#482 add skip missing fonts option and pass parsing options to content stream processor

this doesn't fix the reported issue since the pdf itself is corrupted on page 8 however it will
allow recovery in some scenarios where text content isn't important.

also adds more informative error when stream unintentionally passed with non zero offset

src/UglyToad.PdfPig/Content/IPageFactory.cs

@@ -4,6 +4,10 @@
 
     internal interface IPageFactory
     {
-        Page Create(int number, DictionaryToken dictionary, PageTreeMembers pageTreeMembers, bool clipPaths);
+        Page Create(
+            int number,
+            DictionaryToken dictionary,
+            PageTreeMembers pageTreeMembers,
+            InternalParsingOptions parsingOptions);
     }
 }

src/UglyToad.PdfPig/Content/Pages.cs

@@ -23,10 +23,12 @@
             Count = catalog.PagesDictionary.GetIntOrDefault(NameToken.Count);
         }
         
-        public Page GetPage(int pageNumber, bool clipPaths)
+        public Page GetPage(int pageNumber, InternalParsingOptions parsingOptions)
         {
             if (pageNumber <= 0 || pageNumber > Count)
             {
+                parsingOptions.Logger.Error($"Page {pageNumber} requested but is out of range.");
+
                 throw new ArgumentOutOfRangeException(nameof(pageNumber), 
                     $"Page number {pageNumber} invalid, must be between 1 and {Count}.");
             }
@@ -63,7 +65,11 @@
                 }
             }
 
-            var page = pageFactory.Create(pageNumber, pageNode.NodeDictionary, pageTreeMembers, clipPaths);
+            var page = pageFactory.Create(
+                pageNumber,
+                pageNode.NodeDictionary,
+                pageTreeMembers,
+                parsingOptions);
             
             return page;
         }

src/UglyToad.PdfPig/Graphics/ContentStreamProcessor.cs

@@ -5,7 +5,6 @@
     using Core;
     using Filters;
     using Geometry;
-    using Logging;
     using Operations;
     using Parser;
     using PdfFonts;
@@ -49,9 +48,8 @@
         private readonly IPdfTokenScanner pdfScanner;
         private readonly IPageContentParser pageContentParser;
         private readonly ILookupFilterProvider filterProvider;
-        private readonly ILog log;
-        private readonly bool clipPaths;
         private readonly PdfVector pageSize;
+        private readonly InternalParsingOptions parsingOptions;
         private readonly MarkedContentStack markedContentStack = new MarkedContentStack();
 
         private Stack<CurrentGraphicsState> graphicsStack = new Stack<CurrentGraphicsState>();
@@ -90,9 +88,8 @@
             IPdfTokenScanner pdfScanner,
             IPageContentParser pageContentParser,
             ILookupFilterProvider filterProvider,
-            ILog log,
-            bool clipPaths,
-            PdfVector pageSize)
+            PdfVector pageSize,
+            InternalParsingOptions parsingOptions)
         {
             this.resourceStore = resourceStore;
             this.userSpaceUnit = userSpaceUnit;
@@ -100,9 +97,8 @@
             this.pdfScanner = pdfScanner ?? throw new ArgumentNullException(nameof(pdfScanner));
             this.pageContentParser = pageContentParser ?? throw new ArgumentNullException(nameof(pageContentParser));
             this.filterProvider = filterProvider ?? throw new ArgumentNullException(nameof(filterProvider));
-            this.log = log;
-            this.clipPaths = clipPaths;
             this.pageSize = pageSize;
+            this.parsingOptions = parsingOptions;
 
             // initiate CurrentClippingPath to cropBox
             var clippingSubpath = new PdfSubpath();
@@ -230,6 +226,15 @@
 
             if (font == null)
             {
+                if (parsingOptions.SkipMissingFonts)
+                {
+                    parsingOptions.Logger.Warn($"Skipping a missing font with name {currentState.FontState.FontName} " +
+                                               $"since it is not present in the document and {nameof(InternalParsingOptions.SkipMissingFonts)} " +
+                                               "is set to true. This may result in some text being skipped and not included in the output.");
+
+                    return;
+                }
+
                 throw new InvalidOperationException($"Could not find the font with name {currentState.FontState.FontName} in the resource store. It has not been loaded yet.");
             }
 
@@ -253,7 +258,8 @@
 
                 if (!foundUnicode || unicode == null)
                 {
-                    log.Warn($"We could not find the corresponding character with code {code} in font {font.Name}.");
+                    parsingOptions.Logger.Warn($"We could not find the corresponding character with code {code} in font {font.Name}.");
+
                     // Try casting directly to string as in PDFBox 1.8.
                     unicode = new string((char)code, 1);
                 }
@@ -494,7 +500,7 @@
 
             var contentStream = formStream.Decode(filterProvider, pdfScanner);
 
-            var operations = pageContentParser.Parse(pageNumber, new ByteArrayInputBytes(contentStream), log);
+            var operations = pageContentParser.Parse(pageNumber, new ByteArrayInputBytes(contentStream), parsingOptions.Logger);
 
             // 3. We don't respect clipping currently.
 
@@ -677,7 +683,7 @@
 
             if (CurrentPath.IsClipping)
             {
-                if (!clipPaths)
+                if (!parsingOptions.ClipPaths)
                 {
                     // if we don't clip paths, add clipping path to paths
                     paths.Add(CurrentPath);
@@ -717,9 +723,9 @@
                 CurrentPath.FillColor = currentState.CurrentNonStrokingColor;
             }
 
-            if (clipPaths)
+            if (parsingOptions.ClipPaths)
             {
-                var clippedPath = currentState.CurrentClippingPath.Clip(CurrentPath, log);
+                var clippedPath = currentState.CurrentClippingPath.Clip(CurrentPath, parsingOptions.Logger);
                 if (clippedPath != null)
                 {
                     paths.Add(clippedPath);
@@ -745,15 +751,15 @@
             AddCurrentSubpath();
             CurrentPath.SetClipping(clippingRule);
 
-            if (clipPaths)
+            if (parsingOptions.ClipPaths)
             {
                 var currentClipping = GetCurrentState().CurrentClippingPath;
                 currentClipping.SetClipping(clippingRule);
 
-                var newClippings = CurrentPath.Clip(currentClipping, log);
+                var newClippings = CurrentPath.Clip(currentClipping, parsingOptions.Logger);
                 if (newClippings == null)
                 {
-                    log.Warn("Empty clipping path found. Clipping path not updated.");
+                    parsingOptions.Logger.Warn("Empty clipping path found. Clipping path not updated.");
                 }
                 else
                 {
@@ -796,7 +802,7 @@
         {
             if (inlineImageBuilder != null)
             {
-                log?.Error("Begin inline image (BI) command encountered while another inline image was active.");
+                parsingOptions.Logger.Error("Begin inline image (BI) command encountered while another inline image was active.");
             }
 
             inlineImageBuilder = new InlineImageBuilder();
@@ -806,7 +812,7 @@
         {
             if (inlineImageBuilder == null)
             {
-                log?.Error("Begin inline image data (ID) command encountered without a corresponding begin inline image (BI) command.");
+                parsingOptions.Logger.Error("Begin inline image data (ID) command encountered without a corresponding begin inline image (BI) command.");
                 return;
             }
 
@@ -817,7 +823,7 @@
         {
             if (inlineImageBuilder == null)
             {
-                log?.Error("End inline image (EI) command encountered without a corresponding begin inline image (BI) command.");
+                parsingOptions.Logger.Error("End inline image (EI) command encountered without a corresponding begin inline image (BI) command.");
                 return;
             }
 

src/UglyToad.PdfPig/InternalParsingOptions.cs

@@ -0,0 +1,34 @@
+namespace UglyToad.PdfPig;
+
+using Logging;
+using System.Collections.Generic;
+
+/// <summary>
+/// <see cref="ParsingOptions"/> but without being a public API/
+/// </summary>
+internal class InternalParsingOptions
+{
+    public IReadOnlyList<string> Passwords { get; }
+
+    public bool UseLenientParsing { get; }
+
+    public bool ClipPaths { get; }
+
+    public bool SkipMissingFonts { get; }
+
+    public ILog Logger { get; }
+
+    public InternalParsingOptions(
+        IReadOnlyList<string> passwords,
+        bool useLenientParsing,
+        bool clipPaths,
+        bool skipMissingFonts,
+        ILog logger)
+    {
+        Passwords = passwords;
+        UseLenientParsing = useLenientParsing;
+        ClipPaths = clipPaths;
+        SkipMissingFonts = skipMissingFonts;
+        Logger = logger;
+    }
+}

src/UglyToad.PdfPig/Parser/PageFactory.cs

@@ -21,20 +21,20 @@
         private readonly IResourceStore resourceStore;
         private readonly ILookupFilterProvider filterProvider;
         private readonly IPageContentParser pageContentParser;
-        private readonly ILog log;
 
