A common data manipulation language for nested data in heterogeneous environments

Capability-based localization of distributed and heterogeneous queries

“…We improve and refine the process presented in (Seco et al, 2015) and present a query transformation process consisting of three separate and orthogonal steps. The first phase corresponds to the pruning of the query expression based on the result usage type.…”

“…The initial labelled expression is therefore (db (t, e) ), where corresponds to the location where the computation of arguments should be performed, and location corresponds only to the site where data is located and from where data transmission occurs. In this intermediate form, we also introduce a specialized and flexible labelling scheme for expression binders in foreach expressions (following the earlier approach of Seco et al, 2015). This includes a partition of binders by location, and an association of Boolean conditions -parts of the original condition in the conjunctive form -to each one of the binder partitions.…”

“…This paper extends and refines the approach presented in Seco et al (2015). Our construction and query combination model is designed from first principles, targeting a general model of data sources, from relational data to nested collections (e.g., Colby, 1989;Cheney et al, 2014).…”

“…Our construction and query combination model is designed from first principles, targeting a general model of data sources, from relational data to nested collections (e.g., Colby, 1989;Cheney et al, 2014). We explore a novel language operation, introduced in Seco et al (2015), whose semantics is the in-place modification of nested data, given a tree-like path (cf. XPath; Clark & DeRose, 1999;Cheney et al, 2014).…”

“…We refine the query transformation process presented in Seco et al (2015) by decomposing it into three separate phases, and analysing the contribution of each step in a precise and separate way. We also add extra formal results and proofs.…”