From ed97fc27dbc09c0cc01047f62e73e688b6bfdec0 Mon Sep 17 00:00:00 2001
From: David Newhall II <david@sleepers.pro>
Date: Sat, 12 Apr 2025 15:52:43 -0700
Subject: [PATCH 1/5] allow providing custom file name encoding

---
 7z.go         |  6 +++++-
 ar.go         |  6 +++++-
 cpio.go       |  6 +++++-
 decompress.go | 60 ++++++++++++++++++++++++++++++++++++++++-----------
 files.go      | 28 ++++++++++++++++++++++--
 iso.go        |  7 ++++--
 rar.go        |  6 +++++-
 tar.go        |  6 +++++-
 zip.go        |  6 +++++-
 9 files changed, 109 insertions(+), 22 deletions(-)

diff --git a/7z.go b/7z.go
index 02dc3c9..3374eab 100644
--- a/7z.go
+++ b/7z.go
@@ -84,7 +84,11 @@ func extract7z(xFile *XFile) (int64, []string, []string, error) {
 }
 
 func (x *XFile) un7zip(zipFile *sevenzip.File) (int64, string, error) {
-	wfile := x.clean(zipFile.Name)
+	wfile, err := x.clean(zipFile.Name)
+	if err != nil {
+		return 0, wfile, err
+	}
+
 	if !strings.HasPrefix(wfile, x.OutputDir) {
 		// The file being written is trying to write outside of our base path. Malicious archive?
 		return 0, wfile, fmt.Errorf("%s: %w: %s (from: %s)", zipFile.FileInfo().Name(), ErrInvalidPath, wfile, zipFile.Name)
diff --git a/ar.go b/ar.go
index 93469f6..86ff11d 100644
--- a/ar.go
+++ b/ar.go
@@ -38,7 +38,11 @@ func (x *XFile) unAr(reader io.Reader) (int64, []string, error) {
 			return size, files, fmt.Errorf("%w: %s", ErrInvalidHead, x.FilePath)
 		}
 
-		wfile := x.clean(header.Name)
+		wfile, err := x.clean(header.Name)
+		if err != nil {
+			return 0, files, err
+		}
+
 		if !strings.HasPrefix(wfile, x.OutputDir) {
 			// The file being written is trying to write outside of our base path. Malicious archive?
 			return size, files, fmt.Errorf("%s: %w: %s (from: %s)", x.FilePath, ErrInvalidPath, wfile, header.Name)
diff --git a/cpio.go b/cpio.go
index fd882d0..f1ae106 100644
--- a/cpio.go
+++ b/cpio.go
@@ -64,7 +64,11 @@ func (x *XFile) uncpio(reader io.Reader) (int64, []string, error) {
 }
 
 func (x *XFile) uncpioFile(cpioFile *cpio.Header, cpioReader *cpio.Reader) (int64, error) {
-	wfile := x.clean(cpioFile.Name)
+	wfile, err := x.clean(cpioFile.Name)
+	if err != nil {
+		return 0, err
+	}
+
 	if !strings.HasPrefix(wfile, x.OutputDir) {
 		// The file being written is trying to write outside of the base path. Malicious archive?
 		return 0, fmt.Errorf("%s: %w: %s (from: %s)", cpioFile.FileInfo().Name(), ErrInvalidPath, wfile, cpioFile.Name)
diff --git a/decompress.go b/decompress.go
index 0526334..0e724ae 100644
--- a/decompress.go
+++ b/decompress.go
@@ -31,7 +31,10 @@ func ExtractXZ(xFile *XFile) (size int64, filesList []string, err error) {
 	}
 
 	// Get the absolute path of the file being written.
-	wfile := xFile.clean(xFile.FilePath, ".xz")
+	wfile, err := xFile.clean(xFile.FilePath, ".xz")
+	if err != nil {
+		return 0, nil, err
+	}
 
 	size, err = writeFile(wfile, zipReader, xFile.FileMode, xFile.DirMode)
 	if err != nil {
@@ -56,7 +59,10 @@ func ExtractZlib(xFile *XFile) (size int64, filesList []string, err error) {
 	defer zipReader.Close()
 
 	// Get the absolute path of the file being written.
-	wfile := xFile.clean(xFile.FilePath, ".zz", ".zlib")
+	wfile, err := xFile.clean(xFile.FilePath, ".zz", ".zlib")
+	if err != nil {
+		return 0, nil, err
+	}
 
 	size, err = writeFile(wfile, zipReader, xFile.FileMode, xFile.DirMode)
 	if err != nil {
@@ -80,7 +86,10 @@ func ExtractLZMA(xFile *XFile) (size int64, filesList []string, err error) {
 	}
 
