Apache POI - 读取修改 Excel 文件

Question

每当我使用Apatche POI打开Excel文件时，文件都会被修改，即使我只是在读取文件而不进行任何修改。

以这样的测试代码为例。

public class ApachePoiTest {

    @Test
    public void readingShouldNotModifyFile() throws Exception {
        final File testFile = new File("C:/work/src/test/resources/Book2.xlsx");
        final byte[] originalChecksum = calculateChecksum(testFile);
        Assert.assertTrue("Calculating checksum modified file",
            MessageDigest.isEqual(originalChecksum, calculateChecksum(testFile)));
        try (Workbook wb = WorkbookFactory.create(testFile)) {
            Assert.assertNotNull("Reading file with Apache POI", wb);
        }
        Assert.assertTrue("Reading file with Apache POI modified file",
            MessageDigest.isEqual(originalChecksum, calculateChecksum(testFile)));
    }

    @Test
    public void readingInputStreamShouldNotModifyFile() throws Exception {
        final File testFile = new File("C:/work/src/test/resources/Book2.xlsx");
        final byte[] originalChecksum = calculateChecksum(testFile);
        Assert.assertTrue("Calculating checksum modified file",
            MessageDigest.isEqual(originalChecksum, calculateChecksum(testFile)));
        try (InputStream is = new FileInputStream(testFile); Workbook wb = WorkbookFactory.create(is)) {
            Assert.assertNotNull("Reading file with Apache POI", wb);
        }
        Assert.assertTrue("Reading file with Apache POI modified file",
            MessageDigest.isEqual(originalChecksum, calculateChecksum(testFile)));
    }

    private byte[] calculateChecksum(final File file) throws Exception {
        final MessageDigest md = MessageDigest.getInstance("MD5");
        md.reset();
        try (InputStream is = new FileInputStream(file)) {
            final byte[] bytes = new byte[2048];
            int numBytes;
            while ((numBytes = is.read(bytes)) != -1) {
                md.update(bytes, 0, numBytes);
            }
            return md.digest();
        }
    }
}

Test readingShouldNotModifyFile 总是失败，因为 Apache POI 总是会修改该文件。在测试使用MS Office新创建的空白Excel文件时，Apache POI将文件从8.1 kb切割到6.2 kb并损坏文件。

测试对象:

<dependency>
    <groupId>org.apache.poi</groupId>
    <artifactId>poi-ooxml</artifactId>
    <version>3.15</version>
</dependency>

以及版本3.12

我是否可以防止Apache POI通过其他方式修改我的文件，然后传递< code>InputStream而不是< code>File。我不想传递< code>InputStream，因为我担心Apache警告它会占用更多内存，并且对< code>InputStream有一些特定的要求。

Answer 1

您的问题是您没有传入只读标志，因此Apache POI默认打开文件读/写。

您需要使用重载的 WorkbookFactory.create 方法，该方法采用只读标志集，该只读标志设置为 true

改变路线

try (InputStream is = new FileInputStream(testFile); Workbook wb = WorkbookFactory.create(is)) {

到

try (IWorkbook wb = WorkbookFactory.create(testFile,null,true)) {

并且您的文件将以只读方式打开，没有任何更改