feat: don't parse pdf asap

Benchmark/Program.cs

@@ -7,7 +7,6 @@
 using BenchmarkDotNet.Running;
 using FacturXDotNet;
 using FacturXDotNet.Models.CII;
-using FacturXDotNet.Parsing;
 using FacturXDotNet.Validation;
 
 BenchmarkRunner.Run<BenchmarkCii>();
@@ -44,8 +43,8 @@ public async Task ExtractCiiXml()
             string sourceFilePath = GetSourceFilePath();
             await using FileStream file = File.OpenRead(sourceFilePath);
 
-            FacturXParser parser = new();
-            _ = await parser.ParseFacturXPdfAsync(file);
+            FacturXDocument document = await FacturXDocument.FromFileAsync(GetSourceFilePath());
+            _ = document.GetCrossIndustryInvoiceAttachmentAsync();
         }
 
         [Benchmark]
@@ -54,11 +53,9 @@ public async Task ValidateFacturX()
             string sourceFilePath = GetSourceFilePath();
             await using FileStream file = File.OpenRead(sourceFilePath);
 
-            FacturXParser parser = new();
-            FacturXDocument document = await parser.ParseFacturXPdfAsync(file);
-
+            FacturXDocument document = await FacturXDocument.FromFileAsync(GetSourceFilePath());
             FacturXValidator validator = new();
-            _ = validator.IsValid(document);
+            _ = await validator.IsValidAsync(document);
         }
 
         string GetSourceFilePath()

FacturXDotNet.CLI/Extract/ExtractCommand.cs

@@ -115,15 +115,19 @@ await AnsiConsole.Status()
         return 0;
     }
 
-    static async Task ExtractCii(FileInfo pdfPath, string? ciiAttachment, string outputPath, CancellationToken cancellationToken)
+    static async Task ExtractCii(FileInfo pdfPath, string? ciiAttachmentName, string outputPath, CancellationToken cancellationToken)
     {
         await using FileStream stream = pdfPath.OpenRead();
+        FacturXDocument document = await FacturXDocument.FromStream(stream, cancellationToken);
 
-        FacturXCrossIndustryInvoiceExtractor extractor = new(new FacturXCrossIndustryInvoiceExtractorOptions { CiiXmlAttachmentName = ciiAttachment });
-        await using Stream xmpStream = extractor.ExtractCiiAsync(stream);
+        CrossIndustryInvoiceAttachment? ciiAttachment = await document.GetCrossIndustryInvoiceAttachmentAsync(ciiAttachmentName, cancellationToken: cancellationToken);
+        if (ciiAttachment == null)
+        {
+            return;
+        }
 
         await using FileStream xmpFile = File.Open(outputPath, FileMode.Create);
-        await xmpStream.CopyToAsync(xmpFile, cancellationToken);
+        await ciiAttachment.CopyToAsync(xmpFile, cancellationToken: cancellationToken);
     }
 
     static async Task ExtractXmp(FileInfo pdfPath, string outputPath, CancellationToken cancellationToken)

FacturXDotNet.CLI/Validate/ValidateCommand.cs

@@ -2,7 +2,6 @@
 using System.CommandLine.Parsing;
 using System.Diagnostics;
 using FacturXDotNet.Models;
-using FacturXDotNet.Parsing;
 using FacturXDotNet.Validation;
 using Humanizer;
 using Spectre.Console;
@@ -83,13 +82,6 @@ protected override async Task<int> RunImplAsync(ValidateCommandOptions options,
         ShowOptions(options);
         AnsiConsole.WriteLine();
 
-        FacturXParserOptions parserOptions = new()
-        {
-            CiiXmlAttachmentName = options.CiiAttachment
-        };
-
-        FacturXParser parser = new(parserOptions);
-
         FacturXDocument? facturX = null;
         await AnsiConsole.Status()
             .Spinner(Spinner.Known.Default)
@@ -101,7 +93,7 @@ await AnsiConsole.Status()
                     sw.Start();
 
                     await using FileStream stream = options.Path.OpenRead();
-                    facturX = await parser.ParseFacturXPdfAsync(stream);
+                    facturX = await FacturXDocument.FromStream(stream, cancellationToken);
 
                     sw.Stop();
 