-        public PageFactory(IPdfTokenScanner pdfScanner, IResourceStore resourceStore, ILookupFilterProvider filterProvider,
-            IPageContentParser pageContentParser,
-            ILog log)
+        public PageFactory(
+            IPdfTokenScanner pdfScanner,
+            IResourceStore resourceStore,
+            ILookupFilterProvider filterProvider,
+            IPageContentParser pageContentParser)
         {
             this.resourceStore = resourceStore;
             this.filterProvider = filterProvider;
             this.pageContentParser = pageContentParser;
-            this.log = log;
             this.pdfScanner = pdfScanner;
         }
 
-        public Page Create(int number, DictionaryToken dictionary, PageTreeMembers pageTreeMembers, bool clipPaths)
+        public Page Create(int number, DictionaryToken dictionary, PageTreeMembers pageTreeMembers, InternalParsingOptions parsingOptions)
         {
             if (dictionary == null)
             {
@@ -45,11 +45,11 @@
 
             if (type != null && !type.Equals(NameToken.Page))
             {
-                log?.Error($"Page {number} had its type specified as {type} rather than 'Page'.");
+                parsingOptions.Logger.Error($"Page {number} had its type specified as {type} rather than 'Page'.");
             }
 
-            MediaBox mediaBox = GetMediaBox(number, dictionary, pageTreeMembers);
-            CropBox cropBox = GetCropBox(dictionary, pageTreeMembers, mediaBox);
+            MediaBox mediaBox = GetMediaBox(number, dictionary, pageTreeMembers, parsingOptions.Logger);
+            CropBox cropBox = GetCropBox(dictionary, pageTreeMembers, mediaBox, parsingOptions.Logger);
 
             var rotation = new PageRotationDegrees(pageTreeMembers.Rotation);
             if (dictionary.TryGet(NameToken.Rotate, pdfScanner, out NumericToken rotateToken))
@@ -130,7 +130,7 @@
                     }
                 }
 
-                content = GetContent(number, bytes, cropBox, userSpaceUnit, rotation, clipPaths, mediaBox);
+                content = GetContent(number, bytes, cropBox, userSpaceUnit, rotation, mediaBox, parsingOptions);
             }
             else
             {
@@ -143,7 +143,7 @@
 
                 var bytes = contentStream.Decode(filterProvider, pdfScanner);
 
-                content = GetContent(number, bytes, cropBox, userSpaceUnit, rotation, clipPaths, mediaBox);
+                content = GetContent(number, bytes, cropBox, userSpaceUnit, rotation, mediaBox, parsingOptions);
             }
 
             var page = new Page(number, dictionary, mediaBox, cropBox, rotation, content, 
@@ -158,18 +158,28 @@
             return page;
         }
 
-        private PageContent GetContent(int pageNumber, IReadOnlyList<byte> contentBytes, CropBox cropBox, UserSpaceUnit userSpaceUnit,
-            PageRotationDegrees rotation, bool clipPaths, MediaBox mediaBox)
+        private PageContent GetContent(
+            int pageNumber,
+            IReadOnlyList<byte> contentBytes,
+            CropBox cropBox,
+            UserSpaceUnit userSpaceUnit,
+            PageRotationDegrees rotation,
+            MediaBox mediaBox,
+            InternalParsingOptions parsingOptions)
         {
             var operations = pageContentParser.Parse(pageNumber, new ByteArrayInputBytes(contentBytes),
-                log);
+                parsingOptions.Logger);
 
-            var context = new ContentStreamProcessor(cropBox.Bounds, resourceStore, userSpaceUnit, rotation, pdfScanner,
+            var context = new ContentStreamProcessor(
+                cropBox.Bounds,
+                resourceStore,
+                userSpaceUnit,
+                rotation,
+                pdfScanner,
                 pageContentParser,
                 filterProvider,
-                log,
-                clipPaths,
-                new PdfVector(mediaBox.Bounds.Width, mediaBox.Bounds.Height));
+                new PdfVector(mediaBox.Bounds.Width, mediaBox.Bounds.Height),
+                parsingOptions);
 
             return context.Process(pageNumber, operations);
         }
@@ -185,7 +195,11 @@
             return spaceUnits;
         }
 
