BackDoorBuster/crawler/transfer.py

"""转换原有的漏洞文件格式"""

import re
from packaging.specifiers import SpecifierSet


def load_vulnerable_packages(filename):
    """从文件加载有漏洞的包信息"""
    with open(filename, "r", encoding="utf-8") as file:
        content = file.read()
    vulnerabilities = {}
    blocks = content.split("--------------------------------------------------")
    range_pattern = re.compile(r"\[(.*?),\s*(.*?)\)")

    for block in blocks:
        name_match = re.search(r"Package Name: (.+)", block)
        if name_match:
            package_name = name_match.group(1).strip()
            ranges = range_pattern.findall(block)
            specifier_list = []
            for start, end in ranges:
                if start and end:
                    specifier_list.append(f">={start},<{end}")
                elif start:
                    specifier_list.append(f">={start}")
                elif end:
                    specifier_list.append(f"<{end}")
            if specifier_list:
                vulnerabilities[package_name] = SpecifierSet(",".join(specifier_list))
    return vulnerabilities


def save_vulnerabilities_to_file(vuln_packages, filename):
    """将漏洞信息写入到文件中"""
    with open(filename, "w", encoding="utf-8") as file:
        for package, specifiers in vuln_packages.items():
            file.write(f"Package Name: {package}\n")
            file.write(f"Version Ranges: {specifiers}\n")
            file.write("-" * 50 + "\n")


def main():
    vulnerabilities = load_vulnerable_packages("extracted_data.txt")
    save_vulnerabilities_to_file(vulnerabilities, "trans_extracted_data.txt")


if __name__ == "__main__":
    main()