Home | History | Annotate | Download | only in zlib
      1 // Copyright 2009 The Go Authors. All rights reserved.
      2 // Use of this source code is governed by a BSD-style
      3 // license that can be found in the LICENSE file.
      4 
      5 /*
      6 Package zlib implements reading and writing of zlib format compressed data,
      7 as specified in RFC 1950.
      8 
      9 The implementation provides filters that uncompress during reading
     10 and compress during writing.  For example, to write compressed data
     11 to a buffer:
     12 
     13 	var b bytes.Buffer
     14 	w := zlib.NewWriter(&b)
     15 	w.Write([]byte("hello, world\n"))
     16 	w.Close()
     17 
     18 and to read that data back:
     19 
     20 	r, err := zlib.NewReader(&b)
     21 	io.Copy(os.Stdout, r)
     22 	r.Close()
     23 */
     24 package zlib
     25 
     26 import (
     27 	"bufio"
     28 	"compress/flate"
     29 	"errors"
     30 	"hash"
     31 	"hash/adler32"
     32 	"io"
     33 )
     34 
     35 const zlibDeflate = 8
     36 
     37 var (
     38 	// ErrChecksum is returned when reading ZLIB data that has an invalid checksum.
     39 	ErrChecksum = errors.New("zlib: invalid checksum")
     40 	// ErrDictionary is returned when reading ZLIB data that has an invalid dictionary.
     41 	ErrDictionary = errors.New("zlib: invalid dictionary")
     42 	// ErrHeader is returned when reading ZLIB data that has an invalid header.
     43 	ErrHeader = errors.New("zlib: invalid header")
     44 )
     45 
     46 type reader struct {
     47 	r            flate.Reader
     48 	decompressor io.ReadCloser
     49 	digest       hash.Hash32
     50 	err          error
     51 	scratch      [4]byte
     52 }
     53 
     54 // Resetter resets a ReadCloser returned by NewReader or NewReaderDict to
     55 // to switch to a new underlying Reader. This permits reusing a ReadCloser
     56 // instead of allocating a new one.
     57 type Resetter interface {
     58 	// Reset discards any buffered data and resets the Resetter as if it was
     59 	// newly initialized with the given reader.
     60 	Reset(r io.Reader, dict []byte) error
     61 }
     62 
     63 // NewReader creates a new ReadCloser.
     64 // Reads from the returned ReadCloser read and decompress data from r.
     65 // The implementation buffers input and may read more data than necessary from r.
     66 // It is the caller's responsibility to call Close on the ReadCloser when done.
     67 //
     68 // The ReadCloser returned by NewReader also implements Resetter.
     69 func NewReader(r io.Reader) (io.ReadCloser, error) {
     70 	return NewReaderDict(r, nil)
     71 }
     72 
     73 // NewReaderDict is like NewReader but uses a preset dictionary.
     74 // NewReaderDict ignores the dictionary if the compressed data does not refer to it.
     75 // If the compressed data refers to a different dictionary, NewReaderDict returns ErrDictionary.
     76 //
     77 // The ReadCloser returned by NewReaderDict also implements Resetter.
     78 func NewReaderDict(r io.Reader, dict []byte) (io.ReadCloser, error) {
     79 	z := new(reader)
     80 	err := z.Reset(r, dict)
     81 	if err != nil {
     82 		return nil, err
     83 	}
     84 	return z, nil
     85 }
     86 
     87 func (z *reader) Read(p []byte) (n int, err error) {
     88 	if z.err != nil {
     89 		return 0, z.err
     90 	}
     91 	if len(p) == 0 {
     92 		return 0, nil
     93 	}
     94 
     95 	n, err = z.decompressor.Read(p)
     96 	z.digest.Write(p[0:n])
     97 	if n != 0 || err != io.EOF {
     98 		z.err = err
     99 		return
    100 	}
    101 
    102 	// Finished file; check checksum.
    103 	if _, err := io.ReadFull(z.r, z.scratch[0:4]); err != nil {
    104 		z.err = err
    105 		return 0, err
    106 	}
    107 	// ZLIB (RFC 1950) is big-endian, unlike GZIP (RFC 1952).
    108 	checksum := uint32(z.scratch[0])<<24 | uint32(z.scratch[1])<<16 | uint32(z.scratch[2])<<8 | uint32(z.scratch[3])
    109 	if checksum != z.digest.Sum32() {
    110 		z.err = ErrChecksum
    111 		return 0, z.err
    112 	}
    113 	return
    114 }
    115 
    116 // Calling Close does not close the wrapped io.Reader originally passed to NewReader.
    117 func (z *reader) Close() error {
    118 	if z.err != nil {
    119 		return z.err
    120 	}
    121 	z.err = z.decompressor.Close()
    122 	return z.err
    123 }
    124 
    125 func (z *reader) Reset(r io.Reader, dict []byte) error {
    126 	if fr, ok := r.(flate.Reader); ok {
    127 		z.r = fr
    128 	} else {
    129 		z.r = bufio.NewReader(r)
    130 	}
    131 	_, err := io.ReadFull(z.r, z.scratch[0:2])
    132 	if err != nil {
    133 		return err
    134 	}
    135 	h := uint(z.scratch[0])<<8 | uint(z.scratch[1])
    136 	if (z.scratch[0]&0x0f != zlibDeflate) || (h%31 != 0) {
    137 		return ErrHeader
    138 	}
    139 	haveDict := z.scratch[1]&0x20 != 0
    140 	if haveDict {
    141 		_, err = io.ReadFull(z.r, z.scratch[0:4])
    142 		if err != nil {
    143 			return err
    144 		}
    145 		checksum := uint32(z.scratch[0])<<24 | uint32(z.scratch[1])<<16 | uint32(z.scratch[2])<<8 | uint32(z.scratch[3])
    146 		if checksum != adler32.Checksum(dict) {
    147 			return ErrDictionary
    148 		}
    149 	}
    150 	if z.decompressor == nil {
    151 		if haveDict {
    152 			z.decompressor = flate.NewReaderDict(z.r, dict)
    153 		} else {
    154 			z.decompressor = flate.NewReader(z.r)
    155 		}
    156 	} else {
    157 		z.decompressor.(flate.Resetter).Reset(z.r, dict)
    158 	}
    159 	z.digest = adler32.New()
    160 	return nil
    161 }
    162