use alpha_id::normalize::normalize;

#[test]
fn lowercases_and_collapses_whitespace() {
    assert_eq!(normalize("  Diabetes   Mellitus\tTyp 2 "), "diabetes mellitus typ 2");
}

#[test]
fn expands_common_medical_abbreviations() {
    assert_eq!(normalize("V.a. art. Hypertonie"), "verdacht auf arterielle hypertonie");
    assert_eq!(normalize("Z.n. OP"), "zustand nach operation");
}

// The system must match regardless of whether the dictation uses real
// umlauts (ä, ö, ü, ß) or their ASCII digraphs (ae, oe, ue, ss). Both the
// indexed corpus text and the query pass through normalize(), so folding
// both forms to one canonical (ASCII-digraph) representation makes the
// lexical path symmetric — "Hyperlipidämie" and "Hyperlipidaemie" match.
#[test]
fn folds_umlauts_and_eszett_to_ascii_digraphs() {
    assert_eq!(normalize("Ödem"), "oedem");
    assert_eq!(normalize("Fußödem"), "fussoedem");
    assert_eq!(normalize("über"), "ueber");
}

#[test]
fn umlaut_and_digraph_forms_normalize_identically() {
    assert_eq!(normalize("Hyperlipidämie"), normalize("Hyperlipidaemie"));
    assert_eq!(normalize("Ödem"), normalize("Oedem"));
    assert_eq!(normalize("Fußödem"), normalize("Fussoedem"));
    assert_eq!(normalize("über"), normalize("ueber"));
}

#[test]
fn folding_is_idempotent_on_already_ascii_text() {
    // ASCII-only medical text must be unchanged by the fold.
    assert_eq!(normalize("hyperlipidaemie"), "hyperlipidaemie");
    assert_eq!(normalize("Diabetes Mellitus"), "diabetes mellitus");
}