@@ -122,17 +114,17 @@ await AnsiConsole.Status()
         validationOptions.RulesToSkip.AddRange(options.RulesToSkip);
 
         FacturXValidationResult result = default;
-        AnsiConsole.Status()
+        await AnsiConsole.Status()
             .Spinner(Spinner.Known.Default)
-            .Start(
+            .StartAsync(
                 "Validating...",
-                _ =>
+                async _ =>
                 {
                     Stopwatch sw = new();
                     sw.Start();
 
                     FacturXValidator validator = new(validationOptions);
-                    result = validator.GetValidationResult(facturX);
+                    result = await validator.GetValidationResultAsync(facturX, options.CiiAttachment, cancellationToken: cancellationToken);
 
                     sw.Stop();
 
@@ -184,7 +176,7 @@ static void ShowOptions(ValidateCommandOptions options)
 
     static void ShowFinalResult(FacturXDocument document, FacturXValidationResult result)
     {
-        FacturXProfile documentProfile = document.CrossIndustryInvoice.ExchangedDocumentContext.GuidelineSpecifiedDocumentContextParameterId.ToFacturXProfile();
+        FacturXProfile documentProfile = result.ExpectedProfile;
         FacturXProfile detectedProfile = result.ValidProfiles.GetMaxProfile();
         if (result.Success)
         {
@@ -200,7 +192,11 @@ static void ShowFinalResult(FacturXDocument document, FacturXValidationResult re
         {
             AnsiConsole.MarkupLine("[red]:cross_mark: The Factur-X docuemnt is invalid.[/]");
             AnsiConsole.MarkupLine($"[red]:cross_mark: Document profile: {documentProfile}.[/]");
-            AnsiConsole.MarkupLine($"[red]:cross_mark: Detected profile: {detectedProfile}.[/]");
+
+            if (documentProfile != detectedProfile)
+            {
+                AnsiConsole.MarkupLine($"[red]:cross_mark: Detected profile: {detectedProfile}.[/]");
+            }
         }
     }
 }

