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
|
'use strict';
var utils = require('./utils');
var support = require('./support');
var nodeBuffer = require('./nodeBuffer');
/**
* The following functions come from pako, from pako/lib/utils/strings
* released under the MIT license, see pako https://github.com/nodeca/pako/
*/
// Table with utf8 lengths (calculated by first byte of sequence)
// Note, that 5 & 6-byte values and some 4-byte values can not be represented in JS,
// because max possible codepoint is 0x10ffff
var _utf8len = new Array(256);
for (var i=0; i<256; i++) {
_utf8len[i] = (i >= 252 ? 6 : i >= 248 ? 5 : i >= 240 ? 4 : i >= 224 ? 3 : i >= 192 ? 2 : 1);
}
_utf8len[254]=_utf8len[254]=1; // Invalid sequence start
// convert string to array (typed, when possible)
var string2buf = function (str) {
var buf, c, c2, m_pos, i, str_len = str.length, buf_len = 0;
// count binary size
for (m_pos = 0; m_pos < str_len; m_pos++) {
c = str.charCodeAt(m_pos);
if ((c & 0xfc00) === 0xd800 && (m_pos+1 < str_len)) {
c2 = str.charCodeAt(m_pos+1);
if ((c2 & 0xfc00) === 0xdc00) {
c = 0x10000 + ((c - 0xd800) << 10) + (c2 - 0xdc00);
m_pos++;
}
}
buf_len += c < 0x80 ? 1 : c < 0x800 ? 2 : c < 0x10000 ? 3 : 4;
}
// allocate buffer
if (support.uint8array) {
buf = new Uint8Array(buf_len);
} else {
buf = new Array(buf_len);
}
// convert
for (i=0, m_pos = 0; i < buf_len; m_pos++) {
c = str.charCodeAt(m_pos);
if ((c & 0xfc00) === 0xd800 && (m_pos+1 < str_len)) {
c2 = str.charCodeAt(m_pos+1);
if ((c2 & 0xfc00) === 0xdc00) {
c = 0x10000 + ((c - 0xd800) << 10) + (c2 - 0xdc00);
m_pos++;
}
}
if (c < 0x80) {
/* one byte */
buf[i++] = c;
} else if (c < 0x800) {
/* two bytes */
buf[i++] = 0xC0 | (c >>> 6);
buf[i++] = 0x80 | (c & 0x3f);
} else if (c < 0x10000) {
/* three bytes */
buf[i++] = 0xE0 | (c >>> 12);
buf[i++] = 0x80 | (c >>> 6 & 0x3f);
buf[i++] = 0x80 | (c & 0x3f);
} else {
/* four bytes */
buf[i++] = 0xf0 | (c >>> 18);
buf[i++] = 0x80 | (c >>> 12 & 0x3f);
buf[i++] = 0x80 | (c >>> 6 & 0x3f);
buf[i++] = 0x80 | (c & 0x3f);
}
}
return buf;
};
// Calculate max possible position in utf8 buffer,
// that will not break sequence. If that's not possible
// - (very small limits) return max size as is.
//
// buf[] - utf8 bytes array
// max - length limit (mandatory);
var utf8border = function(buf, max) {
var pos;
max = max || buf.length;
if (max > buf.length) { max = buf.length; }
// go back from last position, until start of sequence found
pos = max-1;
while (pos >= 0 && (buf[pos] & 0xC0) === 0x80) { pos--; }
// Fuckup - very small and broken sequence,
// return max, because we should return something anyway.
if (pos < 0) { return max; }
// If we came to start of buffer - that means vuffer is too small,
// return max too.
if (pos === 0) { return max; }
return (pos + _utf8len[buf[pos]] > max) ? pos : max;
};
// convert array to string
var buf2string = function (buf) {
var str, i, out, c, c_len;
var len = buf.length;
// Reserve max possible length (2 words per char)
// NB: by unknown reasons, Array is significantly faster for
// String.fromCharCode.apply than Uint16Array.
var utf16buf = new Array(len*2);
for (out=0, i=0; i<len;) {
c = buf[i++];
// quick process ascii
if (c < 0x80) { utf16buf[out++] = c; continue; }
c_len = _utf8len[c];
// skip 5 & 6 byte codes
if (c_len > 4) { utf16buf[out++] = 0xfffd; i += c_len-1; continue; }
// apply mask on first byte
c &= c_len === 2 ? 0x1f : c_len === 3 ? 0x0f : 0x07;
// join the rest
while (c_len > 1 && i < len) {
c = (c << 6) | (buf[i++] & 0x3f);
c_len--;
}
// terminated by end of string?
if (c_len > 1) { utf16buf[out++] = 0xfffd; continue; }
if (c < 0x10000) {
utf16buf[out++] = c;
} else {
c -= 0x10000;
utf16buf[out++] = 0xd800 | ((c >> 10) & 0x3ff);
utf16buf[out++] = 0xdc00 | (c & 0x3ff);
}
}
// shrinkBuf(utf16buf, out)
if (utf16buf.length !== out) {
if(utf16buf.subarray) {
utf16buf = utf16buf.subarray(0, out);
} else {
utf16buf.length = out;
}
}
// return String.fromCharCode.apply(null, utf16buf);
return utils.applyFromCharCode(utf16buf);
};
// That's all for the pako functions.
/**
* Transform a javascript string into an array (typed if possible) of bytes,
* UTF-8 encoded.
* @param {String} str the string to encode
* @return {Array|Uint8Array|Buffer} the UTF-8 encoded string.
*/
exports.utf8encode = function utf8encode(str) {
if (support.nodebuffer) {
return nodeBuffer(str, "utf-8");
}
return string2buf(str);
};
/**
* Transform a bytes array (or a representation) representing an UTF-8 encoded
* string into a javascript string.
* @param {Array|Uint8Array|Buffer} buf the data de decode
* @return {String} the decoded string.
*/
exports.utf8decode = function utf8decode(buf) {
if (support.nodebuffer) {
return utils.transformTo("nodebuffer", buf).toString("utf-8");
}
buf = utils.transformTo(support.uint8array ? "uint8array" : "array", buf);
// return buf2string(buf);
// Chrome prefers to work with "small" chunks of data
// for the method buf2string.
// Firefox and Chrome has their own shortcut, IE doesn't seem to really care.
var result = [], k = 0, len = buf.length, chunk = 65536;
while (k < len) {
var nextBoundary = utf8border(buf, Math.min(k + chunk, len));
if (support.uint8array) {
result.push(buf2string(buf.subarray(k, nextBoundary)));
} else {
result.push(buf2string(buf.slice(k, nextBoundary)));
}
k = nextBoundary;
}
return result.join("");
};
// vim: set shiftwidth=4 softtabstop=4:
|
