java实现excel文件转换为xml文件

一、导包：

<dependency>
<groupid>org.apache.poi</groupid>
<artifactid>poi</artifactid>
<version>3.17</version>
</dependency>
<dependency>
<groupid>org.apache.poi</groupid>
<artifactid>poi-examples</artifactid>
<version>3.16</version>
</dependency>
<dependency>
<groupid>org.jdom</groupid>
<artifactid>jdom</artifactid>
<version>1.1.3</version>
</dependency>

二、代码部分：

public class analysisetox {

    public static void main(string[] args) {
        try {
            system.out.println("=============");
            // 用输入流从本地拿到对应的excel文件
            inputstream stream = new fileinputstream("c:\\users\\administrator\\desktop\\试题.xlsx");
            // 指定要生成的xml的路径，并构建文件对象
            file f = new file("test.xml");// 新建个file对象把解析之后得到的xml存入改文件中
            writerxml(stream, f);// 将数据以xml形式写入文本
        } catch (filenotfoundexception e) {
            system.out.println("未找到指定路径的文件!");
            e.printstacktrace();
        } catch (ioexception e) {
            // todo auto-generated catch block
            e.printstacktrace();
        }
    }

    private static void writerxml(inputstream stream, file f) throws ioexception {
        system.out.println("into writerxml");
        fileoutputstream fo = new fileoutputstream(f);// 得到输入流
        document doc = readexcell(stream);// 读取excel函数
        format format = format.getcompactformat().setencoding("utf-8").setindent("");
        xmloutputter xmlout = new xmloutputter(format);// 在元素后换行，每一层元素缩排四格
        xmlout.output(doc, fo);
        fo.close();
    }

    private static document readexcell(inputstream stream) {
        system.out.println("into readexcell");
        // 设置根<tax_institutions></tax_institutions>元素
        element root = new element("tax_institutions");
        document doc = new document(root);
        try {
            hssfworkbook hw = new hssfworkbook(stream);
            // 获取工作薄的个数，即一个excel文件中包含了多少个sheet工作簿
            int wblength = hw.getnumberofsheets();
            system.out.println("wblength=" + wblength);

            for (int i = 0; i < wblength; i++) {
                hssfsheet shee = hw.getsheetat(i);
                int length = shee.getlastrownum();
                system.out.println("行数：" + length);
                for (int j = 1; j <= length; j++) {
                    hssfrow row = shee.getrow(j);
                    if (row == null) {
                        continue;
                    }
                    int cellnum = row.getphysicalnumberofcells();// 获取一行中最后一个单元格的位置
                    system.out.println("列数cellnum：" + cellnum);
                    element e = null;
                    // 设置根元素下的并列元素<tax_institution></tax_institution>
                    e = new element("tax_institution");
                    // element[] es = new element[16];
                    for (int k = 0; k < cellnum; k++) {
                        hssfcell cell = row.getcell((short) k);
                        string temp = get(k);
                        system.out.print(k + " " + temp + ":");
                        element item = new element(temp);
                        if (cell == null) {
                            item.settext("");
                            e.addcontent(item);
                            cellnum++;// 如果存在空列，那么cellnum增加1，这一步很重要。
                            continue;
                        }

                        else {
                            string cellvalue = "";
                            switch (cell.getcelltype()) {
                            // 如果当前cell的type为numeric
                            case hssfcell.cell_type_numeric:
                            case hssfcell.cell_type_formula: {
                                // 判断当前的cell是否为date
                                if (hssfdateutil.iscelldateformatted(cell)) {
                                    // 如果是date类型则，转化为data格式

                                    // 方法1：这样子的data格式是带时分秒的：2011-10-12 0:00:00
                                    // cellvalue =
                                    cell.getdatecellvalue().tolocalestring();

                                    // 方法2：这样子的data格式是不带带时分秒的：2011-10-12
                                    date date = cell.getdatecellvalue();
                                    simpledateformat sdf = new simpledateformat("yyyy-mm-dd");
                                    cellvalue = sdf.format(date);
                                    item.settext(cellvalue);

                                }
                                // 如果是纯数字
                                else {
                                    // 取得当前cell的数值
                                    cellvalue = string.valueof((int) cell.getnumericcellvalue());
                                    item.settext(cellvalue);
                                }
                                break;
                            }
                            // 如果当前cell的type为strin
                            case hssfcell.cell_type_string:
                                // 取得当前的cell字符串
                                cellvalue = cell.getrichstringcellvalue().getstring();
                                item.settext(cellvalue);
                                break;
                            // 默认的cell值
                            default:
                                cellvalue = " ";
                                item.settext(cellvalue);
                            }
                            e.addcontent(item);
                            system.out.println(cellvalue);
                        }
                    }
                    root.addcontent(e);

                }

            }
        } catch (exception e) {
            e.printstacktrace();
        } finally {
            try {
                stream.close();
            } catch (ioexception e1) {
                e1.printstacktrace();
            }

        }
        return doc;
    }

    // 设置并列元素里的子元素名
    private static string get(int k) {
        string test = "";
        switch (k) {
        case 0:
            test = "org_name";
            break;
        case 1:
            test = "legal_mobile_phone";
            break;
        case 2:
            test = "org_address";
            break;
        case 3:
            test = "cert_type";
            break;
        case 4:
            test = "postal_code";
            break;
        case 5:
            test = "reg_sum";
            break;
        case 6:
            test = "business_scope";
            break;
        case 7:
            test = "social_credit_code";
            break;
        case 8:
            test = "reg_type";
            break;
        case 9:
            test = "legal_person_name";
            break;
        case 10:
            test = "cert_number";
            break;
        case 11:
            test = "found_time";
            break;
        case 12:
            test = "service_status";
            break;
        case 13:
            test = "staff_sum";
            break;
        case 14:
            test = "partner_sum";
            break;
        case 15:
            test = "is_branch_org";
            break;
        default:
        }
        return test;

    }

}

转载地址csdn：https://blog.csdn.net/weixin_40420734/article/details/79538772

 

注意：

excel版本保存文件的后缀名问题，如果是.xls，使用hssfworkbook；如果是xlsx，使用xssfworkbook。

版本错误会抛出officexmlfileexception！！！