Module casefold 

Case-insensitive matching (FR-A10) as closure of symbol sets.

A CaseFold is a canonicalisation canon: code point -> code point (stored sparsely as the pairs where it changes a character). Two characters match case-insensitively when their canonical forms are equal, so the set matched by a class S is { c : canon(c) in canon(S) }.

Dialects differ only in the canonicalisation (RULE-10):

• ECMAScript without u: toUpperCase when it yields one character and does not map non-ASCII into ASCII (ECMA-262 Canonicalize);
• ECMAScript with u, Go, Rust, PCRE2: Unicode simple case folding;
• Python re: simple lowercase plus the interpreter’s equivalence keys (e.g. i also matches dotless ı), or ASCII-only under re.ASCII.

No supported dialect defines full case folding (ß vs ss) for matching, so none is implemented.

Structs

CaseFold

Sparse canonicalisation and the closure operations built on it.

Enums

CaseModel

Which canonicalisation to use.