Selenium做自動化測試當然不能避免和Excel打交道。
由於Excel版本的關系，文件擴展名分xls和xlsx，
以往的經驗都是使用HSSFWorkbook和XSSFWorkbook來分別處理。具體的方式就是先判斷文件的類型，然后根據文件擴展名來選擇方法。
大概處理方式如下：
String extention= getExtention(path);
if (!EMPTY.equals(extention)) {
if (XLS.equals(extention)) {
return readXlsForAllSheets(path);
} else if (XLSX.equals(extention)) {
return readXlsxForAllSheets(path);
}
} else {
System.out.println(path + " is not a excel file.");
}
再接着實現readXlsForAllSheets和readXlsxForAllSheets兩個方法。
public Object[][] readXlsxForAllSheets(String path) throws IOException{
System.out.println(path);
FileInputStream is = new FileInputStream(path);
XSSFWorkbook xssfWorkbook = new XSSFWorkbook(is);
System.out.println("There are totally "+xssfWorkbook.getNumberOfSheets()+" sheets in the workbook.");
// Read the Sheet
List<Object[]> records1=new ArrayList<Object[]>();
for (int numSheet = 0; numSheet < xssfWorkbook.getNumberOfSheets(); numSheet++) {
XSSFSheet xssfSheet = xssfWorkbook.getSheetAt(numSheet);
int rowCount=xssfSheet.getLastRowNum()-xssfSheet.getFirstRowNum();
List<Object[]> records=new ArrayList<Object[]>();
String[] separative={"This is sheet "+xssfWorkbook.getSheetName(numSheet)};
records.add(separative);
for(int rowNum =1;rowNum<rowCount+1; rowNum++){
XSSFRow xssfRow=xssfSheet.getRow(rowNum);
String fields[]=new String[xssfRow.getLastCellNum()];
for (int colNum=0;colNum<xssfRow.getLastCellNum();colNum++){
XSSFCell xssfCell=xssfRow.getCell(colNum);
fields[colNum]=this.getXssfCellValue(xssfCell);
}
records.add(fields);
}
records1.addAll(records);
}
Object[][] results=new Object[records1.size()][];
for(int i=0;i<records1.size();i++){
results[i]=records1.get(i);
}
if (xssfWorkbook!=null){xssfWorkbook.close();}
return results;
}
public Object[][] readXlsForAllSheets(String path) throws IOException{
System.out.println(PROCESSING + path);
FileInputStream is = new FileInputStream(path);
HSSFWorkbook hssfWorkbook = new HSSFWorkbook(is);
System.out.println("There are totally "+hssfWorkbook.getNumberOfSheets()+" sheets in the workbook.");
// Read the Sheet
List<Object[]> records1=new ArrayList<Object[]>();
for (int numSheet = 0; numSheet < hssfWorkbook.getNumberOfSheets(); numSheet++) {
HSSFSheet hssfSheet = hssfWorkbook.getSheetAt(numSheet);
int rowCount=hssfSheet.getLastRowNum()-hssfSheet.getFirstRowNum();
List<Object[]> records=new ArrayList<Object[]>();
String[] separative={"This is sheet "+hssfWorkbook.getSheetName(numSheet)};
records.add(separative);
for(int rowNum =1;rowNum<rowCount+1; rowNum++){
HSSFRow xssfRow=hssfSheet.getRow(rowNum);
String fields[]=new String[xssfRow.getLastCellNum()];
for (int colNum=0;colNum<xssfRow.getLastCellNum();colNum++){
HSSFCell xssfCell=xssfRow.getCell(colNum);
fields[colNum]=this.getHssfCellValue(xssfCell);
}
records.add(fields);
}
records1.addAll(records);
}
Object[][] results=new Object[records1.size()][];
for(int i=0;i<records1.size();i++){
results[i]=records1.get(i);
}
if (hssfWorkbook!=null){hssfWorkbook.close();}
return results;
}
再實現上兩個方法中調用的getXssfCellValue和getHssfCellValue方法。
private String getXssfCellValue(XSSFCell xssfCell) {
String cellvalue="";
DataFormatter formatter = new DataFormatter();
if (null != xssfCell) {
switch (xssfCell.getCellType()) {
case XSSFCell.CELL_TYPE_NUMERIC: // 數字
if (org.apache.poi.ss.usermodel.DateUtil.isCellDateFormatted(xssfCell)) {
cellvalue = formatter.formatCellValue(xssfCell);
} else {
double value = xssfCell.getNumericCellValue();
int intValue = (int) value;
cellvalue = value - intValue == 0 ? String.valueOf(intValue) : String.valueOf(value);
}
break;
case XSSFCell.CELL_TYPE_STRING: // 字符串
cellvalue=xssfCell.getStringCellValue();
break;
case XSSFCell.CELL_TYPE_BOOLEAN: // Boolean
cellvalue=String.valueOf(xssfCell.getBooleanCellValue());
break;
case XSSFCell.CELL_TYPE_FORMULA: // 公式
cellvalue=String.valueOf(xssfCell.getCellFormula());
break;
case XSSFCell.CELL_TYPE_BLANK: // 空值
cellvalue="";
break;
case XSSFCell.CELL_TYPE_ERROR: // 故障
cellvalue="";
break;
default:
cellvalue="UNKNOWN TYPE";
break;
}
} else {
System.out.print("-");
}
return cellvalue.trim();
}
private String getHssfCellValue(HSSFCell hssfCell) {
String cellvalue="";
DataFormatter formatter = new DataFormatter();
if (null != hssfCell) {
switch (hssfCell.getCellType()) {
case HSSFCell.CELL_TYPE_NUMERIC: // 數字
if (org.apache.poi.ss.usermodel.DateUtil.isCellDateFormatted(hssfCell)) {
cellvalue = formatter.formatCellValue(hssfCell);
} else {
double value = hssfCell.getNumericCellValue();
int intValue = (int) value;
cellvalue = value - intValue == 0 ? String.valueOf(intValue) : String.valueOf(value);
}
break;
case HSSFCell.CELL_TYPE_STRING: // 字符串
cellvalue=hssfCell.getStringCellValue();
break;
case HSSFCell.CELL_TYPE_BOOLEAN: // Boolean
cellvalue=String.valueOf(hssfCell.getBooleanCellValue());
break;
case HSSFCell.CELL_TYPE_FORMULA: // 公式
cellvalue=String.valueOf(hssfCell.getCellFormula());
break;
case HSSFCell.CELL_TYPE_BLANK: // 空值
cellvalue="";
break;
case HSSFCell.CELL_TYPE_ERROR: // 故障
cellvalue="";
break;
default:
cellvalue="UNKNOWN TYPE";
break;
}
} else {
System.out.print("-");
}
return cellvalue.trim();
}
最終整個解析Excel文件的功能才算完成，我們需要實現4個方法readXlsForAllSheets和readXlsxForAllSheets，getXssfCellValue和getHssfCellValue，那么有沒有更加簡單實用的方法呢？
下面要介紹的是POI jar包提供的WorkbookFactory類。需要加載poi-ooxm-3.15.jar到build path。
只需要兩行就可以實例化workbook，而不用管它是xls還是xlsx。
inStream = new FileInputStream(new File(filePath));
Workbook workBook = WorkbookFactory.create(inStream);
后續可以直接操作sheet，Row,Cell,也不用管文件類型。
目前還沒有發現這種方法的缺點。