Add inclusive words filter

[DavidOgunsAWS]

Issue #, if available:

• Adding validator to catch usage of non-inclusive words in the text of a Smithy model

Description of changes:

• Since the list of non-inclusive words is statically small, those are hard coded
• The full text scanner internal to the validator can be abstracted further so a single pass over the model runs multiple validators more efficiently.

TODO:
Testing so far has been local. Follow up commit will add checked-in unit tests

By submitting this pull request, I confirm that my contribution is made under the terms of the Apache 2.0 license.

[kstich]

Are either of these properties required? What's the default value for appendDefaults if it's not required?

[DavidOgunsAWS]

Will add documentation. It is worth noting there is a non-trivial, maybe counter intuitive defaults behavior:

Though appendDefaults defaults to false, if noninclusiveTerms mappings is entirely unset or empty, appendDefaults behaves as if it were true -- the built in mappings are present. noninclusiveTerms has to be non-empty before appendDefaults behavior applies. If this behavior is acceptable, then I'll focus on clear and concise documentation for it. If not, then I should change the implementation.

[DavidOgunsAWS]

Property structure have changed a bit. But current properties are documented for required or not, along with what the default values are

[kstich]

Is the reasoning for this structure change because you weren't happy with the behavior? I feel like a terms map and a includeDefaults-like boolean that defaults to true would be pretty clear and usable for customers.

[kstich]

This looks like it returned or wasn't fixed initially.

[kstich]

Is the reasoning for this structure change because you weren't happy with the behavior? I feel like a terms map and a includeDefaults-like boolean that defaults to true would be pretty clear and usable for customers.

[kstich]

Let's squash this PR on merge if it's viable.