Metal-organic frameworks (MOFs) are nanoporous crystals which are widely used as selective adsorbents, separation membranes, catalysts, gas and energy storage media, and drug delivery vehicles. The unique adsorption and transport properties of MOFs are determined by their complex three-dimensional (3D) networks of pores, cages, and channels that differ in size, shape, and chemical composition. While the morphological structure of MOF crystals is known, practical MOF materials are rarely ideal crystals. They contain secondary phases, binders, residual chemicals, and various types of defects. It is of paramount importance to evaluate the degree of crystallinity and accessibility of different pore compartments to adsorb guest molecules. To this end, we recently suggested the method of fingerprint isotherms based on the comparison of the experimentally measured adsorption isotherms and theoretical isotherms on ideal MOF crystals produced by Monte Carlo (MC) simulations and decomposed with respect to different pore compartments [Parashar, S. ACS Appl. Nano Mater. 2021, 4, 5531-5540 and Dantas, S.; Neimark, A. V. ACS Appl. Mater. Interfaces 2020, 12, 15595-15605]. In this work, we develop an automated algorithm for pore network compartmentalization that is a prerequisite for calculations of the fingerprint isotherms. The proposed algorithm partitions the unit cell into realistically shaped compartments based on the geometric pore size distribution. The proposed method is demonstrated on several characteristic systems, including Cu-BTC, IRMOF-1, UiO-66, PCN-224, ZIF-412, and 56 structures from the CoRE MOF database.