poi-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From talli...@apache.org
Subject svn commit: r1845138 - in /poi: site/src/documentation/content/xdocs/changes.xml trunk/src/java/org/apache/poi/poifs/macros/VBAMacroReader.java trunk/src/testcases/org/apache/poi/poifs/macros/TestVBAMacroReader.java trunk/test-data/spreadsheet/62624.bin
Date Mon, 29 Oct 2018 15:48:44 GMT
Author: tallison
Date: Mon Oct 29 15:48:44 2018
New Revision: 1845138

URL: http://svn.apache.org/viewvc?rev=1845138&view=rev
Log:
bug 62624 -- fix npe by adding processing of module name mapping in VBAMacroReader

Added:
    poi/trunk/test-data/spreadsheet/62624.bin   (with props)
Modified:
    poi/site/src/documentation/content/xdocs/changes.xml
    poi/trunk/src/java/org/apache/poi/poifs/macros/VBAMacroReader.java
    poi/trunk/src/testcases/org/apache/poi/poifs/macros/TestVBAMacroReader.java

Modified: poi/site/src/documentation/content/xdocs/changes.xml
URL: http://svn.apache.org/viewvc/poi/site/src/documentation/content/xdocs/changes.xml?rev=1845138&r1=1845137&r2=1845138&view=diff
==============================================================================
--- poi/site/src/documentation/content/xdocs/changes.xml (original)
+++ poi/site/src/documentation/content/xdocs/changes.xml Mon Oct 29 15:48:44 2018
@@ -95,6 +95,7 @@
         <summary-item>Upgrade to XMLBeans 3.0.2</summary-item>
       </summary>
       <actions>
+        <action dev="PD" type="fix" fixes-bug="62624" context="POI_Overall">Handle
module name mapping in VBAMacroReader</action>
         <action dev="PD" type="fix" fixes-bug="62859" context="XWPF">Rare NPE while
creating XWPFSDTContent</action>
         <action dev="PD" type="add" fixes-bug="62373" context="SS_Common">Support for
FREQUENCY function</action>
         <action dev="PD" type="fix" fixes-bug="62831" context="POI_Overall">WorkbookFactory.create
support for subclass of File, eg from JFileChooser</action>

Modified: poi/trunk/src/java/org/apache/poi/poifs/macros/VBAMacroReader.java
URL: http://svn.apache.org/viewvc/poi/trunk/src/java/org/apache/poi/poifs/macros/VBAMacroReader.java?rev=1845138&r1=1845137&r2=1845138&view=diff
==============================================================================
--- poi/trunk/src/java/org/apache/poi/poifs/macros/VBAMacroReader.java (original)
+++ poi/trunk/src/java/org/apache/poi/poifs/macros/VBAMacroReader.java Mon Oct 29 15:48:44
2018
@@ -20,9 +20,19 @@ package org.apache.poi.poifs.macros;
 import static org.apache.poi.util.StringUtil.endsWithIgnoreCase;
 import static org.apache.poi.util.StringUtil.startsWithIgnoreCase;
 
-import java.io.*;
+import java.io.ByteArrayInputStream;
+import java.io.ByteArrayOutputStream;
+import java.io.Closeable;
+import java.io.EOFException;
+import java.io.File;
+import java.io.FileInputStream;
+import java.io.IOException;
+import java.io.InputStream;
+import java.io.InputStreamReader;
 import java.nio.charset.Charset;
+import java.nio.charset.StandardCharsets;
 import java.util.HashMap;
+import java.util.LinkedHashMap;
 import java.util.Map;
 import java.util.zip.ZipEntry;
 import java.util.zip.ZipInputStream;
@@ -39,6 +49,8 @@ import org.apache.poi.util.CodePageUtil;
 import org.apache.poi.util.HexDump;
 import org.apache.poi.util.IOUtils;
 import org.apache.poi.util.LittleEndian;
+import org.apache.poi.util.POILogFactory;
+import org.apache.poi.util.POILogger;
 import org.apache.poi.util.RLEDecompressingInputStream;
 import org.apache.poi.util.StringUtil;
 
@@ -56,6 +68,8 @@ import org.apache.poi.util.StringUtil;
  * @since 3.15-beta2
  */
 public class VBAMacroReader implements Closeable {
+    private static final POILogger LOGGER = POILogFactory.getLogger(VBAMacroReader.class);
+
     protected static final String VBA_PROJECT_OOXML = "vbaProject.bin";
     protected static final String VBA_PROJECT_POIFS = "VBA";
 
@@ -111,8 +125,13 @@ public class VBAMacroReader implements C
 
     public Map<String, Module> readMacroModules() throws IOException {
         final ModuleMap modules = new ModuleMap();
+        //ascii -> unicode mapping for module names
+        //preserve insertion order
+        final Map<String, String> moduleNameMap = new LinkedHashMap<>();
+
         findMacros(fs.getRoot(), modules);
-        findProjectProperties(fs.getRoot(), modules);
+        findModuleNameMap(fs.getRoot(), moduleNameMap, modules);
+        findProjectProperties(fs.getRoot(), moduleNameMap, modules);
 
