fix(archive): split internal and driver extraction api

go.mod

@@ -54,6 +54,7 @@ require (
 	github.com/pkg/sftp v1.13.6
 	github.com/pquerna/otp v1.4.0
 	github.com/rclone/rclone v1.67.0
+	github.com/saintfish/chardet v0.0.0-20230101081208-5e3ef4b5456d
 	github.com/sirupsen/logrus v1.9.3
 	github.com/spf13/afero v1.11.0
 	github.com/spf13/cobra v1.8.1
@@ -243,7 +244,7 @@ require (
 	golang.org/x/sync v0.10.0 // indirect
 	golang.org/x/sys v0.28.0 // indirect
 	golang.org/x/term v0.27.0 // indirect
-	golang.org/x/text v0.21.0 // indirect
+	golang.org/x/text v0.21.0
 	golang.org/x/tools v0.24.0 // indirect
 	google.golang.org/api v0.169.0 // indirect
 	google.golang.org/genproto/googleapis/rpc v0.0.0-20240604185151-ef581f913117 // indirect

go.sum

@@ -534,6 +534,8 @@ github.com/russross/blackfriday/v2 v2.1.0/go.mod h1:+Rmxgy9KzJVeS9/2gXHxylqXiyQD
 github.com/rwcarlsen/goexif v0.0.0-20190401172101-9e8deecbddbd/go.mod h1:hPqNNc0+uJM6H+SuU8sEs5K5IQeKccPqeSjfgcKGgPk=
 github.com/ryszard/goskiplist v0.0.0-20150312221310-2dfbae5fcf46 h1:GHRpF1pTW19a8tTFrMLUcfWwyC0pnifVo2ClaLq+hP8=
 github.com/ryszard/goskiplist v0.0.0-20150312221310-2dfbae5fcf46/go.mod h1:uAQ5PCi+MFsC7HjREoAz1BU+Mq60+05gifQSsHSDG/8=
+github.com/saintfish/chardet v0.0.0-20230101081208-5e3ef4b5456d h1:hrujxIzL1woJ7AwssoOcM/tq5JjjG2yYOc8odClEiXA=
+github.com/saintfish/chardet v0.0.0-20230101081208-5e3ef4b5456d/go.mod h1:uugorj2VCxiV1x+LzaIdVa9b4S4qGAcH6cbhh4qVxOU=
 github.com/secsy/goftp v0.0.0-20200609142545-aa2de14babf4 h1:PT+ElG/UUFMfqy5HrxJxNzj3QBOf7dZwupeVC+mG1Lo=
 github.com/secsy/goftp v0.0.0-20200609142545-aa2de14babf4/go.mod h1:MnkX001NG75g3p8bhFycnyIjeQoOjGL6CEIsdE/nKSY=
 github.com/shabbyrobe/gocovmerge v0.0.0-20230507112040-c3350d9342df h1:S77Pf5fIGMa7oSwp8SQPp7Hb4ZiI38K3RNBKD2LLeEM=

internal/archive/archives/archives.go

@@ -5,6 +5,7 @@ import (
 	"github.com/alist-org/alist/v3/internal/model"
 	"github.com/alist-org/alist/v3/internal/stream"
 	"github.com/alist-org/alist/v3/pkg/utils"
+	"io"
 	"io/fs"
 	"os"
 	stdpath "path"
@@ -29,15 +30,22 @@ func (_ *Archives) GetMeta(ss *stream.SeekableStream, args model.ArchiveArgs) (m
 	if err != nil {
 		return nil, filterPassword(err)
 	}
-	return &tool.EmptyMeta{}, nil
+	return &model.ArchiveMetaInfo{
+		Comment:   "",
+		Encrypted: false,
+	}, nil
 }
 
 func (_ *Archives) List(ss *stream.SeekableStream, args model.ArchiveInnerArgs) ([]model.Obj, error) {
 	fsys, err := getFs(ss, args.ArchiveArgs)
 	if err != nil {
 		return nil, err
 	}
-	obj, err := fsys.ReadDir(strings.TrimPrefix(args.InnerPath, "/"))
+	innerPath := strings.TrimPrefix(args.InnerPath, "/")
+	if innerPath == "" {
+		innerPath = "."
+	}
+	obj, err := fsys.ReadDir(innerPath)
 	if err != nil {
 		return nil, filterPassword(err)
 	}
@@ -50,16 +58,20 @@ func (_ *Archives) List(ss *stream.SeekableStream, args model.ArchiveInnerArgs)
 	})
 }
 