-        private CropBox GetCropBox(DictionaryToken dictionary, PageTreeMembers pageTreeMembers, MediaBox mediaBox)
+        private CropBox GetCropBox(
+            DictionaryToken dictionary,
+            PageTreeMembers pageTreeMembers,
+            MediaBox mediaBox,
+            ILog log)
         {
             CropBox cropBox;
             if (dictionary.TryGet(NameToken.CropBox, out var cropBoxObject) &&
@@ -210,7 +224,11 @@
             return cropBox;
         }
 
-        private MediaBox GetMediaBox(int number, DictionaryToken dictionary, PageTreeMembers pageTreeMembers)
+        private MediaBox GetMediaBox(
+            int number,
+            DictionaryToken dictionary,
+            PageTreeMembers pageTreeMembers,
+            ILog log)
         {
             MediaBox mediaBox;
             if (dictionary.TryGet(NameToken.MediaBox, out var mediaboxObject) 

src/UglyToad.PdfPig/Parser/PdfDocumentFactory.cs

@@ -45,9 +45,25 @@
 
         internal static PdfDocument Open(Stream stream, ParsingOptions options)
         {
+            var initialPosition = stream.Position;
+
             var streamInput = new StreamInputBytes(stream, false);
 
-            return Open(streamInput, options);
+            try
+            {
+                return Open(streamInput, options);
+            }
+            catch (Exception ex)
+            {
+                if (initialPosition != 0)
+                {
+                    throw new InvalidOperationException(
+                        "Could not parse document due to an error, the input stream was not at position zero when provided to the Open method.",
+                        ex);
+                }
+
+                throw;
+            }
         }
 
         private static PdfDocument Open(IInputBytes inputBytes, ParsingOptions options = null)
@@ -75,19 +91,28 @@
                 passwords.Add(string.Empty);
             }
 
-            var document = OpenDocument(inputBytes, tokenScanner, options?.Logger ?? new NoOpLog(), isLenientParsing, passwords, clipPaths);
+            var finalOptions = new InternalParsingOptions(
+                passwords,
+                isLenientParsing,
+                clipPaths,
+                options?.SkipMissingFonts ?? false,
+                options?.Logger ?? new NoOpLog());
+
+            var document = OpenDocument(inputBytes, tokenScanner, finalOptions);
 
             return document;
         }
 
-        private static PdfDocument OpenDocument(IInputBytes inputBytes, ISeekableTokenScanner scanner, ILog log, bool isLenientParsing,
-            IReadOnlyList<string> passwords, bool clipPaths)
+        private static PdfDocument OpenDocument(
+            IInputBytes inputBytes,
+            ISeekableTokenScanner scanner,
+            InternalParsingOptions parsingOptions)
         {
             var filterProvider = new FilterProviderWithLookup(DefaultFilterProvider.Instance);
 
             CrossReferenceTable crossReferenceTable = null;
 
-            var xrefValidator = new XrefOffsetValidator(log);
+            var xrefValidator = new XrefOffsetValidator(parsingOptions.Logger);
 
             // We're ok with this since our intent is to lazily load the cross reference table.
             // ReSharper disable once AccessToModifiedClosure
@@ -95,30 +120,39 @@
             var pdfScanner = new PdfTokenScanner(inputBytes, locationProvider, filterProvider, NoOpEncryptionHandler.Instance);
 
             var crossReferenceStreamParser = new CrossReferenceStreamParser(filterProvider);
-            var crossReferenceParser = new CrossReferenceParser(log, xrefValidator, crossReferenceStreamParser);
+            var crossReferenceParser = new CrossReferenceParser(parsingOptions.Logger, xrefValidator, crossReferenceStreamParser);
 
-            var version = FileHeaderParser.Parse(scanner, inputBytes, isLenientParsing, log);
+            var version = FileHeaderParser.Parse(scanner, inputBytes, parsingOptions.UseLenientParsing, parsingOptions.Logger);
 
