Database Reference
In-Depth Information
n . This concrete DB-syntax in Def-
inition 53 is provided by a bijection between the set of abstract and these concrete
terms,
2 by the corresponding DB n -ary coproducts
+
T : T P (X) T P (X) , such that
0 ,
T (nil) =⊥
T (a i .t) = a i ( T (t)) ,
n (t 1 ,...,t n ))
n ((
and
T (
=+
T (t 1 )),...,(
T (t 2 ))) .
Consequently, for the set of abstract trees
T
we define its corresponding set of
T : T P (X) T P (X) and (for the ground
concrete trees
T
, with the bijection
terms only)
T : T →T . Now we are able to define the semantics of process-
programs based on the coalgebras, both for the abstract
T P and concrete DB-syntax
T P :
Proposition 37
T P (X) be a guarded system of equations for a
database mapping process-program P ( in Proposition 34 ). Thus , the coalgebraic
semantics for the database mapping process-programs can be given by the follow-
ing commutative diagram in Set :
Let g E :
X
where
T is the isomorphism between the abstract and DB-terms in Definition 53 ,
so that the terms in
T
are the objects in DB , that is , the instance-databases . The
isomorphism
F is the concrete interpretation of the Lambek's final coalgebra iso-
morphism
Σ . Consequently , we are able to use
T
Ob DB as a DB-denotational
semantics for the abstract Σ P -algebras .
Proof From [ 4 ], the inner commutative diagram (1) represents the final coalgebra
semantics for Σ P , while the diagram (0) is composed of only isomorphisms, so
that
T
Ob DB can be used as an adequate DB-denotational semantics algebra h
:
Σ P (
→T , represented by the (syntax) initial algebra commutative diagram
in Sect. 7.3.1 , where Z
T )
=T
, such that Σ P (
T
)
=
Act
∪{
nil
}∪{T ×{
i
}|
a i
} n 2 T
n
0 , and for t 1 ,t 2 ∈T
Act
, with h being the identity for Act , h(nil)
=⊥
,
n (t 1 ,...,t n ) .
So we obtain the following initial algebra semantics commutative diagram:
h(t 1 ,i)
=
a i
t 1 , h(t 1 ,...,t n )
=+
Search WWH ::




Custom Search