1. 1
2. 2
3. 3
4. 4
5. 5
6. 6
7. 7
8. 8
9. 9
10. 10
11. 11
12. 12
13. 13
14. 14
15. 15
16. 16
17. 17
18. 18
19. 19
20. 20
21. 21
22. 22
23. 23
24. 24
25. 25
26. 26
27. 27
28. 28
29. 29
30. 30
31. 31
32. 32
33. 33
34. 34
35. 35
36. 36
37. 37
38. 38
39. 39
40. 40
41. 41
42. 42
43. 43
44. 44
45. 45
46. 46
47. 47
48. 48
49. 49
50. 50
51. 51
52. 52
53. 53
54. 54
55. 55
56. 56
57. 57
58. 58
59. 59
60. 60
61. 61
62. 62
63. 63
64. 64
65. 65
66. 66
67. 67
68. 68
69. 69
70. 70
71. 71
72. 72
73. 73
74. 74
75. 75
76. 76
77. 77
78. 78
79. 79
80. 80
81. 81
82. 82
83. 83
84. 84
85. 85
86. 86
87. 87
88. 88
89. 89
90. 90
91. 91
92. 92
93. 93
94. 94
95. 95
96. 96
97. 97
98. 98
99. 99
100. 100
101. 101
102. 102
103. 103
104. 104
105. 105
106. 106
107. 107
108. 108
109. 109
110. 110
111. 111
112. 112
113. 113
114. 114
115. 115
116. 116
117. 117
118. 118
119. 119
120. 120
121. 121
122. 122
123. 123
124. 124
125. 125
126. 126
127. 127
128. 128
129. 129
130. 130
131. 131
132. 132
133. 133
134. 134
135. 135
136. 136
137. 137
138. 138
139. 139
140. 140
141. 141
142. 142
143. 143
144. 144
145. 145
146. 146
147. 147
148. 148
149. 149
150. 150
151. 151
152. 152
153. 153
154. 154
155. 155
156. 156
157. 157
158. 158
159. 159
160. 160
161. 161
162. 162
163. 163
164. 164
165. 165
166. 166
167. 167
168. 168
169. 169
170. 170
171. 171
172. 172
173. 173
174. 174
175. 175
176. 176
177. 177
178. 178
179. 179
180. 180
181. 181
182. 182
183. 183
184. 184

// Copyright 2016 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.

package flate

// dictDecoder implements the LZ77 sliding dictionary as used in decompression.
// LZ77 decompresses data through sequences of two forms of commands:
//
//	* Literal insertions: Runs of one or more symbols are inserted into the data
//	stream as is. This is accomplished through the writeByte method for a
//	single symbol, or combinations of writeSlice/writeMark for multiple symbols.
//	Any valid stream must start with a literal insertion if no preset dictionary
//	is used.
//
//	* Backward copies: Runs of one or more symbols are copied from previously
//	emitted data. Backward copies come as the tuple (dist, length) where dist
//	determines how far back in the stream to copy from and length determines how
//	many bytes to copy. Note that it is valid for the length to be greater than
//	the distance. Since LZ77 uses forward copies, that situation is used to
//	perform a form of run-length encoding on repeated runs of symbols.
//	The writeCopy and tryWriteCopy are used to implement this command.
//
// For performance reasons, this implementation performs little to no sanity
// checks about the arguments. As such, the invariants documented for each
// method call must be respected.
type dictDecoder struct {
	hist []byte // Sliding window history

	// Invariant: 0 <= rdPos <= wrPos <= len(hist)
	wrPos int  // Current output position in buffer
	rdPos int  // Have emitted hist[:rdPos] already
	full  bool // Has a full window length been written yet?
}

// init initializes dictDecoder to have a sliding window dictionary of the given
// size. If a preset dict is provided, it will initialize the dictionary with
// the contents of dict.
func (dd *dictDecoder) init(size int, dict []byte) {
	*dd = dictDecoder{hist: dd.hist}

	if cap(dd.hist) < size {
		dd.hist = make([]byte, size)
	}
	dd.hist = dd.hist[:size]

	if len(dict) > len(dd.hist) {
		dict = dict[len(dict)-len(dd.hist):]
	}
	dd.wrPos = copy(dd.hist, dict)
	if dd.wrPos == len(dd.hist) {
		dd.wrPos = 0
		dd.full = true
	}
	dd.rdPos = dd.wrPos
}

