当标题的单元格包含换行符时，Apache POI createTable 会生成损坏的文件

Question

I am using Apache POI 4.1.2 to create Excel files in Java. I have a piece of code that creates a table from existing cells and everything used to work fine, untill I had a linebreak inside a header's cell.

I tried to change the table's column name afterward but it didn't fix anything.

Below is a minimal piece of code to reproduce the problem:

    public void test() throws IOException {
        XSSFWorkbook wb = new XSSFWorkbook();
        XSSFSheet sheet = wb.createSheet();

        // headers
        XSSFRow headersRow = sheet.createRow(0);
        headersRow.createCell(0).setCellValue("Column1");
        headersRow.createCell(1).setCellValue("Column2");

        // a second row
        XSSFRow row = sheet.createRow(1);
        row.createCell(0).setCellValue(1);
        row.createCell(1).setCellValue(2);

        // create a table
        AreaReference area = wb.getCreationHelper().createAreaReference(
                new CellReference(sheet.getRow(0).getCell(0)),
                new CellReference(sheet.getRow(1).getCell(1))
        );
        XSSFTable table = sheet.createTable(area);

        // styling (no problem here)
        sheet.setColumnWidth(0, 5000);
        sheet.setColumnWidth(1, 5000);
        CTTable cttable = table.getCTTable();
        cttable.addNewTableStyleInfo();
        XSSFTableStyleInfo style = (XSSFTableStyleInfo) table.getStyle();
        style.setName("TableStyleMedium6");
        style.setShowColumnStripes(false);
        style.setShowRowStripes(true);
        cttable.addNewAutoFilter().setRef(area.formatAsString());
        CellStyle cellStyle = wb.createCellStyle();
        cellStyle.setWrapText(true);
        headersRow.getCell(0).setCellStyle(cellStyle);

        // this file is OK
        try (FileOutputStream outputStream = new FileOutputStream("C:\\tmp\\test.xlsx")) {
            wb.write(outputStream);
        }

        // add a line break in a header's cell
        headersRow.getCell(0).setCellValue("Column1\nwith a line break");
        // this file has a problem
        try (FileOutputStream outputStream = new FileOutputStream("C:\\tmp\\test2.xlsx")) {
            wb.write(outputStream);
        }

        // this doesn't fix anything
        table.getColumns().get(0).setName("Column1");
        try (FileOutputStream outputStream = new FileOutputStream("C:\\tmp\\test3.xlsx")) {
            wb.write(outputStream);
        }

        // neither does this
        cttable.getTableColumns().getTableColumnList().get(0).setName("Column1");
        try (FileOutputStream outputStream = new FileOutputStream("C:\\tmp\\test4.xlsx")) {
            wb.write(outputStream);
        }
    }

Excel loads text.xlsx properly, but complains about all other files:
We found a problem with some content...

After Excel fixes the files, everything is OK but I would like to get rid of the warning message.

Any help will be appreciated. Thanks

Answer 1

This is an inaccuracy with XSSFTable.updateHeaders . This method gets called while the table's XML gets written. This is because the table column names always must be synchronized with the cell contents. For example if the cell content is "Column1" and this cell is a column header of a table, then this tables column name also must be "Column1" (XML: <tableColumn id="1" name="Column1"/> ).

But for line feeds in column headers, there is a specialty. If the cell content is "Column1\\nwith a line break" and this cell is a column header of a table, then this tables column name must be XML as <tableColumn id="1" name="Column1_x000a_with a line break"/> . So "\\n" is replaced by " x000a ". Also "\\r" would must be replaced by " x000d ". This is because "\\r\\n" line breaks will not have the meaning of line break in XML .

So XSSFTable.java - updateHeaders would must be patched that way that "\\n" gets replaced by " x000a " and "\\r" gets replaced by " x000d ".

...
    public void updateHeaders() {
        XSSFSheet sheet = (XSSFSheet)getParent();
        CellReference ref = getStartCellReference();
        if (ref == null) return;

        int headerRow = ref.getRow();
        int firstHeaderColumn = ref.getCol();
        XSSFRow row = sheet.getRow(headerRow);
        DataFormatter formatter = new DataFormatter();

        if (row != null && row.getCTRow().validate()) {
            int cellnum = firstHeaderColumn;
            CTTableColumns ctTableColumns = getCTTable().getTableColumns();
            if(ctTableColumns != null) {
                for (CTTableColumn col : ctTableColumns.getTableColumnList()) {
                    XSSFCell cell = row.getCell(cellnum);
                    if (cell != null) {
                        String colName = formatter.formatCellValue(cell);
                        colName = colName.replace("\n", "_x000a_");
                        colName = colName.replace("\r", "_x000d_");
                        col.setName(colName);
                    }
                    cellnum++;
                }
            }
        }
        tableColumns = null;
        columnMap = null;
        xmlColumnPrs = null;
        commonXPath = null;
    }
...

Since XSSFTable.updateHeaders gets called while the table's XML gets written while XSSFWorkbook.write , there is no other way than patching this method. One does not have any chance to change table's XML while XSSFWorkbook.write .