FacturXDotNet.sln.DotSettings.user

@@ -20,6 +20,7 @@
 	<s:String x:Key="/Default/CodeInspection/ExcludedFiles/FilesAndFoldersToSkip2/=7020124F_002D9FFC_002D4AC3_002D8F3D_002DAAB8E0240759_002Ff_003APdfNameTree_002Ecs_002Fl_003AC_0021_003FUsers_003Flahki_003FAppData_003FRoaming_003FJetBrains_003FRider2024_002E3_003Fresharper_002Dhost_003FDecompilerCache_003Fdecompiler_003Fcbf868f2490746fbab2087b4b2ac8198ff600_003F6f_003F97381f3b_003FPdfNameTree_002Ecs/@EntryIndexedValue">ForceIncluded</s:String>
 	<s:String x:Key="/Default/CodeInspection/ExcludedFiles/FilesAndFoldersToSkip2/=7020124F_002D9FFC_002D4AC3_002D8F3D_002DAAB8E0240759_002Ff_003APdfString_002Ecs_002Fl_003AC_0021_003FUsers_003Flahki_003FAppData_003FRoaming_003FJetBrains_003FRider2024_002E3_003Fresharper_002Dhost_003FDecompilerCache_003Fdecompiler_003Fcbf868f2490746fbab2087b4b2ac8198ff600_003F7e_003Fddd2c685_003FPdfString_002Ecs/@EntryIndexedValue">ForceIncluded</s:String>
 	<s:String x:Key="/Default/CodeInspection/ExcludedFiles/FilesAndFoldersToSkip2/=7020124F_002D9FFC_002D4AC3_002D8F3D_002DAAB8E0240759_002Ff_003AReaderProperties_002Ecs_002Fl_003AC_0021_003FUsers_003Flahki_003FAppData_003FRoaming_003FJetBrains_003FRider2024_002E3_003Fresharper_002Dhost_003FDecompilerCache_003Fdecompiler_003Fcbf868f2490746fbab2087b4b2ac8198ff600_003F0c_003F131a4d44_003FReaderProperties_002Ecs/@EntryIndexedValue">ForceIncluded</s:String>
+	<s:String x:Key="/Default/CodeInspection/ExcludedFiles/FilesAndFoldersToSkip2/=7020124F_002D9FFC_002D4AC3_002D8F3D_002DAAB8E0240759_002Ff_003AReadOnlyMemoryExtensions_002Ecs_002Fl_003AC_0021_003FUsers_003Flahki_003FAppData_003FRoaming_003FJetBrains_003FRider2024_002E3_003Fresharper_002Dhost_003FSourcesCache_003F205abcd42df15dacc85872d4a43d4d724cfda7324c41952763ef28513b8594_003FReadOnlyMemoryExtensions_002Ecs/@EntryIndexedValue">ForceIncluded</s:String>
 	<s:String x:Key="/Default/CodeInspection/ExcludedFiles/FilesAndFoldersToSkip2/=7020124F_002D9FFC_002D4AC3_002D8F3D_002DAAB8E0240759_002Ff_003ARootCommand_002Ecs_002Fl_003AC_0021_003FUsers_003Flahki_003FAppData_003FRoaming_003FJetBrains_003FRider2024_002E3_003Fresharper_002Dhost_003FSourcesCache_003Ffc4ac513e724771c6a8c5af24a64c6f0818fe32965f637b47eb49bb25f569f_003FRootCommand_002Ecs/@EntryIndexedValue">ForceIncluded</s:String>
 	<s:String x:Key="/Default/CodeInspection/ExcludedFiles/FilesAndFoldersToSkip2/=7020124F_002D9FFC_002D4AC3_002D8F3D_002DAAB8E0240759_002Ff_003ASpinner_002EGenerated_002Ecs_002Fl_003AC_0021_003FUsers_003Flahki_003FAppData_003FRoaming_003FJetBrains_003FRider2024_002E3_003Fresharper_002Dhost_003FSourcesCache_003F2785637749a235e9a285d99c8285b3ce12be9e079838976c3d8b3551a6eb6_003FSpinner_002EGenerated_002Ecs/@EntryIndexedValue">ForceIncluded</s:String>
 	<s:String x:Key="/Default/CodeInspection/ExcludedFiles/FilesAndFoldersToSkip2/=7020124F_002D9FFC_002D4AC3_002D8F3D_002DAAB8E0240759_002Ff_003ASymbolResult_002Ecs_002Fl_003AC_0021_003FUsers_003Flahki_003FAppData_003FRoaming_003FJetBrains_003FRider2024_002E3_003Fresharper_002Dhost_003FSourcesCache_003F21317424d3675027d3784fccc07d603d802830f9542ce12a7ec94118e59e0_003FSymbolResult_002Ecs/@EntryIndexedValue">ForceIncluded</s:String>

FacturXDotNet/CrossIndustryInvoiceAttachment.cs

@@ -0,0 +1,39 @@
+using FacturXDotNet.Parsing.CII;
+
+namespace FacturXDotNet;
+
+/// <summary>
+///     The Cross-Industry Invoice attachment.
+/// </summary>
+public class CrossIndustryInvoiceAttachment : FacturXDocumentAttachment
+{
+    /// <summary>
+    ///     The Cross-Industry Invoice attachment.
+    /// </summary>
+    /// <param name="facturX">The Factur-X document.</param>
+    /// <param name="name">The name of the attachment in the Factur-X document.</param>
+    internal CrossIndustryInvoiceAttachment(FacturXDocument facturX, string name) : base(facturX, name)
+    {
+    }
+
+    /// <summary>
+    ///     Get the parsed Cross-Industry Invoice.
+    /// </summary>
+    /// <param name="password">The password to open the PDF document.</param>
+    /// <param name="options">The options to parse the Cross-Industry Invoice.</param>
+    /// <param name="cancellationToken">The cancellation token.</param>
+    /// <returns>The parsed Cross-Industry Invoice.</returns>
+    public async Task<CrossIndustryInvoice> GetCrossIndustryInvoiceAsync(
+        string? password = null,
+        CrossIndustryInvoiceParserOptions? options = null,
+        CancellationToken cancellationToken = default
+    )
+    {
+        options ??= new CrossIndustryInvoiceParserOptions();
+
+        await using Stream stream = await FindAttachmentStreamAsync(password, cancellationToken);
+
+        CrossIndustryInvoiceParser parser = new(options);
+        return parser.ParseCiiXml(stream);
+    }
+}

