From 680feb50ee6f9dbb26dc7bfb50a65f3347c3b7a8 Mon Sep 17 00:00:00 2001
From: Nick Burch <nick@apache.org>
Date: Sun, 30 Nov 2014 17:09:03 +0000
Subject: [PATCH] Correct older biff detection, and add unit tests for
 HSSFWorkbook giving helpful exceptions on the older formats

git-svn-id: https://svn.apache.org/repos/asf/poi/trunk@1642568 13f79535-47bb-0310-9956-ffa450edef68
---
 .../apache/poi/poifs/storage/HeaderBlock.java |  6 ++--
 .../poi/hssf/usermodel/TestHSSFWorkbook.java  | 36 +++++++++++++++++++
 2 files changed, 39 insertions(+), 3 deletions(-)

diff --git a/src/java/org/apache/poi/poifs/storage/HeaderBlock.java b/src/java/org/apache/poi/poifs/storage/HeaderBlock.java
index 560111fa0..06e9c4a1b 100644
--- a/src/java/org/apache/poi/poifs/storage/HeaderBlock.java
+++ b/src/java/org/apache/poi/poifs/storage/HeaderBlock.java
@@ -136,7 +136,7 @@ public final class HeaderBlock implements HeaderBlockConstants {
             if (_data[0] == 0x09 && _data[1] == 0x00 && // sid=0x0009
                 _data[2] == 0x04 && _data[3] == 0x00 && // size=0x0004
                 _data[4] == 0x00 && _data[5] == 0x00 && // unused
-               (_data[6] == 0x01 || _data[6] == 0x02 || _data[6] == 0x04) &&
+               (_data[6] == 0x10 || _data[6] == 0x20 || _data[6] == 0x40) &&
                 _data[7] == 0x00) {
                 // BIFF2 raw stream
                 throw new OldExcelFormatException("The supplied data appears to be in BIFF2 format. " +
@@ -145,7 +145,7 @@ public final class HeaderBlock implements HeaderBlockConstants {
             if (_data[0] == 0x09 && _data[1] == 0x02 && // sid=0x0209
                 _data[2] == 0x06 && _data[3] == 0x00 && // size=0x0006
                 _data[4] == 0x00 && _data[5] == 0x00 && // unused
-               (_data[6] == 0x01 || _data[6] == 0x02 || _data[6] == 0x04) &&
+               (_data[6] == 0x10 || _data[6] == 0x20 || _data[6] == 0x40) &&
                 _data[7] == 0x00) {
                 // BIFF3 raw stream
                 throw new OldExcelFormatException("The supplied data appears to be in BIFF3 format. " +
@@ -154,7 +154,7 @@ public final class HeaderBlock implements HeaderBlockConstants {
             if (_data[0] == 0x09 && _data[1] == 0x04 && // sid=0x0409
                 _data[2] == 0x06 && _data[3] == 0x00 && // size=0x0006
                 _data[4] == 0x00 && _data[5] == 0x00) { // unused
-                if (((_data[6] == 0x01 || _data[6] == 0x02 || _data[6] == 0x04) &&
+                if (((_data[6] == 0x10 || _data[6] == 0x20 || _data[6] == 0x40) &&
                       _data[7] == 0x00) ||
                     (_data[6] == 0x00 && _data[7] == 0x01)) {
                     // BIFF4 raw stream
diff --git a/src/testcases/org/apache/poi/hssf/usermodel/TestHSSFWorkbook.java b/src/testcases/org/apache/poi/hssf/usermodel/TestHSSFWorkbook.java
index a37751dec..2fabba9de 100644
--- a/src/testcases/org/apache/poi/hssf/usermodel/TestHSSFWorkbook.java
+++ b/src/testcases/org/apache/poi/hssf/usermodel/TestHSSFWorkbook.java
@@ -23,6 +23,7 @@ import static org.junit.Assert.assertNotNull;
 import static org.junit.Assert.assertNull;
 import static org.junit.Assert.assertTrue;
 import static org.junit.Assert.fail;
+import static org.apache.poi.POITestCase.assertContains;
 
 import java.io.ByteArrayInputStream;
 import java.io.ByteArrayOutputStream;
@@ -40,6 +41,7 @@ import org.apache.poi.ddf.EscherBSERecord;
 import org.apache.poi.hpsf.ClassID;
 import org.apache.poi.hssf.HSSFITestDataProvider;
 import org.apache.poi.hssf.HSSFTestDataSamples;
+import org.apache.poi.hssf.OldExcelFormatException;
 import org.apache.poi.hssf.model.HSSFFormulaParser;
 import org.apache.poi.hssf.model.InternalSheet;
 import org.apache.poi.hssf.model.InternalWorkbook;
@@ -561,6 +563,40 @@ public final class TestHSSFWorkbook extends BaseTestWorkbook {
 
         assertTrue(clsid1.equals(clsid2));
     }
+    
+    /**
+     * If we try to open an old (pre-97) workbook, we get a helpful
+     *  Exception give to explain what we've done wrong
+     */
+    @Test
+    public void helpfulExceptionOnOldFiles() throws Exception {
+        InputStream excel4 = POIDataSamples.getSpreadSheetInstance().openResourceAsStream("testEXCEL_4.xls");
+        try {
+            new HSSFWorkbook(excel4);
+            fail("Shouldn't be able to load an Excel 4 file");
+        } catch (OldExcelFormatException e) {
+            assertContains(e.getMessage(), "BIFF4");
+        }
+        excel4.close();
+        
+        InputStream excel5 = POIDataSamples.getSpreadSheetInstance().openResourceAsStream("testEXCEL_5.xls");
+        try {
+            new HSSFWorkbook(excel5);
+            fail("Shouldn't be able to load an Excel 5 file");
+        } catch (OldExcelFormatException e) {
+            assertContains(e.getMessage(), "BIFF5");
+        }
+        excel5.close();
+        
+        InputStream excel95 = POIDataSamples.getSpreadSheetInstance().openResourceAsStream("testEXCEL_95.xls");
+        try {
+            new HSSFWorkbook(excel95);
+            fail("Shouldn't be able to load an Excel 95 file");
+        } catch (OldExcelFormatException e) {
+            assertContains(e.getMessage(), "BIFF5");
+        }
+        excel95.close();
+    }
 
     /**
      * Tests that we can work with both {@link POIFSFileSystem}