支持多种文件格式的上传和解析, 实现自定义知识库的功能

pom.xml

@@ -101,6 +101,12 @@
             <artifactId>spring-ai-starter-vector-store-pgvector</artifactId>
             <version>1.0.0-M7</version>
         </dependency>-->
+        <!-- spring-ai-tika-document-reader -->
+        <dependency>
+            <groupId>org.springframework.ai</groupId>
+            <artifactId>spring-ai-tika-document-reader</artifactId>
+            <version>1.0.0</version>
+        </dependency>
         <dependency>
             <groupId>com.baomidou</groupId>
             <artifactId>mybatis-plus-spring-boot3-starter</artifactId>

src/main/java/com/wok/supportbot/controller/AiController.java

@@ -18,7 +18,8 @@ import reactor.core.publisher.Flux;
 import java.io.IOException;
 
 
-
+@RestController
+@RequestMapping("/ai")
 public class AiController {
 
     @Resource

src/main/java/com/wok/supportbot/controller/DocumentController.java

@@ -0,0 +1,232 @@
+package com.wok.supportbot.controller;
+
+import com.wok.supportbot.document.extract.JsonDocumentLoader;
+import com.wok.supportbot.document.extract.MarkdownDocumentLoader;
+import com.wok.supportbot.document.extract.SimpleStringDocumentReader;
+import com.wok.supportbot.document.extract.TikaDocumentReader;
+import com.wok.supportbot.document.transform.MyKeywordEnricher;
+import com.wok.supportbot.document.transform.MyTokenTextSplitter;
+import org.springframework.ai.document.Document;
+import org.springframework.ai.vectorstore.VectorStore;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.http.ResponseEntity;
+import org.springframework.web.bind.annotation.*;
+import org.springframework.web.multipart.MultipartFile;
+
+import java.util.List;
+import java.util.Map;
+
+@RestController
+@RequestMapping("/document")
+public class DocumentController {
+
+    @Autowired
+    private TikaDocumentReader tikaDocumentReader;
+
+    @Autowired
+    private SimpleStringDocumentReader simpleStringDocumentReader;
+
+    @Autowired
+    private MarkdownDocumentLoader markdownDocumentLoader;
+
+    @Autowired
+    private JsonDocumentLoader jsonDocumentLoader;
+
+    @Autowired
+    private MyTokenTextSplitter myTokenTextSplitter;
+
+    @Autowired
+    private MyKeywordEnricher myKeywordEnricher;
+
+    @Autowired
+    private VectorStore pgVectorVectorStore;
+
+    /**
+     * 上传普通文件（支持多种格式），用 Tika 解析
+     */
+    @PostMapping("/upload/file")
+    public ResponseEntity<Map<String, Object>> uploadFile(@RequestParam("file") MultipartFile file) {
+        try {
+            List<Document> documents = tikaDocumentReader.read(file);
+            
+            // 拆分文档
+            List<Document> splitDocuments = myTokenTextSplitter.splitDocuments(documents);
+            
+            // 添加元数据
+            List<Document> enrichedDocuments = myKeywordEnricher.enrichDocuments(splitDocuments);
+            
+            // 转成向量并存入数据库
+            pgVectorVectorStore.add(enrichedDocuments);
+            
+            return ResponseEntity.ok(Map.of(
+                "success", true,
+                "message", "文件上传并向量化成功",
+                "documentCount", enrichedDocuments.size()
+            ));
+        } catch (Exception e) {
+            return ResponseEntity.status(500).body(Map.of(
+                "success", false,
+                "message", "上传失败：" + e.getMessage()
+            ));
+        }
+    }
+
+    /**
+     * 上传字符串内容
+     */
+    @PostMapping("/upload/string")
+    public ResponseEntity<Map<String, Object>> uploadString(@RequestBody String content) {
+        try {
+            List<Document> documents = simpleStringDocumentReader.read(content);
+            
+            // 拆分文档
+            List<Document> splitDocuments = myTokenTextSplitter.splitDocuments(documents);
+            
+            // 添加元数据
+            List<Document> enrichedDocuments = myKeywordEnricher.enrichDocuments(splitDocuments);
+            
+            // 转成向量并存入数据库
+            pgVectorVectorStore.add(enrichedDocuments);
+            
+            return ResponseEntity.ok(Map.of(
+                "success", true,
+                "message", "文本内容上传并向量化成功",
+                "documentCount", enrichedDocuments.size()
+            ));
+        } catch (Exception e) {
+            return ResponseEntity.status(500).body(Map.of(
+                "success", false,
+                "message", "上传失败：" + e.getMessage()
+            ));
+        }
+    }
+
+    /**
+     * 上传 Markdown 文件
+     */
+    @PostMapping("/upload/markdown")
+    public ResponseEntity<Map<String, Object>> uploadMarkdown(@RequestParam("file") MultipartFile file) {
