scoredb | simple database optimized for returning results | Database library

As shown above, queries are expressed as JSON expressions and then url encoded into the "score" query parameter. Each expression takes a lisp-like form: [<function name>, <argument 1>, <argument 2>, ...]. These are the supported functions:. Simply produces the value of <field_name> as a score. Takes the result of <subexpression> and multiplies it by <factor>. <factor> may be negative. Sums the results of each <subexpression>. Multiplies the result of each <subexpression> together. For bounding reasons, negative inputs are not allowed. Takes the least score resulting from all <subexpression>s. ####["diff", <subexpression 1>, <subexpression 2>] Returns the absolute difference between the values produced by both subexpressions. Raises the result from the given subexpression to the <exponent> power. <exponent> may be fractional (for Nth roots) or negative. However, for bounding reasons, the subexpression may not produce negative values. Establishes a user-defined function using a set of linearly interpolated [x, y] points. Inputs smaller than the smallest X value or larger than the largest X value get the closest specified Y value. Returns the distance to a fixed point in kilometers as a score. This is experimental: may be inaccurate for large distances, and fails across the prime meridian. Since you typically want smaller distances to have higher scores, you'll probably want to wrap the "scale" or "custom_linear" functions around this one to invert it.