// histSize reports the total amount of historical data in the dictionary.
func (dd *dictDecoder) histSize() int {
	if dd.full {
		return len(dd.hist)
	}
	return dd.wrPos
}

// availRead reports the number of bytes that can be flushed by readFlush.
func (dd *dictDecoder) availRead() int {
	return dd.wrPos - dd.rdPos
}

// availWrite reports the available amount of output buffer space.
func (dd *dictDecoder) availWrite() int {
	return len(dd.hist) - dd.wrPos
}

// writeSlice returns a slice of the available buffer to write data to.
//
// This invariant will be kept: len(s) <= availWrite()
func (dd *dictDecoder) writeSlice() []byte {
	return dd.hist[dd.wrPos:]
}

// writeMark advances the writer pointer by cnt.
//
// This invariant must be kept: 0 <= cnt <= availWrite()
func (dd *dictDecoder) writeMark(cnt int) {
	dd.wrPos += cnt
}

// writeByte writes a single byte to the dictionary.
//
// This invariant must be kept: 0 < availWrite()
func (dd *dictDecoder) writeByte(c byte) {
	dd.hist[dd.wrPos] = c
	dd.wrPos++
}

// writeCopy copies a string at a given (dist, length) to the output.
// This returns the number of bytes copied and may be less than the requested
// length if the available space in the output buffer is too small.
//
// This invariant must be kept: 0 < dist <= histSize()
func (dd *dictDecoder) writeCopy(dist, length int) int {
	dstBase := dd.wrPos
	dstPos := dstBase
	srcPos := dstPos - dist
	endPos := dstPos + length
	if endPos > len(dd.hist) {
		endPos = len(dd.hist)
	}

	// Copy non-overlapping section after destination position.
	//
	// This section is non-overlapping in that the copy length for this section
	// is always less than or equal to the backwards distance. This can occur
	// if a distance refers to data that wraps-around in the buffer.
	// Thus, a backwards copy is performed here; that is, the exact bytes in
	// the source prior to the copy is placed in the destination.
	if srcPos < 0 {
		srcPos += len(dd.hist)
		dstPos += copy(dd.hist[dstPos:endPos], dd.hist[srcPos:])
		srcPos = 0
	}

	// Copy possibly overlapping section before destination position.
	//
	// This section can overlap if the copy length for this section is larger
	// than the backwards distance. This is allowed by LZ77 so that repeated
	// strings can be succinctly represented using (dist, length) pairs.
	// Thus, a forwards copy is performed here; that is, the bytes copied is
	// possibly dependent on the resulting bytes in the destination as the copy
	// progresses along. This is functionally equivalent to the following:
	//
	//	for i := 0; i < endPos-dstPos; i++ {
	//		dd.hist[dstPos+i] = dd.hist[srcPos+i]
	//	}
	//	dstPos = endPos
	//
	for dstPos < endPos {
		dstPos += copy(dd.hist[dstPos:endPos], dd.hist[srcPos:dstPos])
	}

	dd.wrPos = dstPos
	return dstPos - dstBase
}

// tryWriteCopy tries to copy a string at a given (distance, length) to the
// output. This specialized version is optimized for short distances.
//
// This method is designed to be inlined for performance reasons.
//
// This invariant must be kept: 0 < dist <= histSize()
func (dd *dictDecoder) tryWriteCopy(dist, length int) int {
	dstPos := dd.wrPos
	endPos := dstPos + length
	if dstPos < dist || endPos > len(dd.hist) {
		return 0
	}
	dstBase := dstPos
	srcPos := dstPos - dist

	// Copy possibly overlapping section before destination position.
loop:
	dstPos += copy(dd.hist[dstPos:endPos], dd.hist[srcPos:dstPos])
	if dstPos < endPos {
		goto loop // Avoid for-loop so that this function can be inlined
	}

	dd.wrPos = dstPos
	return dstPos - dstBase
}

// readFlush returns a slice of the historical buffer that is ready to be
// emitted to the user. The data returned by readFlush must be fully consumed
// before calling any other dictDecoder methods.
func (dd *dictDecoder) readFlush() []byte {
	toRead := dd.hist[dd.rdPos:dd.wrPos]
	dd.rdPos = dd.wrPos
	if dd.wrPos == len(dd.hist) {
		dd.wrPos, dd.rdPos = 0, 0
		dd.full = true
	}
	return toRead
}