pub const RESERVED_CONTEXTUAL: &[&str];Expand description
The reserved lexer tokens the parser deliberately re-admits as identifiers
outside their one keyword position (expect_ident, ADR-tracked at
parser.rs). Unlike CONTEXTUAL_KEYWORDS these words are real
#[token]s and are members of KEYWORDS (so the lexer↔registry drift
guard sees them) — they simply are not rejected in identifier position. The
keyword reference renders them as a distinct tier so the page no longer
claims, falsely, that every listed word is unusable as an identifier.
Single source of truth: the expect_ident exemption arm (via
is_reserved_contextual) and the
is_reserved_keyword_covers_every_lexer_keyword drift guard both defer to
this list, so adding a word here is enough to make the parser admit it.