selector syntax regression

[robdockins]

Recent changes in the selector syntax seem to have broken some previous code. This is from the saw-script repository, in the example ECDSA verification.

https://github.com/GaloisInc/saw-script/blob/6247f9ac936628cb6f9e74f33738e059eb32582b/examples/ecdsa/cryptol-spec/p384_ec_mul.cry#L212

Maybe we accidentally removed prime characters from the lexical syntax for selectors?

$ ./cry run ~/code/saw-script/examples/ecdsa/cryptol-spec/p384_ec_mul.cry
┏━╸┏━┓╻ ╻┏━┓╺┳╸┏━┓╻
┃  ┣┳┛┗┳┛┣━┛ ┃ ┃ ┃┃
┗━╸╹┗╸ ╹ ╹   ╹ ┗━┛┗━╸
version 2.9.1.99 (96fcce9, modified)
https://cryptol.net  :? for help

Parse error at /Users/rdockins/code/saw-script/examples/ecdsa/cryptol-spec/p384_ec_mul.cry:212:35--212:44
  malformed selector: .tma2_c0'

[yav]

A yes, I forgot about those.