Previously, if the tokenizer encountered the sequence of characters ''\$, next_string would consume the first 3 and then loop back around and see the $ next, which it treated as the start of an interpolation.
With this change, it will instead always immediately consume the next character after ''\ before looping back. This is correct, because ''\x is a valid escape sequence for any x.
Summary & Motivation
Previously, if the tokenizer encountered the sequence of characters
''\$
,next_string
would consume the first 3 and then loop back around and see the$
next, which it treated as the start of an interpolation.With this change, it will instead always immediately consume the next character after
''\
before looping back. This is correct, because''\x
is a valid escape sequence for anyx
.Further context
Fixes #100