-package: com.proposal.**.domain.entity
+
+milvus:
+  host: 127.0.0.1
+  port: 19530
+analysis:
+  host: http://127.0.0.1
+  port: 8888
+  url: /word/parse
+ding:
+  appKey: dinghgexcbos9dorl3kb
+  appSecret: bSNX1ZOUTk2A4FGwQ1U6fyHlFssh9IZtFO6FuKx8A97PE1VZL9lBTJapwzDGCS7v
+#  appKey: ding4eootvq4jzas96lr #保伦架构
+#  appSecret: UI6XcD8GFPE_W_yo2eQkMuoSsTEf1whpZYEXrsDA7CV7bkJp40B3VNcETWS1aGgg
+  agentId: 4418045751
+  corpId: ding13a2ff8718e6bcdd
+#  corpId: 3318151640
+  noticeUrlPrefix: https://107pm707566hq.vicp.fun/ai-evaluator-test
+algo:
+  clusterUrl: /semantic-cluster
+  baseUrl: http://127.0.0.1:8002
+  generateQuestionUrl: /v1/generate/questions_from_cluster
+  apiUrl: http://127.0.0.1:8000
+jwt:
+  secret: "my-very-fixed-and-secure-secret-key-1234567890"
+
+# 题目生成相关配置
+question:
+  generation:
+    # 限流速率：每秒允许的API请求数
+    rate-limit: 4
+    downgrade: false
+  # 队列配置
+  queue:
+    # 重试间隔（秒）
+    retry-interval: 60
+
+scheduled-task:
+  owner: test

src/main/resources/application-test.yml

@@ -9,7 +9,7 @@ spring:
         master:
           driverClassName: com.mysql.cj.jdbc.Driver
           password: Itc@123456
-          url: jdbc:mysql://8.129.84.155:3306/ai_evaluator?serverTimezone=GMT%2B8&useUnicode=true&characterEncoding=utf-8&allowMultiQueries=true
+          url: jdbc:mysql://8.129.84.155:3306/ai_evaluator_test?serverTimezone=GMT%2B8&useUnicode=true&characterEncoding=utf-8&allowMultiQueries=true
           username: root
   # Redis
   data:
@@ -47,10 +47,10 @@ spring:
     generator:
       write-numbers-as-strings: true
 minio:
-  endpoint: ${MINIO_ENDPOINT:http://localhost:9000}
+  endpoint: ${MINIO_ENDPOINT:http://8.129.84.155:9000}
   accessKey: ${MINIO_ASSESSKEY:DTKYZDZM1i31XOvd24SP}
   secretKey: ${MINIO_SECRETKEY:PnfLPcJbvaUboZIwYZAADPB0pDtPZgbi0QiLSs3C}
-  bucket: ${MINIO_BUCKET:ai-evaluator}
+  bucket: ${MINIO_BUCKET:ai-evaluator-test}
   tempAccessFileUrl: http://8.129.84.155/minio-api
 mybatis-plus:
   configuration: