Source file src/compress/flate/inflate.go

     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  // Package flate implements the DEFLATE compressed data format, described in
     6  // RFC 1951.  The gzip and zlib packages implement access to DEFLATE-based file
     7  // formats.
     8  package flate
     9  
    10  import (
    11  	"bufio"
    12  	"io"
    13  	"math/bits"
    14  	"strconv"
    15  	"sync"
    16  )
    17  
    18  const (
    19  	maxCodeLen = 16 // max length of Huffman code
    20  	// The next three numbers come from the RFC section 3.2.7, with the
    21  	// additional proviso in section 3.2.5 which implies that distance codes
    22  	// 30 and 31 should never occur in compressed data.
    23  	maxNumLit  = 286
    24  	maxNumDist = 30
    25  	numCodes   = 19 // number of codes in Huffman meta-code
    26  )
    27  
    28  // Initialize the fixedHuffmanDecoder only once upon first use.
    29  var fixedOnce sync.Once
    30  var fixedHuffmanDecoder huffmanDecoder
    31  
    32  // A CorruptInputError reports the presence of corrupt input at a given offset.
    33  type CorruptInputError int64
    34  
    35  func (e CorruptInputError) Error() string {
    36  	return "flate: corrupt input before offset " + strconv.FormatInt(int64(e), 10)
    37  }
    38  
    39  // An InternalError reports an error in the flate code itself.
    40  type InternalError string
    41  
    42  func (e InternalError) Error() string { return "flate: internal error: " + string(e) }
    43  
    44  // A ReadError reports an error encountered while reading input.
    45  //
    46  // Deprecated: No longer returned.
    47  type ReadError struct {
    48  	Offset int64 // byte offset where error occurred
    49  	Err    error // error returned by underlying Read
    50  }
    51  
    52  func (e *ReadError) Error() string {
    53  	return "flate: read error at offset " + strconv.FormatInt(e.Offset, 10) + ": " + e.Err.Error()
    54  }
    55  
    56  // A WriteError reports an error encountered while writing output.
    57  //
    58  // Deprecated: No longer returned.
    59  type WriteError struct {
    60  	Offset int64 // byte offset where error occurred
    61  	Err    error // error returned by underlying Write
    62  }
    63  
    64  func (e *WriteError) Error() string {
    65  	return "flate: write error at offset " + strconv.FormatInt(e.Offset, 10) + ": " + e.Err.Error()
    66  }
    67  
    68  // Resetter resets a ReadCloser returned by NewReader or NewReaderDict
    69  // to switch to a new underlying Reader. This permits reusing a ReadCloser
    70  // instead of allocating a new one.
    71  type Resetter interface {
    72  	// Reset discards any buffered data and resets the Resetter as if it was
    73  	// newly initialized with the given reader.
    74  	Reset(r io.Reader, dict []byte) error
    75  }
    76  
    77  // The data structure for decoding Huffman tables is based on that of
    78  // zlib. There is a lookup table of a fixed bit width (huffmanChunkBits),
    79  // For codes smaller than the table width, there are multiple entries
    80  // (each combination of trailing bits has the same value). For codes
    81  // larger than the table width, the table contains a link to an overflow
    82  // table. The width of each entry in the link table is the maximum code
    83  // size minus the chunk width.
    84  //
    85  // Note that you can do a lookup in the table even without all bits
    86  // filled. Since the extra bits are zero, and the DEFLATE Huffman codes
    87  // have the property that shorter codes come before longer ones, the
    88  // bit length estimate in the result is a lower bound on the actual
    89  // number of bits.
    90  //
    91  // See the following:
    92  //	https://github.com/madler/zlib/raw/master/doc/algorithm.txt
    93  
    94  // chunk & 15 is number of bits
    95  // chunk >> 4 is value, including table link
    96  
    97  const (
    98  	huffmanChunkBits  = 9
    99  	huffmanNumChunks  = 1 << huffmanChunkBits
   100  	huffmanCountMask  = 15
   101  	huffmanValueShift = 4
   102  )
   103  
   104  type huffmanDecoder struct {
   105  	min      int                      // the minimum code length
   106  	chunks   [huffmanNumChunks]uint32 // chunks as described above
   107  	links    [][]uint32               // overflow links
   108  	linkMask uint32                   // mask the width of the link table
   109  }
   110  
   111  // Initialize Huffman decoding tables from array of code lengths.
   112  // Following this function, h is guaranteed to be initialized into a complete
   113  // tree (i.e., neither over-subscribed nor under-subscribed). The exception is a
   114  // degenerate case where the tree has only a single symbol with length 1. Empty
   115  // trees are permitted.
   116  func (h *huffmanDecoder) init(lengths []int) bool {
   117  	// Sanity enables additional runtime tests during Huffman
   118  	// table construction. It's intended to be used during
   119  	// development to supplement the currently ad-hoc unit tests.
   120  	const sanity = false
   121  
   122  	if h.min != 0 {
   123  		*h = huffmanDecoder{}
   124  	}
   125  
   126  	// Count number of codes of each length,
   127  	// compute min and max length.
   128  	var count [maxCodeLen]int
   129  	var min, max int
   130  	for _, n := range lengths {
   131  		if n == 0 {
   132  			continue
   133  		}
   134  		if min == 0 || n < min {
   135  			min = n
   136  		}
   137  		if n > max {
   138  			max = n
   139  		}
   140  		count[n]++
   141  	}
   142  
   143  	// Empty tree. The decompressor.huffSym function will fail later if the tree
   144  	// is used. Technically, an empty tree is only valid for the HDIST tree and
   145  	// not the HCLEN and HLIT tree. However, a stream with an empty HCLEN tree
   146  	// is guaranteed to fail since it will attempt to use the tree to decode the
   147  	// codes for the HLIT and HDIST trees. Similarly, an empty HLIT tree is
   148  	// guaranteed to fail later since the compressed data section must be
   149  	// composed of at least one symbol (the end-of-block marker).
   150  	if max == 0 {
   151  		return true
   152  	}
   153  
   154  	code := 0
   155  	var nextcode [maxCodeLen]int
   156  	for i := min; i <= max; i++ {
   157  		code <<= 1
   158  		nextcode[i] = code
   159  		code += count[i]
   160  	}
   161  
   162  	// Check that the coding is complete (i.e., that we've
   163  	// assigned all 2-to-the-max possible bit sequences).
   164  	// Exception: To be compatible with zlib, we also need to
   165  	// accept degenerate single-code codings. See also
   166  	// TestDegenerateHuffmanCoding.
   167  	if code != 1<<uint(max) && !(code == 1 && max == 1) {
   168  		return false
   169  	}
   170  
   171  	h.min = min
   172  	if max > huffmanChunkBits {
   173  		numLinks := 1 << (uint(max) - huffmanChunkBits)
   174  		h.linkMask = uint32(numLinks - 1)
   175  
   176  		// create link tables
   177  		link := nextcode[huffmanChunkBits+1] >> 1
   178  		h.links = make([][]uint32, huffmanNumChunks-link)
   179  		for j := uint(link); j < huffmanNumChunks; j++ {
   180  			reverse := int(bits.Reverse16(uint16(j)))
   181  			reverse >>= uint(16 - huffmanChunkBits)
   182  			off := j - uint(link)
   183  			if sanity && h.chunks[reverse] != 0 {
   184  				panic("impossible: overwriting existing chunk")
   185  			}
   186  			h.chunks[reverse] = uint32(off<<huffmanValueShift | (huffmanChunkBits + 1))
   187  			h.links[off] = make([]uint32, numLinks)
   188  		}
   189  	}
   190  
   191  	for i, n := range lengths {
   192  		if n == 0 {
   193  			continue
   194  		}
   195  		code := nextcode[n]
   196  		nextcode[n]++
   197  		chunk := uint32(i<<huffmanValueShift | n)
   198  		reverse := int(bits.Reverse16(uint16(code)))
   199  		reverse >>= uint(16 - n)
   200  		if n <= huffmanChunkBits {
   201  			for off := reverse; off < len(h.chunks); off += 1 << uint(n) {
   202  				// We should never need to overwrite
   203  				// an existing chunk. Also, 0 is
   204  				// never a valid chunk, because the
   205  				// lower 4 "count" bits should be
   206  				// between 1 and 15.
   207  				if sanity && h.chunks[off] != 0 {
   208  					panic("impossible: overwriting existing chunk")
   209  				}
   210  				h.chunks[off] = chunk
   211  			}
   212  		} else {
   213  			j := reverse & (huffmanNumChunks - 1)
   214  			if sanity && h.chunks[j]&huffmanCountMask != huffmanChunkBits+1 {
   215  				// Longer codes should have been
   216  				// associated with a link table above.
