132 lines
3.7 KiB
JavaScript
132 lines
3.7 KiB
JavaScript
"use strict";
|
|
|
|
var utf8 = exports;
|
|
var replacementChar = "\ufffd";
|
|
|
|
function utf8Read(buffer, start, end, fatal) {
|
|
var str = "";
|
|
for (var i = start; i < end;) {
|
|
var first = buffer[i++];
|
|
if (first <= 0x7F) {
|
|
str += String.fromCharCode(first);
|
|
continue;
|
|
}
|
|
|
|
var needed = 0;
|
|
var codePoint = 0;
|
|
var minCodePoint = 0;
|
|
|
|
if (first >= 0xC2 && first <= 0xDF) {
|
|
needed = 1;
|
|
codePoint = first & 0x1F;
|
|
minCodePoint = 0x80;
|
|
} else if (first >= 0xE0 && first <= 0xEF) {
|
|
needed = 2;
|
|
codePoint = first & 0x0F;
|
|
minCodePoint = 0x800;
|
|
} else if (first >= 0xF0 && first <= 0xF4) {
|
|
needed = 3;
|
|
codePoint = first & 0x07;
|
|
minCodePoint = 0x10000;
|
|
} else {
|
|
if (fatal) throw Error("invalid UTF-8 encoding");
|
|
str += replacementChar;
|
|
continue;
|
|
}
|
|
|
|
if (i + needed > end) {
|
|
if (fatal) throw Error("invalid UTF-8 encoding");
|
|
str += replacementChar;
|
|
break;
|
|
}
|
|
|
|
var valid = true;
|
|
for (var j = 0; j < needed; ++j) {
|
|
var next = buffer[i + j];
|
|
if ((next & 0xC0) !== 0x80) {
|
|
valid = false;
|
|
break;
|
|
}
|
|
codePoint = codePoint << 6 | next & 0x3F;
|
|
}
|
|
|
|
if (
|
|
!valid ||
|
|
codePoint < minCodePoint ||
|
|
codePoint > 0x10FFFF ||
|
|
codePoint >= 0xD800 && codePoint <= 0xDFFF
|
|
) {
|
|
if (fatal) throw Error("invalid UTF-8 encoding");
|
|
str += replacementChar;
|
|
continue;
|
|
}
|
|
|
|
i += needed;
|
|
if (codePoint <= 0xFFFF) {
|
|
str += String.fromCharCode(codePoint);
|
|
} else {
|
|
codePoint -= 0x10000;
|
|
str += String.fromCharCode(0xD800 + (codePoint >> 10));
|
|
str += String.fromCharCode(0xDC00 + (codePoint & 0x3FF));
|
|
}
|
|
}
|
|
return str;
|
|
}
|
|
|
|
utf8.length = function utf8_length(string) {
|
|
var len = 0,
|
|
c = 0;
|
|
for (var i = 0; i < string.length; ++i) {
|
|
c = string.charCodeAt(i);
|
|
if (c < 128)
|
|
len += 1;
|
|
else if (c < 2048)
|
|
len += 2;
|
|
else if ((c & 0xFC00) === 0xD800 && (string.charCodeAt(i + 1) & 0xFC00) === 0xDC00) {
|
|
++i;
|
|
len += 4;
|
|
} else
|
|
len += 3;
|
|
}
|
|
return len;
|
|
};
|
|
|
|
utf8.read = function utf8_read(buffer, start, end) {
|
|
if (end - start < 1)
|
|
return "";
|
|
return utf8Read(buffer, start, end, false);
|
|
};
|
|
|
|
utf8.readStrict = function utf8_read_strict(buffer, start, end) {
|
|
if (end - start < 1)
|
|
return "";
|
|
return utf8Read(buffer, start, end, true);
|
|
};
|
|
|
|
utf8.write = function utf8_write(string, buffer, offset) {
|
|
var start = offset,
|
|
c1,
|
|
c2;
|
|
for (var i = 0; i < string.length; ++i) {
|
|
c1 = string.charCodeAt(i);
|
|
if (c1 < 128) {
|
|
buffer[offset++] = c1;
|
|
} else if (c1 < 2048) {
|
|
buffer[offset++] = c1 >> 6 | 192;
|
|
buffer[offset++] = c1 & 63 | 128;
|
|
} else if ((c1 & 0xFC00) === 0xD800 && ((c2 = string.charCodeAt(i + 1)) & 0xFC00) === 0xDC00) {
|
|
c1 = 0x10000 + ((c1 & 0x03FF) << 10) + (c2 & 0x03FF);
|
|
++i;
|
|
buffer[offset++] = c1 >> 18 | 240;
|
|
buffer[offset++] = c1 >> 12 & 63 | 128;
|
|
buffer[offset++] = c1 >> 6 & 63 | 128;
|
|
buffer[offset++] = c1 & 63 | 128;
|
|
} else {
|
|
buffer[offset++] = c1 >> 12 | 224;
|
|
buffer[offset++] = c1 >> 6 & 63 | 128;
|
|
buffer[offset++] = c1 & 63 | 128;
|
|
}
|
|
}
|
|
return offset - start;
|
|
};
|