 	// Get the absolute path of the file being written.
-	wfile := xFile.clean(xFile.FilePath, ".lzma", ".lz", ".lzip")
+	wfile, err := xFile.clean(xFile.FilePath, ".lzma", ".lz", ".lzip")
+	if err != nil {
+		return 0, nil, err
+	}
 
 	size, err = writeFile(wfile, zipReader, xFile.FileMode, xFile.DirMode)
 	if err != nil {
@@ -104,7 +113,10 @@ func ExtractLZMA2(xFile *XFile) (size int64, filesList []string, err error) {
 	}
 
 	// Get the absolute path of the file being written.
-	wfile := xFile.clean(xFile.FilePath, ".lzma", ".lzma2")
+	wfile, err := xFile.clean(xFile.FilePath, ".lzma", ".lzma2")
+	if err != nil {
+		return 0, nil, err
+	}
 
 	size, err = writeFile(wfile, zipReader, xFile.FileMode, xFile.DirMode)
 	if err != nil {
@@ -129,7 +141,10 @@ func ExtractZstandard(xFile *XFile) (size int64, filesList []string, err error)
 	defer zipReader.Close()
 
 	// Get the absolute path of the file being written.
-	wfile := xFile.clean(xFile.FilePath, ".zstd", ".zst")
+	wfile, err := xFile.clean(xFile.FilePath, ".zstd", ".zst")
+	if err != nil {
+		return 0, nil, err
+	}
 
 	size, err = writeFile(wfile, zipReader, xFile.FileMode, xFile.DirMode)
 	if err != nil {
@@ -153,7 +168,10 @@ func ExtractLZW(xFile *XFile) (size int64, filesList []string, err error) {
 	}
 
 	// Get the absolute path of the file being written.
-	wfile := xFile.clean(xFile.FilePath, ".Z")
+	wfile, err := xFile.clean(xFile.FilePath, ".Z")
+	if err != nil {
+		return 0, nil, err
+	}
 
 	size, err = writeFile(wfile, zipReader, xFile.FileMode, xFile.DirMode)
 	if err != nil {
@@ -172,7 +190,10 @@ func ExtractLZ4(xFile *XFile) (size int64, filesList []string, err error) {
 	defer compressedFile.Close()
 
 	// Get the absolute path of the file being written.
-	wfile := xFile.clean(xFile.FilePath, ".lz4")
+	wfile, err := xFile.clean(xFile.FilePath, ".lz4")
+	if err != nil {
+		return 0, nil, err
+	}
 
 	size, err = writeFile(wfile, lz4.NewReader(compressedFile), xFile.FileMode, xFile.DirMode)
 	if err != nil {
@@ -191,7 +212,10 @@ func ExtractSnappy(xFile *XFile) (size int64, filesList []string, err error) {
 	defer compressedFile.Close()
 
 	// Get the absolute path of the file being written.
-	wfile := xFile.clean(xFile.FilePath, ".snappy", ".sz")
+	wfile, err := xFile.clean(xFile.FilePath, ".snappy", ".sz")
+	if err != nil {
+		return 0, nil, err
+	}
 
 	size, err = writeFile(wfile, snappy.NewReader(compressedFile), xFile.FileMode, xFile.DirMode)
 	if err != nil {
@@ -210,7 +234,10 @@ func ExtractS2(xFile *XFile) (size int64, filesList []string, err error) {
 	defer compressedFile.Close()
 
 	// Get the absolute path of the file being written.
-	wfile := xFile.clean(xFile.FilePath, ".s2")
+	wfile, err := xFile.clean(xFile.FilePath, ".s2")
+	if err != nil {
+		return 0, nil, err
+	}
 
 	size, err = writeFile(wfile, s2.NewReader(compressedFile), xFile.FileMode, xFile.DirMode)
 	if err != nil {
@@ -229,7 +256,10 @@ func ExtractBrotli(xFile *XFile) (size int64, filesList []string, err error) {
 	defer compressedFile.Close()
 
 	// Get the absolute path of the file being written.
-	wfile := xFile.clean(xFile.FilePath, ".brotli", ".br")
+	wfile, err := xFile.clean(xFile.FilePath, ".brotli", ".br")
+	if err != nil {
+		return 0, nil, err
+	}
 
 	size, err = writeFile(wfile, brotli.NewReader(compressedFile), xFile.FileMode, xFile.DirMode)
 	if err != nil {
@@ -248,7 +278,10 @@ func ExtractBzip(xFile *XFile) (size int64, filesList []string, err error) {
 	defer compressedFile.Close()
 
