aboutsummaryrefslogtreecommitdiff
path: root/vendor/github.com/json-iterator/go/iter_str.go
blob: adc487ea80483cbc892f0438c3133705cbcb46cd (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
package jsoniter

import (
	"fmt"
	"unicode/utf16"
)

// ReadString read string from iterator
func (iter *Iterator) ReadString() (ret string) {
	c := iter.nextToken()
	if c == '"' {
		for i := iter.head; i < iter.tail; i++ {
			c := iter.buf[i]
			if c == '"' {
				ret = string(iter.buf[iter.head:i])
				iter.head = i + 1
				return ret
			} else if c == '\\' {
				break
			} else if c < ' ' {
				iter.ReportError("ReadString",
					fmt.Sprintf(`invalid control character found: %d`, c))
				return
			}
		}
		return iter.readStringSlowPath()
	} else if c == 'n' {
		iter.skipThreeBytes('u', 'l', 'l')
		return ""
	}
	iter.ReportError("ReadString", `expects " or n, but found `+string([]byte{c}))
	return
}

func (iter *Iterator) readStringSlowPath() (ret string) {
	var str []byte
	var c byte
	for iter.Error == nil {
		c = iter.readByte()
		if c == '"' {
			return string(str)
		}
		if c == '\\' {
			c = iter.readByte()
			str = iter.readEscapedChar(c, str)
		} else {
			str = append(str, c)
		}
	}
	iter.ReportError("readStringSlowPath", "unexpected end of input")
	return
}

func (iter *Iterator) readEscapedChar(c byte, str []byte) []byte {
	switch c {
	case 'u':
		r := iter.readU4()
		if utf16.IsSurrogate(r) {
			c = iter.readByte()
			if iter.Error != nil {
				return nil
			}
			if c != '\\' {
				iter.unreadByte()
				str = appendRune(str, r)
				return str
			}
			c = iter.readByte()
			if iter.Error != nil {
				return nil
			}
			if c != 'u' {
				str = appendRune(str, r)
				return iter.readEscapedChar(c, str)
			}
			r2 := iter.readU4()
			if iter.Error != nil {
				return nil
			}
			combined := utf16.DecodeRune(r, r2)
			if combined == '\uFFFD' {
				str = appendRune(str, r)
				str = appendRune(str, r2)
			} else {
				str = appendRune(str, combined)
			}
		} else {
			str = appendRune(str, r)
		}
	case '"':
		str = append(str, '"')
	case '\\':
		str = append(str, '\\')
	case '/':
		str = append(str, '/')
	case 'b':
		str = append(str, '\b')
	case 'f':
		str = append(str, '\f')
	case 'n':
		str = append(str, '\n')
	case 'r':
		str = append(str, '\r')
	case 't':
		str = append(str, '\t')
	default:
		iter.ReportError("readEscapedChar",
			`invalid escape char after \`)
		return nil
	}
	return str
}

// ReadStringAsSlice read string from iterator without copying into string form.
// The []byte can not be kept, as it will change after next iterator call.
func (iter *Iterator) ReadStringAsSlice() (ret []byte) {
	c := iter.nextToken()
	if c == '"' {
		for i := iter.head; i < iter.tail; i++ {
			// require ascii string and no escape
			// for: field name, base64, number
			if iter.buf[i] == '"' {
				// fast path: reuse the underlying buffer
				ret = iter.buf[iter.head:i]
				iter.head = i + 1
				return ret
			}
		}
		readLen := iter.tail - iter.head
		copied := make([]byte, readLen, readLen*2)
		copy(copied, iter.buf[iter.head:iter.tail])
		iter.head = iter.tail
		for iter.Error == nil {
			c := iter.readByte()
			if c == '"' {
				return copied
			}
			copied = append(copied, c)
		}
		return copied
	}
	iter.ReportError("ReadStringAsSlice", `expects " or n, but found `+string([]byte{c}))
	return
}

func (iter *Iterator) readU4() (ret rune) {
	for i := 0; i < 4; i++ {
		c := iter.readByte()
		if iter.Error != nil {
			return
		}
		if c >= '0' && c <= '9' {
			ret = ret*16 + rune(c-'0')
		} else if c >= 'a' && c <= 'f' {
			ret = ret*16 + rune(c-'a'+10)
		} else if c >= 'A' && c <= 'F' {
			ret = ret*16 + rune(c-'A'+10)
		} else {
			iter.ReportError("readU4", "expects 0~9 or a~f, but found "+string([]byte{c}))
			return
		}
	}
	return ret
}

const (
	t1 = 0x00 // 0000 0000
	tx = 0x80 // 1000 0000
	t2 = 0xC0 // 1100 0000
	t3 = 0xE0 // 1110 0000
	t4 = 0xF0 // 1111 0000
	t5 = 0xF8 // 1111 1000

	maskx = 0x3F // 0011 1111
	mask2 = 0x1F // 0001 1111
	mask3 = 0x0F // 0000 1111
	mask4 = 0x07 // 0000 0111

	rune1Max = 1<<7 - 1
	rune2Max = 1<<11 - 1
	rune3Max = 1<<16 - 1

	surrogateMin = 0xD800
	surrogateMax = 0xDFFF

	maxRune   = '\U0010FFFF' // Maximum valid Unicode code point.
	runeError = '\uFFFD'     // the "error" Rune or "Unicode replacement character"
)

func appendRune(p []byte, r rune) []byte {
	// Negative values are erroneous. Making it unsigned addresses the problem.
	switch i := uint32(r); {
	case i <= rune1Max:
		p = append(p, byte(r))
		return p
	case i <= rune2Max:
		p = append(p, t2|byte(r>>6))
		p = append(p, tx|byte(r)&maskx)
		return p
	case i > maxRune, surrogateMin <= i && i <= surrogateMax:
		r = runeError
		fallthrough
	case i <= rune3Max:
		p = append(p, t3|byte(r>>12))
		p = append(p, tx|byte(r>>6)&maskx)
		p = append(p, tx|byte(r)&maskx)
		return p
	default:
		p = append(p, t4|byte(r>>18))
		p = append(p, tx|byte(r>>12)&maskx)
		p = append(p, tx|byte(r>>6)&maskx)
		p = append(p, tx|byte(r)&maskx)
		return p
	}
}