Re: Academic name for associative array when used to pair column names with data

On Thursday, March 20, 2014 9:00:30 AM UTC-7, william.d..._at_gmail.com wrote:

> I'm thinking about this more generally than relational databases in the vein of Codd. It seems like there is a similar concept of pairing column names with rows of data in other means of storing or transferring data.

> Is there a generic name for a hash when it contains the pairings of column names with the values from a given row regardless of what the underlying storage mechanism is?

A perl hash already is that. My answer wasn't limited to a relational or database context. Other common row terms are "record", "struct", "dictionary", "map" or "associative array". https://en.wikipedia.org/wiki/Associative_array

It's not clear whether you understood my comment about overall abstraction. The common operator name doesn't matter. What is an abstract description of handled things? A sequence of a sequence of key-value pairs? A set of a set of key-value pairs? Do all rows have the same key set? Is such an abstract thing also what an input source denotes? (Apparently, from your more recent examples a CSV file comes with a header and JSON string also come with keys; what about arrays?) Notice that you've only mentioned parts (ie pairs and bunches of pairs) of the actual handled abstraction (ie (iterator for) bunches of bunches of pairs). And even though you are asking for a term for an unordered row, is that what your handled things have?

philip Received on Sun Mar 30 2014 - 00:27:07 CET