Overcoming 64kb data size limit in handling large spatial data in GISNIC while cleaning and building topology

Abstract

When asked if topology is a key concept of Geographic Information Systems (GIS), most GIS users will nod their heads in agreement. But when asked how topology is handled in GIS the nodding heads give way to shrugging shoulders. Why should GIS users care about topology? What are the advantages and disadvantages of storing polygon (topology) data? This paper attempts to analyse the algorithm used in GISNIC, an in-house developed software, to clean the spatial data and then build topology from line (segment) data and associate this topology with point data. Suppose there is a structure of 32 bytes size, and it is required to store, say, 11,000 data of that structure. It is obvious by simple arithmetic that we need nearly 333KB of memory by dynamic allocation. In a 16-bit application, only power (2, 16) memory units (byte in byte addressable memory system) can be addressed. Thus, simple allocation of memory for 11,000 data units of the said size cannot work in this case. This paper discusses the method to overcome this constraint. The problem can be solved by giving precedence to the first index and then to the secondary index. For instance, to address the location where 3335th item could be residing the following should be done: LPCNT = int(3335/2000) REMAINDER=3335 LPCNT * 2000. Then variable, say region, can be accessed as: region[LPCNT])[REMAINDER].