+        try {
+            List<Document> documents = markdownDocumentLoader.loadMarkdownFromFile(file);
+            
+            // 拆分文档
+            List<Document> splitDocuments = myTokenTextSplitter.splitDocuments(documents);
+            
+            // 添加元数据
+            List<Document> enrichedDocuments = myKeywordEnricher.enrichDocuments(splitDocuments);
+            
+            // 转成向量并存入数据库
+            pgVectorVectorStore.add(enrichedDocuments);
+            
+            return ResponseEntity.ok(Map.of(
+                "success", true,
+                "message", "Markdown文件上传并向量化成功",
+                "documentCount", enrichedDocuments.size()
+            ));
+        } catch (Exception e) {
+            return ResponseEntity.status(500).body(Map.of(
+                "success", false,
+                "message", "上传失败：" + e.getMessage()
+            ));
+        }
+    }
+
+    /**
+     * 上传 JSON 文件（基本方式）
+     * 把 JSON 根节点当成一个整体文档
+     */
+    @PostMapping("/upload/json/basic")
+    public ResponseEntity<Map<String, Object>> uploadJsonBasic(@RequestParam("file") MultipartFile file) {
+        try {
+            List<Document> documents = jsonDocumentLoader.loadBasicJson(file);
+            
+            // 拆分文档
+            List<Document> splitDocuments = myTokenTextSplitter.splitDocuments(documents);
+            
+            // 添加元数据
+            List<Document> enrichedDocuments = myKeywordEnricher.enrichDocuments(splitDocuments);
+            
+            // 转成向量并存入数据库
+            pgVectorVectorStore.add(enrichedDocuments);
+            
+            return ResponseEntity.ok(Map.of(
+                "success", true,
+                "message", "JSON文件（基本方式）上传并向量化成功",
+                "documentCount", enrichedDocuments.size()
+            ));
+        } catch (Exception e) {
+            return ResponseEntity.status(500).body(Map.of(
+                "success", false,
+                "message", "上传失败：" + e.getMessage()
+            ));
+        }
+    }
+
+    /**
+     * 上传 JSON 文件（按字段提取）
+     * 用于提取指定字段文本
+     */
+    @PostMapping("/upload/json/fields")
+    public ResponseEntity<Map<String, Object>> uploadJsonWithFields(
+            @RequestParam("file") MultipartFile file,
+            @RequestParam("fields") List<String> fields) {
+        try {
+            List<Document> documents = jsonDocumentLoader.loadJsonByFields(file, fields.toArray(new String[0]));
+            
+            // 拆分文档
+            List<Document> splitDocuments = myTokenTextSplitter.splitDocuments(documents);
+            
+            // 添加元数据
+            List<Document> enrichedDocuments = myKeywordEnricher.enrichDocuments(splitDocuments);
+            
+            // 转成向量并存入数据库
+            pgVectorVectorStore.add(enrichedDocuments);
+            
+            return ResponseEntity.ok(Map.of(
+                "success", true,
+                "message", "JSON文件（按字段）上传并向量化成功",
+                "documentCount", enrichedDocuments.size(),
+                "extractedFields", fields
+            ));
+        } catch (Exception e) {
+            return ResponseEntity.status(500).body(Map.of(
+                "success", false,
+                "message", "上传失败：" + e.getMessage()
+            ));
+        }
+    }
+
+    /**
+     * 上传 JSON 文件（按指针拆分）
+     * 用于拆分数组元素，常用来分段成多文档
+     */
+    @PostMapping("/upload/json/pointer")
+    public ResponseEntity<Map<String, Object>> uploadJsonWithPointer(
+            @RequestParam("file") MultipartFile file,
+            @RequestParam("pointer") String pointer) {
+        try {
+            List<Document> documents = jsonDocumentLoader.loadJsonByPointer(file, pointer);
+            
+            // 拆分文档
+            List<Document> splitDocuments = myTokenTextSplitter.splitDocuments(documents);
+            
+            // 添加元数据
+            List<Document> enrichedDocuments = myKeywordEnricher.enrichDocuments(splitDocuments);
+            
+            // 转成向量并存入数据库
+            pgVectorVectorStore.add(enrichedDocuments);
+            
+            return ResponseEntity.ok(Map.of(
+                "success", true,
+                "message", "JSON文件（按指针）上传并向量化成功",
+                "documentCount", enrichedDocuments.size(),
+                "pointer", pointer
+            ));
+        } catch (Exception e) {
+            return ResponseEntity.status(500).body(Map.of(
+                "success", false,
+                "message", "上传失败：" + e.getMessage()
+            ));
+        }
+    }
+}

