storj.io/minio@v0.0.0-20230509071714-0cbc90f649b1/cmd/untar.go (about)

     1  /*
     2   * MinIO Cloud Storage, (C) 2021 MinIO, Inc.
     3   *
     4   * Licensed under the Apache License, Version 2.0 (the "License");
     5   * you may not use this file except in compliance with the License.
     6   * You may obtain a copy of the License at
     7   *
     8   *     http://www.apache.org/licenses/LICENSE-2.0
     9   *
    10   * Unless required by applicable law or agreed to in writing, software
    11   * distributed under the License is distributed on an "AS IS" BASIS,
    12   * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
    13   * See the License for the specific language governing permissions and
    14   * limitations under the License.
    15   */
    16  
    17  package cmd
    18  
    19  import (
    20  	"archive/tar"
    21  	"bufio"
    22  	"bytes"
    23  	"compress/bzip2"
    24  	"fmt"
    25  	"io"
    26  	"os"
    27  	"path"
    28  
    29  	"github.com/klauspost/compress/s2"
    30  	"github.com/klauspost/compress/zstd"
    31  	gzip "github.com/klauspost/pgzip"
    32  	"github.com/pierrec/lz4"
    33  )
    34  
    35  func detect(r *bufio.Reader) format {
    36  	z, err := r.Peek(4)
    37  	if err != nil {
    38  		return formatUnknown
    39  	}
    40  	for _, f := range magicHeaders {
    41  		if bytes.Equal(f.header, z[:len(f.header)]) {
    42  			return f.f
    43  		}
    44  	}
    45  	return formatUnknown
    46  }
    47  
    48  //go:generate stringer -type=format -trimprefix=format $GOFILE
    49  type format int
    50  
    51  const (
    52  	formatUnknown format = iota
    53  	formatGzip
    54  	formatZstd
    55  	formatLZ4
    56  	formatS2
    57  	formatBZ2
    58  )
    59  
    60  var magicHeaders = []struct {
    61  	header []byte
    62  	f      format
    63  }{
    64  	{
    65  		header: []byte{0x1f, 0x8b, 8},
    66  		f:      formatGzip,
    67  	},
    68  	{
    69  		// Zstd default header.
    70  		header: []byte{0x28, 0xb5, 0x2f, 0xfd},
    71  		f:      formatZstd,
    72  	},
    73  	{
    74  		// Zstd skippable frame header.
    75  		header: []byte{0x2a, 0x4d, 0x18},
    76  		f:      formatZstd,
    77  	},
    78  	{
    79  		// LZ4
    80  		header: []byte{0x4, 0x22, 0x4d, 0x18},
    81  		f:      formatLZ4,
    82  	},
    83  	{
    84  		// Snappy/S2 stream
    85  		header: []byte{0xff, 0x06, 0x00, 0x00},
    86  		f:      formatS2,
    87  	},
    88  	{
    89  		header: []byte{0x42, 0x5a, 'h'},
    90  		f:      formatBZ2,
    91  	},
    92  }
    93  
    94  func untar(r io.Reader, putObject func(reader io.Reader, info os.FileInfo, name string)) error {
    95  	bf := bufio.NewReader(r)
    96  	switch f := detect(bf); f {
    97  	case formatGzip:
    98  		gz, err := gzip.NewReader(bf)
    99  		if err != nil {
   100  			return err
   101  		}
   102  		defer gz.Close()
   103  		r = gz
   104  	case formatS2:
   105  		r = s2.NewReader(bf)
   106  	case formatZstd:
   107  		dec, err := zstd.NewReader(bf)
   108  		if err != nil {
   109  			return err
   110  		}
   111  		defer dec.Close()
   112  		r = dec
   113  	case formatBZ2:
   114  		r = bzip2.NewReader(bf)
   115  	case formatLZ4:
   116  		r = lz4.NewReader(bf)
   117  	case formatUnknown:
   118  		r = bf
   119  	default:
   120  		return fmt.Errorf("Unsupported format %s", f)
   121  	}
   122  	tarReader := tar.NewReader(r)
   123  	for {
   124  		header, err := tarReader.Next()
   125  
   126  		switch {
   127  
   128  		// if no more files are found return
   129  		case err == io.EOF:
   130  			return nil
   131  
   132  		// return any other error
   133  		case err != nil:
   134  			return err
   135  
   136  		// if the header is nil, just skip it (not sure how this happens)
   137  		case header == nil:
   138  			continue
   139  		}
   140  
   141  		name := header.Name
   142  		if name == slashSeparator {
   143  			continue
   144  		}
   145  
   146  		switch header.Typeflag {
   147  		case tar.TypeDir: // = directory
   148  			putObject(tarReader, header.FileInfo(), trimLeadingSlash(pathJoin(name, slashSeparator)))
   149  		case tar.TypeReg, tar.TypeChar, tar.TypeBlock, tar.TypeFifo, tar.TypeGNUSparse: // = regular
   150  			putObject(tarReader, header.FileInfo(), trimLeadingSlash(path.Clean(name)))
   151  		default:
   152  			// ignore symlink'ed
   153  			continue
   154  		}
   155  	}
   156  }