Extraire le contenu étiqueté d'un PDF

Dans cet article, vous apprendrez comment extraire le contenu étiqueté d’un document PDF en utilisant C#.

Le code suivant fonctionne également avec la bibliothèque Aspose.PDF.Drawing.

Obtenir le contenu PDF étiqueté

Pour obtenir le contenu d’un document PDF avec du texte étiqueté, Aspose.PDF propose la propriété TaggedContent de la classe Document.

Le code suivant montre comment obtenir le contenu d’un document PDF avec du texte étiqueté :

.NET Core 3.1

// For complete examples and data files, visit https://github.com/aspose-pdf/Aspose.PDF-for-.NET
private static void GetTaggedContent()
{
    // The path to the documents directory
    var dataDir = RunExamples.GetDataDir_AsposePdf_WorkingDocuments();

    // Create PDF Document
    using (var document = new Aspose.Pdf.Document())
    {
        // Get Content for work with Tagged PDF
        Aspose.Pdf.Tagged.ITaggedContent taggedContent = document.TaggedContent;

        // Work with Tagged PDF content
        // Set Title and Language for Document
        taggedContent.SetTitle("Simple Tagged Pdf Document");
        taggedContent.SetLanguage("en-US");

        // Save Tagged PDF Document
        document.Save(dataDir + "TaggedPDFContent_out.pdf");
    }
}

.NET 8

// For complete examples and data files, visit https://github.com/aspose-pdf/Aspose.PDF-for-.NET
private static void GetTaggedContent()
{
    // The path to the documents directory
    var dataDir = RunExamples.GetDataDir_AsposePdf_WorkingDocuments();

    // Create PDF Document
    using var document = new Aspose.Pdf.Document();

    // Get Content for work with TaggedPdf
    Aspose.Pdf.Tagged.ITaggedContent taggedContent = document.TaggedContent;

    // Work with Tagged PDF content
    // Set Title and Language for Document
    taggedContent.SetTitle("Simple Tagged Pdf Document");
    taggedContent.SetLanguage("en-US");

    // Save Tagged PDF Document
    document.Save(dataDir + "TaggedPDFContent_out.pdf");
}

Obtenir la structure racine

Pour obtenir la structure racine d’un document PDF étiqueté, Aspose.PDF propose la propriété StructTreeRootElement de l’interface ITaggedContent et StructureElement. Le code suivant montre comment obtenir la structure racine d’un document PDF étiqueté :

.NET Core 3.1

// For complete examples and data files, visit https://github.com/aspose-pdf/Aspose.PDF-for-.NET
private static void GetRootStructure()
{
    // Create PDF Document
    using (var document = new Aspose.Pdf.Document())
    {
        // Get Content for work with Tagged PDF
        Aspose.Pdf.Tagged.ITaggedContent taggedContent = document.TaggedContent;

        // Set Title and Language for Document
        taggedContent.SetTitle("Tagged Pdf Document");
        taggedContent.SetLanguage("en-US");

        // Properties StructTreeRootElement and RootElement are used for access to
        // StructTreeRoot object of pdf document and to root structure element (Document structure element).
        Aspose.Pdf.LogicalStructure.StructTreeRootElement structTreeRootElement = taggedContent.StructTreeRootElement;
        Aspose.Pdf.LogicalStructure.StructureElement rootElement = taggedContent.RootElement;
    }
}

.NET 8

// For complete examples and data files, visit https://github.com/aspose-pdf/Aspose.PDF-for-.NET
private static void GetRootStructure()
{
    // Create PDF Document
    using var document = new Aspose.Pdf.Document();

    // Get Content for work with Tagged PDF
    Aspose.Pdf.Tagged.ITaggedContent taggedContent = document.TaggedContent;

    // Set Title and Language for Document
    taggedContent.SetTitle("Tagged Pdf Document");
    taggedContent.SetLanguage("en-US");

    // Properties StructTreeRootElement and RootElement are used for access to
    // StructTreeRoot object of pdf document and to root structure element (Document structure element).
    Aspose.Pdf.LogicalStructure.StructTreeRootElement structTreeRootElement = taggedContent.StructTreeRootElement;
    Aspose.Pdf.LogicalStructure.StructureElement rootElement = taggedContent.RootElement;
}

Accéder aux éléments enfants

Pour accéder aux éléments enfants d’un document PDF étiqueté, Aspose.PDF propose la classe ElementList. Le code suivant montre comment accéder aux éléments enfants d’un document PDF étiqueté :

.NET Core 3.1

// For complete examples and data files, visit https://github.com/aspose-pdf/Aspose.PDF-for-.NET
private static void AccessChildElements()
{
    // The path to the documents directory
    var dataDir = RunExamples.GetDataDir_AsposePdf_WorkingDocuments();

    // Open PDF Document
    using (var document = new Aspose.Pdf.Document(dataDir + "StructureElementsTree.pdf"))
    {
        // Get Content for work with Tagged PDF
        Aspose.Pdf.Tagged.ITaggedContent taggedContent = document.TaggedContent;