FacturXDotNet/FacturXDocument.cs

@@ -1,22 +1,155 @@
-namespace FacturXDotNet;
+using System.Runtime.CompilerServices;
+using System.Text.RegularExpressions;
+using CommunityToolkit.HighPerformance;
+using FacturXDotNet.Parsing;
+using FacturXDotNet.Parsing.XMP;
+using PdfSharp.Pdf;
+using PdfSharp.Pdf.IO;
+
+namespace FacturXDotNet;
 
 /// <summary>
 ///     A Factur-X document.
 /// </summary>
-public class FacturXDocument
+public partial class FacturXDocument
 {
     /// <summary>
-    ///     The XMP metadata of the Factur-X document.
+    ///     Create a new Factur-X document.
+    /// </summary>
+    public FacturXDocument(ReadOnlyMemory<byte> data)
+    {
+        Data = data;
+    }
+
+    /// <summary>
+    ///     The raw document.
     /// </summary>
-    public required XmpMetadata XmpMetadata { get; set; }
+    public ReadOnlyMemory<byte> Data { get; }
 
     /// <summary>
-    ///     Information about the Cross-Industry Invoice file that was found in the Factur-X document.
+    ///     Get the XMP metadata of the Factur-X document.
     /// </summary>
-    public required CrossIndustryInvoiceFileInformation CrossIndustryInvoiceFileInformation { get; set; }
+    /// <param name="password">The password to open the PDF document.</param>
+    /// <param name="xmpParserOptions">The options to parse the XMP metadata.</param>
+    /// <param name="cancellationToken">The cancellation token.</param>
+    /// <returns>The XMP metadata of the Factur-X document.</returns>
+    public async Task<XmpMetadata?> GetXmpMetadataAsync(string? password = null, XmpMetadataParserOptions? xmpParserOptions = null, CancellationToken cancellationToken = default)
+    {
+        using PdfDocument pdfDocument = await OpenPdfDocumentAsync(password, cancellationToken);
+
+        ExtractXmpFromFacturX extractor = new();
+        if (!extractor.TryExtractXmpMetadata(pdfDocument, out Stream? xmpStream))
+        {
+            return null;
+        }
+
+        await using Stream _ = xmpStream;
+
+        // TODO: avoid these two extra copies, it is only required because TurboXML doesn't support the <?xpacket...?> processing instructions
+        // an issue has been opened to address this: https://github.com/xoofx/TurboXml/issues/6
+        // I need to fix this in the library, but it will take some time
+
+        using StreamReader reader = new(xmpStream);
+        string content = await reader.ReadToEndAsync(cancellationToken);
+        string transformedContent = PacketInstructions().Replace(content, string.Empty);
+
+        await using MemoryStream transformedStream = new(transformedContent.Length + 54);
+        await using StreamWriter writer = new(transformedStream);
+        await writer.WriteAsync("<?xml version=\"1.0\" encoding=\"UTF-8\" standalone=\"no\"?>");
+        await writer.WriteAsync(transformedContent);
+        await writer.FlushAsync(cancellationToken);
+        transformedStream.Seek(0, SeekOrigin.Begin);
+
+        XmpMetadataParser xmpParser = new(xmpParserOptions);
+        return xmpParser.ParseXmpMetadata(transformedStream);
+    }
 