   217  				panic("impossible: not an indirect chunk")
   218  			}
   219  			value := h.chunks[j] >> huffmanValueShift
   220  			linktab := h.links[value]
   221  			reverse >>= huffmanChunkBits
   222  			for off := reverse; off < len(linktab); off += 1 << uint(n-huffmanChunkBits) {
   223  				if sanity && linktab[off] != 0 {
   224  					panic("impossible: overwriting existing chunk")
   225  				}
   226  				linktab[off] = chunk
   227  			}
   228  		}
   229  	}
   230  
   231  	if sanity {
   232  		// Above we've sanity checked that we never overwrote
   233  		// an existing entry. Here we additionally check that
   234  		// we filled the tables completely.
   235  		for i, chunk := range h.chunks {
   236  			if chunk == 0 {
   237  				// As an exception, in the degenerate
   238  				// single-code case, we allow odd
   239  				// chunks to be missing.
   240  				if code == 1 && i%2 == 1 {
   241  					continue
   242  				}
   243  				panic("impossible: missing chunk")
   244  			}
   245  		}
   246  		for _, linktab := range h.links {
   247  			for _, chunk := range linktab {
   248  				if chunk == 0 {
   249  					panic("impossible: missing chunk")
   250  				}
   251  			}
   252  		}
   253  	}
   254  
   255  	return true
   256  }
   257  
   258  // The actual read interface needed by NewReader.
   259  // If the passed in io.Reader does not also have ReadByte,
   260  // the NewReader will introduce its own buffering.
   261  type Reader interface {
   262  	io.Reader
   263  	io.ByteReader
   264  }
   265  
   266  // Decompress state.
   267  type decompressor struct {
   268  	// Input source.
   269  	r       Reader
   270  	roffset int64
   271  
   272  	// Input bits, in top of b.
   273  	b  uint32
   274  	nb uint
   275  
   276  	// Huffman decoders for literal/length, distance.
   277  	h1, h2 huffmanDecoder
   278  
   279  	// Length arrays used to define Huffman codes.
   280  	bits     *[maxNumLit + maxNumDist]int
   281  	codebits *[numCodes]int
   282  
   283  	// Output history, buffer.
   284  	dict dictDecoder
   285  
   286  	// Temporary buffer (avoids repeated allocation).
   287  	buf [4]byte
   288  
   289  	// Next step in the decompression,
   290  	// and decompression state.
   291  	step      func(*decompressor)
   292  	stepState int
   293  	final     bool
   294  	err       error
   295  	toRead    []byte
   296  	hl, hd    *huffmanDecoder
   297  	copyLen   int
   298  	copyDist  int
   299  }
   300  
   301  func (f *decompressor) nextBlock() {
   302  	for f.nb < 1+2 {
   303  		if f.err = f.moreBits(); f.err != nil {
   304  			return
   305  		}
   306  	}
   307  	f.final = f.b&1 == 1
   308  	f.b >>= 1
   309  	typ := f.b & 3
   310  	f.b >>= 2
   311  	f.nb -= 1 + 2
   312  	switch typ {
   313  	case 0:
   314  		f.dataBlock()
   315  	case 1:
   316  		// compressed, fixed Huffman tables
   317  		f.hl = &fixedHuffmanDecoder
   318  		f.hd = nil
   319  		f.huffmanBlock()
   320  	case 2:
   321  		// compressed, dynamic Huffman tables
   322  		if f.err = f.readHuffman(); f.err != nil {
   323  			break
   324  		}
   325  		f.hl = &f.h1
   326  		f.hd = &f.h2
   327  		f.huffmanBlock()
   328  	default:
   329  		// 3 is reserved.
