Merge branch 'develop' into devsecops

itext.tests/itext.io.tests/itext/io/source/PdfTokenizerTest.cs

@@ -21,6 +21,7 @@ You should have received a copy of the GNU Affero General Public License
 along with this program.  If not, see <https://www.gnu.org/licenses/>.
 */
 using System;
+using System.Text;
 using iText.Commons.Utils;
 using iText.IO.Exceptions;
 using iText.Test;
@@ -167,6 +168,44 @@ public virtual void ReadFullyThenReadStringTest() {
             NUnit.Framework.Assert.AreEqual("15", tok.ReadString(data.Length));
         }
 
+        [NUnit.Framework.Test]
+        public virtual void GetNextEofShortTextTest() {
+            String data = "some text to test \ngetting end of\n file logic%%EOF";
+            RandomAccessSourceFactory factory = new RandomAccessSourceFactory();
+            using (PdfTokenizer tok = new PdfTokenizer(new RandomAccessFileOrArray(factory.CreateSource(data.GetBytes(
+                iText.Commons.Utils.EncodingUtil.ISO_8859_1))))) {
+                long eofPosition = tok.GetNextEof();
+                NUnit.Framework.Assert.AreEqual(data.Length + 1, eofPosition);
+            }
+        }
+
+        [NUnit.Framework.Test]
+        public virtual void GetNextEofLongTextTest() {
+            String data = "some text to test \ngetting end of\n file logic";
+            StringBuilder stringBuilder = new StringBuilder();
+            for (int i = 0; i < 20; ++i) {
+                stringBuilder.Append(data);
+            }
+            stringBuilder.Append("%%EOF");
+            RandomAccessSourceFactory factory = new RandomAccessSourceFactory();
+            using (PdfTokenizer tok = new PdfTokenizer(new RandomAccessFileOrArray(factory.CreateSource(stringBuilder.
+                ToString().GetBytes(iText.Commons.Utils.EncodingUtil.ISO_8859_1))))) {
+                long eofPosition = tok.GetNextEof();
+                NUnit.Framework.Assert.AreEqual(data.Length * 20 + 6, eofPosition);
+            }
+        }
+
+        [NUnit.Framework.Test]
+        public virtual void GetNextEofSeveralEofTest() {
+            String data = "some text %%EOFto test \nget%%EOFting end of\n fil%%EOFe logic%%EOF";
+            RandomAccessSourceFactory factory = new RandomAccessSourceFactory();
+            using (PdfTokenizer tok = new PdfTokenizer(new RandomAccessFileOrArray(factory.CreateSource(data.GetBytes(
+                iText.Commons.Utils.EncodingUtil.ISO_8859_1))))) {
+                long eofPosition = tok.GetNextEof();
+                NUnit.Framework.Assert.AreEqual(data.IndexOf("%%EOF", StringComparison.Ordinal) + 6, eofPosition);
+            }
+        }
+
         [NUnit.Framework.Test]
         public virtual void GetDecodedStringContentTest() {
             String data = "/Name1 15";

