Change enum attribute parsing to handle special characters and
multi-word
identifiers. This allows enum attrs to use symbols like "+" and strings
with separators like "dash-separated-sentence" instead of being limited
to
valid identifiers.
This also aligns enum attribute parsing with how enums are already
handled
by the `FieldParser`:
https://github.com/llvm/llvm-project/blob/main/mlir/tools/mlir-tblgen/EnumsGen.cpp#L108
Signed-off-by: Fabian Mora <fabian.mora-cordero@amd.com>
913d44da83