Fix attribute ordering in as_case_data() to match test snapshots

[Copilot]

When input data already carries custom attributes (e.g. output of sim_case_data()), tibble::as_tibble() reorders the attribute list and pushes class to the end. The test snapshot expects class to precede the custom attributes (id_var, biomarker_var, timeindays, value_var), causing consistent snapshot failures.

Changes

• Strip inherited custom attributes before re-adding them, using a loop over the known attribute names
• Set class first via class(result) <- union("case_data", class(data)) to anchor it in the correct position
• Re-add custom attributes in order: id_var (via serocalculator::set_id_var()), then biomarker_var, timeindays, value_var

This replaces the previous structure() call at the end of the pipe, which could not override attribute positions already established by as_tibble().

# Before: as_tibble() on a case_data object pushes `class` to the end
# names(attributes(result))
# [1] "names"         "id_var"        "biomarker_var" "timeindays"    "value_var"     "class"

# After: explicit ordering ensures class appears before custom attrs
# [1] "names"         "class"         "id_var"        "biomarker_var" "timeindays"    "value_var"

[sschildhauer]

Looks good