itext.tests/itext.kernel.tests/itext/kernel/pdf/PdfRevisionsReaderTest.cs

@@ -0,0 +1,188 @@
+/*
+This file is part of the iText (R) project.
+Copyright (c) 1998-2024 Apryse Group NV
+Authors: Apryse Software.
+
+This program is offered under a commercial and under the AGPL license.
+For commercial licensing, contact us at https://itextpdf.com/sales.  For AGPL licensing, see below.
+
+AGPL licensing:
+This program is free software: you can redistribute it and/or modify
+it under the terms of the GNU Affero General Public License as published by
+the Free Software Foundation, either version 3 of the License, or
+(at your option) any later version.
+
+This program is distributed in the hope that it will be useful,
+but WITHOUT ANY WARRANTY; without even the implied warranty of
+MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+GNU Affero General Public License for more details.
+
+You should have received a copy of the GNU Affero General Public License
+along with this program.  If not, see <https://www.gnu.org/licenses/>.
+*/
+using System;
+using System.Collections.Generic;
+using System.Linq;
+using iText.Test;
+
+namespace iText.Kernel.Pdf {
+    [NUnit.Framework.Category("UnitTest")]
+    public class PdfRevisionsReaderTest : ExtendedITextTest {
+        private static readonly String SOURCE_FOLDER = iText.Test.TestUtil.GetParentProjectDirectory(NUnit.Framework.TestContext
+            .CurrentContext.TestDirectory) + "/resources/itext/kernel/pdf/PdfRevisionsReaderTest/";
+
+        [NUnit.Framework.Test]
+        public virtual void SingleRevisionDocumentTest() {
+            String filename = SOURCE_FOLDER + "singleRevisionDocument.pdf";
+            using (PdfReader reader = new PdfReader(filename)) {
+                PdfRevisionsReader revisionsReader = new PdfRevisionsReader(reader);
+                IList<DocumentRevision> documentRevisions = revisionsReader.GetAllRevisions();
+                NUnit.Framework.Assert.AreEqual(1, documentRevisions.Count);
+                DocumentRevision firstRevision = documentRevisions[0];
+                AssertResultingRevision(firstRevision, 1, 2, 3, 4, 5, 6);
+                NUnit.Framework.Assert.AreEqual(929, firstRevision.GetEofOffset());
+            }
+        }
+
+        [NUnit.Framework.Test]
+        public virtual void SingleRevisionWithXrefStreamTest() {
+            String filename = SOURCE_FOLDER + "singleRevisionWithXrefStream.pdf";
+            using (PdfReader reader = new PdfReader(filename)) {
+                PdfRevisionsReader revisionsReader = new PdfRevisionsReader(reader);
+                IList<DocumentRevision> documentRevisions = revisionsReader.GetAllRevisions();
+                NUnit.Framework.Assert.AreEqual(1, documentRevisions.Count);
+                DocumentRevision firstRevision = documentRevisions[0];
+                AssertResultingRevision(firstRevision, 1, 2, 3, 4, 5, 6, 7, 8);
+                NUnit.Framework.Assert.AreEqual(1085, firstRevision.GetEofOffset());
+            }
+        }
+
+        [NUnit.Framework.Test]
+        public virtual void MultipleRevisionsDocument() {
+            String filename = SOURCE_FOLDER + "multipleRevisionsDocument.pdf";
+            using (PdfReader reader = new PdfReader(filename)) {
+                PdfRevisionsReader revisionsReader = new PdfRevisionsReader(reader);
+                IList<DocumentRevision> documentRevisions = revisionsReader.GetAllRevisions();
+                NUnit.Framework.Assert.AreEqual(3, documentRevisions.Count);
+                DocumentRevision firstRevision = documentRevisions[0];
+                AssertResultingRevision(firstRevision, 1, 2, 3, 4, 5, 6);
+                NUnit.Framework.Assert.AreEqual(929, firstRevision.GetEofOffset());
+                DocumentRevision secondRevision = documentRevisions[1];
+                AssertResultingRevision(secondRevision, 1, 3, 4, 7, 8, 9, 10, 11, 12, 13, 14, 15);
+                NUnit.Framework.Assert.AreEqual(28119, secondRevision.GetEofOffset());