   330  		f.err = CorruptInputError(f.roffset)
   331  	}
   332  }
   333  
   334  func (f *decompressor) Read(b []byte) (int, error) {
   335  	for {
   336  		if len(f.toRead) > 0 {
   337  			n := copy(b, f.toRead)
   338  			f.toRead = f.toRead[n:]
   339  			if len(f.toRead) == 0 {
   340  				return n, f.err
   341  			}
   342  			return n, nil
   343  		}
   344  		if f.err != nil {
   345  			return 0, f.err
   346  		}
   347  		f.step(f)
   348  		if f.err != nil && len(f.toRead) == 0 {
   349  			f.toRead = f.dict.readFlush() // Flush what's left in case of error
   350  		}
   351  	}
   352  }
   353  
   354  func (f *decompressor) Close() error {
   355  	if f.err == io.EOF {
   356  		return nil
   357  	}
   358  	return f.err
   359  }
   360  
   361  // RFC 1951 section 3.2.7.
   362  // Compression with dynamic Huffman codes
   363  
   364  var codeOrder = [...]int{16, 17, 18, 0, 8, 7, 9, 6, 10, 5, 11, 4, 12, 3, 13, 2, 14, 1, 15}
   365  
   366  func (f *decompressor) readHuffman() error {
   367  	// HLIT[5], HDIST[5], HCLEN[4].
   368  	for f.nb < 5+5+4 {
   369  		if err := f.moreBits(); err != nil {
   370  			return err
   371  		}
   372  	}
   373  	nlit := int(f.b&0x1F) + 257
   374  	if nlit > maxNumLit {
   375  		return CorruptInputError(f.roffset)
   376  	}
   377  	f.b >>= 5
   378  	ndist := int(f.b&0x1F) + 1
   379  	if ndist > maxNumDist {
   380  		return CorruptInputError(f.roffset)
   381  	}
   382  	f.b >>= 5
   383  	nclen := int(f.b&0xF) + 4
   384  	// numCodes is 19, so nclen is always valid.
   385  	f.b >>= 4
   386  	f.nb -= 5 + 5 + 4
   387  
   388  	// (HCLEN+4)*3 bits: code lengths in the magic codeOrder order.
   389  	for i := 0; i < nclen; i++ {
   390  		for f.nb < 3 {
   391  			if err := f.moreBits(); err != nil {
   392  				return err
   393  			}
   394  		}
   395  		f.codebits[codeOrder[i]] = int(f.b & 0x7)
   396  		f.b >>= 3
   397  		f.nb -= 3
   398  	}
   399  	for i := nclen; i < len(codeOrder); i++ {
   400  		f.codebits[codeOrder[i]] = 0
   401  	}
   402  	if !f.h1.init(f.codebits[0:]) {
   403  		return CorruptInputError(f.roffset)
   404  	}
   405  
   406  	// HLIT + 257 code lengths, HDIST + 1 code lengths,
   407  	// using the code length Huffman code.
   408  	for i, n := 0, nlit+ndist; i < n; {
   409  		x, err := f.huffSym(&f.h1)
   410  		if err != nil {
   411  			return err
   412  		}
   413  		if x < 16 {
   414  			// Actual length.
   415  			f.bits[i] = x
   416  			i++
   417  			continue
   418  		}
   419  		// Repeat previous length or zero.
   420  		var rep int
   421  		var nb uint
   422  		var b int
   423  		switch x {
   424  		default:
   425  			return InternalError("unexpected length code")
   426  		case 16:
   427  			rep = 3
   428  			nb = 2
   429  			if i == 0 {
   430  				return CorruptInputError(f.roffset)
   431  			}
   432  			b = f.bits[i-1]
   433  		case 17:
   434  			rep = 3
   435  			nb = 3
   436  			b = 0
   437  		case 18:
   438  			rep = 11
   439  			nb = 7
   440  			b = 0
   441  		}
   442  		for f.nb < nb {
   443  			if err := f.moreBits(); err != nil {
   444  				return err
   445  			}
   446  		}
   447  		rep += int(f.b & uint32(1<<nb-1))
   448  		f.b >>= nb
   449  		f.nb -= nb
   450  		if i+rep > n {
   451  			return CorruptInputError(f.roffset)
   452  		}
   453  		for j := 0; j < rep; j++ {
   454  			f.bits[i] = b
   455  			i++
   456  		}
   457  	}
   458  
   459  	if !f.h1.init(f.bits[0:nlit]) || !f.h2.init(f.bits[nlit:nlit+ndist]) {
   460  		return CorruptInputError(f.roffset)
   461  	}
   462  
   463  	// As an optimization, we can initialize the min bits to read at a time
   464  	// for the HLIT tree to the length of the EOB marker since we know that
   465  	// every block must terminate with one. This preserves the property that
   466  	// we never read any extra bytes after the end of the DEFLATE stream.
