This is not done and perhaps never will be in its current form. It is just too complex, bloats the number of states in the parser, and is tedious to write. It is also not entirely clear how to correctly model the returned tokens.
A better solution (which will be taken at some point) might be to have a separate parser just for doctype stuff which can work alongside Scanner for people who really need it.
This PR unfortunately never stood any chance of being merged, and now there are too many conflicts anyways. A better approach will be considered later.
Closes #9
This is not done and perhaps never will be in its current form. It is just too complex, bloats the number of states in the parser, and is tedious to write. It is also not entirely clear how to correctly model the returned tokens.
A better solution (which will be taken at some point) might be to have a separate parser just for doctype stuff which can work alongside
Scanner
for people who really need it.