 	// Get the absolute path of the file being written.
-	wfile := xFile.clean(xFile.FilePath, ".bz", ".bz2")
+	wfile, err := xFile.clean(xFile.FilePath, ".bz", ".bz2")
+	if err != nil {
+		return 0, nil, err
+	}
 
 	size, err = writeFile(wfile, bzip2.NewReader(compressedFile), xFile.FileMode, xFile.DirMode)
 	if err != nil {
@@ -273,7 +306,10 @@ func ExtractGzip(xFile *XFile) (size int64, filesList []string, err error) {
 	defer zipReader.Close()
 
 	// Get the absolute path of the file being written.
-	wfile := xFile.clean(xFile.FilePath, ".gz")
+	wfile, err := xFile.clean(xFile.FilePath, ".gz")
+	if err != nil {
+		return 0, nil, err
+	}
 
 	size, err = writeFile(wfile, zipReader, xFile.FileMode, xFile.DirMode)
 	if err != nil {
diff --git a/files.go b/files.go
index 711f517..77a7f56 100644
--- a/files.go
+++ b/files.go
@@ -9,6 +9,8 @@ import (
 	"path/filepath"
 	"regexp"
 	"strings"
+
+	"golang.org/x/text/encoding"
 )
 
 // ArchiveList is the value returned when searching for compressed files.
@@ -111,6 +113,8 @@ type XFile struct {
 	Password string
 	// (RAR/7z) Archive passwords (to try multiple).
 	Passwords []string
+	// If file names are not UTF8 encoded, pass your own encoder here.
+	Encoding encoding.Encoding
 	// Logger allows printing debug messages.
 	log Logger
 }
@@ -446,8 +450,9 @@ func (x *Xtractr) Rename(oldpath, newpath string) error {
 }
 
 // clean returns an absolute path for a file inside the OutputDir.
+// clean also decodes the file name using a provided decoder.
 // If trim length is > 0, then the suffixes are trimmed, and filepath removed.
-func (x *XFile) clean(filePath string, trim ...string) string {
+func (x *XFile) clean(filePath string, trim ...string) (string, error) {
 	if len(trim) != 0 {
 		filePath = filepath.Base(filePath)
 		for _, suffix := range trim {
@@ -455,7 +460,26 @@ func (x *XFile) clean(filePath string, trim ...string) string {
 		}
 	}
 
-	return filepath.Clean(filepath.Join(x.OutputDir, filePath))
+	decoded, err := x.decode(filePath)
+	if err != nil {
+		return "", err
+	}
+
+	return filepath.Clean(filepath.Join(x.OutputDir, decoded)), nil
+}
+
+// decode a string using the provided decoder.
+func (x *XFile) decode(input string) (string, error) {
+	if x.Encoding == nil {
+		return input, nil
+	}
+
+	output, err := x.Encoding.NewDecoder().String(input)
+	if err != nil {
+		return "", fmt.Errorf("decoding file name: %w", err)
+	}
+
+	return output, nil
 }
 
 // AllExcept can be used as an input to ExcludeSuffix in a Filter.
diff --git a/iso.go b/iso.go
index 4137bde..0e756d1 100644
--- a/iso.go
+++ b/iso.go
@@ -67,8 +67,11 @@ func (x *XFile) uniso(isoFile *iso9660.File, parent string) (int64, []string, er
 	return size, files, nil
 }
 
-func (x *XFile) unisofile(isoFile *iso9660.File, wfile string) (int64, []string, error) {
-	wfile = x.clean(wfile)
+func (x *XFile) unisofile(isoFile *iso9660.File, fileName string) (int64, []string, error) {
+	wfile, err := x.clean(fileName)
+	if err != nil {
+		return 0, nil, err
+	}
 
 	//nolint:gocritic // this 1-argument filepath.Join removes a ./ prefix should there be one.
 	if !strings.HasPrefix(wfile, filepath.Join(x.OutputDir)) {
diff --git a/rar.go b/rar.go
index 166935d..0a02a6d 100644
--- a/rar.go
+++ b/rar.go
@@ -92,7 +92,11 @@ func (x *XFile) unrar(rarReader *rardecode.ReadCloser) (int64, []string, error)
 			return size, files, fmt.Errorf("%w: %s", ErrInvalidHead, x.FilePath)
 		}
 
