VulTrigger

compile_and_clang.py

import os
import glob
from pydriller import Repository, Git
from unidiff import PatchSet
import subprocess

REPOS = {"Augeas": "https://github.com/hercules-team/augeas", "BitlBee": "https://github.com/bitlbee/bitlbee", "Cantata": "https://github.com/CDrummond/cantata", "ImageMagick": "https://github.com/ImageMagick/ImageMagick"
         ,"JPEGSnoop": "https://github.com/ImpulseAdventure/JPEGsnoop", "JasPer": "https://github.com/jasper-software/jasper", "LibRaw": "https://github.com/LibRaw/LibRaw", "PHP": "https://github.com/php/php-src", "WavPack": "https://github.com/dbry/WavPack",
         "Ceph": "https://github.com/ceph/ceph", "FFmpeg": "https://github.com/FFmpeg/FFmpeg", "file":"https://github.com/file/file", "freetype2":"https://github.com/aseprite/freetype2", "libgd": "https://github.com/libgd/libgd", "libpng": "https://github.com/glennrp/libpng",
         "libsndfile": "https://github.com/libsndfile/libsndfile", "libtiff":"https://github.com/libsdl-org/libtiff", "libxml2": "https://github.com/GNOME/libxml2", "lua": "https://github.com/lua/lua", "miniupnp": "https://github.com/miniupnp/miniupnp", "mupdf": "https://github.com/ArtifexSoftware/mupdf", "openjpeg": "https://github.com/uclouvain/openjpeg", "openssl": "https://github.com/openssl/openssl", "qemu": "https://github.com/qemu/qemu", "radare2": "https://github.com/radareorg/radare2", "sqlite3": "https://github.com/sqlite/sqlite", "tcpdump": "https://github.com/the-tcpdump-group/tcpdump"}



def git_clone_coverity(github_url, git_hash, cve):
    repo_path = "data/tmp/" + github_url.split("/")[-1]
    if not os.path.exists(repo_path):
        os.system("git clone {} {}".format(github_url, repo_path))
    build = "/Applications/cov-analysis-macosx-2022.12.1/bin/cov-run-desktop --disconnected --build --auth-key-file /Applications/cov-analysis-macosx-2022.12.1/bin/auth-key.txt make"
    command = "/Applications/cov-analysis-macosx-2022.12.1/bin/cov-run-desktop --disconnected --auth-key-file /Applications/cov-analysis-macosx-2022.12.1/bin/auth-key.txt --analyze-captured-source"
    os.chdir(repo_path)
    os.system("git checkout {}^".format(git_hash))
    os.system("autoreconf -i && ./configure && make clean && {} || true && {} || true >> {}.txt 2>&1".format(build, command, cve))
    os.chdir("/Users/USER/Projects/VulTrigger/dataset")

def git_clone(github_url, git_hash, cve):
    repo_path = "data/tmp/" + github_url.split("/")[-1]
    if not os.path.exists(repo_path):
        os.system("git clone {} {}".format(github_url, repo_path))
    os.chdir(repo_path)
    os.system("git checkout {}^".format(git_hash))
    os.system("autoreconf -i && ./configure &&  scan-build make >> {}.txt 2>&1 || true".format(cve))
    os.chdir("/Users/USER/Projects/VulTrigger/dataset")

for project in glob.glob("DIFF_NEW_OLD/*"):
    splitted = project.split("/")[1]
    if splitted in REPOS:
        github_url = REPOS[splitted]
        print(github_url, splitted, project)
        for cve in glob.glob(project + "/*/"):
            cve_name = cve.split("/")[-2].replace("/", "")
            print("cve", cve_name)
            if not os.path.exists("data/{}.txt".format(cve_name)):
                continue
            for diff in glob.glob(cve + "/*.diff"):
                git_hash = diff.split("/")[-1].split("_")[2]
                print(git_hash)
                try:
                    patchset = PatchSet(open(diff, "r").read())
                except:
                    continue
                files = []
                for patch in patchset.modified_files:
                    files.append("/".join(patch.source_file.split("/")[1:]))
                    git_clone(github_url, git_hash, cve)