- Often a validator may result in multiple errors of different types,
  but still being part of the same validator / validation testing task.
  This identification can help categorize and probably provide some
  filtering mechanism and client do not have to parse the error strings
  to distinguish various error types from the same validator /
    - ErrorIdentifier is constructed from the validator name which generated
      the error and the error name.
        - A Coding error is thrown if a client tries to call GetIdentifier
          on an UsdValidationError which isn't generated via a call to
          UsdValidator::Validate(), which is what sets the _validator on the
          error.
        - Also note if validator writer doesn't provide an error name for an
          error, Error Identifier will be same as the name of the validator
          which generated the error. validationt task.

- The Change here provides static tokens for all the validators
  currently in our codebase: usd, usdGeom and usdShade.

- Updates tests and python wrappings to take care of the newly added
  identifier

Additional things:
- Cleans many 80 column with violations
- Cleans many indentations to be uniform with surrounding code.

Fixes #3241

(Internal change: 2339935)