-		wfile := x.clean(header.Name)
+		wfile, err := x.clean(header.Name)
+		if err != nil {
+			return 0, nil, err
+		}
+
 		//nolint:gocritic // this 1-argument filepath.Join removes a ./ prefix should there be one.
 		if !strings.HasPrefix(wfile, filepath.Join(x.OutputDir)) {
 			// The file being written is trying to write outside of our base path. Malicious archive?
diff --git a/tar.go b/tar.go
index 69f42fd..37aa804 100644
--- a/tar.go
+++ b/tar.go
@@ -119,7 +119,11 @@ func (x *XFile) untar(reader io.Reader) (int64, []string, error) {
 			return size, files, fmt.Errorf("%w: %s", ErrInvalidHead, x.FilePath)
 		}
 
-		wfile := x.clean(header.Name)
+		wfile, err := x.clean(header.Name)
+		if err != nil {
+			return 0, nil, err
+		}
+
 		if !strings.HasPrefix(wfile, x.OutputDir) {
 			// The file being written is trying to write outside of our base path. Malicious archive?
 			return size, files, fmt.Errorf("%s: %w: %s (from: %s)", x.FilePath, ErrInvalidPath, wfile, header.Name)
diff --git a/zip.go b/zip.go
index 9e35a9a..85ab46c 100644
--- a/zip.go
+++ b/zip.go
@@ -37,7 +37,11 @@ func ExtractZIP(xFile *XFile) (size int64, filesList []string, err error) {
 }
 
 func (x *XFile) unzip(zipFile *zip.File) (int64, string, error) {
-	wfile := x.clean(zipFile.Name)
+	wfile, err := x.clean(zipFile.Name)
+	if err != nil {
+		return 0, wfile, err
+	}
+
 	if !strings.HasPrefix(wfile, x.OutputDir) {
 		// The file being written is trying to write outside of our base path. Malicious archive?
 		return 0, wfile, fmt.Errorf("%s: %w: %s (from: %s)", zipFile.FileInfo().Name(), ErrInvalidPath, wfile, zipFile.Name)

From bb205fdbd2da8c0aafc4b0024be086ec41e85488 Mon Sep 17 00:00:00 2001
From: David Newhall II <david@sleepers.pro>
Date: Sat, 12 Apr 2025 17:13:19 -0700
Subject: [PATCH 2/5] attempt to detect charset using an oooold library

---
 files.go | 12 ++++++++++--
 go.mod   |  5 ++++-
 go.sum   |  8 +++++++-
 3 files changed, 21 insertions(+), 4 deletions(-)

diff --git a/files.go b/files.go
index 77a7f56..962caee 100644
--- a/files.go
+++ b/files.go
@@ -10,6 +10,8 @@ import (
 	"regexp"
 	"strings"
 
+	"github.com/saintfish/chardet"
+	"golang.org/x/net/html/charset"
 	"golang.org/x/text/encoding"
 )
 
@@ -470,11 +472,17 @@ func (x *XFile) clean(filePath string, trim ...string) (string, error) {
 
 // decode a string using the provided decoder.
 func (x *XFile) decode(input string) (string, error) {
-	if x.Encoding == nil {
+	encoding := x.Encoding
+	if encoding == nil {
+		res, _ := chardet.NewTextDetector().DetectBest([]byte(input))
+		encoding, _ = charset.Lookup(res.Charset)
+	}
+
+	if encoding == nil {
 		return input, nil
 	}
 
-	output, err := x.Encoding.NewDecoder().String(input)
+	output, err := encoding.NewDecoder().String(input)
 	if err != nil {
 		return "", fmt.Errorf("decoding file name: %w", err)
 	}
diff --git a/go.mod b/go.mod
index 9f2fdd1..b4c5bf4 100644
--- a/go.mod
+++ b/go.mod
@@ -15,10 +15,13 @@ require (
 	github.com/nwaples/rardecode v1.1.3
 	github.com/peterebden/ar v0.0.0-20241106141004-20dc11b778e8
 	github.com/pierrec/lz4/v4 v4.1.22
+	github.com/saintfish/chardet v0.0.0-20230101081208-5e3ef4b5456d
 	github.com/sshaman1101/dcompress v0.0.0-20200109162717-50436a6332de
 	github.com/stretchr/testify v1.10.0
 	github.com/therootcompany/xz v1.0.1
 	github.com/ulikunitz/xz v0.5.12
+	golang.org/x/net v0.39.0
+	golang.org/x/text v0.24.0
 )
 
