Java解析word中的表格或者文本

接到一个需求，要将word里面表格某些列的字段解析出来，这些列名是固定的，但是表格可能增加列，因此不能直接用列对应的索引去直接取列，要做成动态的。

整体思路就是先解析出word里面的表格，因为表格可能有多个，找到自己要的那个表格，再根据要的列名找到这些列的索引，找到之后遍历表格设置对象属性

1、maven依赖

<dependency><groupId>org.apache.poi</groupId><artifactId>poi-scratchpad</artifactId><version>3.17</version></dependency><!--POI包 --><dependency><groupId>org.apache.poi</groupId><artifactId>poi-ooxml</artifactId><version>3.17</version></dependency>

2、读取流数据

@PostMapping(value = "/read")
public void upload(@RequestPart("file") MultipartFile multipartFile) throws IOException {handlerDocx(multipartFile.getInputStream());
}

 3、处理逻辑

    //解析字段关联的对象字段private static final Map<String, String> needColumnsMap = new HashMap<>();static {needColumnsMap.put("姓名", "name");needColumnsMap.put("单位", "unit");needColumnsMap.put("月数", "month");}private void handlerDocx(InputStream inputStream) throws IOException {XWPFDocument xwpf = new XWPFDocument(inputStream);// 获取word中的所有段落与表格List<IBodyElement> elements = xwpf.getBodyElements();// 解析表格后续不解析for (IBodyElement element : elements) {// 段落if (element instanceof XWPFParagraph) {String paragraphText = DocUtils.getParagraphText((XWPFParagraph) element);log.info("paragraphText = {}", paragraphText);} else if (element instanceof XWPFTable) {// 表格List<List<String>> tableText= DocUtils.getTableText((XWPFTable) element);log.info("tableText = {}", tableText);//列索引关联的字段Map<Integer, String> needColumnIndexRelaName = new HashMap<>();for (List<String> strings : tableText) {if (strings.containsAll(needColumnsMap.keySet())) {//找出每个字段的索引，再找出这些索引对应的值for (String needColumn : needColumnsMap.keySet()) {needColumnIndexRelaName.put(strings.indexOf(needColumn), needColumnsMap.get(needColumn));}}}if (CollUtil.isNotEmpty(needColumnIndexRelaName)) {List<UserInfo> userInfos = new ArrayList<>();for (List<String> strings : tableText) {//过滤掉表头if (strings.containsAll(needColumnsMap.keySet())) {continue;}UserInfo info = new UserInfo();//根据列索引关联的字段设置对象属性for (Integer needColumnIndex : needColumnIndexRelaName.keySet()) {ReflectUtil.setFieldValue(info, needColumnIndexRelaName.get(needColumnIndex), strings.get(needColumnIndex));}userInfos.add(info);}log.info("{}", userInfos);}} else {log.info("其他内容");}}}}@Data
class UserInfo {@TableField(value = "姓名")private String name;@TableField(value = "单位")private String unit;@TableField(value = "月数")private String month;
}

4、工具类

public class DocUtils {/*** 获取表格内容* @param table* @return*/public static List<List<String>> getTableText(XWPFTable table) {List<List<String>> result = new ArrayList<>();List<XWPFTableRow> rows = table.getRows();for (XWPFTableRow row : rows) {String key = null;List<String> list = new ArrayList<>();int i = 0;List<XWPFTableCell> cells = row.getTableCells();for (XWPFTableCell cell : cells) {// 简单获取内容（简单方式是不能获取字体对齐方式的）StringBuffer sb = new StringBuffer();// 一个单元格可以理解为一个word文档，单元格里也可以加段落与表格List<XWPFParagraph> paragraphs = cell.getParagraphs();for (XWPFParagraph paragraph : paragraphs) {sb.append(DocUtils.getParagraphText(paragraph));}if (i == 0) {key = sb.toString();list.add(key == null || Objects.deepEquals(key, "") ? null : key.replace(",", ""));} else {String value = sb.toString();list.add(value == null || Objects.deepEquals(value, "") ? null : value.replace(",", ""));}i++;}result.add(list);}return result;}/*** docx 获取段落字符串* 获取段落内容** @param paragraph*/public static String getParagraphText(XWPFParagraph paragraph) {StringBuffer runText = new StringBuffer();// 获取段落中所有内容List<XWPFRun> runs = paragraph.getRuns();if (runs.size() == 0) {return runText.toString();}for (XWPFRun run : runs) {runText.append(run.text());}return runText.toString();}/*** doc 格式的字段解析表格* @param tb* @return*/public static Map<String, List<String>> getTabelDocText(Table tb) {Map<String, List<String>> result = new HashMap<>();//迭代行，默认从0开始,可以依据需要设置i的值,改变起始行数，也可设置读取到那行，只需修改循环的判断条件即可for (int i = 0; i < tb.numRows(); i++) {List<String> list = new ArrayList<>();int x = 0;TableRow tr = tb.getRow(i);String key = null;//迭代列，默认从0开始for (int j = 0; j < tr.numCells(); j++) {//取得单元格TableCell td = tr.getCell(j);StringBuffer sb = new StringBuffer();//取得单元格的内容for (int k = 0; k < td.numParagraphs(); k++) {Paragraph paragraph = td.getParagraph(k);String s = paragraph.text();//去除后面的特殊符号if (null != s && !"".equals(s)) {s = s.substring(0, s.length() - 1);}sb.append(s);}if (x == 0) {key = sb.toString();} else {String value = sb.toString();list.add(value == null || Objects.deepEquals(value, "") ? null : value.replace(",", ""));}x++;}result.put(key, list);}return result;}
}

5、最后表格里的字段映射到实体类的结果

 [UserInfo(name=王, unit=电科院, month=4), UserInfo(name=邵, unit=电科院, month=4), UserInfo(name=王, unit=电科院, month=12), UserInfo(name=张, unit=有限公司, month=4) ]