-func (_ *Archives) Extract(ss *stream.SeekableStream, args model.ArchiveInnerArgs) (*model.Link, error) {
+func (_ *Archives) Extract(ss *stream.SeekableStream, args model.ArchiveInnerArgs) (io.ReadCloser, int64, error) {
 	fsys, err := getFs(ss, args.ArchiveArgs)
 	if err != nil {
-		return nil, err
+		return nil, 0, err
 	}
 	file, err := fsys.Open(strings.TrimPrefix(args.InnerPath, "/"))
 	if err != nil {
-		return nil, filterPassword(err)
+		return nil, 0, filterPassword(err)
+	}
+	stat, err := file.Stat()
+	if err != nil {
+		return nil, 0, filterPassword(err)
 	}
-	return &model.Link{MFile: &tool.SequentialFile{Reader: file}}, nil
+	return file, stat.Size(), nil
 }
 
 func (_ *Archives) Decompress(ss *stream.SeekableStream, outputPath string, args model.ArchiveInnerArgs, up model.UpdateProgress) error {

internal/archive/tool/base.go

@@ -3,12 +3,13 @@ package tool
 import (
 	"github.com/alist-org/alist/v3/internal/model"
 	"github.com/alist-org/alist/v3/internal/stream"
+	"io"
 )
 
 type Tool interface {
 	AcceptedExtensions() []string
 	GetMeta(ss *stream.SeekableStream, args model.ArchiveArgs) (model.ArchiveMeta, error)
 	List(ss *stream.SeekableStream, args model.ArchiveInnerArgs) ([]model.Obj, error)
-	Extract(ss *stream.SeekableStream, args model.ArchiveInnerArgs) (*model.Link, error)
+	Extract(ss *stream.SeekableStream, args model.ArchiveInnerArgs) (io.ReadCloser, int64, error)
 	Decompress(ss *stream.SeekableStream, outputPath string, args model.ArchiveInnerArgs, up model.UpdateProgress) error
 }

internal/archive/tool/utils.go

@@ -2,8 +2,6 @@ package tool
 
 import (
 	"github.com/alist-org/alist/v3/internal/errs"
-	"github.com/alist-org/alist/v3/internal/model"
-	"io"
 )
 
 var (
@@ -23,37 +21,3 @@ func GetArchiveTool(ext string) (Tool, error) {
 	}
 	return t, nil
 }
-
-type SequentialFile struct {
-	Reader io.ReadCloser
-}
-
-func (s *SequentialFile) Read(p []byte) (n int, err error) {
-	return s.Reader.Read(p)
-}
-
-func (s *SequentialFile) ReadAt(p []byte, off int64) (n int, err error) {
-	return 0, errs.NotSupport
-}
-
-func (s *SequentialFile) Seek(offset int64, whence int) (int64, error) {
-	return 0, errs.NotSupport
-}
-
-func (s *SequentialFile) Close() error {
-	return s.Reader.Close()
-}
-
-type EmptyMeta struct{}
-
-func (e *EmptyMeta) GetComment() string {
-	return ""
-}
-
-func (e *EmptyMeta) IsEncrypted() bool {
-	return false
-}
-
-func (e *EmptyMeta) GetTree() []model.ObjTree {
-	return nil
-}

internal/archive/zip/utils.go

@@ -1,34 +1,30 @@
 package zip
 
 import (
+	"bytes"
+	"github.com/alist-org/alist/v3/internal/errs"
 	"github.com/alist-org/alist/v3/internal/model"
 	"github.com/alist-org/alist/v3/internal/stream"
+	"github.com/saintfish/chardet"
 	"github.com/yeka/zip"
+	"golang.org/x/text/encoding"
+	"golang.org/x/text/encoding/charmap"
+	"golang.org/x/text/encoding/japanese"
+	"golang.org/x/text/encoding/korean"
+	"golang.org/x/text/encoding/simplifiedchinese"
+	"golang.org/x/text/encoding/traditionalchinese"
+	"golang.org/x/text/encoding/unicode"
+	"golang.org/x/text/encoding/unicode/utf32"
+	"golang.org/x/text/transform"
 	"io"
 	"os"
 	stdpath "path"
+	"strings"
 )
 
