fix chinese file name garbled when file compressed on windows

pkg/open-hydra/dataset-handler.go

@@ -2,6 +2,7 @@ package openhydra
 
 import (
  "archive/zip"
+ "bytes"
  stdErr "errors"
  "fmt"
  "io"
@@ -18,6 +19,8 @@ import (
  xDatasetV1 "open-hydra/pkg/apis/open-hydra-api/dataset/core/v1"
 
  "github.com/emicklei/go-restful/v3"
+ simpleChinese "golang.org/x/text/encoding/simplifiedchinese"
+ "golang.org/x/text/transform"
 )
 
 /*
@@ -255,7 +258,14 @@ func unzipTo(src io.ReaderAt, size int64, destPath string) error {
  }
  defer os.RemoveAll(tempPath)
  for _, file := range reader.File {
- err := writeZipFile(filepath.Join(tempPath, file.Name), file)
+ // for chinese file name that compressed on windows might cause name is garbled
+ // to fix this we have to always convert file name to format B18030
+ decodedFileName, err := decodeGB18030ToString([]byte(file.Name))
+ if err != nil {
+ return errors.NewBadRequest(fmt.Sprintf("Failed decoded file name '%s' to format GB18030 due to error: %v", file.Name, err.Error()))
+ }
+
+ err = writeZipFile(filepath.Join(tempPath, decodedFileName), file)
  if err != nil {
  return errors.NewInternalError(err)
  }
@@ -292,3 +302,12 @@ func writeZipFile(destPah string, zipFile *zip.File) error {
  _, err = io.Copy(fh, zh)
  return err
 }
+
+func decodeGB18030ToString(data []byte) (string, error) {
+ reader := transform.NewReader(bytes.NewReader(data), simpleChinese.GB18030.NewDecoder())
+ decoded, err := io.ReadAll(reader)
+ if err != nil {
+ return "", err
+ }
+ return string(decoded), nil
+}