        // Access to root element(s)
        Aspose.Pdf.LogicalStructure.ElementList elementList = taggedContent.StructTreeRootElement.ChildElements;

        foreach (Aspose.Pdf.LogicalStructure.Element element in elementList)
        {
            if (element is Aspose.Pdf.LogicalStructure.StructureElement)
            {
                var structureElement = element as Aspose.Pdf.LogicalStructure.StructureElement;

                // Get properties
                string title = structureElement.Title;
                string language = structureElement.Language;
                string actualText = structureElement.ActualText;
                string expansionText = structureElement.ExpansionText;
                string alternativeText = structureElement.AlternativeText;
            }
        }

        // Access to child elements of first element in root element
        elementList = taggedContent.RootElement.ChildElements[1].ChildElements;

        foreach (Aspose.Pdf.LogicalStructure.Element element in elementList)
        {
            if (element is Aspose.Pdf.LogicalStructure.StructureElement)
            {
                var structureElement = element as Aspose.Pdf.LogicalStructure.StructureElement;

                // Set properties
                structureElement.Title = "title";
                structureElement.Language = "fr-FR";
                structureElement.ActualText = "actual text";
                structureElement.ExpansionText = "exp";
                structureElement.AlternativeText = "alt";
            }
        }

        // Save Tagged PDF Document
        document.Save(dataDir + "AccessChildElements_out.pdf");
    }
}

.NET 8

// For complete examples and data files, visit https://github.com/aspose-pdf/Aspose.PDF-for-.NET
private static void AccessChildElements()
{
    // The path to the documents directory
    var dataDir = RunExamples.GetDataDir_AsposePdf_WorkingDocuments();

    // Open PDF Document
    using var document = new Aspose.Pdf.Document(dataDir + "StructureElementsTree.pdf");

    // Get Content for work with Tagged PDF
    Aspose.Pdf.Tagged.ITaggedContent taggedContent = document.TaggedContent;

    // Access to root element(s)
    Aspose.Pdf.LogicalStructure.ElementList elementList = taggedContent.StructTreeRootElement.ChildElements;

    foreach (Aspose.Pdf.LogicalStructure.Element element in elementList)
    {
        if (element is Aspose.Pdf.LogicalStructure.StructureElement)
        {
            var structureElement = element as Aspose.Pdf.LogicalStructure.StructureElement;

            // Get properties
            string title = structureElement.Title;
            string language = structureElement.Language;
            string actualText = structureElement.ActualText;
            string expansionText = structureElement.ExpansionText;
            string alternativeText = structureElement.AlternativeText;
        }
    }

    // Access to child elements of first element in root element
    elementList = taggedContent.RootElement.ChildElements[1].ChildElements;

    foreach (Aspose.Pdf.LogicalStructure.Element element in elementList)
    {
        if (element is Aspose.Pdf.LogicalStructure.StructureElement)
        {
            var structureElement = element as Aspose.Pdf.LogicalStructure.StructureElement;

            // Set properties
            structureElement.Title = "title";
            structureElement.Language = "fr-FR";
            structureElement.ActualText = "actual text";
            structureElement.ExpansionText = "exp";
            structureElement.AlternativeText = "alt";
        }
    }

    // Save Tagged PDF Document
    document.Save(dataDir + "AccessChildElements_out.pdf");
}

Étiquetage des images dans un PDF existant

Pour étiqueter des images dans un document PDF existant, Aspose.PDF propose la méthode FindElements de la classe StructureElement. Vous pouvez ajouter un texte alternatif pour les figures en utilisant la propriété AlternativeText de la classe FigureElement.

Le code suivant montre comment étiqueter des images dans un document PDF existant :

.NET Core 3.1

// For complete examples and data files, visit https://github.com/aspose-pdf/Aspose.PDF-for-.NET
private static void TagImages()
{
    // The path to the documents directory
    var dataDir = RunExamples.GetDataDir_AsposePdf_WorkingDocuments();

    // Open PDF document
    using (var document1 = new Aspose.Pdf.Document(dataDir + "TH.pdf"))
    {
        // Gets tagged content and root structure element
        Aspose.Pdf.Tagged.ITaggedContent taggedContent = document1.TaggedContent;
        Aspose.Pdf.LogicalStructure.StructureElement rootElement = taggedContent.RootElement;

        // Set title for tagged PDF document
        taggedContent.SetTitle("Document with images");

        foreach (Aspose.Pdf.LogicalStructure.FigureElement figureElement in rootElement.FindElements<Aspose.Pdf.LogicalStructure.FigureElement>(true))
        {
            // Set AlternativeText for Figure
            figureElement.AlternativeText = "Figure alternative text (technique 2)";

