opinionated, standards non-compliant HTML parser meant to consume handwritten HTML.
Not particularly fast, probably.

- all tags must close *(even `<br/>`, `<meta/>`, `<link/>`)*
- tag names must be separated from the tag-body *(where the attributes go)* by a space character *(` `, 0x20)*
- self-closing tags must have the closing `/` at the ver y end of the body *(directly before the `>`)*
- no > in tags except at the end (not even in attributes)
- inline `<script>` and `<style>` must have their closing-tag be
  first-of-line *(excluding whitespace)*