peComplexityAnalyzer.py is a Python script designed to estimate the cyclomatic complexity of PE (Portable Executable) files, commonly used in Windows environments.

peComplexityAnalyzer.py

__author__ = 'Matthew Schwartz'
'''
This script utilizes pefile for parsing PE file structures and capstone for disassembling the binary code. The primary focus is on accurately extracting and analyzing the assembly code from executable sections of the PE file to calculate cyclomatic complexity, a metric that provides insight into the code's complexity and potential maintainability issues.
'''

import sys
import pefile
import capstone

def disassemble_pe(input_file):
    try:
        pe = pefile.PE(input_file)
        if pe.FILE_HEADER.Machine == pefile.MACHINE_TYPE['IMAGE_FILE_MACHINE_I386']:
            md = capstone.Cs(capstone.CS_ARCH_X86, capstone.CS_MODE_32)
        elif pe.FILE_HEADER.Machine == pefile.MACHINE_TYPE['IMAGE_FILE_MACHINE_AMD64']:
            md = capstone.Cs(capstone.CS_ARCH_X86, capstone.CS_MODE_64)
        else:
            raise Exception("Unsupported architecture")

        instructions = []

        for section in pe.sections:
            if section.IMAGE_SCN_MEM_EXECUTE:
                base_addr = section.VirtualAddress
                code = section.get_data()
                disassembly = md.disasm(bytes(code), base_addr)

                for insn in disassembly:
                    instructions.append(f"{insn.mnemonic} {insn.op_str}")

        return instructions

    except Exception as e:
        print(f"Error during disassembly: {e}")
        return []

def estimate_cyclomatic_complexity(assembly_instructions):
    conditional_jumps = [
        'ja', 'jae', 'jb', 'jbe', 'jc', 'jcxz', 'jecxz',
        'jrcxz', 'je', 'jg', 'jge', 'jl', 'jle', 'jna',
        'jnae', 'jnb', 'jnbe', 'jnc', 'jne', 'jng', 'jnge',
        'jnl', 'jnle', 'jno', 'jnp', 'jns', 'jnz', 'jo', 
        'jp', 'jpe', 'jpo', 'js', 'jz'
    ]

    count = sum(1 for instruction in assembly_instructions if instruction.split()[0] in conditional_jumps)
    return count + 1  # Adding 1 to account for the initial path

def main():
    if len(sys.argv) != 2:
        print("Usage: python script.py <binary_file>")
        sys.exit(1)

    binary_file = sys.argv[1]
    assembly_instructions = disassemble_pe(binary_file)

    if assembly_instructions:
        complexity = estimate_cyclomatic_complexity(assembly_instructions)
        print("Cyclomatic Complexity:", complexity)
    else:
        print("No executable sections found or unable to disassemble.")

if __name__ == "__main__":
    main()