 require (
@@ -28,9 +31,9 @@ require (
 	github.com/hashicorp/errwrap v1.1.0 // indirect
 	github.com/hashicorp/go-multierror v1.1.1 // indirect
 	github.com/hashicorp/golang-lru/v2 v2.0.7 // indirect
+	github.com/kr/text v0.2.0 // indirect
 	github.com/pmezard/go-difflib v1.0.0 // indirect
 	go4.org v0.0.0-20230225012048-214862532bf5 // indirect
 	golang.org/x/crypto v0.37.0 // indirect
-	golang.org/x/text v0.24.0 // indirect
 	gopkg.in/yaml.v3 v3.0.1 // indirect
 )
diff --git a/go.sum b/go.sum
index c70a504..ae26d59 100644
--- a/go.sum
+++ b/go.sum
@@ -34,6 +34,7 @@ github.com/chzyer/logex v1.1.10/go.mod h1:+Ywpsq7O8HXn0nuIou7OrIPyXbp3wmkHB+jjWR
 github.com/chzyer/readline v0.0.0-20180603132655-2972be24d48e/go.mod h1:nSuG5e5PlCu98SY8svDHJxuZscDgtXS6KTTbou5AhLI=
 github.com/chzyer/test v0.0.0-20180213035817-a1ea475d72b1/go.mod h1:Q3SI9o4m/ZMnBNeIyt5eFwwo7qiLfzFZmjNmxjkiQlU=
 github.com/client9/misspell v0.3.4/go.mod h1:qj6jICC3Q7zFZvVWo7KLAzC3yx5G7kyvSDkc90ppPyw=
+github.com/creack/pty v1.1.9/go.mod h1:oKZEueFk5CKHvIhNR5MUki03XCEU+Q6VDXinZuGJ33E=
 github.com/davecgh/go-spew v1.1.0/go.mod h1:J7Y8YcW2NihsgmVo/mv3lAwl/skON4iLHjSsI+c5H38=
 github.com/davecgh/go-spew v1.1.1 h1:vj9j/u1bqnvCEfJOwUhtlOARqs3+rkHYY13jYWTU97c=
 github.com/davecgh/go-spew v1.1.1/go.mod h1:J7Y8YcW2NihsgmVo/mv3lAwl/skON4iLHjSsI+c5H38=
@@ -91,8 +92,9 @@ github.com/klauspost/cpuid v1.2.0/go.mod h1:Pj4uuM528wm8OyEC2QMXAi2YiTZ96dNQPGgo
 github.com/kr/pretty v0.1.0 h1:L/CwN0zerZDmRFUapSPitk6f+Q3+0za1rQkzVuMiMFI=
 github.com/kr/pretty v0.1.0/go.mod h1:dAy3ld7l9f0ibDNOQOHHMYYIIbhfbHSm3C4ZsoJORNo=
 github.com/kr/pty v1.1.1/go.mod h1:pFQYn66WHrOpPYNljwOMqo10TkYh1fy3cYio2l3bCsQ=
-github.com/kr/text v0.1.0 h1:45sCR5RtlFHMR4UwH9sdQ5TC8v0qDQCHnXt+kaKSTVE=
 github.com/kr/text v0.1.0/go.mod h1:4Jbv+DJW3UT/LiOwJeYQe1efqtUx/iVham/4vfdArNI=
+github.com/kr/text v0.2.0 h1:5Nx0Ya0ZqY2ygV366QzturHI13Jq95ApcVaJBhpS+AY=
+github.com/kr/text v0.2.0/go.mod h1:eLer722TekiGuMkidMxC/pM04lWEeraHUUmBw8l2grE=
 github.com/nwaples/rardecode v1.1.3 h1:cWCaZwfM5H7nAD6PyEdcVnczzV8i/JtotnyW/dD9lEc=
 github.com/nwaples/rardecode v1.1.3/go.mod h1:5DzqNKiOdpKKBH87u8VlvAnPZMXcGRhxWkRpHbbfGS0=
 github.com/peterebden/ar v0.0.0-20241106141004-20dc11b778e8 h1:27L3dHkYbeWGU3/5NasAzVDgXG9QzlfKCvcl4cdNW6c=
@@ -104,6 +106,8 @@ github.com/pmezard/go-difflib v1.0.0/go.mod h1:iKH77koFhYxTK1pcRnkKkqfTogsbg7gZN
 github.com/prometheus/client_model v0.0.0-20190812154241-14fe0d1b01d4/go.mod h1:xMI15A0UPsDsEKsMN9yxemIoYk6Tm2C1GtYGdfGttqA=
 github.com/rogpeppe/go-internal v1.3.0/go.mod h1:M8bDsm7K2OlrFYOpmOWEs/qY81heoFRclV5y23lUDJ4=
 github.com/rwcarlsen/goexif v0.0.0-20190401172101-9e8deecbddbd/go.mod h1:hPqNNc0+uJM6H+SuU8sEs5K5IQeKccPqeSjfgcKGgPk=
+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/sshaman1101/dcompress v0.0.0-20200109162717-50436a6332de h1:uIeuAon/xwRdiZaCmEd5mocquesYkWCf71WBO7obTmA=
 github.com/sshaman1101/dcompress v0.0.0-20200109162717-50436a6332de/go.mod h1:XIUpD+1rteMazWrMFjNSpM6TocSHxDYXk6UEgBb5+F0=
 github.com/stretchr/objx v0.1.0/go.mod h1:HFkY916IF+rwdDfMAkV7OtwuqBVzrE8GR6GFx+wExME=
@@ -178,6 +182,8 @@ golang.org/x/net v0.0.0-20200202094626-16171245cfb2/go.mod h1:z5CRVTTTmAJ677TzLL
 golang.org/x/net v0.0.0-20210226172049-e18ecbb05110/go.mod h1:m0MpNAwzfU5UDzcl9v0D8zg8gWTRqZa9RBIspLL5mdg=
 golang.org/x/net v0.0.0-20220722155237-a158d28d115b/go.mod h1:XRhObCWvk6IyKnWLug+ECip1KBveYUHfp+8e9klMJ9c=
 golang.org/x/net v0.7.0/go.mod h1:2Tu9+aMcznHK/AK1HMvgo6xiTLG5rD5rZLDS+rp2Bjs=
+golang.org/x/net v0.39.0 h1:ZCu7HMWDxpXpaiKdhzIfaltL9Lp31x/3fCP11bc6/fY=
+golang.org/x/net v0.39.0/go.mod h1:X7NRbYVEA+ewNkCNyJ513WmMdQ3BineSwVtN2zD/d+E=
 golang.org/x/oauth2 v0.0.0-20180821212333-d2e6202438be/go.mod h1:N/0e6XlmueqKjAGxoOufVs8QHGRruUQn6yWY3a++T0U=
 golang.org/x/oauth2 v0.0.0-20190226205417-e64efc72b421/go.mod h1:gOpvHmFTYa4IltrdGE7lF6nIHvwfUNPOp7c8zoXwtLw=
 golang.org/x/oauth2 v0.0.0-20190604053449-0f29369cfe45/go.mod h1:gOpvHmFTYa4IltrdGE7lF6nIHvwfUNPOp7c8zoXwtLw=

