Node blacklist / whitelist

[roydejong]

The parser library currently operates mostly without any config, and it uses good guesswork to parse anything that looks like a shortcode.

That works well with some cases, but anywhere with user input, you'll run into problems sooner or later without being able to whitelist or blacklist tags.

For example, consider this text:

An Iraqi battalion has consumed [sic] [recte assumed] control of the former American military base, and our forces are now about 40 minutes outside the city.

The tree parser will break because [sic] and [recte (..)] look like an openings tag, but lack a corresponding closing tag. However, this is just plain text and the parser shouldn't even mess with it.

One solution would be to allow the parser to operate in blacklist / whitelist mode so only pre-approved tags are processed, or so certain tags are ignored.

(Marking als bug because the behavior is unintended, and enhancement because the config system needs to be improved to cover these use cases)