            // Create and Set BBox Attribute
            var bboxAttribute = new Aspose.Pdf.LogicalStructure.StructureAttribute(Aspose.Pdf.LogicalStructure.AttributeKey.BBox);
            bboxAttribute.SetRectangleValue(new Aspose.Pdf.Rectangle(0.0, 0.0, 100.0, 100.0));

            Aspose.Pdf.LogicalStructure.StructureAttributes figureLayoutAttributes = figureElement.Attributes.GetAttributes(Aspose.Pdf.LogicalStructure.AttributeOwnerStandard.Layout);
            figureLayoutAttributes.SetAttribute(bboxAttribute);
        }

        // Move Span Element into Paragraph (find wrong span and paragraph in first TD)
        Aspose.Pdf.LogicalStructure.TableElement tableElement = rootElement.FindElements<Aspose.Pdf.LogicalStructure.TableElement>(true)[0];
        Aspose.Pdf.LogicalStructure.SpanElement spanElement = tableElement.FindElements<Aspose.Pdf.LogicalStructure.SpanElement>(true)[0];
        Aspose.Pdf.LogicalStructure.TableTDElement firstTdElement = tableElement.FindElements<Aspose.Pdf.LogicalStructure.TableTDElement>(true)[0];
        Aspose.Pdf.LogicalStructure.ParagraphElement paragraph = firstTdElement.FindElements<Aspose.Pdf.LogicalStructure.ParagraphElement>(true)[0];

        // Move Span Element into Paragraph
        spanElement.ChangeParentElement(paragraph);

        // Save PDF document
        document1.Save(dataDir + "TH_out.pdf");
    }

    // Check PDF/UA Compliance for out document
    using (var document2 = new Aspose.Pdf.Document(dataDir + "TH_out.pdf"))
    {
        bool isPdfUaCompliance = document2.Validate(dataDir + "TH_out.xml", Aspose.Pdf.PdfFormat.PDF_UA_1);
        Console.WriteLine(String.Format("PDF/UA compliance: {0}", isPdfUaCompliance));
    }
}

.NET 8

// For complete examples and data files, visit https://github.com/aspose-pdf/Aspose.PDF-for-.NET
private static void TagImages()
{
    // The path to the documents directory
    var dataDir = RunExamples.GetDataDir_AsposePdf_WorkingDocuments();

    // Open PDF document
    using var document1 = new Aspose.Pdf.Document(dataDir + "TH.pdf");

    // Gets tagged content and root structure element
    Aspose.Pdf.Tagged.ITaggedContent taggedContent = document1.TaggedContent;
    Aspose.Pdf.LogicalStructure.StructureElement rootElement = taggedContent.RootElement;

    // Set title for tagged PDF document
    taggedContent.SetTitle("Document with images");

    foreach (Aspose.Pdf.LogicalStructure.FigureElement figureElement in rootElement.FindElements<Aspose.Pdf.LogicalStructure.FigureElement>(true))
    {
        // Set AlternativeText for Figure
        figureElement.AlternativeText = "Figure alternative text (technique 2)";

        // Create and Set BBox Attribute
        var bboxAttribute = new Aspose.Pdf.LogicalStructure.StructureAttribute(Aspose.Pdf.LogicalStructure.AttributeKey.BBox);
        bboxAttribute.SetRectangleValue(new Aspose.Pdf.Rectangle(0.0, 0.0, 100.0, 100.0));

        Aspose.Pdf.LogicalStructure.StructureAttributes figureLayoutAttributes = figureElement.Attributes.GetAttributes(Aspose.Pdf.LogicalStructure.AttributeOwnerStandard.Layout);
        figureLayoutAttributes.SetAttribute(bboxAttribute);
    }

    // Move Span Element into Paragraph (find wrong span and paragraph in first TD)
    Aspose.Pdf.LogicalStructure.TableElement tableElement = rootElement.FindElements<Aspose.Pdf.LogicalStructure.TableElement>(true)[0];
    Aspose.Pdf.LogicalStructure.SpanElement spanElement = tableElement.FindElements<Aspose.Pdf.LogicalStructure.SpanElement>(true)[0];
    Aspose.Pdf.LogicalStructure.TableTDElement firstTdElement = tableElement.FindElements<Aspose.Pdf.LogicalStructure.TableTDElement>(true)[0];
    Aspose.Pdf.LogicalStructure.ParagraphElement paragraph = firstTdElement.FindElements<Aspose.Pdf.LogicalStructure.ParagraphElement>(true)[0];

    // Move Span Element into Paragraph
    spanElement.ChangeParentElement(paragraph);

    // Save PDF document
    document1.Save(dataDir + "TH_out.pdf");

    // Check PDF/UA Compliance for out document
    using var document2 = new Aspose.Pdf.Document(dataDir + "TH_out.pdf");

    bool isPdfUaCompliance = document2.Validate(dataDir + "TH_out.pdf", Aspose.Pdf.PdfFormat.PDF_UA_1);
    Console.WriteLine(String.Format("PDF/UA compliance: {0}", isPdfUaCompliance));
}