Type inference, principal typings, and let-polymorphism for first-class mixin modules
Proceedings of the tenth ACM SIGPLAN international conference on Functional programming
A crash course on database queries
Proceedings of the twenty-sixth ACM SIGMOD-SIGACT-SIGART symposium on Principles of database systems
Polymorphic type inference for the named nested relational calculus
ACM Transactions on Computational Logic (TOCL)
Proceedings of the 1st International Workshop on Workflow Approaches to New Data-centric Science
| Hi-index | 0.00 |
We investigate the complexity of the typability problem for the relational algebra. This problem consists of deciding, for a given relational algebra expression, whether there exists an assignment of types to variables occurring in the expression such that the expression is well-typed under the assignment. We obtain that the problem is NP-complete in general. In particular, we show that the problem becomes NP-hard due to (1) the cartesian product operator, (2) the selection operator on arbitrary sets of typed predicates, (3) the selection operator on “well-behaved” sets of typed predicates together with join and projection or renaming. However, the problem is in P when (1) we only allow union, difference, join and selection on “well-behaved” sets of typed predicates, or (2) we allow all operators except cartesian product, where the set of selection predicates can mention at most one base type. Most of these results follow from a close connection of the typability problem to non-uniform constraint satisfaction.