     /// <summary>
-    ///     The Cross-Industry Invoice of the Factur-X document.
+    ///     Get the Cross-Industry Invoice of the Factur-X document.
     /// </summary>
-    public required CrossIndustryInvoice CrossIndustryInvoice { get; set; }
+    /// <param name="attachmentFileName">The name of the attachment containing the Cross-Industry Invoice XML file. If not specified, the default name 'factur-x.xml' will be used.</param>
+    /// <param name="password">The password to open the PDF document.</param>
+    /// <param name="cancellationToken">The cancellation token.</param>
+    /// <returns>The Cross-Industry Invoice of the Factur-X document.</returns>
+    public async Task<CrossIndustryInvoiceAttachment?> GetCrossIndustryInvoiceAttachmentAsync(
+        string? attachmentFileName = null,
+        string? password = null,
+        CancellationToken cancellationToken = default
+    )
+    {
+        attachmentFileName ??= "factur-x.xml";
+        using PdfDocument pdfDocument = await OpenPdfDocumentAsync(password, cancellationToken);
+
+        ExtractAttachmentsFromFacturX extractor = new();
+        foreach ((string Name, Stream Content) attachment in extractor.ExtractFacturXAttachments(pdfDocument))
+        {
+            if (attachment.Name != attachmentFileName)
+            {
+                continue;
+            }
+
+            return new CrossIndustryInvoiceAttachment(this, attachmentFileName);
+        }
+
+        return null;
+    }
+
+    /// <summary>
+    ///     Get the attachments of the Factur-X document.
+    /// </summary>
+    /// <param name="password">The password to open the PDF document.</param>
+    /// <param name="cancellationToken">The cancellation token.</param>
+    /// <returns>The attachments of the Factur-X document.</returns>
+    public async IAsyncEnumerable<FacturXDocumentAttachment> GetAttachmentsAsync(string? password = null, [EnumeratorCancellation] CancellationToken cancellationToken = default)
+    {
+        using PdfDocument pdfDocument = await OpenPdfDocumentAsync(password, cancellationToken);
+
+        ExtractAttachmentsFromFacturX extractor = new();
+        foreach ((string Name, Stream Content) attachment in extractor.ExtractFacturXAttachments(pdfDocument))
+        {
+            byte[] attachmentBytes = new byte[attachment.Content.Length];
+            attachment.Content.ReadExactly(attachmentBytes);
+            yield return new FacturXDocumentAttachment(this, attachment.Name);
+        }
+    }
+
+    internal async Task<PdfDocument> OpenPdfDocumentAsync(string? password, CancellationToken _ = default)
+    {
+        await using Stream stream = Data.AsStream();
+        PdfDocument document;
+
+        if (password != null)
+        {
+            document = PdfReader.Open(stream, PdfDocumentOpenMode.Import, args => args.Password = password);
+        }
+        else
+        {
+            document = PdfReader.Open(stream, PdfDocumentOpenMode.Import);
+        }
+
+        return document;
+    }
+
+    /// <summary>
+    ///     Create a new Factur-X document from a file.
+    /// </summary>
+    public static async Task<FacturXDocument> FromFileAsync(string filePath, CancellationToken cancellationToken = default)
+    {
+        byte[] buffer = await File.ReadAllBytesAsync(filePath, cancellationToken);
+        return new FacturXDocument(buffer);
+    }
+
+    /// <summary>
+    ///     Create a new Factur-X document from a stream.
+    /// </summary>
+    /// <remarks>This method will copy the entire stream. Consider using the <see cref="FacturXDocument(ReadOnlyMemory{byte})" /> constructor if the data is already in memory.</remarks>
+    public static async Task<FacturXDocument> FromStream(Stream stream, CancellationToken cancellationToken = default)
+    {
+        byte[] buffer = new byte[stream.Length];
+        await stream.ReadExactlyAsync(buffer, cancellationToken);
+        return new FacturXDocument(buffer);
+    }
+
+    [GeneratedRegex("<\\?xpacket.*?\\?>")]
+    private static partial Regex PacketInstructions();
 }