当前位置: 首页 > 图文教程 > 网络编程 > Javascript > lib.utf.js

Javascript
IE bug table元素的innerHTML
Javascript学习笔记1 数据类型
Javascript学习笔记2 函数
Javascript学习笔记3 作用域
Javascript学习笔记4 Eval函数
Javascript学习笔记5 类和对象
Javascript学习笔记6 prototype的提出
Javascript学习笔记7 原型链的原理
Javascript学习笔记8 用JSON做原型
Javascript学习笔记9 prototype封装继承
Javascript 读书笔记索引贴
JavaScript 事件冒泡应用实例分析
JavaScript 事件冒泡简介及应用
Extjs学习笔记之九 数据模型(上)
javascript Object与Function使用
取选中的radio的值
JavaScript 学习笔记(九)call和apply方法
javascript下判断一个对象是否具有指定名称的属性的的代码
Javascript 实现TreeView CheckBox全选效果
jquery插件 cluetip 关键词注释

Javascript 中的 lib.utf.js


出处:互联网   整理: 软晨网(RuanChen.com)   发布: 2009-09-12   浏览: 132 ::
收藏到网摘: n/a

/* UTF-8 <=> UTF-16 convertion library.
*
/* Copyright (C) 1999 Masanao Izumo <[email protected]>
* 2007 Ma Bingyao <[email protected]>
* Version: 2.1
* LastModified: Feb 25, 2007
* This library is free. You can redistribute it and/or modify it.
*/


/*
* Interfaces:
* utf8 = utf16to8(utf16);
* utf16 = utf16to8(utf8);
*/


function utf16to8(str) {
if (str.match(/^[\x00-\x7f]*$/) != null) {
return str;
}
var out, i, j, len, c, c2;
out = [];
len = str.length;
for (i = 0, j = 0; i < len; i++, j++) {
c = str.charCodeAt(i);
if (c <= 0x7f) {
out[j] = str.charAt(i);
}
else if (c <= 0x7ff) {
out[j] = String.fromCharCode(0xc0 | (c >>> 6),
0x80 | (c & 0x3f));
}
else if (c < 0xd800 || c > 0xdfff) {
out[j] = String.fromCharCode(0xe0 | (c >>> 12),
0x80 | ((c >>> 6) & 0x3f),
0x80 | (c & 0x3f));
}
else {
if (++i < len) {
c2 = str.charCodeAt(i);
if (c <= 0xdbff && 0xdc00 <= c2 && c2 <= 0xdfff) {
c = ((c & 0x03ff) << 10 | (c2 & 0x03ff)) + 0x010000;
if (0x010000 <= c && c <= 0x10ffff) {
out[j] = String.fromCharCode(0xf0 | ((c >>> 18) & 0x3f),
0x80 | ((c >>> 12) & 0x3f),
0x80 | ((c >>> 6) & 0x3f),
0x80 | (c & 0x3f));
}
else {
out[j] = '?';
}
}
else {
i--;
out[j] = '?';
}
}
else {
i--;
out[j] = '?';
}
}
}
return out.join('');
}


function utf8to16(str) {
if ((str.match(/^[\x00-\x7f]*$/) != null) ||
(str.match(/^[\x00-\xff]*$/) == null)) {
return str;
}
var out, i, j, len, c, c2, c3, c4, s;


out = [];
len = str.length;
i = j = 0;
while (i < len) {
c = str.charCodeAt(i++);
switch (c >> 4) {
case 0: case 1: case 2: case 3: case 4: case 5: case 6: case 7:
// 0xxx xxxx
out[j++] = str.charAt(i - 1);
break;
case 12: case 13:
// 110x xxxx 10xx xxxx
c2 = str.charCodeAt(i++);
out[j++] = String.fromCharCode(((c & 0x1f) << 6) |
(c2 & 0x3f));
break;
case 14:
// 1110 xxxx 10xx xxxx 10xx xxxx
c2 = str.charCodeAt(i++);
c3 = str.charCodeAt(i++);
out[j++] = String.fromCharCode(((c & 0x0f) << 12) |
((c2 & 0x3f) << 6) |
(c3 & 0x3f));
break;
case 15:
switch (c & 0xf) {
case 0: case 1: case 2: case 3: case 4: case 5: case 6: case 7:
// 1111 0xxx 10xx xxxx 10xx xxxx 10xx xxxx
c2 = str.charCodeAt(i++);
c3 = str.charCodeAt(i++);
c4 = str.charCodeAt(i++);
s = ((c & 0x07) << 18) |
((c2 & 0x3f) << 12) |
((c3 & 0x3f) << 6) |
(c4 & 0x3f) - 0x10000;
if (0 <= s && s <= 0xfffff) {
out[j] = String.fromCharCode(((s >>> 10) & 0x03ff) | 0xd800,
(s & 0x03ff) | 0xdc00);
}
else {
out[j] = '?';
}
break;
case 8: case 9: case 10: case 11:
// 1111 10xx 10xx xxxx 10xx xxxx 10xx xxxx 10xx xxxx
i+=4;
out[j] = '?';
break;
case 12: case 13:
// 1111 110x 10xx xxxx 10xx xxxx 10xx xxxx 10xx xxxx 10xx xxxx
i+=5;
out[j] = '?';
break;
}
}
j++;
}
return out.join('');
}