Support Grammar Fragments by disallowing rules

[mvcisback]

It would be nice to disallow the use of certain rules when parsing.

There is already the ability to parse the grammar starting on particular rule:

GRAMMAR[rule].parse(foo)

What I would like is the ability to also disallow rules from being used:

GRAMMAR[rule1, (rule3, False)].parse(foo)

Or maybe

GRAMMAR[rule1].disallow(rule2, rule3).parse(foo)

Use case:

I have been writing DSL to act the API for a set of tools. Some tools only support a subset of the grammar. I would like to just parse that fragment and otherwise error.

The subsets are not hierarchical so simple inheritance would work, and thinking about it compositionally is slightly unnatural.

[mvcisback]

Also, my current work around is to filter the lines the start with a disallowed rule when creating the Grammar object.

It works, but it's hard to justify packaging it put

[erikrose]

If you want to whitelist, @mvcisback's solution is the thing. But if you want to blacklist, I think the ultimate solution is to make a subgrammar (once #30 is implemented) and override the rules you hate with a rule that throws an explicit error. I eventually want to implement "naughty ORs" in Parsimonious to improve error reporting, for just such situations as yours.