- <para>Declares <command>u</command> as an implicit coercion.
- If the type of <command>u</command> is
- <command>∀x1:T1. … ∀x(n-1):T(n-1).Tn</command> the coercion target is
- <command>T(n - ariety)</command> while its source is
- <command>T(n - ariety - saturation - 1)</command>.
- Every time a term <command>x</command>
- of type source is used with expected type target, Matita
- automatically replaces <command>x</command> with
- <command>(u ? … ? x ? … ?)</command> to avoid a typing error.</para>
- Note that the number of <command>?</command> added after
- <command>x</command> is saturation.
- <para>Implicit coercions are not displayed to the user:
- <command>(u ? … ? x)</command> is rendered simply
- as <command>x</command>.</para>
- <para>When a coercion <command>u</command> is declared
- from source <command>s</command> to target <command>t</command>
- and there is already a coercion <command>u'</command> of
- target <command>s</command> or source <command>t</command>,
- a composite implicit coercion is automatically computed
- by Matita unless <emphasis role="bold">nocomposites</emphasis>
- is specified.</para>
+ <para>Declares the hint at precedence <command>n</command></para>
+ <para>The file <command>hints_declaration.ma</command> must be
+ included to declare hints with that syntax.</para>
+ <para>Unification hints are described in the paper
+ "Hints in unification" by
+ Asperti, Ricciotti, Sacerdoti and Tassi.
+ </para>