moparisthebest/poi
1
0
Fork 0
Code Issues Pull Requests Releases 1 Wiki Activity

add convenience method for files to FileMagic

Browse Source 
git-svn-id: https://svn.apache.org/repos/asf/poi/trunk@1830782 13f79535-47bb-0310-9956-ffa450edef68
This commit is contained in:
Andreas Beeker 2018-05-02 18:33:23 +00:00
parent 17fdb3188c
commit 5961b640a5
2 changed files with 92 additions and 65 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

19
src/java/org/apache/poi/poifs/filesystem/FileMagic.java
View File

@ -22,6 +22,8 @@ import static org.apache.poi.poifs.common.POIFSConstants.RAW_XML_FILE_HEADER;
import static java.nio.charset.StandardCharsets.UTF_8;
import java.io.BufferedInputStream;
import java.io.File;
import java.io.FileInputStream;
import java.io.IOException;
import java.io.InputStream;
@ -117,6 +119,23 @@ public enum FileMagic {
return UNKNOWN;
}
/**
* Get the file magic of the supplied {@link File}<p>
*
* Even if this method returns {@link FileMagic#UNKNOWN} it could potentially mean,
* that the ZIP stream has leading junk bytes
*
* @param inp a file to be identified
*/
public static FileMagic valueOf(final File inp) throws IOException {
try (FileInputStream fis = new FileInputStream(inp)) {
final byte[] data = IOUtils.toByteArray(fis, 8);
return FileMagic.valueOf(data);
}
}
/**
* Get the file magic of the supplied InputStream (which MUST
* support mark and reset).<p>

80
src/testcases/org/apache/poi/poifs/filesystem/TestOfficeXMLException.java
View File

@ -18,8 +18,12 @@
package org.apache.poi.poifs.filesystem;
import static org.apache.poi.POITestCase.assertContains;
import static org.junit.Assert.assertArrayEquals;
import static org.junit.Assert.assertEquals;
import static org.junit.Assert.assertNotEquals;
import java.io.ByteArrayInputStream;
import java.io.File;
import java.io.IOException;
import java.io.InputStream;
import java.util.Arrays;
@ -27,90 +31,94 @@ import java.util.Arrays;
import org.apache.poi.hssf.HSSFTestDataSamples;
import junit.framework.TestCase;
import org.junit.Rule;
import org.junit.Test;
import org.junit.rules.ExpectedException;
/**
* Class to test that POIFS complains when given an Office 2003 XML
* of Office Open XML (OOXML, 2007+) document
*/
public class TestOfficeXMLException extends TestCase {
public class TestOfficeXMLException {
@Rule
public ExpectedException thrown = ExpectedException.none();
private static InputStream openSampleStream(String sampleFileName) {
return HSSFTestDataSamples.openSampleFileStream(sampleFileName);
}
public void testOOXMLException() throws IOException
{
InputStream in = openSampleStream("sample.xlsx");
try {
new POIFSFileSystem(in).close();
fail("expected exception was not thrown");
} catch(OfficeXmlFileException e) {
// expected during successful test
assertContains(e.getMessage(), "The supplied data appears to be in the Office 2007+ XML");
assertContains(e.getMessage(), "You are calling the part of POI that deals with OLE2 Office Documents");
@Test
public void testOOXMLException() throws IOException {
thrown.expect(OfficeXmlFileException.class);
thrown.expectMessage("You are calling the part of POI that deals with OLE2 Office Documents");
try (InputStream in = openSampleStream("sample.xlsx");
POIFSFileSystem fs = new POIFSFileSystem(in)) {
}
}
@Test
public void test2003XMLException() throws IOException {
InputStream in = openSampleStream("SampleSS.xml");
thrown.expect(NotOLE2FileException.class);
thrown.expectMessage("The supplied data appears to be a raw XML file");
try {
new POIFSFileSystem(in).close();
fail("expected exception was not thrown");
} catch(NotOLE2FileException e) {
// expected during successful test
assertContains(e.getMessage(), "The supplied data appears to be a raw XML file");
assertContains(e.getMessage(), "Formats such as Office 2003 XML");
try (InputStream in = openSampleStream("SampleSS.xml");
POIFSFileSystem fs = new POIFSFileSystem(in)) {
}
}
@Test
public void testDetectAsPOIFS() throws IOException {
// ooxml file isn't
confirmIsPOIFS("SampleSS.xlsx", false);
confirmIsPOIFS("SampleSS.xlsx", FileMagic.OOXML);
// 2003 xml file isn't
confirmIsPOIFS("SampleSS.xml", false);
confirmIsPOIFS("SampleSS.xml", FileMagic.XML);
// xls file is
confirmIsPOIFS("SampleSS.xls", true);
confirmIsPOIFS("SampleSS.xls", FileMagic.OLE2);
// older biff formats aren't
confirmIsPOIFS("testEXCEL_3.xls", false);
confirmIsPOIFS("testEXCEL_4.xls", false);
confirmIsPOIFS("testEXCEL_3.xls", FileMagic.BIFF3);
confirmIsPOIFS("testEXCEL_4.xls", FileMagic.BIFF4);
// newer excel formats are
confirmIsPOIFS("testEXCEL_5.xls", true);
confirmIsPOIFS("testEXCEL_95.xls", true);
confirmIsPOIFS("testEXCEL_5.xls", FileMagic.OLE2);
confirmIsPOIFS("testEXCEL_95.xls", FileMagic.OLE2);
// text file isn't
confirmIsPOIFS("SampleSS.txt", false);
confirmIsPOIFS("SampleSS.txt", FileMagic.UNKNOWN);
}
private void confirmIsPOIFS(String sampleFileName, boolean expectedResult) throws IOException {
try (InputStream in = FileMagic.prepareToCheckMagic(openSampleStream(sampleFileName))) {
assertEquals(expectedResult, FileMagic.valueOf(in) == FileMagic.OLE2);
}
private void confirmIsPOIFS(String sampleFileName, FileMagic expected) throws IOException {
final File file = HSSFTestDataSamples.getSampleFile(sampleFileName);
assertEquals(expected, FileMagic.valueOf(file));
}
@Test
public void testFileCorruption() throws Exception {
// create test InputStream
byte[] testData = { (byte)1, (byte)2, (byte)3 };
byte[] testData = {1, 2, 3};
InputStream testInput = new ByteArrayInputStream(testData);
// detect header
InputStream in = FileMagic.prepareToCheckMagic(testInput);
assertFalse(FileMagic.valueOf(in) == FileMagic.OLE2);
assertNotEquals(FileMagic.OLE2, FileMagic.valueOf(in));
// check if InputStream is still intact
byte[] test = new byte[3];
assertEquals(3, in.read(test));
assertTrue(Arrays.equals(testData, test));
assertArrayEquals(testData, test);
assertEquals(-1, in.read());
}
@Test
public void testFileCorruptionOPOIFS() throws Exception {
// create test InputStream
@ -119,13 +127,13 @@ public class TestOfficeXMLException extends TestCase {
// detect header
InputStream in = FileMagic.prepareToCheckMagic(testInput);
assertFalse(FileMagic.valueOf(in) == FileMagic.OLE2);
assertNotEquals(FileMagic.OLE2, FileMagic.valueOf(in));
assertEquals(FileMagic.UNKNOWN, FileMagic.valueOf(in));
// check if InputStream is still intact
byte[] test = new byte[3];
assertEquals(3, in.read(test));
assertTrue(Arrays.equals(testData, test));
assertArrayEquals(testData, test);
assertEquals(-1, in.read());
}
}