-type zipMeta struct {
-	Comment   string
-	Encrypted bool
-}
-
-func (m *zipMeta) GetComment() string {
-	return m.Comment
-}
-
-func (m *zipMeta) IsEncrypted() bool {
-	return m.Encrypted
-}
-
-func (m *zipMeta) GetTree() []model.ObjTree {
-	return nil
-}
-
 func toModelObj(file os.FileInfo) *model.Object {
 	return &model.Object{
-		Name:     file.Name(),
+		Name:     decodeName(file.Name()),
 		Size:     file.Size(),
 		Modified: file.ModTime(),
 		IsFolder: file.IsDir(),
@@ -80,3 +76,81 @@ func _decompress(file *zip.File, targetPath, password string, up model.UpdatePro
 	}
 	return nil
 }
+
+func filterPassword(err error) error {
+	if err != nil && strings.Contains(err.Error(), "password") {
+		return errs.WrongArchivePassword
+	}
+	return err
+}
+
+func decodeName(name string) string {
+	b := []byte(name)
+	detector := chardet.NewTextDetector()
+	result, err := detector.DetectBest(b)
+	if err != nil {
+		return name
+	}
+	enc := getEncoding(result.Charset)
+	if enc == nil {
+		return name
+	}
+	i := bytes.NewReader(b)
+	decoder := transform.NewReader(i, enc.NewDecoder())
+	content, _ := io.ReadAll(decoder)
+	return string(content)
+}
+
+func getEncoding(name string) (enc encoding.Encoding) {
+	switch name {
+	case "UTF-16BE":
+		enc = unicode.UTF16(unicode.BigEndian, unicode.IgnoreBOM)
+	case "UTF-16LE":
+		enc = unicode.UTF16(unicode.LittleEndian, unicode.IgnoreBOM)
+	case "UTF-32BE":
+		enc = utf32.UTF32(utf32.BigEndian, utf32.IgnoreBOM)
+	case "UTF-32LE":
+		enc = utf32.UTF32(utf32.LittleEndian, utf32.IgnoreBOM)
+	case "ISO-8859-1":
+		enc = charmap.ISO8859_1
+	case "ISO-8859-2":
+		enc = charmap.ISO8859_2
+	case "ISO-8859-3":
+		enc = charmap.ISO8859_3
+	case "ISO-8859-4":
+		enc = charmap.ISO8859_4
+	case "ISO-8859-5":
+		enc = charmap.ISO8859_5
+	case "ISO-8859-6":
+		enc = charmap.ISO8859_6
+	case "ISO-8859-7":
+		enc = charmap.ISO8859_7
+	case "ISO-8859-8":
+		enc = charmap.ISO8859_8
+	case "ISO-8859-8-I":
+		enc = charmap.ISO8859_8I
+	case "ISO-8859-9":
+		enc = charmap.ISO8859_9
+	case "windows-1251":
+		enc = charmap.Windows1251
+	case "windows-1256":
+		enc = charmap.Windows1256
+	case "KOI8-R":
+		enc = charmap.KOI8R
+	case "Shift_JIS":
+		enc = japanese.ShiftJIS
+	case "GB-18030":
+		enc = simplifiedchinese.GB18030
+	case "EUC-JP":
+		enc = japanese.EUCJP
+	case "EUC-KR":
+		enc = korean.EUCKR
+	case "Big5":
+		enc = traditionalchinese.Big5
+	case "ISO-2022-JP":
+		enc = japanese.ISO2022JP
+	default:
+		enc = nil
+	}
+	return
+}