   467  	if f.h1.min < f.bits[endBlockMarker] {
   468  		f.h1.min = f.bits[endBlockMarker]
   469  	}
   470  
   471  	return nil
   472  }
   473  
   474  // Decode a single Huffman block from f.
   475  // hl and hd are the Huffman states for the lit/length values
   476  // and the distance values, respectively. If hd == nil, using the
   477  // fixed distance encoding associated with fixed Huffman blocks.
   478  func (f *decompressor) huffmanBlock() {
   479  	const (
   480  		stateInit = iota // Zero value must be stateInit
   481  		stateDict
   482  	)
   483  
   484  	switch f.stepState {
   485  	case stateInit:
   486  		goto readLiteral
   487  	case stateDict:
   488  		goto copyHistory
   489  	}
   490  
   491  readLiteral:
   492  	// Read literal and/or (length, distance) according to RFC section 3.2.3.
   493  	{
   494  		v, err := f.huffSym(f.hl)
   495  		if err != nil {
   496  			f.err = err
   497  			return
   498  		}
   499  		var n uint // number of bits extra
   500  		var length int
   501  		switch {
   502  		case v < 256:
   503  			f.dict.writeByte(byte(v))
   504  			if f.dict.availWrite() == 0 {
   505  				f.toRead = f.dict.readFlush()
   506  				f.step = (*decompressor).huffmanBlock
   507  				f.stepState = stateInit
   508  				return
   509  			}
   510  			goto readLiteral
   511  		case v == 256:
   512  			f.finishBlock()
   513  			return
   514  		// otherwise, reference to older data
   515  		case v < 265:
   516  			length = v - (257 - 3)
   517  			n = 0
   518  		case v < 269:
   519  			length = v*2 - (265*2 - 11)
   520  			n = 1
   521  		case v < 273:
   522  			length = v*4 - (269*4 - 19)
   523  			n = 2
   524  		case v < 277:
   525  			length = v*8 - (273*8 - 35)
   526  			n = 3
   527  		case v < 281:
   528  			length = v*16 - (277*16 - 67)
   529  			n = 4
   530  		case v < 285:
   531  			length = v*32 - (281*32 - 131)
   532  			n = 5
   533  		case v < maxNumLit:
   534  			length = 258
   535  			n = 0
   536  		default:
   537  			f.err = CorruptInputError(f.roffset)
   538  			return
   539  		}
   540  		if n > 0 {
   541  			for f.nb < n {
   542  				if err = f.moreBits(); err != nil {
   543  					f.err = err
   544  					return
   545  				}
   546  			}
   547  			length += int(f.b & uint32(1<<n-1))
   548  			f.b >>= n
   549  			f.nb -= n
   550  		}
   551  
   552  		var dist int
   553  		if f.hd == nil {
   554  			for f.nb < 5 {
   555  				if err = f.moreBits(); err != nil {
   556  					f.err = err
   557  					return
   558  				}
   559  			}
   560  			dist = int(bits.Reverse8(uint8(f.b & 0x1F << 3)))
   561  			f.b >>= 5
   562  			f.nb -= 5
   563  		} else {
   564  			if dist, err = f.huffSym(f.hd); err != nil {
   565  				f.err = err
   566  				return
   567  			}
   568  		}
   569  
   570  		switch {
   571  		case dist < 4:
   572  			dist++
   573  		case dist < maxNumDist:
   574  			nb := uint(dist-2) >> 1
   575  			// have 1 bit in bottom of dist, need nb more.
   576  			extra := (dist & 1) << nb
   577  			for f.nb < nb {
   578  				if err = f.moreBits(); err != nil {
   579  					f.err = err
   580  					return
   581  				}
   582  			}
   583  			extra |= int(f.b & uint32(1<<nb-1))
   584  			f.b >>= nb
   585  			f.nb -= nb
   586  			dist = 1<<(nb+1) + 1 + extra
   587  		default:
   588  			f.err = CorruptInputError(f.roffset)
   589  			return
   590  		}
   591  
   592  		// No check on length; encoding can be prescient.
   593  		if dist > f.dict.histSize() {
   594  			f.err = CorruptInputError(f.roffset)
   595  			return
   596  		}
   597  
   598  		f.copyLen, f.copyDist = length, dist
   599  		goto copyHistory
   600  	}
   601  
   602  copyHistory:
   603  	// Perform a backwards copy according to RFC section 3.2.3.