src/main/java/com/wok/supportbot/document/extract/JsonDocumentLoader.java

@@ -0,0 +1,59 @@
+package com.wok.supportbot.document.extract;
+
+import lombok.extern.slf4j.Slf4j;
+import org.springframework.ai.document.Document;
+import org.springframework.ai.reader.JsonReader;
+import org.springframework.core.io.FileSystemResource;
+import org.springframework.core.io.Resource;
+import org.springframework.stereotype.Component;
+import org.springframework.web.multipart.MultipartFile;
+
+import java.io.File;
+import java.io.IOException;
+import java.util.List;
+
+@Component
+@Slf4j
+public class JsonDocumentLoader {
+
+    /**
+     * 基本读取方式
+     */
+    public List<Document> loadBasicJson(MultipartFile file) {
+        Resource resource = toResource(file);
+        JsonReader reader = new JsonReader(resource);
+        return reader.get();
+    }
+
+    /**
+     * 指定字段读取方式（例如 description、features 字段）
+     */
+    public List<Document> loadJsonByFields(MultipartFile file, String... fields) {
+        Resource resource = toResource(file);
+        JsonReader reader = new JsonReader(resource, fields);
+        return reader.get();
+    }
+
+    /**
+     * 使用 JSON Pointer 提取数组路径内容（如 /items）
+     */
+    public List<Document> loadJsonByPointer(MultipartFile file, String pointer) {
+        Resource resource = toResource(file);
+        JsonReader reader = new JsonReader(resource);
+        return reader.get(pointer);
+    }
+
+    /**
+     * 将 MultipartFile 转换为 Resource
+     */
+    private Resource toResource(MultipartFile file) {
+        try {
+            File temp = File.createTempFile("upload-", file.getOriginalFilename());
+            file.transferTo(temp);
+            return new FileSystemResource(temp);
+        } catch (IOException e) {
+            log.error("JSON 文件转换失败", e);
+            throw new RuntimeException("JSON 文件转换失败", e);
+        }
+    }
+}

src/main/java/com/wok/supportbot/document/extract/MarkdownDocumentLoader.java

@@ -2,52 +2,46 @@ package com.wok.supportbot.document.extract;
 
 import lombok.extern.slf4j.Slf4j;
 import org.springframework.ai.document.Document;
-
 import org.springframework.ai.reader.markdown.MarkdownDocumentReader;
 import org.springframework.ai.reader.markdown.config.MarkdownDocumentReaderConfig;
+import org.springframework.core.io.FileSystemResource;
 import org.springframework.core.io.Resource;
-import org.springframework.core.io.support.ResourcePatternResolver;
 import org.springframework.stereotype.Component;
+import org.springframework.web.multipart.MultipartFile;
 
+import java.io.File;
 import java.io.IOException;
-import java.util.ArrayList;
 import java.util.List;
 
 /**
- * 文档加载器
+ * Markdown 文件上传解析器
  */
 @Component
 @Slf4j
 public class MarkdownDocumentLoader {
 
-    private final ResourcePatternResolver resourcePatternResolver;
-
-    public MarkdownDocumentLoader(ResourcePatternResolver resourcePatternResolver) {
-        this.resourcePatternResolver = resourcePatternResolver;
-    }
-
-    /**
-     * 加载多篇 Markdown 文档
-     * @return
-     */
-    public List<Document> loadMarkdowns() {
-        List<Document> allDocuments = new ArrayList<>();
+    public List<Document> loadMarkdownFromFile(MultipartFile file) {
         try {
-            Resource[] resources = resourcePatternResolver.getResources("classpath:document/*.md");
-            for (Resource resource : resources) {
-                String filename = resource.getFilename();
+            // 将 MultipartFile 保存为临时文件
+            File temp = File.createTempFile("upload-", file.getOriginalFilename());
+            file.transferTo(temp);
+            Resource resource = new FileSystemResource(temp);
+
+            // 配置文档解析
             MarkdownDocumentReaderConfig config = MarkdownDocumentReaderConfig.builder()
                     .withHorizontalRuleCreateDocument(true)
                     .withIncludeCodeBlock(false)
                     .withIncludeBlockquote(false)
-                        .withAdditionalMetadata("filename", filename)
+                    .withAdditionalMetadata("filename", file.getOriginalFilename())
                     .build();
-                MarkdownDocumentReader markdownDocumentReader = new MarkdownDocumentReader(resource, config);
-                allDocuments.addAll(markdownDocumentReader.get());
-            }
+
+            // 读取文档内容
+            MarkdownDocumentReader reader = new MarkdownDocumentReader(resource, config);
+            return reader.get();
+
         } catch (IOException e) {
-           log.error("Markdown 文档加载失败", e);
+            log.error("Markdown 文件解析失败", e);
+            throw new RuntimeException("Markdown 文件解析失败", e);
         }
-        return allDocuments;
     }
 }