From b33c9282111d3f169fc0ec11adccdb5bcd4b34b4 Mon Sep 17 00:00:00 2001
From: David Newhall II <david@sleepers.pro>
Date: Sat, 19 Apr 2025 10:45:34 -0700
Subject: [PATCH 3/5] Add encoder input interface

---
 encoding.go | 30 ++++++++++++++++++++++++++++++
 files.go    | 37 ++++++++-----------------------------
 go.mod      |  2 --
 go.sum      |  4 ----
 4 files changed, 38 insertions(+), 35 deletions(-)
 create mode 100644 encoding.go

diff --git a/encoding.go b/encoding.go
new file mode 100644
index 0000000..85ed061
--- /dev/null
+++ b/encoding.go
@@ -0,0 +1,30 @@
+package xtractr
+
+import "fmt"
+
+/* This file will surely grow when someone writes a proper character encoding detector. */
+
+// EncoderInput is used as input for a custom encoder procedure.
+type EncoderInput struct {
+	FileName string
+	XFile    *XFile
+}
+
+// decode a string using the provided decoder.
+func (x *XFile) decode(input string) (string, error) {
+	if x.Encoder == nil {
+		return input, nil
+	}
+
+	encoding := x.Encoder(&EncoderInput{FileName: input, XFile: x})
+	if encoding == nil {
+		return input, nil
+	}
+
+	output, err := encoding.String(input)
+	if err != nil {
+		return "", fmt.Errorf("decoding file name: %w", err)
+	}
+
+	return output, nil
+}
diff --git a/files.go b/files.go
index 962caee..716f368 100644
--- a/files.go
+++ b/files.go
@@ -10,8 +10,6 @@ import (
 	"regexp"
 	"strings"
 
-	"github.com/saintfish/chardet"
-	"golang.org/x/net/html/charset"
 	"golang.org/x/text/encoding"
 )
 
@@ -116,7 +114,8 @@ type XFile struct {
 	// (RAR/7z) Archive passwords (to try multiple).
 	Passwords []string
 	// If file names are not UTF8 encoded, pass your own encoder here.
-	Encoding encoding.Encoding
+	// Provide a function that takes in a file name and returns an encoder for it.
+	Encoder func(input *EncoderInput) *encoding.Decoder
 	// Logger allows printing debug messages.
 	log Logger
 }
@@ -455,6 +454,11 @@ func (x *Xtractr) Rename(oldpath, newpath string) error {
 // clean also decodes the file name using a provided decoder.
 // If trim length is > 0, then the suffixes are trimmed, and filepath removed.
 func (x *XFile) clean(filePath string, trim ...string) (string, error) {
+	filePath, err := x.decode(filePath)
+	if err != nil {
+		return "", err
+	}
+
 	if len(trim) != 0 {
 		filePath = filepath.Base(filePath)
 		for _, suffix := range trim {
@@ -462,32 +466,7 @@ func (x *XFile) clean(filePath string, trim ...string) (string, error) {
 		}
 	}
 