+                DocumentRevision thirdRevision = documentRevisions[2];
+                AssertResultingRevision(thirdRevision, 1, 3, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28);
+                NUnit.Framework.Assert.AreEqual(36207, thirdRevision.GetEofOffset());
+            }
+        }
+
+        [NUnit.Framework.Test]
+        public virtual void FreeReferencesDocument() {
+            String filename = SOURCE_FOLDER + "freeReferencesDocument.pdf";
+            using (PdfReader reader = new PdfReader(filename)) {
+                PdfRevisionsReader revisionsReader = new PdfRevisionsReader(reader);
+                IList<DocumentRevision> documentRevisions = revisionsReader.GetAllRevisions();
+                NUnit.Framework.Assert.AreEqual(5, documentRevisions.Count);
+                DocumentRevision firstRevision = documentRevisions[0];
+                AssertResultingRevision(firstRevision, 1, 2, 3, 4, 5, 6);
+                NUnit.Framework.Assert.AreEqual(929, firstRevision.GetEofOffset());
+                DocumentRevision secondRevision = documentRevisions[1];
+                AssertResultingRevision(secondRevision, 1, 3, 4, 7, 8, 9, 10, 11, 12, 13, 14, 15);
+                NUnit.Framework.Assert.AreEqual(28119, secondRevision.GetEofOffset());
+                DocumentRevision thirdRevision = documentRevisions[2];
+                AssertResultingRevision(thirdRevision, 1, 3, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28);
+                NUnit.Framework.Assert.AreEqual(36207, thirdRevision.GetEofOffset());
+                DocumentRevision fourthRevision = documentRevisions[3];
+                AssertResultingRevision(fourthRevision, new int[] { 1, 3, 23, 24 }, new int[] { 0, 0, 1, 1 });
+                NUnit.Framework.Assert.AreEqual(37006, fourthRevision.GetEofOffset());
+                DocumentRevision fifthRevision = documentRevisions[4];
+                AssertResultingRevision(fifthRevision, new int[] { 1, 3, 19, 20, 21, 22, 23, 25 }, new int[] { 0, 0, 1, 1, 
+                    1, 1, 1, 1 });
+                NUnit.Framework.Assert.AreEqual(38094, fifthRevision.GetEofOffset());
+            }
+        }
+
+        [NUnit.Framework.Test]
+        public virtual void MultipleRevisionsWithXrefStreamTest() {
+            String filename = SOURCE_FOLDER + "multipleRevisionsWithXrefStream.pdf";
+            using (PdfReader reader = new PdfReader(filename)) {
+                PdfRevisionsReader revisionsReader = new PdfRevisionsReader(reader);
+                IList<DocumentRevision> documentRevisions = revisionsReader.GetAllRevisions();
+                NUnit.Framework.Assert.AreEqual(3, documentRevisions.Count);
+                DocumentRevision firstRevision = documentRevisions[0];
+                AssertResultingRevision(firstRevision, 1, 2, 3, 4, 5, 6, 7, 8);
+                NUnit.Framework.Assert.AreEqual(1085, firstRevision.GetEofOffset());
+                DocumentRevision secondRevision = documentRevisions[1];
+                AssertResultingRevision(secondRevision, 1, 3, 4, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19);
+                NUnit.Framework.Assert.AreEqual(28137, secondRevision.GetEofOffset());
+                DocumentRevision thirdRevision = documentRevisions[2];
+                AssertResultingRevision(thirdRevision, 1, 3, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34);
+                NUnit.Framework.Assert.AreEqual(36059, thirdRevision.GetEofOffset());
+            }
+        }
+
+        [NUnit.Framework.Test]
+        public virtual void FreeReferencesWithXrefStream() {
+            String filename = SOURCE_FOLDER + "freeReferencesWithXrefStream.pdf";
+            using (PdfReader reader = new PdfReader(filename)) {
+                PdfRevisionsReader revisionsReader = new PdfRevisionsReader(reader);
+                IList<DocumentRevision> documentRevisions = revisionsReader.GetAllRevisions();
+                NUnit.Framework.Assert.AreEqual(5, documentRevisions.Count);