src/main/java/com/wok/supportbot/document/extract/MyJsonReader.java

@@ -1,37 +0,0 @@
-package com.wok.supportbot.document.extract;
-
-import org.springframework.ai.document.Document;
-import org.springframework.ai.reader.JsonReader;
-import org.springframework.beans.factory.annotation.Value;
-import org.springframework.core.io.Resource;
-import org.springframework.stereotype.Component;
-
-import java.util.List;
-
-// 从 classpath 下的 JSON 文件中读取文档
- @Component
- class MyJsonReader {
-     private final Resource resource;
-
-     MyJsonReader(@Value("classpath:products.json") Resource resource) {
-         this.resource = resource;
-     }
-
-     // 基本用法
-     List<Document> loadBasicJsonDocuments() {
-         JsonReader jsonReader = new JsonReader(this.resource);
-         return jsonReader.get();
-     }
-
-     // 指定使用哪些 JSON 字段作为文档内容
-     List<Document> loadJsonWithSpecificFields() {
-         JsonReader jsonReader = new JsonReader(this.resource, "description", "features");
-         return jsonReader.get();
-     }
-
-     // 使用 JSON 指针精确提取文档内容
-     List<Document> loadJsonWithPointer() {
-         JsonReader jsonReader = new JsonReader(this.resource);
-         return jsonReader.get("/items"); // 提取 items 数组内的内容
-     }
- }

src/main/java/com/wok/supportbot/document/extract/SimpleStringDocumentReader.java

@@ -0,0 +1,20 @@
+package com.wok.supportbot.document.extract;
+
+import org.springframework.ai.document.Document;
+import org.springframework.stereotype.Component;
+
+import java.util.Collections;
+import java.util.List;
+import java.util.UUID;
+
+@Component
+public class SimpleStringDocumentReader {
+
+    public List<Document> read(String content) {
+        Document doc = Document.builder()
+                .id(UUID.randomUUID().toString())
+                .text(content)
+                .build();
+        return Collections.singletonList(doc);
+    }
+}

src/main/java/com/wok/supportbot/document/extract/TikaDocumentReader.java

@@ -0,0 +1,44 @@
+package com.wok.supportbot.document.extract;
+
+import lombok.extern.slf4j.Slf4j;
+import org.apache.tika.Tika;
+import org.apache.tika.exception.TikaException;
+import org.springframework.ai.document.Document;
+import org.springframework.core.io.Resource;
+import org.springframework.core.io.FileSystemResource;
+import org.springframework.stereotype.Component;
+import org.springframework.web.multipart.MultipartFile;
+
+import java.io.File;
+import java.io.IOException;
+import java.util.Collections;
+import java.util.List;
+import java.util.UUID;
+
+@Component
+@Slf4j
+public class TikaDocumentReader {
+
+    public List<Document> read(MultipartFile file) {
+        try {
+            // MultipartFile 转 Resource
+            File tempFile = File.createTempFile("upload-", file.getOriginalFilename());
+            file.transferTo(tempFile);
+            Resource resource = new FileSystemResource(tempFile);
+
+            Tika tika = new Tika();
+            String text = tika.parseToString(resource.getInputStream());
+
+            Document doc = Document.builder()
+                    .id(UUID.randomUUID().toString())
+                    .text(text)
+                    .build();
+
+            return Collections.singletonList(doc);
+
+        } catch (IOException | TikaException e) {
+            log.error("Tika 文件解析失败", e);
+            throw new RuntimeException("Tika 文件解析失败", e);
+        }
+    }
+}

src/main/java/com/wok/supportbot/document/transform/MyTokenTextSplitter.java

@@ -10,7 +10,7 @@ import java.util.List;
  * 自定义基于 Token 的切词器
  */
 @Component
-class MyTokenTextSplitter {
+public class MyTokenTextSplitter {
 
     /**
      * 使用默认设置创建分割器。