def extractJsFromHtmlDir(inputdir, outputdir): if not utils.checkdir(inputdir, outputdir): return infiles = utils.lsresult(inputdir) for infile in infiles: outfile = outputdir + '/' + infile infile = inputdir + infile if utils.checkfile(infile, outfile): extractJsFromHtml(infile, outfile)
def deobFiles(self, regex, order = -1): # deobfuscate files that match regex and generate output to ../../data/JSAdd2Tree files = utils.lsresult(regex) num_files = len(files) if order == -1: for i in range(num_files-1, -1, -1): print 'Processing file: ' + files[i] self.deobFile(files[i]) elif order == 1: for i in range(0, num_files): print 'Processing file: ' + files[i] self.deobFile(files[i])
def jsdeobfuscatedir(inputdir, outputdir): if os.path.exists(outputdir): if os.path.isfile(outputdir): print 'The out dir is conflict with other filename' return shutil.rmtree(outputdir) os.makedirs(outputdir) # if not utils.checkdir(inputdir, outputdir): # return infiles = utils.lsresult(inputdir) for infile in infiles: outfile = outputdir + '/' + infile infile = inputdir + '/' + infile if utils.checkfile(infile, outfile): jsdeobfuscatefile(infile, outfile)
def getFilenameForClass(self, classname, count):
# When selecting the benign examples, reInit
if classname == 'B1':
self.reInit()
# Randomly select count urls
self.groundTruthPath = '../../data/JSAdd2Tree/'
filenames = utils.lsresult("-d {0}/{1}.*".format(self.groundTruthPath, classname))
random.shuffle(filenames)
# Handle invalid urls here
progress = 0
result = list()
for filename in filenames:
if progress >= count:
return result
if self.isValid(filename):
result.append(filename)
progress = progress + 1
return result
def postReqFiles(self, inregex, indir, outdir):
infiles = utils.lsresult(inregex)
infiles = [infile.split('/')[-1] for infile in infiles]
for infile in infiles:
print "Processing {0}".format(infile)
self.postReqFile(indir+infile, outdir+infile)
def rmCommentDir(indir, outdir):
infiles = utils.lsresult(indir)
for infile in infiles:
outfile = outdir + infile
infile = indir + infile
rmCommentFile(infile, outfile)
