We investigate the relationship between programming with constraints and database query languages. We show that efficient, declarative database programming can be combined with efficient constraint solving. The key intuition is that the generalization of a ground fact, or tuple, is a conjunction of constraints over a small number of variables. We describe the basic Constraint Query Language design principles and illustrate them with four classes of constraints: real polynomial inequalities, dense linear order inequalities, equalities over an infinite domain, and boolean equalities. For the analysis, we use quantifier elimination techniques from logic and the concept of data complexity from database theory. This framework is applicable to managing spatial data and can be combined with existing multidimensional searching algorithms and data structures.

. The declarative programming paradigms used in constraint languages can lead to powerful extensions of Codd's relational data model. The development of constraint database query languages from logical database query languages has many similarities with the development of constraint logic programming from logic programming, but with the additional requirements of data efficient, set-at-a-time, and bottomup evaluation. In this overview of constraint query languages (CQLs) we first present the framework of [41]. The principal idea is that: "the k-tuple (or record) data type can be generalized by a conjunction of quantifier-free constraints over k variables". The generalization must preserve various language properties of the relational data model, e.g., the calculus/algebra equivalence, and have time complexity polynomial in the size of the data. We next present an algebra for dense order constraints that is simpler to evaluate than the calculus described in [41], and we sharpen some of...

The set-height of a complex object type is defined to be its level of nesting of the set construct. In a query of the complex object calculus which maps a database D to an output type T,anintermediate type is a type which is used by some variable of the query, but which is not present in D or T.Foreachk, i ≥ 0 we define CALCk,i to be the family of calculus queries mapping from and to types with set-height ≤ k and using intermediate types with set-height ≤ i. In particular, CALC0,0 is the classical relational calculus, and CALC0,1 is equivalent to the family of secondorder (relational) queries. Several results concerning these families of languages are obtained. A primary focus is on the families CALC0,i, which map relations to relations. Upper and lower bounds in terms of hyper-exponential time and space on the complexity of these families are provided. The CALC0,i hierarchy does not collapse with respect to expressive power. The union ∪0≤iCALC0,i is exactly the family of elementary queries, i.e., queries with hyper-exponential complexity. The expressive power of queries from the complex object calculus interpreted using semantics based on the use of arbitrarily large finite or infinite set of invented values is studied. Under these semantics, the expressive power of the relational calculus is not increased, and the CALC0,i hierarchy collapses at CALC0,1. In general, queries with these semantics may not be computable. We also consider an alternative semantics which yields a family of queries equivalent to the computable queries. 1

We investigate properties of finite relational structures over the reals expressed by first-order sentences whose predicates are the relations of the structure plus arbitrary polynomial inequalities, and whose quantifiers can range over the whole set of reals. In constraint programming terminology, this corresponds to Boolean real polynomial constraint queries on finite structures. The fact that quantifiers range over all reals seems crucial; however, we observe that each sentence in the first-order theory of the reals can be evaluated by letting each quantifier range over only a finite set of real numbers without changing its truth value. Inspired by this observation, we then show that when all polynomials used are linear, each query can be expressed uniformly on all finite structures by a sentence of which the quantifiers range only over the finite domain of the structure. In other words, linear constraint programming on finite structures can be reduced to ordinary query evaluation a...

: We provide a generalization of Datalog based on generalizing databases by adding integer order constraints to relational tuples. For Datalog queries with integer (gap)-order constraints (denoted Datalog !Z ) we show that there is a closed form evaluation. We also show that the tuple recognition problem can be done in PTIME in the size of the generalized database, assuming that the size of the constants in the query is logarithmic in the size of the database. Note that the absence of negation is critical, Datalog : queries with integer order constraints can express any Turing computable function. 1 Introduction In this paper we consider a generalization of Datalog based on the notion of a constraint tuple. The important idea of a constraint tuple comes from constraint logic programming systems, e.g. CLP [14], Prolog III [4], and CHIP [8], and it generalizes the notion of a ground fact. This allows the declarative programming of new applications, including various combinatorial se...

ing with credit is permitted. To copy otherwise, to republish, to post on servers, or to redistribute to lists, requires prior specific permission and/or a fee. Request permissions from Publications Dept, ACM Inc., fax +1 (212) 869-0481, or permissions@acm.org. Safe Constraint Queries Michael Benedikt Bell Laboratories 1000 E Warrenville Rd Naperville, IL 60566 E-mail: benedikt@research.bell-labs.com Leonid Libkin Bell Laboratories 600 Mountain Avenue Murray Hill, NJ 07974 E-mail: libkin@research.bell-labs.com Abstract We extend some of the classical characterization theorems of the relational theory --- particularly those related to query safety --- to the context where database elements come with fixed interpreted structure, and where formulae over elements of that structure can be used in queries. We show that the addition of common interpreted functions such as real addition and multiplication to the relational calculus preserves important characterization theorems ...

. Constraint databases generalize relational databases by finitely representable infinite relations. This paper surveys the state of the art in constraint databases: known results, remaining open problems and current research directions. The paper also describes a new algebra for databases with integer order constraints and a complexity analysis of evaluating queries in this algebra. In memory of Paris C. Kanellakis 1 Introduction There is a growing interest in recent years among database researchers in constraint databases, which are a generalization of relational databases by finitely representable infinite relations. Constraint databases are parametrized by the type of constraint domains and constraint used. The good news is that for many parameters constraint databases leave intact most of the fundamental assumptions of the relational database framework proposed by Codd. In particular, 1. Constraint databases can be queried by constraint query languages that (a) have a semantics ba...

We rework parts of the classical relational theory when the underlying domain is a structure with some interpreted operations that can be used in queries. We identify parts of the classical theory that go through `as before' when interpreted structure is present, parts that go through only for classes of nicely-behaved structures, and parts that only arise in the interpreted case. The first category includes a number of results on language equivalence and expressive power characterizations for the active-domain semantics for a variety of logics. Under this semantics, quantifiers range over elements of a relational database. The main kind of results we prove here are generic collapse results: for generic queries, adding operations beyond order, does not give us extra power. The second category includes results on the natural semantics, under which quantifiers range over the entire interpreted structure. We prove, for a variety of structures, natural-active collapse results, s...

. Constraint query languages are natural extensions of relational database query languages. A framework for their declarative specification (constraint calculi) and efficient implementation (low data complexity and secondary storage indexing) was presented in Kanellakis et al., 1995. Constraint query algebras form a procedural language layer between high-level declarative calculi and low-level indexing methods. Just like the relational algebra, this intermediate layer can be very useful for program optimization. In this paper, we study properties of constraint query algebras, which we present through three concrete examples. The dense order constraint algebra illustrates how the appropriate canonical form can simplify expensive operations, such as projection, and facilitate interaction with updates. The monotone two-variable linear constraint algebra illustrates the concept of strongly polynomial operations. Finally, the lazy evaluation of (non)linear constraint algebras illustrates ho...

Several alternative semantics (or interpretations) of the relational (domain) calculus are studied here. It is shown that they all have the same expressive power, i.e., the selection of any of the semantics neither gains nor loses expressive power. Since the domain is potentially infinite, the answer to a relational calculus query is sometimes infinite (and hence not a relation). The following approaches which guarantee the finiteness of answers to queries are studied here: output-restricted unlimited interpretation, domain independent queries, output-restricted finite and countable invention, and limited interpretation. Of particular interest is the output-restricted unlimited interpretation -- although the output is restricted to the active domain of the input and query, the quantified variables range over the infinite underlying domain. While this is close to the intuitive interpretation given to calculus formulas, the naive approach to evaluating queries under this semantics calls ...