+                DocumentRevision firstRevision = documentRevisions[0];
+                AssertResultingRevision(firstRevision, 1, 2, 3, 4, 5, 6, 7, 8);
+                NUnit.Framework.Assert.AreEqual(1085, firstRevision.GetEofOffset());
+                DocumentRevision secondRevision = documentRevisions[1];
+                AssertResultingRevision(secondRevision, 1, 3, 4, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19);
+                NUnit.Framework.Assert.AreEqual(28137, secondRevision.GetEofOffset());
+                DocumentRevision thirdRevision = documentRevisions[2];
+                AssertResultingRevision(thirdRevision, 1, 3, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34);
+                NUnit.Framework.Assert.AreEqual(36059, thirdRevision.GetEofOffset());
+                DocumentRevision fourthRevision = documentRevisions[3];
+                AssertResultingRevision(fourthRevision, new int[] { 1, 3, 27, 28, 35 }, new int[] { 0, 0, 1, 1, 0 });
+                NUnit.Framework.Assert.AreEqual(36975, fourthRevision.GetEofOffset());
+                DocumentRevision fifthRevision = documentRevisions[4];
+                AssertResultingRevision(fifthRevision, new int[] { 1, 3, 23, 24, 25, 26, 27, 29, 36 }, new int[] { 0, 0, 1
+                    , 1, 1, 1, 1, 1, 0 });
+                NUnit.Framework.Assert.AreEqual(38111, fifthRevision.GetEofOffset());
+            }
+        }
+
+        [NUnit.Framework.Test]
+        public virtual void DocumentWithStreamAndTableXref() {
+            String filename = SOURCE_FOLDER + "documentWithStreamAndTableXref.pdf";
+            using (PdfReader reader = new PdfReader(filename)) {
+                PdfRevisionsReader revisionsReader = new PdfRevisionsReader(reader);
+                IList<DocumentRevision> documentRevisions = revisionsReader.GetAllRevisions();
+                NUnit.Framework.Assert.AreEqual(3, documentRevisions.Count);
+                DocumentRevision thirdRevision = revisionsReader.GetAllRevisions()[0];
+                // xref was broken in this revision and fixed in the next one
+                AssertResultingRevision(thirdRevision, new int[] { 1, 2, 3, 4, 5, 6, 7, 8, 9 }, new int[] { 0, 0, 0, 0, 0, 
+                    0, 65535, 0, 0 });
+                NUnit.Framework.Assert.AreEqual(1381, thirdRevision.GetEofOffset());
+                DocumentRevision secondRevision = revisionsReader.GetAllRevisions()[1];
+                AssertResultingRevision(secondRevision, 1, 2, 3, 4, 5, 6, 7, 8);
+                NUnit.Framework.Assert.AreEqual(1550, secondRevision.GetEofOffset());
+                DocumentRevision firstRevision = revisionsReader.GetAllRevisions()[2];
+                AssertResultingRevision(firstRevision);
+                NUnit.Framework.Assert.AreEqual(1550, firstRevision.GetEofOffset());
+            }
+        }
+
+        private void AssertResultingRevision(DocumentRevision documentRevision, params int[] objNumbers) {
+            AssertResultingRevision(documentRevision, objNumbers, new int[objNumbers.Length]);
+        }
+
+        private void AssertResultingRevision(DocumentRevision documentRevision, int[] objNumbers, int[] objGens) {
+            NUnit.Framework.Assert.AreEqual(objNumbers.Length, objGens.Length);
+            NUnit.Framework.Assert.AreEqual(objNumbers.Length + 1, documentRevision.GetModifiedObjects().Count);
+            for (int i = 0; i < objNumbers.Length; ++i) {
+                int objNumber = objNumbers[i];
+                int objGen = objGens[i];
+                NUnit.Framework.Assert.IsTrue(documentRevision.GetModifiedObjects().Any((reference) => reference.GetObjNumber
+                    () == objNumber && reference.GetGenNumber() == objGen));
+            }
+            NUnit.Framework.Assert.IsTrue(documentRevision.GetModifiedObjects().Any((reference) => reference.GetObjNumber
+                () == 0 && reference.GetGenNumber() == 65535 && reference.IsFree()));
+        }
+    }
+}

