- Added additional lexer / parser for UTF-8 Identifiers
- Added TF setting to scope changes behind TF_UTF8_IDENTIFIERS
- Made first pass at Tf string methods to switch on setting
  and use ASCII vs. Unicode validation rules
- Enabled different rules for identifiers vs. prim names
  (validation for prim names falls back to identifiers
  when under ASCII rules)
- Added UCA sorting for UTF-8 strings and tests against
  Unicode conformance tests
- Pulled common parse layer methods out of textFileFormat
  parser to enable use by both ASCII and Unicode flavors
- Added support for Unicode UTF-8 identifiers / prim names
  in usdGenSchema

- Added missing files generated by new scripts to cache
  unicode character class data

- Implemented chosen sort algorithm
  Lexicographic ASCII + UTF-8 Byte Ordering
- Removed all functionality related to UCA
- Removed all functionality for case mapping
- Adjusted tests to reflect original ordering that did not change
  with new algorithm
- Added documentation notes where appropriate for Tf string methods
  that should only be used in ASCII-only circumstances

- Lofted up TfMakeValidPrimName / TfIsValidPrimName to Sdf

- Removed overly restrictive comments on certain TfString
  methods and documented specific restrictions
- Added TF_DEV_AXIOM statements where required to validate
  those specific restrictions

- Changed default setting for TF_UTF8_IDENTIFIERS to `true`, enabling
  new Unicode code paths by default

- Changed logic for ASCII checks to check leading bit
- Leveraged ASCII check in _LessImpl
- Removed TF_UTF8_IDENTIFIERS from header - all access goes through
  UseUTF8Identifiers which checks value of TF_UTF8_IDENTIFIERS
- Fixed typo on TfDictionaryLessThan
- Removed specific overload for checking valid prim name regardless
  of TF_UTF8_IDENTIFIERS value

- Placed appropriate parser objects back in PXR namespace

- Placed appropriate objects back in PXR namespace
- Modified yy files to ensure they emit proper parser context object
  for yyparse

- Removed empty line in comment

- Changed IsValidUTF8PrimName to IsValidUTF8Name
- Removed TF_API from several inlined methods

- Refactored naming methods to streamline implementation
- Added comment to test to indicate long term treatment
  if runtime switch is removed

- Refactored XID_Start / Continue methods to take code point
- Refactored usage of these methods to pass code point
  via dereferenced iterator