An approach to fuzz testing data generation for applied and system software

The problems related to input data generation for fuzz testing are discussed in the article. Existing fuzzers mainly realize the generators depending on target software. The creation of more universal generators requires data structures descriptive tools, which would have more expressive description possibilities for data structures internal relations.