撸个插件给你word-to-html

最近遇到一个需求，需要将非常多内容的（文字/表格）word文档展示出来，这个需求出现在pc端就用插件好了或者直接下载文件？如果需求是在移动端呢？怎么办？转成html吧。。。几十页的word怎么搞？为了造福大家，花了几天时间撸了一个插件word-to-html,可以转嵌套的表格，合并单元格的表格，github地址.

emmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmm!
安利一波其他项目：

NodeNote，持续更新中react相关库源码浅析，
react ts3 项目

---

时间仓促，代码写得有点乱，这里贴一下reamde吧，谁用谁知道哈，如果解决给位的痛点，希望不要吝啬您的star，非常欢迎提issue，大家一起讨论完善。由于用到了jsdom，这个库模拟出来的DOMpaser有点弱，如果你选择在浏览器中用我的给的方法，你甚至能将word中每一行不同文字的字体字号都转成对应的html，借助浏览器的js调试面板的源码我放在了github上对应项目的test/browser文件夹中了。

下面是readme：

word-to-html

A tiny tool to convert Microsoft Word document to HTML in Nodejs and in chrome,
you can use the tool convert tables with merged cells and nested tables to html file in Nodejs or chrome, the online tool wordhtml can not do this.
Beyond that, you can convert words with different font-family or font-size in a line to html string in chrome.

table example

example.html(mobile)

attention

If a line of words have different font-family or font-size in your .docx, it can not convert
your .docx to html expectly in nodejs, but this can fix in the browsers such as chrome. because
the npm package jsdom can not realize the DOMParser's function perfectly.
So if you want to convert the font-family and font-size exactly, you can see how to use word2html.js in browsers!

Install

npm i word-to-html --save-dev

or

yarn add word-to-html

api in nodejs: word2html(absPath [,options])

absPath: string | Array<string>

absPath is your file's absolute path

options: {tdTextAlign:string,tdVerticalAlign:string}

tdTextAlign controls the <td> tag's text-align

tdVerticalAlign controls the <td> tag's vertical-align

Usage in nodejs

var path = require('path');
var word2html = require('word-to-html');
//Word document's absolute path
var absPath = path.join(__dirname,'test.docx');
word2html(absPath,{tdVerticalAlign:'top'})

the html generated in your WorkSpace.

Usage in browsers

details in my github

step 1: Take the code in your html or your console panel as the global functions

<script type = ' text/javascript'> 
        loadXML  =   function(xmlString){
            var  xmlDoc = null ; 
            if ( ! window.DOMParser  &&  window.ActiveXObject){    // window.DOMParser 判断是否是非ie浏览器 
                var  xmlDomVersions  =  [ ' MSXML.2.DOMDocument.6.0 ' , ' MSXML.2.DOMDocument.3.0 ' , ' Microsoft.XMLDOM ' ];
                for ( var  i = 0 ;i < xmlDomVersions.length;i ++ ){
                    try {
                        xmlDoc  =   new  ActiveXObject(xmlDomVersions[i]);
                        xmlDoc.async  =   false ;
                        xmlDoc.loadXML(xmlString);  // loadXML方法载入xml字符串 
                        break ;
                    } catch (e){
                    }
                }
            }
            else if (window.DOMParser  &&  document.implementation  &&  document.implementation.createDocument){
                try {
                    /*  DOMParser 对象解析 XML 文本并返回一个 XML Document 对象。
                    * 要使用 DOMParser，使用不带参数的构造函数来实例化它，然后调用其 parseFromString() 方法
                    * parseFromString(text, contentType) 参数text:要解析的 XML 标记 参数contentType文本的内容类型
                    * 可能是 "text/xml" 、"application/xml" 或 "application/xhtml+xml" 中的一个。注意，不支持 "text/html"。
                    */ 
                    domParser  =   new DOMParser();
                    xmlDoc  =  domParser.parseFromString(xmlString,  "application/xml" );
                } catch (e){

                }
            }
            else {
                return   null ;
            }
            return  xmlDoc;
        }
        
        var getDirectDomsByTagName = function(dom, tagName){
            var childs = Array.prototype.slice.call(dom.children);
            var doms = childs.filter((item,index)=>{
                return item.tagName === tagName
            })
            return doms
        }
        
        // amd-zip将docx格式的文件转换成xml的规则是：
        // table规则：
        // <w:tbl></w:tbl>表示整个表格                                                          tblFn:需要<table>包裹
        // <w:tr></w:tr>表示表格的一行                                                          trFn:需要<tr>包裹
        // <w:tc></w:tc>表示表格某一行的一列                                                    tcFn:需要<td></td>包裹
        // 在<w:tc></w:tc>这一列中，对应的word中有多少个回车就会生成多少个<w:p>，
        // 在<w:p></w:p>中，对应的word中有多少个软回车（向下的箭头↓），就会有多少<w:r></w:r>
        // 一般？：在<w:r></w:r>中的<w:t></w:t>就包裹了需要的文字内容
        // 这里需要注意的一个问题是：特殊符号比如上标也会单独成为一个<w:r></w:r>
        
