VulntoES.py

#!/usr/bin/env python

from datetime import datetime
from elasticsearch import Elasticsearch
import json
import time
import codecs
import struct
import locale
import glob
import sys
import getopt
import xml.etree.ElementTree as xml
import re
#import socket
#import pprint



class NessusES:
	"This class will parse an Nessus v2 XML file and send it to Elasticsearch"

	def __init__(self, input_file,es_ip,es_port,index_name, static_fields):
		self.input_file = input_file
		self.tree = self.__importXML()
		self.root = self.tree.getroot()
		self.es = Elasticsearch([{'host':es_ip,'port':es_port}])
		self.index_name = index_name
                self.static_fields = static_fields
                vulnmapping = { "properties": {
                            "pluginName": { "type": "string", "fields": {
                                "raw": { "type": "string", "index": "not_analyzed" } } },
                            "ip": { "type": "string", "fields": {
                                "raw": { "type": "string", "index": "not_analyzed" } } },
                            "risk_factor": { "type": "string", "fields": {
                                "raw": { "type": "string", "index": "not_analyzed" } } },
                            "severity": { "type": "integer" },
                            "port": { "type": "integer" },
                            "pluginFamily": { "type": "string", "fields": {
                                "raw": { "type": "string", "index": "not_analyzed" } } },
                            "plugin_type": { "type": "string", "fields": {
                                "raw": { "type": "string", "index": "not_analyzed" } } },
                            "svc_name": { "type": "string", "fields": {
                                "raw": { "type": "string", "index": "not_analyzed" } } },
                            "svcid": { "type": "string", "fields": {
                                "raw": { "type": "string", "index": "not_analyzed" } } },
                            "synopsis": { "type": "string", "fields": {
                                "raw": { "type": "string", "index": "not_analyzed" } } },
                            "solution": { "type": "string", "fields": {
                                "raw": { "type": "string", "index": "not_analyzed" } } },
                            } }
                mappings = { "mappings": { "vuln": vulnmapping } }
                try:    # try to create index
                    self.es.indices.create(index=index_name, body=json.dumps(mappings))
                except: # if index exists: ensure mapping is created
                    self.es.indices.put_mapping(index=index_name, doc_type="vuln", body=json.dumps(vulnmapping))

	def displayInputFileName(self):
		print self.input_file

	def __importXML(self):
		#Parse XML directly from the file path
		return xml.parse(self.input_file)

	def toES(self):
		"Returns a dict of dictionaries for each issue in the report"
		#Nessus root node only has 2 children. policy and report, we grab report here
		report = self.root.getchildren()[1]
		dict_item={}
		#each child node of report is a report host - rh
		for rh in report:
			ip = rh.attrib['name']
			host_item={}
			#print rh.tag
			#iterate through attributes of ReportHost tags
			for tag in rh.getchildren():
				dict_item={}
				if tag.tag == 'HostProperties':
					for child in tag.getchildren():
						if child.attrib['name'] == 'HOST_END':
							host_item['time'] = child.text
							host_item['time'] = datetime.strptime(host_item['time'], '%a %b %d %H:%M:%S %Y')
							host_item['time'] = datetime.strftime(host_item['time'], '%Y/%m/%d %H:%M:%S')
						if child.attrib['name'] == 'operating-system':
							host_item['operating-system'] = child.text
						if child.attrib['name'] == 'mac-address':
							host_item['mac-address'] = child.text
						if child.attrib['name'] == 'host-fqdn':
							host_item['fqdn'] = child.text
						host_item['ip'] = ip
				elif tag.tag == 'ReportItem':
					dict_item['scanner'] = 'nessus'
					if tag.attrib['port']:
						dict_item['port'] = int(tag.attrib['port'])
					if tag.attrib['svc_name']:
						dict_item['svc_name'] = tag.attrib['svc_name']
					if tag.attrib['protocol']:
						dict_item['protocol'] = tag.attrib['protocol']
					if tag.attrib['severity']:
						dict_item['severity'] = tag.attrib['severity']
					if tag.attrib['pluginID']:
						dict_item['pluginID'] = tag.attrib['pluginID']
					if tag.attrib['pluginName']:
						dict_item['pluginName'] = tag.attrib['pluginName']
					if tag.attrib['pluginFamily']:
						dict_item['pluginFamily'] = tag.attrib['pluginFamily']
					#Iterate through child tags and texts of ReportItems
					#These are necessary because there can be multiple of these tags
					dict_item['cve'] = []
					dict_item['bid'] = []
					dict_item['xref'] = []
					for child in tag.getchildren():
						#print child.tag
						if child.tag == 'solution':
							dict_item[child.tag] = child.text
						if child.tag == 'risk_factor':
							dict_item[child.tag] = child.text
						if child.tag == 'description':
							dict_item[child.tag] = child.text
						if child.tag == 'synopsis':
							dict_item[child.tag] = child.text
						if child.tag == 'plugin_output':
							dict_item[child.tag] = child.text
						if child.tag == 'plugin_version':
							dict_item[child.tag] = child.text
						if child.tag == 'see_also':
							dict_item[child.tag] = child.text
						if child.tag == 'xref':
							dict_item[child.tag].append(child.text)
						if child.tag == 'bid':
							dict_item[child.tag].append(child.text)
						if child.tag == 'cve':
							dict_item[child.tag].append(child.text)
						if child.tag == 'cvss_base_score':
							dict_item[child.tag] = float(child.text)
						if child.tag == 'cvss_temporal_score':
							dict_item[child.tag] = float(child.text)
						if child.tag == 'cvss_vector':
							dict_item[child.tag] = child.text
						if child.tag == 'exploit_available':
							if child.text == 'true':
								dict_item[child.tag] = 1
							else:
								dict_item[child.tag] = 0
						if child.tag == 'plugin_modification_date':
							dict_item[child.tag] = child.text
						if child.tag == 'plugin_type':
							dict_item[child.tag] = child.text
                                try:
                                    ip = host_item['ip']
                                except KeyError:
                                    ip = "-"

