Stropping is to be implemented purely at the lexical level and it must not have any impact on the letters and digits that constitute bold words and tags. Therefore, in principle, of all the components of a compiler or interpreter only the lexical analyzer (or tokenizer) shall concern itself with the stropping regime.
A programming editor can highlight any sequence of two or more consecutive underscore characters as an error: these are not syntactically valid in any standard stropping regime, out of string denotations and pragments.
Implementations are advised to always gate the usage of this extension
through a command-line option. Appropriate PORTCHECK
diagnostics should be emitted whenever appropriate.