UTF8UtilTest.java 文件源码

/**
 * Demonstrates that skipping incorrectly encoded character sequences
 * works because the stream is not checked for well-formedness.
 */
public void testSkippingInvalidEncodingWorks()
        throws IOException {
    // The array contains three valid characters and one invalid three-byte
    // representation that only has two bytes present.
    // When skipping, this sequence is (incorrectly) taken as a sequence of
    // three characters ('a' - some three byte character - 'a').
    // 0xef = 11101111, 0xb8 = 10111000
    byte[] data = {'a', (byte)0xef, (byte)0xb8, 'a', 'a'};
    byte[] dataWithLength =
        {0x0, 0x5, 'a', (byte)0xef, (byte)0xb8, 'a', 'a'};
    InputStream is = new ByteArrayInputStream(data);
    // This is actually incorrect, but does work currently.
    UTF8Util.skipFully(is, 3);
    // Verify that decoding this actually fails.
    DataInputStream dis = new DataInputStream(
                                new ByteArrayInputStream(dataWithLength));
    try {
        dis.readUTF();
        fail("UTF-8 expected to be invalid, read should fail");
    } catch (UTFDataFormatException udfe) {
        // This is expected, since the UTF-8 encoding is invalid
    }
}