Recursive backtracking is a general algorithm for finding solutions to computational problems that incrementally builds partial solutions and drops those which are identified as futile *as early as possible*.
Recursive backtracking is a general algorithm for finding solutions to computational problems that incrementally builds partial solutions and drops as early as possible those partial solutions which are identified as futile i.e. such that can not lead to a correct solution.
It essentially involves building nested loops through recursion. Each loop embodies a series of choices. Each inner loop can accept a partial-solution-so-far and try its own choices in a loop, or reject the partial solution it is presented with, through early exit or an equivalent of the continue or break etc. statement.
The control is inverted in that the full solution is achieved in the innermost loop, which can act on it in a particular manner, like printing it to the output / external file, or calling a callback with the solution as an argument.
The backtracking computational structure is embodied in the states of the nested loops, in each of the nested loops' loop variables.