   604  	{
   605  		cnt := f.dict.tryWriteCopy(f.copyDist, f.copyLen)
   606  		if cnt == 0 {
   607  			cnt = f.dict.writeCopy(f.copyDist, f.copyLen)
   608  		}
   609  		f.copyLen -= cnt
   610  
   611  		if f.dict.availWrite() == 0 || f.copyLen > 0 {
   612  			f.toRead = f.dict.readFlush()
   613  			f.step = (*decompressor).huffmanBlock // We need to continue this work
   614  			f.stepState = stateDict
   615  			return
   616  		}
   617  		goto readLiteral
   618  	}
   619  }
   620  
   621  // Copy a single uncompressed data block from input to output.
   622  func (f *decompressor) dataBlock() {
   623  	// Uncompressed.
   624  	// Discard current half-byte.
   625  	f.nb = 0
   626  	f.b = 0
   627  
   628  	// Length then ones-complement of length.
   629  	nr, err := io.ReadFull(f.r, f.buf[0:4])
   630  	f.roffset += int64(nr)
   631  	if err != nil {
   632  		f.err = noEOF(err)
   633  		return
   634  	}
   635  	n := int(f.buf[0]) | int(f.buf[1])<<8
   636  	nn := int(f.buf[2]) | int(f.buf[3])<<8
   637  	if uint16(nn) != uint16(^n) {
   638  		f.err = CorruptInputError(f.roffset)
   639  		return
   640  	}
   641  
   642  	if n == 0 {
   643  		f.toRead = f.dict.readFlush()
   644  		f.finishBlock()
   645  		return
   646  	}
   647  
   648  	f.copyLen = n
   649  	f.copyData()
   650  }
   651  
   652  // copyData copies f.copyLen bytes from the underlying reader into f.hist.
   653  // It pauses for reads when f.hist is full.
   654  func (f *decompressor) copyData() {
   655  	buf := f.dict.writeSlice()
   656  	if len(buf) > f.copyLen {
   657  		buf = buf[:f.copyLen]
   658  	}
   659  
   660  	cnt, err := io.ReadFull(f.r, buf)
   661  	f.roffset += int64(cnt)
   662  	f.copyLen -= cnt
   663  	f.dict.writeMark(cnt)
   664  	if err != nil {
   665  		f.err = noEOF(err)
   666  		return
   667  	}
   668  
   669  	if f.dict.availWrite() == 0 || f.copyLen > 0 {
   670  		f.toRead = f.dict.readFlush()
   671  		f.step = (*decompressor).copyData
   672  		return
   673  	}
   674  	f.finishBlock()
   675  }
   676  
   677  func (f *decompressor) finishBlock() {
   678  	if f.final {
   679  		if f.dict.availRead() > 0 {
   680  			f.toRead = f.dict.readFlush()
   681  		}
   682  		f.err = io.EOF
   683  	}
   684  	f.step = (*decompressor).nextBlock
   685  }
   686  
   687  // noEOF returns err, unless err == io.EOF, in which case it returns io.ErrUnexpectedEOF.
   688  func noEOF(e error) error {
   689  	if e == io.EOF {
   690  		return io.ErrUnexpectedEOF
   691  	}
   692  	return e
   693  }
   694  
   695  func (f *decompressor) moreBits() error {
   696  	c, err := f.r.ReadByte()
   697  	if err != nil {
   698  		return noEOF(err)
   699  	}
   700  	f.roffset++
   701  	f.b |= uint32(c) << f.nb
   702  	f.nb += 8
   703  	return nil
   704  }
   705  
   706  // Read the next Huffman-encoded symbol from f according to h.
   707  func (f *decompressor) huffSym(h *huffmanDecoder) (int, error) {
   708  	// Since a huffmanDecoder can be empty or be composed of a degenerate tree
   709  	// with single element, huffSym must error on these two edge cases. In both
   710  	// cases, the chunks slice will be 0 for the invalid sequence, leading it
   711  	// satisfy the n == 0 check below.
   712  	n := uint(h.min)
   713  	// Optimization. Compiler isn't smart enough to keep f.b,f.nb in registers,
   714  	// but is smart enough to keep local variables in registers, so use nb and b,
   715  	// inline call to moreBits and reassign b,nb back to f on return.
