I've just had a chance to review LINQ, DLinq, and XLinq, (which I only heard
about last week after the PDC).
The various LINQs actually seem to live up to expectations - Using query
semantics to retrieve a subset of data from arbitrary data-architectures
(though I hope the sub-LINQ distinctions can be held coherent over time by
the same governing architectureess).
I summized pre-familiarily that we would accomplish LINQ support by adapting
attributes, interfaces, or base classes into our datastructures that we wish
to be queried upon - LINQ allowing the general programmer to enable query
support on custom data contraptions (e.g. unforeseen data architectures
foldable into the LINQ query model).
*?* How applicable is this prenotion to the current system *?*
If you need an example, let's say an arbitrary data graph where I want to
query based on connectedness (first neighbor, second neighbor, etc.), or by
shortest-path... Could I define a few utility functions and use their
invocation in a LINQ query, to, let's say, select all nodes reachable in 5
hops from an indicated one?
<side-note>
The video on XLinq showed an apparently different model... using specialized
functions on a specific XML data-architecture. Fortunately, the XLinq
Overview document suggested that those functions might just be a part of the
internal LINQ support, and that the now beloved LINQ semantic model is
coherent in the context of XLinq, thus salvaging that subtechnology.
</side-note>
|
Similar Threads