        // 总之遍历到标签<w:t></w:t>则表示结束
        
        /**
         * 
         * @param {*} tblDom 处理<w:tbl>标签对应的DOM
         * @return {string} tblText 返回table标签对应的html字符串  
         */
        var tblFn = function(tblDom){
            //
            let tblLeft = `<table><tbody>`
            let tblRight = `</tbody></table>`
            let tblText = tblLeft;
        
            let trArray = getDirectDomsByTagName(tblDom,'w:tr'), len = trArray.length;
            for(let i = 0;i<len; i++){
                let tr = trArray[i];
                tblText = tblText + trFn(tr,i,trArray);
            }
        
            tblText = tblText + tblRight;
            return tblText; 
        }
        
        /**
         * @param trDom: 处理<w:tr>标签对应的DOM
         * @param rNum:trDom所处的trArray的第几行
         * @param trArray:表的所有行trArray
         * @return trText： 字符串，表示的是表格的一行的html字符串
         */
        var trFn = function(trDom,rNum,trArray){
            let trStart = `<tr>`, 
                trEnd = `</tr>`, 
                trText = trStart;
            let tcArray = getDirectDomsByTagName(trDom,'w:tc'), len = tcArray.length;
            for(let i = 0;i<len; i++){
                let tc = tcArray[i];
                trText = trText + tcFn(tc,rNum, i,trArray);
            }
            trText = trText + trEnd;
            return trText;
        }
        
        /**
         * @param tcDom: 处理<w:tc>标签对应的DOM
         * @param rNum:trDom所处的trArray的第几行
         * @param cNum: 传入的tcDom处于tr中tcArray的第几个，即第几列
         * @param trArray:表的所有行trArray
         * @return tcText： 字符串，表示的是表格的一行的html字符串
         */
        var tcFn = function(tcDom,rNum,cNum,trArray){
            let {colspan, vMerge, hasT} = getTcDomOptions(tcDom);
            if(!hasT){
                return ''
            }
            // 合并行
            let rowspan;
            if(vMerge === 'restart'){
                let len = trArray.length;
                rowspan =1;
                for(let n = rNum+1;n<len;n++){
                    let  tcArray = getDirectDomsByTagName(trArray[n],'w:tc')
                    if(tcArray.length-1 < cNum) break;
                    let tcPrDom = getDirectDomsByTagName(tcArray[cNum],'w:tcPr')[0];
                    let vMergeDom = getDirectDomsByTagName(tcPrDom,'w:vMerge')[0];
                    if(vMergeDom && vMergeDom.getAttribute('w:val')!=='restart'){
                        rowspan++ 
                    }else{
                        break;
                    }
                }
            } 
            let tdStart = `<td ${colspan?`colspan=${colspan}`:''} ${rowspan?`rowspan=${rowspan}`:''}>`, //合并列
                tdEnd = `</td>`,
                tcText = tdStart;
            
            tcText = tcText + wanderDom(tcDom) + tdEnd;
        
            return tcText;
        }
        
        /**
         * @param tcDom: 处理<w:tc>标签对应的DOM
         * @return ： combinations:{colspan, vMerge, hasT} 对象，表示的是表格的一行的html字符串
         */
        var getTcDomOptions = function(tcDom){
            let tcPrDom = getDirectDomsByTagName(tcDom,'w:tcPr')[0];
            let gridSpanDom = getDirectDomsByTagName(tcPrDom,'w:gridSpan')[0],
                vMergeDom = getDirectDomsByTagName(tcPrDom,'w:vMerge')[0],
                tDom = tcDom.getElementsByTagName('w:t');
            let colspan = gridSpanDom?gridSpanDom.getAttribute('w:val'):'';
            let vMerge = vMergeDom ?
                            vMergeDom.getAttribute('w:val') ? vMergeDom.getAttribute('w:val') : '1'
                            :
                            '' ;
            let hasT = tDom.length?true:false;
            return {colspan,vMerge, hasT}
        }
        