-	decoded, err := x.decode(filePath)
-	if err != nil {
-		return "", err
-	}
-
-	return filepath.Clean(filepath.Join(x.OutputDir, decoded)), nil
-}
-
-// decode a string using the provided decoder.
-func (x *XFile) decode(input string) (string, error) {
-	encoding := x.Encoding
-	if encoding == nil {
-		res, _ := chardet.NewTextDetector().DetectBest([]byte(input))
-		encoding, _ = charset.Lookup(res.Charset)
-	}
-
-	if encoding == nil {
-		return input, nil
-	}
-
-	output, err := encoding.NewDecoder().String(input)
-	if err != nil {
-		return "", fmt.Errorf("decoding file name: %w", err)
-	}
-
-	return output, nil
+	return filepath.Clean(filepath.Join(x.OutputDir, filePath)), nil
 }
 
 // AllExcept can be used as an input to ExcludeSuffix in a Filter.
diff --git a/go.mod b/go.mod
index b4c5bf4..fd1b207 100644
--- a/go.mod
+++ b/go.mod
@@ -15,12 +15,10 @@ require (
 	github.com/nwaples/rardecode v1.1.3
 	github.com/peterebden/ar v0.0.0-20241106141004-20dc11b778e8
 	github.com/pierrec/lz4/v4 v4.1.22
-	github.com/saintfish/chardet v0.0.0-20230101081208-5e3ef4b5456d
 	github.com/sshaman1101/dcompress v0.0.0-20200109162717-50436a6332de
 	github.com/stretchr/testify v1.10.0
 	github.com/therootcompany/xz v1.0.1
 	github.com/ulikunitz/xz v0.5.12
-	golang.org/x/net v0.39.0
 	golang.org/x/text v0.24.0
 )
 
diff --git a/go.sum b/go.sum
index ae26d59..2277822 100644
--- a/go.sum
+++ b/go.sum
@@ -106,8 +106,6 @@ github.com/pmezard/go-difflib v1.0.0/go.mod h1:iKH77koFhYxTK1pcRnkKkqfTogsbg7gZN
 github.com/prometheus/client_model v0.0.0-20190812154241-14fe0d1b01d4/go.mod h1:xMI15A0UPsDsEKsMN9yxemIoYk6Tm2C1GtYGdfGttqA=
 github.com/rogpeppe/go-internal v1.3.0/go.mod h1:M8bDsm7K2OlrFYOpmOWEs/qY81heoFRclV5y23lUDJ4=
 github.com/rwcarlsen/goexif v0.0.0-20190401172101-9e8deecbddbd/go.mod h1:hPqNNc0+uJM6H+SuU8sEs5K5IQeKccPqeSjfgcKGgPk=
-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/sshaman1101/dcompress v0.0.0-20200109162717-50436a6332de h1:uIeuAon/xwRdiZaCmEd5mocquesYkWCf71WBO7obTmA=
 github.com/sshaman1101/dcompress v0.0.0-20200109162717-50436a6332de/go.mod h1:XIUpD+1rteMazWrMFjNSpM6TocSHxDYXk6UEgBb5+F0=
 github.com/stretchr/objx v0.1.0/go.mod h1:HFkY916IF+rwdDfMAkV7OtwuqBVzrE8GR6GFx+wExME=
@@ -182,8 +180,6 @@ golang.org/x/net v0.0.0-20200202094626-16171245cfb2/go.mod h1:z5CRVTTTmAJ677TzLL
 golang.org/x/net v0.0.0-20210226172049-e18ecbb05110/go.mod h1:m0MpNAwzfU5UDzcl9v0D8zg8gWTRqZa9RBIspLL5mdg=
 golang.org/x/net v0.0.0-20220722155237-a158d28d115b/go.mod h1:XRhObCWvk6IyKnWLug+ECip1KBveYUHfp+8e9klMJ9c=
 golang.org/x/net v0.7.0/go.mod h1:2Tu9+aMcznHK/AK1HMvgo6xiTLG5rD5rZLDS+rp2Bjs=
-golang.org/x/net v0.39.0 h1:ZCu7HMWDxpXpaiKdhzIfaltL9Lp31x/3fCP11bc6/fY=
-golang.org/x/net v0.39.0/go.mod h1:X7NRbYVEA+ewNkCNyJ513WmMdQ3BineSwVtN2zD/d+E=
 golang.org/x/oauth2 v0.0.0-20180821212333-d2e6202438be/go.mod h1:N/0e6XlmueqKjAGxoOufVs8QHGRruUQn6yWY3a++T0U=
 golang.org/x/oauth2 v0.0.0-20190226205417-e64efc72b421/go.mod h1:gOpvHmFTYa4IltrdGE7lF6nIHvwfUNPOp7c8zoXwtLw=
 golang.org/x/oauth2 v0.0.0-20190604053449-0f29369cfe45/go.mod h1:gOpvHmFTYa4IltrdGE7lF6nIHvwfUNPOp7c8zoXwtLw=