-            var crossReferenceOffset = FileTrailerParser.GetFirstCrossReferenceOffset(inputBytes, scanner,
-                isLenientParsing) + version.OffsetInFile;
+            var crossReferenceOffset = FileTrailerParser.GetFirstCrossReferenceOffset(
+                inputBytes,
+                scanner,
+                parsingOptions.UseLenientParsing) + version.OffsetInFile;
 
             // TODO: make this use the scanner.
             var validator = new CrossReferenceOffsetValidator(xrefValidator);
 
-            crossReferenceOffset = validator.Validate(crossReferenceOffset, scanner, inputBytes, isLenientParsing);
+            crossReferenceOffset = validator.Validate(crossReferenceOffset, scanner, inputBytes, parsingOptions.UseLenientParsing);
 
-            crossReferenceTable = crossReferenceParser.Parse(inputBytes, isLenientParsing,
+            crossReferenceTable = crossReferenceParser.Parse(
+                inputBytes,
+                parsingOptions.UseLenientParsing,
                 crossReferenceOffset,
                 version.OffsetInFile,
                 pdfScanner,
                 scanner);
 
-            var (rootReference, rootDictionary) = ParseTrailer(crossReferenceTable, isLenientParsing,
+            var (rootReference, rootDictionary) = ParseTrailer(
+                crossReferenceTable,
+                parsingOptions.UseLenientParsing,
                 pdfScanner,
                 out var encryptionDictionary);
 
             var encryptionHandler = encryptionDictionary != null ?
-                (IEncryptionHandler)new EncryptionHandler(encryptionDictionary, crossReferenceTable.Trailer, passwords)
+                (IEncryptionHandler)new EncryptionHandler(
+                    encryptionDictionary,
+                    crossReferenceTable.Trailer,
+                    parsingOptions.Passwords)
                 : NoOpEncryptionHandler.Instance;
 
             pdfScanner.UpdateEncryptionHandler(encryptionHandler);
@@ -128,35 +162,45 @@
 
             var type1Handler = new Type1FontHandler(pdfScanner, filterProvider, encodingReader);
 
-            var fontFactory = new FontFactory(log, new Type0FontHandler(cidFontFactory,
+            var fontFactory = new FontFactory(parsingOptions.Logger, new Type0FontHandler(cidFontFactory,
                 filterProvider, pdfScanner),
-                new TrueTypeFontHandler(log, pdfScanner, filterProvider, encodingReader, SystemFontFinder.Instance,
+                new TrueTypeFontHandler(parsingOptions.Logger, pdfScanner, filterProvider, encodingReader, SystemFontFinder.Instance,
                     type1Handler),
                 type1Handler,
                 new Type3FontHandler(pdfScanner, filterProvider, encodingReader));
 
             var resourceContainer = new ResourceStore(pdfScanner, fontFactory);
 
-            var information = DocumentInformationFactory.Create(pdfScanner, crossReferenceTable.Trailer, isLenientParsing);
+            var information = DocumentInformationFactory.Create(
+                pdfScanner,
+                crossReferenceTable.Trailer,
+                parsingOptions.UseLenientParsing);
 
-            var catalog = CatalogFactory.Create(rootReference, rootDictionary, pdfScanner, isLenientParsing);
+            var catalog = CatalogFactory.Create(
+                rootReference,
+                rootDictionary,
+                pdfScanner,
+                parsingOptions.UseLenientParsing);
 
             var pageFactory = new PageFactory(pdfScanner, resourceContainer, filterProvider,
-                new PageContentParser(new ReflectionGraphicsStateOperationFactory()),
-                log);
-
-            var caching = new ParsingCachingProviders(resourceContainer);
+                new PageContentParser(new ReflectionGraphicsStateOperationFactory()));
 
             var acroFormFactory = new AcroFormFactory(pdfScanner, filterProvider, crossReferenceTable);
-            var bookmarksProvider = new BookmarksProvider(log, pdfScanner);
+            var bookmarksProvider = new BookmarksProvider(parsingOptions.Logger, pdfScanner);
 
