Convert markdown to normal xml #158

Activity

1. carlossanlop

   added

   port-to-tripleslash

   on Feb 3, 2023
2. carlossanlop commented on Feb 7, 2023

   

   carlossanlop

   on Feb 7, 2023

   ContributorAuthor

   More actions

   We have functions that do this task:

   api-docs-sync/src/PortToTripleSlash/src/libraries/RoslynTripleSlash/TripleSlashSyntaxRewriter.cs

   Lines 922 to 990 in cc8e3e8

   	private static XmlNodeSyntax GetTextAsFormatCData(string text, SyntaxTriviaList leadingWhitespace)
   	{
   	XmlTextSyntax remarks = GetTextAsCommentedTokens(text, leadingWhitespace, wrapWithNewLines: true);
   	XmlNameSyntax formatName = SyntaxFactory.XmlName("format");
   	XmlAttributeSyntax formatAttribute = SyntaxFactory.XmlTextAttribute("type", "text/markdown");
   	var formatAttributes = new SyntaxList<XmlAttributeSyntax>(formatAttribute);
   	var formatStart = SyntaxFactory.XmlElementStartTag(formatName, formatAttributes);
   	var formatEnd = SyntaxFactory.XmlElementEndTag(formatName);
   	XmlCDataSectionSyntax cdata = SyntaxFactory.XmlCDataSection(remarks.TextTokens);
   	var cdataList = new SyntaxList<XmlNodeSyntax>(cdata);
   	XmlElementSyntax contents = SyntaxFactory.XmlElement(formatStart, cdataList, formatEnd);
   	return contents;
   	}
   	private static string RemoveUnnecessaryMarkdown(string text)
   	{
   	text = Regex.Replace(text, @"<!\[CDATA\[(\r?\n)*[\t ]*", "");
   	text = Regex.Replace(text, @"\]\]>", "");
   	text = Regex.Replace(text, @"##[ ]?Remarks(\r?\n)*[\t ]*", "");
   	return text;
   	}
   	private static string ReplaceMarkdownWithXmlElements(string text, List<DocsParam> docsParams, List<DocsTypeParam> docsTypeParams)
   	{
   	text = CleanXrefs(text);
   	// commonly used url entities
   	text = Regex.Replace(text, @"%23", "#");
   	text = Regex.Replace(text, @"%28", "(");
   	text = Regex.Replace(text, @"%29", ")");
   	text = Regex.Replace(text, @"%2C", ",");
   	// hyperlinks
   	text = Regex.Replace(text, RegexMarkdownLinkPattern, RegexHtmlLinkReplacement);
   	// bold
   	text = Regex.Replace(text, RegexMarkdownBoldPattern, RegexXmlBoldReplacement);
   	// code snippet
   	text = Regex.Replace(text, RegexMarkdownCodeStartPattern, RegexXmlCodeStartReplacement);
   	text = Regex.Replace(text, RegexMarkdownCodeEndPattern, RegexXmlCodeEndReplacement);
   	// langwords|parameters|typeparams
   	MatchCollection collection = Regex.Matches(text, @"(?<backtickedParam>`(?<paramName>[a-zA-Z0-9_]+)`)");
   	foreach (Match match in collection)
   	{
   	string backtickedParam = match.Groups["backtickedParam"].Value;
   	string paramName = match.Groups["paramName"].Value;
   	if (ReservedKeywords.Any(x => x == paramName))
   	{
   	text = Regex.Replace(text, $"{backtickedParam}", $"<see langword=\"{paramName}\" />");
   	}
   	else if (docsParams.Any(x => x.Name == paramName))
   	{
   	text = Regex.Replace(text, $"{backtickedParam}", $"<paramref name=\"{paramName}\" />");
   	}
   	else if (docsTypeParams.Any(x => x.Name == paramName))
   	{
   	text = Regex.Replace(text, $"{backtickedParam}", $"<typeparamref name=\"{paramName}\" />");
   	}
   	}
   	return text;
   	}

   but still need to confirm with detailed unit tests that it works as expected, especially for items that are not markdown standard like [!NOTE], [!IMPORTANT] and similar.

   cc @smasher164