         Map<String, Module> moduleSources = new HashMap<>();
         for (Map.Entry<String, ModuleImpl> entry : modules.entrySet()) {
@@ -327,16 +346,33 @@ public class VBAMacroReader implements C
         }
     }
 
-    protected void findProjectProperties(DirectoryNode node, ModuleMap modules) throws IOException
{
+    protected void findProjectProperties(DirectoryNode node, Map<String, String> moduleNameMap,
ModuleMap modules) throws IOException {
         for (Entry entry : node) {
             if ("project".equalsIgnoreCase(entry.getName())) {
                 DocumentNode document = (DocumentNode)entry;
                 DocumentInputStream dis = new DocumentInputStream(document);
-                readProjectProperties(dis, modules);
+                readProjectProperties(dis, moduleNameMap, modules);
+            } else {
+                for (Entry child : node) {
+                    if (child instanceof DirectoryNode) {
+                        findProjectProperties((DirectoryNode)child, moduleNameMap, modules);
+                    }
+                }
+
+            }
+        }
+    }
+
+    protected void findModuleNameMap(DirectoryNode node, Map<String, String> moduleNameMap,
ModuleMap modules) throws IOException {
+        for (Entry entry : node) {
+            if ("projectwm".equalsIgnoreCase(entry.getName())) {
+                DocumentNode document = (DocumentNode)entry;
+                DocumentInputStream dis = new DocumentInputStream(document);
+                readNameMapRecords(dis, moduleNameMap, modules.charset);
             } else {
                 for (Entry child : node) {
                     if (child instanceof DirectoryNode) {
-                        findProjectProperties((DirectoryNode)child, modules);
+                        findModuleNameMap((DirectoryNode)child, moduleNameMap, modules);
                     }
                 }
 
@@ -559,6 +595,75 @@ public class VBAMacroReader implements C
         return new ASCIIUnicodeStringPair(ascii, unicode);
     }
 
+    private static void readNameMapRecords(InputStream is, Map<String, String> moduleNames,
Charset charset) throws IOException {
+        //see 2.3.3 PROJECTwm Stream: Module Name Information
+        //multibytecharstring
+        String mbcs = null;
+        String unicode = null;
+        do {
+            try {
+                mbcs = readMBCS(is, charset);
+            } catch (EOFException e) {
+                return;
+            }
+            if (mbcs == null) {
+                return;
+            }
+            try {
+                unicode = readUnicode(is);
+            } catch (EOFException e) {
+                return;
+            }
+            if (mbcs != null && unicode != null) {
+                moduleNames.put(mbcs, unicode);
+            }
+        } while (mbcs != null && unicode != null);
+    }
+
+    private static String readUnicode(InputStream is) throws IOException {
+        //reads null-terminated unicode string
+        ByteArrayOutputStream bos = new ByteArrayOutputStream();
+        int b0 = is.read();
+        int b1 = is.read();
+
+        while ((b0 + b1) != 0) {
+            if (b0 == -1 || b1 == -1) {
+                throw new EOFException();
+            }
+
+            bos.write(b0);
+            bos.write(b1);
+            b0 = is.read();
+            b1 = is.read();
+        }
+        return new String (bos.toByteArray(), StandardCharsets.UTF_16LE);
+    }
+
+    //returns a string if any bytes are read or null if two 0x00 are read
+    private static String readMBCS(InputStream is, Charset charset) throws IOException {
+        ByteArrayOutputStream bos = new ByteArrayOutputStream();
+        int len = 0;
+        int b = is.read();
+        while (b != 0) {
+            ++len;
+            if (b == -1) {
+                throw new EOFException();
+            }
+            bos.write(b);
+            b = is.read();
+        }
+        if (len == 0) {
+            b = is.read();
+            if (b == -1) {
+                throw new EOFException();
+            }
+            if (b != 0) {
+                LOGGER.log(POILogger.WARN, "expected two 0x00 at end of module name map");
+            }
+            return null;
+        }
+        return new String(bos.toByteArray(), charset);
+    }
 
     /**
      * Read <tt>length</tt> bytes of MBCS (multi-byte character set) characters
from the stream
@@ -579,7 +684,8 @@ public class VBAMacroReader implements C
         return new String(buffer, 0, length, charset);
     }
 
