A more exotic non-SQL way of querying databases is the Z39.50 protocol, which is used by the Library of Congress. My memory is a little fuzzy since it’s years since I worked with this, but Z39.50 is a binary protocol, meaning that it’s the client’s responsibility to encode queries into binary structures before sending them to the server.

The syntax of the queries is a little bizarre and non-intuitive. Back when I was writing a client program to query the LC database, I remember writing my parser to translate user queries into “reverse polish notation” (RPN) which is a query syntax where the operators come at the end of the clause. Thus, for example, in the Z39.50 version of RPN, a predicate like

WHERE name CONTAINS ’smith’ AND subject CONTAINS ‘fruit’

would look like this:

smith/1 fruit/21 AND

…1 being the agreed-upon marker for “name” and 21 being the marker for “subject”.

However, I revisited the LC website and now they are accepting queries in something called CQL or Contextual Query Language: http://www.loc.gov/standards/sru/specs/cql.html

You can read more about the wonders of Z39.50 on Wikipedia: http://en.wikipedia.org/wiki/Z39.50

I’m working on a project and am doing some research thanks