word文档----to——html

<html>

<head>

<title>Word2Html</title>

<script language="javascript">

function convert2html(){

var htt;

htt=document.getElementById("word").innerHTML;
console.log(htt);

htt = htt.replace(/</?SPAN[^>]*>/gi, "" );

        // Remove Class attributes

        htt = htt.replace(/<(w[^>]*) class=([^ |>]*)([^>]*)/gi, "<$1$3") ;

        // Remove Style attributes

        htt = htt.replace(/<(w[^>]*) style="([^"]*)"([^>]*)/gi, "<$1$3") ;

        // Remove Lang attributes

        htt = htt.replace(/<(w[^>]*) lang=([^ |>]*)([^>]*)/gi, "<$1$3") ;

        // Remove XML elements and declarations

        htt = htt.replace(/<\??xml[^>]*>/gi, "") ;

        // Remove Tags with XML namespace declarations: <o:p></o:p>

        htt = htt.replace(/</?w+:[^>]*>/gi, "") ;

        // Replace the &nbsp;

        htt= htt.replace(/&nbsp;/, " " );

        //我的replace
        //
        htt= htt.replace(/align="justify"/gi, 'class="justify"');

        // Transform <P> to <DIV>

        var re = new RegExp("(<P)([^>]*>.*?)(</P>)","gi") ;        // Different because of a IE 5.0 error

        htt = htt.replace( re, "<div$2</div>" ) ;

    html.value = htt;     

}

// 清除WORD冗余格式并粘贴

function cleanAndPaste(  ) {
        // Remove all SPAN tags

        htt = htt.replace(/</?SPAN[^>]*>/gi, "" );

        // Remove Class attributes

        htt = htt.replace(/<(w[^>]*) class=([^ |>]*)([^>]*)/gi, "<$1$3") ;

        // Remove Style attributes

        htt = htt.replace(/<(w[^>]*) style="([^"]*)"([^>]*)/gi, "<$1$3") ;

        // Remove Lang attributes

        htt = htt.replace(/<(w[^>]*) lang=([^ |>]*)([^>]*)/gi, "<$1$3") ;

        // Remove XML elements and declarations

        htt = htt.replace(/<\??xml[^>]*>/gi, "") ;

        // Remove Tags with XML namespace declarations: <o:p></o:p>

        htt = htt.replace(/</?w+:[^>]*>/gi, "") ;

        // Replace the &nbsp;

        htt = htt.replace(/&nbsp;/, " " );

        // Transform <P> to <DIV>

        var re = new RegExp("(<P)([^>]*>.*?)(</P>)","gi") ;        // Different because of a IE 5.0 error

        htt = htt.replace( re, "<div$2</div>" ) ;

        html.value = htt;  
}

</script>

</head>

<body>

<p>请在这里贴入WORD文件内容

<div style="border:1 outset #ffffff; overflow:auto;80%;height:50%" id="word" contenteditable></div>

<input type="button" value="转换成HTML" onClick="convert2html()">
<input type="button" value="清除格式" onClick="cleanAndPaste()">点击后把生成的代码,全选后复制,并粘贴到wiki的编辑框中<br>

<textarea cols="115" rows="18" id="html"></textarea><br>

注意:您的浏览器必须是IE5.5后以上，否则无法插入文字!

</body></html>
相关推荐
