Hey folks, So, I've been on a couple of projects now that I've used complex-objects (nee hash/array trees) on Nosql dbs, and have been coding as-FP-as-I-can-get with the processing of the tree, for instance when calculating aggregations against many docs or verifying input. In general my tactic has been to put a new ruby-Class around each "level" of the object tree. So, for instance, in a People object, the :phone field may be a sub-tree and contained in the ruby Phone class.
"Get around to the question already!" -- indeed. I find myself when coding the sub-tree objects, that I often end up referencing "up-tree", especially around any position-order indexing or field-unique constraints. Question : Is there a way to avoid feeding the "up-tree" into the sub-tree object/ruby-Class when manipulating the object in question? To be salient to the example above... if for some reason phone numbers are supposed to be unique across People... how does the Phone class go about that? Is the separation of concerns wrong here? I appreciate any thougths and suggestions. For the polyglots in the room : I have encountered this same issue in Python and JS, so language dependence is not necessarily a constraint in your answer. -- -- SD Ruby mailing list [email protected] http://groups.google.com/group/sdruby --- You received this message because you are subscribed to the Google Groups "SD Ruby" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
