For the first time, a nonlinear Schrödinger equation of the general form is considered, depending on time and two spatial variables, the potential and dispersion of which are specified by two arbitrary functions. This equation naturally generalizes a number of simpler nonlinear partial differential equations encountered in various fields of theoretical physics, including nonlinear optics, superconductivity, and plasma physics. Two- and one-dimensional reductions are described, which reduce the studied nonlinear Schrödinger equation to simpler equations of lower dimension or ordinary differential equations (or systems of ordinary differential equations). In addition to the general Schrödinger equation with two arbitrary functions, related nonlinear partial differential equations are also examined, in which the dispersion function is specified arbitrarily while the potential function is expressed in terms of it. For all considered classes of nonlinear PDEs, using the methods of generalized and functional separation of variables, as well as the semi-inverse approach and the principle of structural analogy of solutions, many new exact solutions have been found, which are expressed in terms of elementary or special functions, or in the form of quadratures. Both Cartesian and polar coordinate systems are employed to analyze the equations under consideration. Special attention is paid to finding solutions with radial symmetry. It is shown that the nonlinear Schrödinger equation, in which the functions defining the potential and dispersion are linearly related (one of these functions can be chosen arbitrarily), can be reduced to a two-dimensional nonlinear PDE that admits exact linearization. The exact solutions obtained in this work can be used as test problems intended for verifying the adequacy and assessing the accuracy of numerical and approximate analytical methods for solving complex nonlinear PDEs of mathematical physics.
