Awesome

Neo.Markdig.Xaml

Is a xaml/wpf extension for lunet-io/markdig

With this library it is possible to create FlowDocument's

var content = File.ReadAllText(@"..\..\..\..\Readme.md");
var doc = MarkdownXaml.ToFlowDocument(content,
	new MarkdownPipelineBuilder()
	.UseXamlSupportedExtensions()
	.Build()
);
flowDocumentViewer.Document = doc;

or Xaml-Source files.

var content = File.ReadAllText(@"..\..\..\..\Readme.md");
var xaml = MarkdownXaml.ToXaml(content,
	new MarkdownPipelineBuilder()
	.UseXamlSupportedExtensions()
	.Build()
);

Features

Supports standard features from Markdig.

Additionally, the following extensions are supported:

AutoLinks (such as https://github.com)
Tables
Extra emphasis

Styles

It is possible to replace the default resource set.

Inherit from XamlMarkdownWriter and override the method GetDefaultStyle.

Difference to markdig.wpf

markdig.wpf implements two renderers, one for xaml-output and one for wpf-objects.

This implementation has only one renderer, that has a xaml-token-stream as output. This stream can written with framework implemented XamlXmlWriter or the XamlObjectWriter.

Numbers:

Markdig.Wpf-toxaml  : 109ms
Markdig.Xaml-toxaml : 850ms
Markdig.Wpf-towpf   : 12.579ms
Markdig.Xaml-towpf  : 6.882ms

If you want to create plain xaml text output, my library is a slower. The reason is, that my implementation validates the output against the wpf-object-schema. That is one expensive step more, but it ensures a valid wpf-xaml.

The object creation should be always faster in Neo.Markdig.Xaml than the Markdig.Wpf approach, because Neo.Markdig.Xaml uses the internal wpf-object creation with all is caching and replay mechanism.

Needs to prove: Neo.Markdig.Xaml should consume less memory on large documents in both cases.

What is missing