        
        /**
         * @param {*} rArray 数组，<w:r>标签对应的DOM组成的数组
         * @return {string} textContent 返回table标签对应的html字符串  
         */
        var rFn = function(rArray){
            var br = `<br>`,textContent = '',rTextArray = [];
                // In browser
                for(let i =0; i<rArray.length;i++){
                    var r = rArray[i];
                    var rFontFamily = r.getElementsByTagName('w:rFonts')[0].getAttribute('w:ascii');
                    var rFontSize = r.getElementsByTagName('w:sz')[0].getAttribute('w:val');
                    var t = r.getElementsByTagName('w:t')[0];
                    let tText = `<span style="font-family:${rFontFamily};font-size:${rFontSize/100}rem">` +
                                t.textContent + 
                                `</span>`;
                    rTextArray.push(tText);
                }
        textContent = rTextArray.join('');
        return textContent
    }
        
        /**
         * 无论是p还是table最终还是会到这个函数，用于取出最后的文字内容
         * @param {*} pDom 处理<w:p>标签对应的DOM,这个标签和tbl是互斥的
         * @return {string} htmlStr 返回table标签对应的html字符串  
         */
        var pFn = function(pDom){
            let rArray = getDirectDomsByTagName(pDom,'w:r');
            return '<p>'+rFn(rArray)+'</p>';
        }
        
        /**
         * 
         * @param {*} dom DOM子树根节点
         * @return htmlStr 字符串 
         */
        var wanderDom = function(dom){
            let htmlStr = '',childrens = dom.children,len = childrens.length;
            for(let i=0; i<len;i++){
                let children = childrens[i];
                let tagName = children.tagName;
                switch(tagName){
                    case 'w:tbl': htmlStr= htmlStr + tblFn(children) ;break;
                    case 'w:p': htmlStr= htmlStr + pFn(children) ;break;
                    default:break ;
                }
            }
            return htmlStr;
        }
        
        
        
        /**
         * 
         * @param {*} xmlDoc 整个XML的DOM树
         * @return htmlStr 字符串 
         */
        let convert = function(xmlDoc){
            let dom = xmlDoc.getElementsByTagName('w:body')[0];
            return wanderDom(dom);
        }
    </script>

step 2: Use the code to convert your .docx to xml sting in your .xml file

var admZip = require('adm-zip');
var fs = require('fs');
var path = require('path');
var resultList = [];
const tableName = [
    'test.docx' // replace your docx name here
];
tableName.forEach((item, index)=>{
    var absPath = path.join(__dirname,item);
    fs.exists(absPath, function(exists){
        if(exists){
            const zip = new admZip(absPath);
            var contentXml = zip.readAsText("word/document.xml");
            var len = item.length-1;
            var name = item.slice(0,len-4) + ".xml"
            fs.writeFileSync(name,contentXml)           
        }else{
            callback(resultList)
        }
    })
});

step 3: Take your .xml file's string in variables str, and excute the code in your browser. the res is your html string, you can put it to the template html.

var str =`${your xml sting}`
var res = convert(loadXML(str));

step4:template example

html-body:

<div class="container">{res}</div>

head:

<script>
        (function (doc, win) {
            var docEl = doc.documentElement,
                resizeEvt = 'orientationchange' in window ? 'orientationchange' : 'resize',
                recalc = function () {
                    var clientWidth = docEl.clientWidth;
                    if (!clientWidth) return;
                    if (clientWidth >= 640) {
                        docEl.style.fontSize = '100px';
                    } else {
                        docEl.style.fontSize = 100 * (clientWidth / 750) + 'px';
                        var div = document.createElement('div');
                        div.style.width = '1.4rem';
                        div.style.height = '0';
                        document.body.appendChild(div);
                        var ideal = 140 * clientWidth / 750;
                        var rmd = (div.clientWidth / ideal);
                        if (rmd > 1.2 || rmd < 0.8) {
                            docEl.style.fontSize = 100 * (clientWidth / 750) / rmd + 'px';
                        }
                        document.body.removeChild(div);
                    }
                };
            if (!doc.addEventListener) return;
            win.addEventListener(resizeEvt, recalc, false);
            doc.addEventListener('DOMContentLoaded', recalc, false);
        })(document, window);
    </script>
    <style>
        body,
        html,
        table {
            padding: 0;
            margin: 0;
        }

        p {
            margin: 0.14rem 0;
            list-style-type: none;
            font-family: PingFang-SC-Medium;
            font-size: 0.28rem;
            color: #848484;
            letter-spacing: 0;
            text-align: justify;
            line-height: 0.48rem;
            padding: 0 0.3rem;
        }

        td {
            border: 1px solid #000;
            text-align: left;
            vertical-align: middle;
            font-family: PingFang-SC-Medium;
            font-size: 0.28rem;
            color: #848484;
        }

        .container>table {
            width: 6.9rem;
        }
        table {
            border-collapse: collapse;
        }

        .container {
            padding: 0.3rem
        }
    </style>