github/JeecgBoot
1
0
Fork 0
mirror of https://github.com/jeecgboot/JeecgBoot.git synced 2026-04-23 10:18:18 +00:00
Code Issues Packages Projects Releases Wiki Activity

PR有问题修复

Browse Source
This commit is contained in:
JEECG 2026-04-15 23:09:28 +08:00
parent 65c466a132
commit fd657a478a
1 changed files with 15 additions and 17 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

32
jeecg-boot/jeecg-boot-module/jeecg-boot-module-airag/src/main/java/org/jeecg/modules/airag/llm/document/TikaDocumentParser.java
View File

@ -71,27 +71,25 @@ public class TikaDocumentParser {
public Document parse(File file) { public Document parse(File file) {
AssertUtils.assertNotEmpty("请选择文件", file); AssertUtils.assertNotEmpty("请选择文件", file);
try { // 使用 Tika 自动检测 MIME 类型
String fileName = file.getName().toLowerCase();
//后缀
String ext = FilenameUtils.getExtension(fileName);
if (fileName.endsWith(".txt")
|| fileName.endsWith(".md")
|| fileName.endsWith(".pdf")) {
// 用于解析(使用FileInputStream避免file.toPath()在Linux非UTF-8环境下中文文件名报错) // 用于解析(使用FileInputStream避免file.toPath()在Linux非UTF-8环境下中文文件名报错)
try (InputStream isForParsing = new FileInputStream(file)) { try (InputStream isForParsing = new FileInputStream(file)) {
// 使用 Tika 自动检测 MIME 类型 return extractByTika(isForParsing);
String fileName = file.getName().toLowerCase(); } catch (IOException e) {
//后缀
String ext = FilenameUtils.getExtension(fileName);
if (fileName.endsWith(".txt")
|| fileName.endsWith(".md")
|| fileName.endsWith(".pdf")) {
return extractByTika(isForParsing);
//update-begin---author:wangshuai---date:2026-01-09---for:【QQYUN-14261】【AI】AI助手，支持多模态能力- 文档---
} else if (FILE_SUFFIX.contains(ext.toLowerCase())) {
return parseDocExcelPdfUsingApachePoi(file);
//update-end---author:wangshuai---date:2026-01-09---for:【QQYUN-14261】【AI】AI助手，支持多模态能力- 文档---
} else {
throw new IllegalArgumentException("不支持的文件格式: " + FilenameUtils.getExtension(fileName));
}
} catch (IOException e) {
throw new RuntimeException(e); throw new RuntimeException(e);
} }
//update-begin---author:wangshuai---date:2026-01-09---for:【QQYUN-14261】【AI】AI助手，支持多模态能力- 文档---
} else if (FILE_SUFFIX.contains(ext.toLowerCase())) {
return parseDocExcelPdfUsingApachePoi(file);
//update-end---author:wangshuai---date:2026-01-09---for:【QQYUN-14261】【AI】AI助手，支持多模态能力- 文档---
} else {
throw new IllegalArgumentException("不支持的文件格式: " + FilenameUtils.getExtension(fileName));
} }
} }