itext/itext.io/itext/io/exceptions/IoExceptionMessageConstant.cs

@@ -208,6 +208,8 @@ public sealed class IoExceptionMessageConstant {
 
         public const String PDF_STARTXREF_NOT_FOUND = "PDF startxref not found.";
 
+        public const String PDF_EOF_NOT_FOUND = "PDF \"%%EOF\" marker is not found.";
+
         public const String PHOTOMETRIC_IS_NOT_SUPPORTED = "Photometric {0} is not supported.";
 
         public const String PLANAR_IMAGES_ARE_NOT_SUPPORTED = "Planar images are not supported.";

itext/itext.io/itext/io/source/PdfTokenizer.cs

@@ -253,6 +253,24 @@ public virtual long GetStartxref() {
             throw new iText.IO.Exceptions.IOException(IoExceptionMessageConstant.PDF_STARTXREF_NOT_FOUND, this);
         }
 
+        /// <summary>Gets next %%EOF marker in current PDF file.</summary>
+        /// <returns>next %%EOF marker position</returns>
+        public virtual long GetNextEof() {
+            int arrLength = 128;
+            String str;
+            do {
+                long currentPosition = file.GetPosition();
+                str = ReadString(arrLength);
+                long eofPosition = str.IndexOf("%%EOF", StringComparison.Ordinal);
+                if (eofPosition >= 0) {
+                    // 6 stands for '%%EOF' length + 1
+                    return currentPosition + eofPosition + 6;
+                }
+            }
+            while (!String.IsNullOrEmpty(str));
+            throw new iText.IO.Exceptions.IOException(IoExceptionMessageConstant.PDF_EOF_NOT_FOUND, this);
+        }
+
         public virtual void NextValidToken() {
             int level = 0;
             byte[] n1 = null;

itext/itext.kernel/itext/kernel/pdf/DocumentRevision.cs

@@ -0,0 +1,69 @@
+/*
+This file is part of the iText (R) project.
+Copyright (c) 1998-2024 Apryse Group NV
+Authors: Apryse Software.
+
+This program is offered under a commercial and under the AGPL license.
+For commercial licensing, contact us at https://itextpdf.com/sales.  For AGPL licensing, see below.
+
+AGPL licensing:
+This program is free software: you can redistribute it and/or modify
+it under the terms of the GNU Affero General Public License as published by
+the Free Software Foundation, either version 3 of the License, or
+(at your option) any later version.
+
+This program is distributed in the hope that it will be useful,
+but WITHOUT ANY WARRANTY; without even the implied warranty of
+MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+GNU Affero General Public License for more details.
+
+You should have received a copy of the GNU Affero General Public License
+along with this program.  If not, see <https://www.gnu.org/licenses/>.
+*/
+using System.Collections.Generic;
+
+namespace iText.Kernel.Pdf {
+    /// <summary>Class which stores information about single PDF document revision.</summary>
+    public class DocumentRevision {
+        private readonly long eofOffset;
+
+        private readonly ICollection<PdfIndirectReference> modifiedObjects;
+
+        /// <summary>
+        /// Creates
+        /// <see cref="DocumentRevision"/>
+        /// from end-of-file byte position and a set of indirect references which were
+        /// modified in this document revision.
+        /// </summary>
+        /// <param name="eofOffset">end-of-file byte position</param>
+        /// <param name="modifiedObjects">
+        /// 
+        /// <see cref="Java.Util.Set{E}"/>
+        /// of
+        /// <see cref="PdfIndirectReference"/>
+        /// objects which were modified
+        /// </param>
+        public DocumentRevision(long eofOffset, ICollection<PdfIndirectReference> modifiedObjects) {
+            this.eofOffset = eofOffset;
+            this.modifiedObjects = modifiedObjects;
+        }
+
+        /// <summary>Gets end-of-file byte position.</summary>
+        /// <returns>end-of-file byte position</returns>
+        public virtual long GetEofOffset() {
+            return eofOffset;
+        }
+
+        /// <summary>Gets objects which were modified in this document revision.</summary>
+        /// <returns>
+        /// 
+        /// <see cref="Java.Util.Set{E}"/>
+        /// of
+        /// <see cref="PdfIndirectReference"/>
+        /// objects which were modified
+        /// </returns>
+        public virtual ICollection<PdfIndirectReference> GetModifiedObjects() {
+            return modifiedObjects;
+        }
+    }
+}