-            return new PdfDocument(log, inputBytes, version, crossReferenceTable, caching, pageFactory, catalog, information,
+            return new PdfDocument(
+                inputBytes,
+                version,
+                crossReferenceTable,
+                pageFactory,
+                catalog,
+                information,
                 encryptionDictionary,
                 pdfScanner,
                 filterProvider,
                 acroFormFactory,
                 bookmarksProvider,
-                clipPaths);
+                parsingOptions);
         }
 
         private static (IndirectReference, DictionaryToken) ParseTrailer(CrossReferenceTable crossReferenceTable, bool isLenientParsing, IPdfTokenScanner pdfTokenScanner,

src/UglyToad.PdfPig/ParsingOptions.cs

@@ -48,5 +48,11 @@
         /// All passwords to try when opening this document, will include any values set for <see cref="Password"/>.
         /// </summary>
         public List<string> Passwords { get; set; } = new List<string>();
+
+        /// <summary>
+        /// Skip extracting content where the font could not be found, will result in some letters being skipped/missed
+        /// but will prevent the library throwing where the source PDF has some corrupted text.
+        /// </summary>
+        public bool SkipMissingFonts { get; set; } = false;
     }
 }

src/UglyToad.PdfPig/PdfDocument.cs

@@ -10,7 +10,6 @@
     using Encryption;
     using Exceptions;
     using Filters;
-    using Logging;
     using Parser;
     using Tokenization.Scanner;
     using Tokens;
@@ -28,16 +27,9 @@
         
         [NotNull]
         private readonly HeaderVersion version;
-        
-        private readonly ILog log;
 
         private readonly IInputBytes inputBytes;
 
-        private readonly bool clipPaths;
-
-        [NotNull]
-        private readonly ParsingCachingProviders cachingProviders;
-
         [CanBeNull]
         private readonly EncryptionDictionary encryptionDictionary;
 
@@ -46,6 +38,7 @@
 
         private readonly ILookupFilterProvider filterProvider;
         private readonly BookmarksProvider bookmarksProvider;
+        private readonly InternalParsingOptions parsingOptions;
 
         [NotNull]
         private readonly Pages pages;
@@ -82,11 +75,10 @@
         /// </summary>
         public bool IsEncrypted => encryptionDictionary != null;
 
-        internal PdfDocument(ILog log, 
+        internal PdfDocument(
             IInputBytes inputBytes,
             HeaderVersion version, 
             CrossReferenceTable crossReferenceTable,
-            ParsingCachingProviders cachingProviders,
             IPageFactory pageFactory,
             Catalog catalog,
             DocumentInformation information, 
@@ -95,17 +87,16 @@
             ILookupFilterProvider filterProvider,
             AcroFormFactory acroFormFactory,
             BookmarksProvider bookmarksProvider,
-            bool clipPaths)
+            InternalParsingOptions parsingOptions)
         {
-            this.log = log;
             this.inputBytes = inputBytes;
             this.version = version ?? throw new ArgumentNullException(nameof(version));
-            this.cachingProviders = cachingProviders ?? throw new ArgumentNullException(nameof(cachingProviders));
             this.encryptionDictionary = encryptionDictionary;
             this.pdfScanner = pdfScanner ?? throw new ArgumentNullException(nameof(pdfScanner));
             this.filterProvider = filterProvider ?? throw new ArgumentNullException(nameof(filterProvider));
             this.bookmarksProvider = bookmarksProvider ?? throw new ArgumentNullException(nameof(bookmarksProvider));
-            this.clipPaths = clipPaths;
+            this.parsingOptions = parsingOptions;
+
             Information = information ?? throw new ArgumentNullException(nameof(information));
             pages = new Pages(catalog, pageFactory, pdfScanner);
             Structure = new Structure(catalog, crossReferenceTable, pdfScanner);
@@ -153,11 +144,11 @@
                 throw new ObjectDisposedException("Cannot access page after the document is disposed.");
             }
 
-            log.Debug($"Accessing page {pageNumber}.");
+            parsingOptions.Logger.Debug($"Accessing page {pageNumber}.");
 
             try
             {
-                return pages.GetPage(pageNumber, clipPaths);
+                return pages.GetPage(pageNumber, parsingOptions);
             }
             catch (Exception ex)
             {
@@ -258,7 +249,7 @@
             }
             catch (Exception ex)
             {
-                log.Error("Failed disposing the PdfDocument due to an error.", ex);
+                parsingOptions.Logger.Error("Failed disposing the PdfDocument due to an error.", ex);
             }
             finally
             {