[Tracking] Parser improvements feedback

[ajaybhargavb]

I'm putting all the miscellaneous improvements we want to do after we get the end-to-end working in the list below,

• Cleanup attribute value (First step: Get rid of GenericBlock)
• Cleanup how CSharpParser.ParseBlock()` works
• Cleanup MarkupElementRewriter (Original comment: Added MarkupElementRewriter aspnet/Razor#2624 (review))
• Should we generate classified spans for a tag helper end tag?
• In cases like <p class= >, the attribute value should contain a marker. Currently it is null.
• Add TagHelperPrefix to generated baselines
• Razor directive descriptors and directive token descriptors should be first class properties on the directive nodes
• Need to rethink what goes inside RazorDirectiveBody (New baselines for RazorDirectivesTest aspnet/Razor#2621 (comment))
• Should we create a new type for significant whitespace? (New baselines for RazorDirectivesTest aspnet/Razor#2621 (comment))
• @: shouldn't be represented as CSharpTemplateBlock (New baselines for CSharpTemplateTest and CSharpVerbatimBlockTest aspnet/Razor#2708 (comment))
• Why does WhitespaceRewriter exist separately?
• Generate TagHelperSpans with more information on attributes
• Trailing extensible directive whitespace is considered markup at design time (https://github.com/aspnet/AspNetCore-Internal/issues/1464)
• Get rid of the remaining few chunk generators

[mkArtakMSFT]

@ajaybhargavb what needs to happen here? Can this be closed already?

[ajaybhargavb]

This is a full list of all the quality of life improvements we wanted to do to the parser. But at this point the ROI on this is low. I suggest keeping this around in the backlog in case we end up needing to do one or more of these in the future to light up other scenarios. At this point, there are none.