-    protected void readProjectProperties(DocumentInputStream dis, ModuleMap modules) throws
IOException {
+    protected void readProjectProperties(DocumentInputStream dis,
+                                         Map<String, String> moduleNameMap, ModuleMap
modules) throws IOException {
         InputStreamReader reader = new InputStreamReader(dis, modules.charset);
         StringBuilder builder = new StringBuilder();
         char[] buffer = new char[512];
@@ -588,6 +694,9 @@ public class VBAMacroReader implements C
             builder.append(buffer, 0, read);
         }
         String properties = builder.toString();
+        //the module name map names should be in exactly the same order
+        //as the module names here. See 2.3.3 PROJECTwm Stream.
+        //At some point, we might want to enforce that.
         for (String line : properties.split("\r\n|\n\r")) {
             if (!line.startsWith("[")) {
                 String[] tokens = line.split("=");
@@ -595,22 +704,40 @@ public class VBAMacroReader implements C
                         && tokens[1].startsWith("\"") && tokens[1].endsWith("\""))
{
                     // Remove any double quotes
                     tokens[1] = tokens[1].substring(1, tokens[1].length() - 1);
-
                 }
-                if ("Document".equals(tokens[0])) {
+                if ("Document".equals(tokens[0]) && tokens.length > 1) {
                     String mn = tokens[1].substring(0, tokens[1].indexOf("/&H"));
-                    ModuleImpl module = modules.get(mn);
-                    module.moduleType = ModuleType.Document;
-                } else if ("Module".equals(tokens[0])) {
-                    ModuleImpl module = modules.get(tokens[1]);
-                    module.moduleType = ModuleType.Module;
-                } else if ("Class".equals(tokens[0])) {
-                    ModuleImpl module = modules.get(tokens[1]);
-                    module.moduleType = ModuleType.Class;
+                    ModuleImpl module = getModule(mn, moduleNameMap, modules);
+                    if (module != null) {
+                        module.moduleType = ModuleType.Document;
+                    } else {
+                        LOGGER.log(POILogger.WARN, "couldn't find module with name: "+mn);
+                    }
+                } else if ("Module".equals(tokens[0]) && tokens.length > 1) {
+                    ModuleImpl module = getModule(tokens[1], moduleNameMap, modules);
+                    if (module != null) {
+                        module.moduleType = ModuleType.Module;
+                    } else {
+                        LOGGER.log(POILogger.WARN, "couldn't find module with name: "+tokens[1]);
+                    }
+                } else if ("Class".equals(tokens[0]) && tokens.length > 1) {
+                    ModuleImpl module = getModule(tokens[1], moduleNameMap, modules);
+                    if (module != null) {
+                        module.moduleType = ModuleType.Class;
+                    } else {
+                        LOGGER.log(POILogger.WARN, "couldn't find module with name: "+tokens[1]);
+                    }
                 }
             }
         }
     }
+    //can return null!
+    private ModuleImpl getModule(String moduleName, Map<String, String> moduleNameMap,
ModuleMap moduleMap) {
+        if (moduleNameMap.containsKey(moduleName)) {
+            return moduleMap.get(moduleNameMap.get(moduleName));
+        }
+        return moduleMap.get(moduleName);
+    }
 
     private String readUnicodeString(RLEDecompressingInputStream in, int unicodeNameRecordLength)
throws IOException {
         byte[] buffer = IOUtils.safelyAllocate(unicodeNameRecordLength, 20000);

Modified: poi/trunk/src/testcases/org/apache/poi/poifs/macros/TestVBAMacroReader.java
URL: http://svn.apache.org/viewvc/poi/trunk/src/testcases/org/apache/poi/poifs/macros/TestVBAMacroReader.java?rev=1845138&r1=1845137&r2=1845138&view=diff
==============================================================================
--- poi/trunk/src/testcases/org/apache/poi/poifs/macros/TestVBAMacroReader.java (original)
+++ poi/trunk/src/testcases/org/apache/poi/poifs/macros/TestVBAMacroReader.java Mon Oct 29
15:48:44 2018
@@ -286,4 +286,18 @@ public class TestVBAMacroReader {
         assertContains(content, "Attribute VB_Customizable = True");
         r.close();
     }
+
+    @Test
+    public void bug62624() throws IOException {
+        //macro comes from Common Crawl: HRLOXHGMGLFIJQQU27RIWXOARRHAAAAS
+        File f = POIDataSamples.getSpreadSheetInstance().getFile("62624.bin");
+        VBAMacroReader r = new VBAMacroReader(f);
+
+        Map<String, Module> macros = r.readMacroModules();
+        assertEquals(13, macros.size());
+        assertNotNull(macros.get("M\u00F3dulo1"));
+        assertContains(macros.get("M\u00F3dulo1").getContent(), "Calcula_tributos");
+        assertEquals(Module.ModuleType.Module, macros.get("M\u00F3dulo1").geModuleType());
+        r.close();
+    }
 }

Added: poi/trunk/test-data/spreadsheet/62624.bin
URL: http://svn.apache.org/viewvc/poi/trunk/test-data/spreadsheet/62624.bin?rev=1845138&view=auto
==============================================================================
Binary file - no diff available.

Propchange: poi/trunk/test-data/spreadsheet/62624.bin
------------------------------------------------------------------------------
    svn:mime-type = application/octet-stream



---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@poi.apache.org
For additional commands, e-mail: commits-help@poi.apache.org


Mime
View raw message