(migrated from the mailing list)
Johannes Koch said:
ED 2014-03-21, section 4.3.11 Consume a name:
When the algorithm is based on the consumed next token, at the end of
this section (when returning the name), the current input code point is
the first after the name. This code point will be missed when e.g.
returning to 4.3.1 Consume a token, which again consumes the next token.
So. e.g., in 4.3.11 the first non-name code point could be reconsumed
before returning the name.
Tab Atkins said:
Thanks for yet another detailed catch! Fixed.
(migrated from the mailing list)
Johannes Koch said:
Tab Atkins said: