Move tokenizer's position to the end of the stream after reading object. Minor refactoring

DEVSIX-6548

Author: ars18wrw

kernel/src/main/java/com/itextpdf/kernel/pdf/PdfReader.java

@@ -892,11 +892,15 @@ protected PdfObject readObject(boolean readAsDirect, boolean objStm) throws IOEx
                     do {
                         ch = tokens.read();
                     } while (ch == 32 || ch == 9 || ch == 0 || ch == 12);
-                    if (ch != '\n')
+                    if (ch != '\n') {
                         ch = tokens.read();
-                    if (ch != '\n')
+                    }
+                    if (ch != '\n') {
                         tokens.backOnePosition(ch);
-                    return new PdfStream(tokens.getPosition(), dict);
+                    }
+                    PdfStream pdfStream = new PdfStream(tokens.getPosition(), dict);
+                    tokens.seek(pdfStream.getOffset() + pdfStream.getLength());
+                    return pdfStream;
                 } else {
                     tokens.seek(pos);
                     return dict;
@@ -953,11 +957,13 @@ protected PdfDictionary readDictionary(boolean objStm) throws IOException {
         PdfDictionary dic = new PdfDictionary();
         while (true) {
             tokens.nextValidToken();
-            if (tokens.getTokenType() == PdfTokenizer.TokenType.EndDic)
+            if (tokens.getTokenType() == PdfTokenizer.TokenType.EndDic) {
                 break;
-            if (tokens.getTokenType() != PdfTokenizer.TokenType.Name)
+            }
+            if (tokens.getTokenType() != PdfTokenizer.TokenType.Name) {
                 tokens.throwError(
                         KernelExceptionMessageConstant.THIS_DICTIONARY_KEY_IS_NOT_A_NAME, tokens.getStringValue());
+            }
             PdfName name = readPdfName(true);
             PdfObject obj = readObject(true, objStm);
             if (obj == null) {
@@ -1302,7 +1308,8 @@ protected void rebuildXref() throws IOException {
         tokens.seek(0);
         trailer = null;
         ByteBuffer buffer = new ByteBuffer(24);
-        PdfTokenizer lineTokeniser = new PdfTokenizer(new RandomAccessFileOrArray(new ReusableRandomAccessSource(buffer)));
+        PdfTokenizer lineTokenizer =
+                new PdfTokenizer(new RandomAccessFileOrArray(new ReusableRandomAccessSource(buffer)));
         for (; ; ) {
             long pos = tokens.getPosition();
             buffer.reset();
@@ -1326,7 +1333,7 @@ protected void rebuildXref() throws IOException {
                     tokens.seek(pos);
                 }
             } else if (buffer.get(0) >= '0' && buffer.get(0) <= '9') {
-                int[] obj = PdfTokenizer.checkObjectStart(lineTokeniser);
+                int[] obj = PdfTokenizer.checkObjectStart(lineTokenizer);
                 if (obj == null)
                     continue;
                 int num = obj[0];

kernel/src/test/java/com/itextpdf/kernel/pdf/PdfReaderTest.java

@@ -50,6 +50,7 @@ This file is part of the iText (R) project.
 import com.itextpdf.io.source.ByteArrayOutputStream;
 import com.itextpdf.io.source.ByteUtils;
 import com.itextpdf.io.source.IRandomAccessSource;
+import com.itextpdf.io.source.PdfTokenizer;
 import com.itextpdf.io.source.RASInputStream;
 import com.itextpdf.io.source.RandomAccessSourceFactory;
 import com.itextpdf.kernel.exceptions.InvalidXRefPrevException;
@@ -2607,6 +2608,26 @@ public void streamWithoutEndstreamKeywordConservativeModeTest() throws IOExcepti
         }
     }
 
+    @Test
+    public void tokensPositionIsNotUpdatedWhileReadingLengthTest() throws IOException {
+        String filename = SOURCE_FOLDER + "simpleDocWithIndirectLength.pdf";
+        try (PdfDocument pdfDoc = new PdfDocument(new PdfReader(filename))) {
+            PdfTokenizer tokenizer = pdfDoc.getReader().tokens;
+
+            // we will try to get the content stream object
+            // since it's not been gotten yet, iText will read this object,
+            // which will change the tokenizer's position
+            PdfStream pageContentStream = (PdfStream) pdfDoc.getPdfObject(5);
+
+            // tokenizer's position after reading object should point to the end of the object's stream
+            Assert.assertEquals(pageContentStream.getOffset() + pageContentStream.getLength(), tokenizer.getPosition());
+
+            // let's read next valid token and check that it means ending stream
+            tokenizer.nextValidToken();
+            tokenizer.tokenValueEqualsTo(ByteUtils.getIsoBytes("endstream"));
+        }
+    }
+
     /**
      * Returns the current memory use.
      *