Error when converting HTML to PDF in Linux (container)

[Dev]

I have a method like so that takes HTML and returns a PDF: 

private static Stream ConvertToPdf(string htmlFileContent)
{
	var htmlProvider = new HtmlFormatProvider();

	var htmlDocument = htmlProvider.Import(htmlFileContent);

	var pdfProvider = new PdfFormatProvider();

	var pdfMemoryStream = new MemoryStream();

	pdfProvider.Export(htmlDocument, pdfMemoryStream);

	pdfMemoryStream.Position = 0;

	return pdfMemoryStream;
}

This method works perfectly on my local Windows dev box, but when I run it in a Fargate container in AWS ECS, I get the following exception if the HTML contains an image:

System.NotSupportedException: Not supported image format.
   at Telerik.Windows.Documents.Fixed.Model.Resources.ImageSource.DoOnUnknownData(Byte[] unknownData, ImageQuality imageQuality, Action`1 doOnEncodedData)
   at Telerik.Windows.Documents.Fixed.Model.Resources.ImageSource.InitializeImageInfoFromUnknownData(Byte[] unknownData, ImageQuality imageQuality)
   at Telerik.Windows.Documents.Fixed.Model.Resources.ImageSource.EnsureImageInfo()
   at Telerik.Windows.Documents.Fixed.Model.Resources.ImageSource.get_Width()
   at Telerik.Windows.Documents.Fixed.FormatProviders.Pdf.Model.Elements.Objects.ImageXObject.CopyPropertiesFrom(IPdfExportContext context, ImageSource imageSource)
   at Telerik.Windows.Documents.Fixed.FormatProviders.Pdf.Export.PdfExportContext.GetResource(ImageSource resource)
   at Telerik.Windows.Documents.Fixed.FormatProviders.Pdf.Export.PdfContentExportContext.GetResource(ImageSource imageSource)
   at Telerik.Windows.Documents.Fixed.FormatProviders.Pdf.Export.ContentElementWriters.ImageWriter.WriteOverride(PdfWriter writer, IPdfContentExportContext context, Image element)
   at Telerik.Windows.Documents.Fixed.FormatProviders.Pdf.Export.ContentElementWriters.MarkableContentElementWriter`1.Write(PdfWriter writer, IPdfContentExportContext context, Object element)
   at Telerik.Windows.Documents.Fixed.FormatProviders.Pdf.Export.ContentElementWriters.ClippingWriter.WriteOverride(PdfWriter writer, IPdfContentExportContext context, Clipping clipping)
   at Telerik.Windows.Documents.Fixed.FormatProviders.Pdf.Export.ContentElementWriters.ContentElementWriter`1.Write(PdfWriter writer, IPdfContentExportContext context, Object element)
   at Telerik.Windows.Documents.Fixed.FormatProviders.Pdf.Export.ContentElementWriters.ContentRootWriter.WriteOverride(PdfWriter writer, IPdfContentExportContext context, IContentRootElement element)
   at Telerik.Windows.Documents.Fixed.FormatProviders.Pdf.Export.ContentElementWriters.ContentElementWriter`1.Write(PdfWriter writer, IPdfContentExportContext context, Object element)
   at Telerik.Windows.Documents.Fixed.FormatProviders.Pdf.Export.ContentElementWriters.ContentElementWriterBase.WriteElement(PdfWriter writer, IPdfContentExportContext context, Object element)
   at Telerik.Windows.Documents.Fixed.FormatProviders.Pdf.Model.Elements.DocumentStructure.ContentStream.BuildContentData(IPdfExportContext context, IResourceHolder resourceHolder, IContentRootElement contentRootElement)
   at Telerik.Windows.Documents.Fixed.FormatProviders.Pdf.Model.Elements.DocumentStructure.Page.CopyPropertiesFrom(IPdfExportContext context, RadFixedPage fixedPage)
   at Telerik.Windows.Documents.Fixed.FormatProviders.Pdf.Model.Elements.DocumentStructure.DocumentCatalog.CopyRadFixedPageProperties(RadFixedPage source, Page destination, IRadFixedDocumentExportContext context)
   at Telerik.Windows.Documents.Fixed.FormatProviders.Pdf.Model.Elements.DocumentStructure.DocumentCatalog.CopyPagePropertiesFrom(IRadFixedDocumentExportContext context)
   at Telerik.Windows.Documents.Fixed.FormatProviders.Pdf.Model.Elements.DocumentStructure.DocumentCatalog.CopyPropertiesFrom(IRadFixedDocumentExportContext context)
   at Telerik.Windows.Documents.Fixed.FormatProviders.Pdf.Export.PdfExporter.Export(IRadFixedDocumentExportContext context, Stream output)
   at Telerik.Windows.Documents.Fixed.FormatProviders.Pdf.PdfFormatProvider.ExportOverride(RadFixedDocument document, Stream output)
   at Telerik.Windows.Documents.Common.FormatProviders.FormatProviderBase`1.Export(T document, Stream output)
   at Telerik.Windows.Documents.Flow.FormatProviders.Pdf.PdfFormatProvider.ExportOverride(RadFlowDocument document, Stream output)
   at Telerik.Windows.Documents.Common.FormatProviders.FormatProviderBase`1.Export(T document, Stream output)

All of those Telerik.Windows.Documents.* namespaces make me skeptical. I am referencing the .NET Standard NuGet packages from the Telerik feed, so they should work on Linux, right?

[Nikolay Demirev]

Hi, 

This usually happens when using the .NET Standard version of the PdfProcessing and working with images different than Jpeg and Jpeg2000. Because .NET Standard doesn't support converting image formats out of the box. Have you provided implementation for the FixedExtensibilityManager.ImagePropertiesResolver and FixedExtensibilityManager.JpegImageConverter?

You can check here for more information: Cross-platform support.

Please don't mind the namespaces. You can use the DPL on Linux if you are using the .NET Standard version.

[Dev]

Due to time constraints, we opted to remove the images altogether, but I'll keep this limitation in mind on future projects.

[Nikolay Demirev]

If you decide to do it, you could overcome this with several lines of code and references. There is a Telerik.Documents.ImageUtils assembly, where you can find default implementations based on SkiaSharp. Here is for the ImagePropertiesResolver, and this is for the JpegImageConverter.