Use file header offset when doing brute force find and fix #1223

src/UglyToad.PdfPig.Tests/Integration/GithubIssuesTests.cs

@@ -11,6 +11,19 @@
 
     public class GithubIssuesTests
     {
+        [Fact]
+        public void Issue1223()
+        {
+            var path = IntegrationHelpers.GetSpecificTestDocumentPath("23056.PMC2132516.pdf");
+            using (var document = PdfDocument.Open(path, new ParsingOptions() { UseLenientParsing = true }))
+            {
+                Assert.NotNull(document);
+                var firstPage = document.GetPage(1);
+                Assert.NotNull(firstPage);
+                Assert.Contains("The Rockefeller University Press", firstPage.Text);
+            }
+        }
+
         [Fact]
         public void Issue1213()
         {

src/UglyToad.PdfPig.Tests/Tokenization/Scanner/PdfTokenScannerTests.cs

@@ -3,6 +3,7 @@
     using System.Text;
     using PdfPig.Core;
     using PdfPig.Encryption;
+    using PdfPig.Parser.FileStructure;
     using PdfPig.Tokenization.Scanner;
     using PdfPig.Tokens;
 
@@ -720,8 +721,12 @@ endobj";
         {
             var input = StringBytesTestConverter.Convert(s, false);
 
-            return new PdfTokenScanner(input.Bytes, locationProvider ?? new TestObjectLocationProvider(),
-                new TestFilterProvider(), NoOpEncryptionHandler.Instance, useLenientParsing ? new ParsingOptions() : ParsingOptions.LenientParsingOff);
+            return new PdfTokenScanner(input.Bytes,
+                locationProvider ?? new TestObjectLocationProvider(),
+                new TestFilterProvider(),
+                NoOpEncryptionHandler.Instance,
+                new FileHeaderOffset(0),
+                useLenientParsing ? new ParsingOptions() : ParsingOptions.LenientParsingOff);
         }
 
         private static IReadOnlyList<ObjectToken> ReadToEnd(PdfTokenScanner scanner)

src/UglyToad.PdfPig/Parser/PdfDocumentFactory.cs

@@ -124,8 +124,10 @@
 
             var version = FileHeaderParser.Parse(scanner, inputBytes, parsingOptions.UseLenientParsing, parsingOptions.Logger);
 
+            var fileHeaderOffset = new FileHeaderOffset((int)version.OffsetInFile);
+
             var initialParse = FirstPassParser.Parse(
-                new FileHeaderOffset((int)version.OffsetInFile),
+                fileHeaderOffset,
                 inputBytes,
                 scanner,
                 parsingOptions.Logger);
@@ -143,7 +145,7 @@
                 initialParse.BruteForceOffsets,
                 inputBytes);
 
-            var pdfScanner = new PdfTokenScanner(inputBytes, locationProvider, filterProvider, NoOpEncryptionHandler.Instance, parsingOptions);
+            var pdfScanner = new PdfTokenScanner(inputBytes, locationProvider, filterProvider, NoOpEncryptionHandler.Instance, fileHeaderOffset, parsingOptions);
 
             var (rootReference, rootDictionary) = ParseTrailer(
                 trailer,

src/UglyToad.PdfPig/Tokenization/Scanner/PdfTokenScanner.cs

@@ -11,6 +11,7 @@
     using Encryption;
     using Filters;
     using Tokens;
+    using UglyToad.PdfPig.Parser.FileStructure;
 
     internal class PdfTokenScanner : IPdfTokenScanner
     {
@@ -23,6 +24,7 @@
         private readonly ILookupFilterProvider filterProvider;
         private readonly CoreTokenScanner coreTokenScanner;
         private readonly ParsingOptions parsingOptions;
+        private readonly FileHeaderOffset fileHeaderOffset;
 
         private IEncryptionHandler encryptionHandler;
         private bool isDisposed;
@@ -54,12 +56,14 @@
             IObjectLocationProvider objectLocationProvider,
             ILookupFilterProvider filterProvider,
             IEncryptionHandler encryptionHandler,
+            FileHeaderOffset fileHeaderOffset,
             ParsingOptions parsingOptions)
         {
             this.inputBytes = inputBytes;
             this.objectLocationProvider = objectLocationProvider;
             this.filterProvider = filterProvider;
             this.encryptionHandler = encryptionHandler;
+            this.fileHeaderOffset = fileHeaderOffset;
             this.parsingOptions = parsingOptions;
             coreTokenScanner = new CoreTokenScanner(inputBytes, true, useLenientParsing: parsingOptions.UseLenientParsing);
         }
@@ -795,8 +799,8 @@
             {
                 // Brute force read the entire file
                 isBruteForcing = true;
-
-                Seek(0);
+                
+                Seek(fileHeaderOffset.Value);
 
                 while (MoveNext())
                 {