school-ai.git - Gitblit

¶Ô±ÈÐÂÎÄ¼þ
			@@ -0,0 +1,41 @@
			package org.ruoyi.chain.loader;

			import dev.langchain4j.data.document.Document;
			import dev.langchain4j.data.document.parser.apache.tika.ApacheTikaDocumentParser;
			import lombok.AllArgsConstructor;
			import lombok.extern.slf4j.Slf4j;
			import org.ruoyi.chain.split.TextSplitter;
			import org.ruoyi.common.core.exception.UtilException;
			import org.springframework.stereotype.Component;

			import java.io.BufferedInputStream;
			import java.io.IOException;
			import java.io.InputStream;
			import java.util.List;
			@Component
			@AllArgsConstructor
			@Slf4j
			public class ExcelFileLoader implements ResourceLoader {
			private final TextSplitter textSplitter;
			private static final int DEFAULT_BUFFER_SIZE = 8192;
			@Override
			public String getContent(InputStream inputStream) {
			// ä½¿ç¨å¸¦ç¼å²çè¾å¥æµåè£ï¼ä¿æåæµä¸èªå¨å³éï¼
			try (InputStream bufferedStream = new BufferedInputStream(inputStream, DEFAULT_BUFFER_SIZE)) {
			ApacheTikaDocumentParser apacheTikaDocumentParser = new ApacheTikaDocumentParser();
			Document document = apacheTikaDocumentParser.parse(bufferedStream);
			return document.text();
			} catch (IOException e) {
			String errorMsg = "Excelæä»¶æµè¯»åå¤±è´¥";
			throw new UtilException(errorMsg, e);
			} catch (RuntimeException e) {
			String errorMsg = "Excelåå®¹è§£æå¼å¸¸";
			throw new UtilException(errorMsg, e);
			}
			}

			@Override
			public List<String> getChunkList(String content, String kid) {
			return textSplitter.split(content, kid);
			}
			}