PDFBOX-6178, PDFBOX-4076: handle #00 as valid; see testfile by Acrobat

git-svn-id: https://svn.apache.org/repos/asf/pdfbox/trunk@1932405 13f79535-47bb-0310-9956-ffa450edef68

Author: Maruan Sahyoun

pdfbox/pom.xml

@@ -933,6 +933,19 @@
                             <sha512>d39486af0614bd099167a6adaab833aed41a0ebec7b85b13b382a2fdb6fddbcaaea9ab26ed0a81b72822258c8dd66dd535fa5c76afd1e5a8b1bff7d81e890274</sha512>
                         </configuration>
                     </execution>
+                    <execution>
+                        <id>PDFBOX-6178-1</id>
+                        <phase>generate-test-resources</phase>
+                        <goals>
+                            <goal>wget</goal>
+                        </goals>
+                        <configuration>
+                            <url>https://issues.apache.org/jira/secure/attachment/13081311/form_selected_ASCII_NUL_acrobat.pdf</url>
+                            <outputDirectory>${project.build.directory}/pdfs</outputDirectory>
+                            <outputFileName>PDFBOX-6178-1.pdf</outputFileName>
+                            <sha512>83bc557e6f7d3e98de6e81168b2e2fb3def5025cc5fab1ddb3ef658505351c615253587f09b93503bb73a6225f3d3898be894e9d05dba8d464f4dd9c54514bc3</sha512>
+                        </configuration>
+                    </execution>
                 </executions>
             </plugin>
         </plugins>

pdfbox/src/main/java/org/apache/pdfbox/cos/COSName.java

@@ -715,15 +715,6 @@ public static COSName getPDFName(String aName)
      */
     public static COSName getPDFName(byte[] bytes)
     {
-        for (byte b : bytes)
-        {
-            if (b == 0)
-            {
-                throw new IllegalArgumentException(
-                        "PDF name bytes must not contain null (0x00) characters");
-            }
-        }
-
         // Wrap for lookup only to avoid unnecessary copying of the byte array for the key.
         ByteBuffer lookupKey = ByteBuffer.wrap(bytes);
 

pdfbox/src/main/java/org/apache/pdfbox/pdfparser/COSParser.java

@@ -1852,19 +1852,7 @@ protected COSName parseCOSName() throws IOException
                     String hex = Character.toString((char) ch1) + (char) ch2;
                     try
                     {
-                        // Beginning with PDF 1.2 a name object is an atomic symbol uniquely defined by a 
-                        // sequence of any characters (8-bit values) except null (character code 0)
-                        // Although not explicitly stated in the PDF specification, it is generally accepted that
-                        // the # escape sequence is used to represent characters that are not allowed in a name object,
-                        // such as the null byte (0x00). Therefore, we will throw an IOException if we encounter a #00 sequence
-                        // in a name object, as this would indicate an invalid name according to the PDF specification.
-                        // ASCII NUL (0x00) is already handled in BaseParser#isEndOfName
-                        int decoded = Integer.parseInt(hex, 16);
-                        if (decoded == 0)
-                        {
-                            throw new IOException("PDF name must not contain null byte (0x00), found #00 at offset " + source.getPosition());
-                        }
-                        buffer.write(decoded);
+                        buffer.write(Integer.parseInt(hex, 16));
                     }
                     catch (NumberFormatException e)
                     {

pdfbox/src/test/java/org/apache/pdfbox/cos/TestCOSName.java

@@ -17,7 +17,6 @@
 package org.apache.pdfbox.cos;
 
 import static org.junit.jupiter.api.Assertions.assertEquals;
-import static org.junit.jupiter.api.Assertions.assertThrows;
 import static org.junit.jupiter.api.Assertions.assertTrue;
 
 import java.io.ByteArrayOutputStream;
@@ -34,16 +33,6 @@ class TestCOSName
 {
     private static final File TARGETPDFDIR = new File("target/pdfs");
 
-    @Test
-    void testNullByteRejection()
-    {
-        // Null bytes should not be allowed in name bytes
-        byte[] bytesWithNull = new byte[] { 'N', 'a', 'm', 'e', 0, 'X' };
-        assertThrows(IllegalArgumentException.class, () -> {
-            COSName.getPDFName(bytesWithNull);
-        }, "getPDFName should reject bytes containing null (0x00)");
-    }
-
     /**
      * PDFBOX-4076: Check that characters outside of US_ASCII are not replaced with "?".
      * 
@@ -72,7 +61,7 @@ void PDFBox4076() throws IOException
     }
 
     /**
-     * PDFBOX-4076: Check that characters outside of US_ASCII are not replaced with "?".
+     * PDFBOX-6178: Ensure that names with escape sequences #xx are written as is.
      * 
      * @throws IOException 
      */
@@ -103,4 +92,35 @@ void PDFBox6178() throws IOException
             System.out.println(writtenKeys);
         }
     }
+
+        /**
+     * PDFBOX-6178: Ensure that names with escape sequences #xx are written as is.
+     * 
+     * @throws IOException 
+     */
+    @Test
+    void NameWithASCII_NUL() throws IOException
+    {
+        ByteArrayOutputStream baos = new ByteArrayOutputStream();
+
+        try (PDDocument document = Loader.loadPDF(new File(TARGETPDFDIR,"PDFBOX-6178-1.pdf"))) {
+            PDField field = document.getDocumentCatalog()
+                .getAcroForm(null)
+                .getField("Geschlecht");
+            
+            field.getWidgets()
+                .get(0).getAppearance().getNormalAppearance().getCOSObject()
+                .keySet().forEach(k -> {
+                    try {
+                        k.writePDF(baos);
+                    } catch (IOException e) {
+                        // ignored
+                    }
+                });
+
+            String writtenKeys = new String(baos.toByteArray(), "UTF-8");
+            assertTrue(writtenKeys.contains("/m#00nnlich"), "Output should be /m#00nnlich (with 0xE4 as hex escape)");
+            System.out.println(writtenKeys);
+        }
+    }
 }