% Input parameters

\label{s:params}
\sectiondept{it}

An implementation \shall support each parameter defined in this section for the
purpose of accepting input data.

An implementation \shallnot ^fail in ^[fail!error] if a parameter is provided
that is not listed within this section.\footnote{For example, this allows data
to be used for multiple suppliers.}

\p{param-fail} An implementation \must ^fail in ^[fail!error] if one or more of
the parameters defined in this section are not provided, unless the parameter is
unused in every calculation that applies to the input data. The exact failure
point is unspecified, but the implementation must not return the value of any
calculation.

\begin{ex}
\index{fail!error|(}
The implementation may check all required parameters before performing any
calculations, or may decide to defer parameter checking until the parameter is
actually used for the first time; the latter would be expected to be more
performant in the case that the input data is expected to be well-formed most of
the time, but will have wasted many cycles in computing calculations that will
be thrown away in event of a failure. In the former case, the upfront validation
cost is more steep, but no cycles are wasted on calculations in the event of an
immediate failure.
\index{fail!error|)}
\end{ex}

An implementation \must ^fail in ^[fail!error] if an ^argument associated with
its parameter does not fall within the ^domain of the parameter, unless the
parameter is unused in every calculation that applies to the input data. The
exact failure point is subject to the same requirements as \pref{param-fail}.

An implementation \mustnot ^fail in ^[fail!error] if an argument associated with
its parameter \emph{does} fall within its ^[domain], but \may fail for other
reasons defined within this specification unrelated to the domain.


%% content set with \inputparams will be outptu at this point