   716  	nb, b := f.nb, f.b
   717  	for {
   718  		for nb < n {
   719  			c, err := f.r.ReadByte()
   720  			if err != nil {
   721  				f.b = b
   722  				f.nb = nb
   723  				return 0, noEOF(err)
   724  			}
   725  			f.roffset++
   726  			b |= uint32(c) << (nb & 31)
   727  			nb += 8
   728  		}
   729  		chunk := h.chunks[b&(huffmanNumChunks-1)]
   730  		n = uint(chunk & huffmanCountMask)
   731  		if n > huffmanChunkBits {
   732  			chunk = h.links[chunk>>huffmanValueShift][(b>>huffmanChunkBits)&h.linkMask]
   733  			n = uint(chunk & huffmanCountMask)
   734  		}
   735  		if n <= nb {
   736  			if n == 0 {
   737  				f.b = b
   738  				f.nb = nb
   739  				f.err = CorruptInputError(f.roffset)
   740  				return 0, f.err
   741  			}
   742  			f.b = b >> (n & 31)
   743  			f.nb = nb - n
   744  			return int(chunk >> huffmanValueShift), nil
   745  		}
   746  	}
   747  }
   748  
   749  func makeReader(r io.Reader) Reader {
   750  	if rr, ok := r.(Reader); ok {
   751  		return rr
   752  	}
   753  	return bufio.NewReader(r)
   754  }
   755  
   756  func fixedHuffmanDecoderInit() {
   757  	fixedOnce.Do(func() {
   758  		// These come from the RFC section 3.2.6.
   759  		var bits [288]int
   760  		for i := 0; i < 144; i++ {
   761  			bits[i] = 8
   762  		}
   763  		for i := 144; i < 256; i++ {
   764  			bits[i] = 9
   765  		}
   766  		for i := 256; i < 280; i++ {
   767  			bits[i] = 7
   768  		}
   769  		for i := 280; i < 288; i++ {
   770  			bits[i] = 8
   771  		}
   772  		fixedHuffmanDecoder.init(bits[:])
   773  	})
   774  }
   775  
   776  func (f *decompressor) Reset(r io.Reader, dict []byte) error {
   777  	*f = decompressor{
   778  		r:        makeReader(r),
   779  		bits:     f.bits,
   780  		codebits: f.codebits,
   781  		dict:     f.dict,
   782  		step:     (*decompressor).nextBlock,
   783  	}
   784  	f.dict.init(maxMatchOffset, dict)
   785  	return nil
   786  }
   787  
   788  // NewReader returns a new ReadCloser that can be used
   789  // to read the uncompressed version of r.
   790  // If r does not also implement io.ByteReader,
   791  // the decompressor may read more data than necessary from r.
   792  // It is the caller's responsibility to call Close on the ReadCloser
   793  // when finished reading.
   794  //
   795  // The ReadCloser returned by NewReader also implements Resetter.
   796  func NewReader(r io.Reader) io.ReadCloser {
   797  	fixedHuffmanDecoderInit()
   798  
   799  	var f decompressor
   800  	f.r = makeReader(r)
   801  	f.bits = new([maxNumLit + maxNumDist]int)
   802  	f.codebits = new([numCodes]int)
   803  	f.step = (*decompressor).nextBlock
   804  	f.dict.init(maxMatchOffset, nil)
   805  	return &f
   806  }
   807  
   808  // NewReaderDict is like NewReader but initializes the reader
   809  // with a preset dictionary. The returned Reader behaves as if
   810  // the uncompressed data stream started with the given dictionary,
   811  // which has already been read. NewReaderDict is typically used
   812  // to read data compressed by NewWriterDict.
   813  //
   814  // The ReadCloser returned by NewReader also implements Resetter.
   815  func NewReaderDict(r io.Reader, dict []byte) io.ReadCloser {
   816  	fixedHuffmanDecoderInit()
   817  
   818  	var f decompressor
   819  	f.r = makeReader(r)
   820  	f.bits = new([maxNumLit + maxNumDist]int)
   821  	f.codebits = new([numCodes]int)
   822  	f.step = (*decompressor).nextBlock
   823  	f.dict.init(maxMatchOffset, dict)
   824  	return &f
   825  }
   826  

View as plain text