From ccca0adf07c201d2749337bcdf4b86f49115842a Mon Sep 17 00:00:00 2001
From: David Newhall II <david@sleepers.pro>
Date: Sat, 19 Apr 2025 11:02:40 -0700
Subject: [PATCH 4/5] consistency

---
 7z.go   | 10 ++++------
 cpio.go | 11 +++++------
 2 files changed, 9 insertions(+), 12 deletions(-)

diff --git a/7z.go b/7z.go
index 2a88e99..5785c07 100644
--- a/7z.go
+++ b/7z.go
@@ -91,13 +91,7 @@ func (x *XFile) un7zip(zipFile *sevenzip.File) (int64, string, error) {
 	}
 	defer zFile.Close()
 
-	wfile, err := x.clean(zipFile.Name)
-	if err != nil {
-		return 0, wfile, err
-	}
-
 	file := &file{
-		Path:     wfile,
 		Data:     zFile,
 		FileMode: zipFile.Mode(),
 		DirMode:  x.DirMode,
@@ -105,6 +99,10 @@ func (x *XFile) un7zip(zipFile *sevenzip.File) (int64, string, error) {
 		Atime:    zipFile.Accessed,
 	}
 
+	if file.Path, err = x.clean(zipFile.Name); err != nil {
+		return 0, file.Path, err
+	}
+
 	if !strings.HasPrefix(file.Path, x.OutputDir) {
 		// The file being written is trying to write outside of our base path. Malicious archive?
 		err := fmt.Errorf("%s: %w: %s (from: %s)", zipFile.FileInfo().Name(), ErrInvalidPath, file.Path, zipFile.Name)
diff --git a/cpio.go b/cpio.go
index 0ab566c..ced8b1d 100644
--- a/cpio.go
+++ b/cpio.go
@@ -64,19 +64,18 @@ func (x *XFile) uncpio(reader io.Reader) (int64, []string, error) {
 }
 
 func (x *XFile) uncpioFile(cpioFile *cpio.Header, cpioReader *cpio.Reader) (int64, error) {
-	wfile, err := x.clean(cpioFile.Name)
-	if err != nil {
-		return 0, err
-	}
-
 	file := &file{
-		Path:     wfile,
 		Data:     cpioReader,
 		FileMode: cpioFile.FileInfo().Mode(),
 		DirMode:  x.DirMode,
 		Mtime:    cpioFile.ModTime,
 	}
 
+	var err error
+	if file.Path, err = x.clean(cpioFile.Name); err != nil {
+		return 0, err
+	}
+
 	if !strings.HasPrefix(file.Path, x.OutputDir) {
 		// The file being written is trying to write outside of the base path. Malicious archive?
 		return 0, fmt.Errorf("%s: %w: %s (from: %s)", cpioFile.FileInfo().Name(), ErrInvalidPath, file.Path, cpioFile.Name)

From 1939370795cda207662cc634cfc2862a4fade752 Mon Sep 17 00:00:00 2001
From: David Newhall II <david@sleepers.pro>
Date: Sat, 19 Apr 2025 11:46:03 -0700
Subject: [PATCH 5/5] remove name

---
 files.go | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/files.go b/files.go
index 8fadce5..a757152 100644
--- a/files.go
+++ b/files.go
@@ -116,7 +116,7 @@ type XFile struct {
 	Passwords []string
 	// If file names are not UTF8 encoded, pass your own encoder here.
 	// Provide a function that takes in a file name and returns an encoder for it.
-	Encoder func(input *EncoderInput) *encoding.Decoder
+	Encoder func(*EncoderInput) *encoding.Decoder
 	// If the archive only has one directory in the root, then setting
 	// this true will cause the extracted content to be moved into the
 	// output folder, and the root folder in the archive to be removed.