                                try:
                                    protocol = dict_item['protocol']
                                except KeyError:
                                    protocol = "-"

                                try:
                                    port = dict_item['port']
                                except KeyError:
                                    port = 0
                                host_item['svcid'] = "%s/%s/%d" % (ip, protocol, port)

                                for name in self.static_fields:
                                    dict_item[name] = self.static_fields[name]

				self.es.index(index=self.index_name,doc_type="vuln", body=json.dumps(dict(host_item.items()+dict_item.items())))


class NmapES:
	"This class will parse an Nmap XML file and send data to Elasticsearch"

	def __init__(self, input_file,es_ip,es_port,index_name):
		self.input_file = input_file
		self.tree = self.__importXML()
		self.root = self.tree.getroot()
		self.es = Elasticsearch([{'host':es_ip,'port':es_port}])
		self.index_name = index_name

	def displayInputFileName(self):
		print self.input_file

	def __importXML(self):
		#Parse XML directly from the file path
		return xml.parse(self.input_file)

	def toES(self):
		"Returns a list of dictionaries (only for open ports) for each host in the report"
		for h in self.root.iter('host'):
			dict_item = {}
			dict_item['scanner'] = 'nmap'
			if h.tag == 'host':
				if h.attrib['endtime']:
					dict_item['time'] = time.strftime('%Y/%m/%d %H:%M:%S', time.gmtime(float(h.attrib['endtime'])))
			
			for c in h:
				if c.tag == 'address':
					if c.attrib['addr']:
						dict_item['ip'] = c.attrib['addr']
				elif c.tag == 'hostnames':
					for names in c.getchildren():
						if names.attrib['name']:
							dict_item['hostname'] = names.attrib['name']
				elif c.tag == 'ports':
					for port in c.getchildren():
						dict_itemb = {}
						if port.tag == 'port':
							dict_item['port'] = port.attrib['portid']
							dict_item['protocol'] = port.attrib['protocol']
							for p in port.getchildren():
								if p.tag == 'state':
									dict_item['state'] = p.attrib['state']
								elif p.tag == 'service':
									dict_item['service'] = p.attrib['name']
                                                                        if p.attrib['banner']:
                                                                            dict_item['banner'] = p.attrib['banner']
								elif p.tag == 'script':
									if p.attrib['id']:
										if p.attrib['output']:
											dict_item[p.attrib['id']] = p.attrib['output']									
							if dict_item['state'] == 'open':
								#Only sends document to ES if the port is open
								self.es.index(index=self.index_name,doc_type="vuln", body=json.dumps(dict_item))

class NiktoES:
	"This class will parse an Nikto XML file and create an object"

	def __init__(self, input_file,es_ip,es_port,index_name):
		self.input_file = input_file
		self.tree = self.__importXML()
		self.root = self.tree.getroot()
		self.es = Elasticsearch([{'host':es_ip,'port':es_port}])
		self.index_name = index_name

	def displayInputFileName(self):
		print self.input_file

	def __importXML(self):
		#Parse XML directly from the file path
		return xml.parse(self.input_file)

	def toES(self):
		"Sends each item to Elasticsearch as a unique document"
		for item in self.root.iter('item'):
			dict_item = {}
			dict_item['scanner'] = 'nikto'
			dict_item['osvdbid'] = item.attrib['osvdbid']
			dict_item['method'] =  item.attrib['method']
			for c in item:
				if c.tag == 'description':
					dict_item['description'] = c.text
				elif c.tag == 'uri':
					dict_item['uri'] = c.text
				elif c.tag == 'namelink':
					#regex = re.compile(":\/\/([\w]*):")
					regex = re.compile("(https?)://([.0-9a-zA-Z-]+)(/?.*?)([^/]*)")
					#print regex.search(c.text).groups()
					dict_item['hostname'] = regex.search(c.text).groups()[1]
					dict_item['srcport'] = regex.search(c.text).groups()[3][1:]
					dict_item['site'] = regex.search(c.text).groups()[0] + '://' +  regex.search(c.text).groups()[1]
				elif c.tag == 'iplink':
					regex = re.compile("((?:[0-9]{1,3}\.){3}[0-9]{1,3})")
					dict_item['srcip'] = regex.search(c.text).groups()[0]
			self.es.index(index=self.index_name,doc_type="vuln", body=json.dumps(dict_item))


class OpenVasES:
	"This class will parse an OpenVAS XML file and send it to Elasticsearch"

	def __init__(self, input_file,es_ip,es_port,index_name):
		self.input_file = input_file
                self.displayInputFileName()
		self.tree = self.__importXML()
		self.root = self.tree.getroot()
		self.issueList = self.__createIssuesList()
		self.portList = self.__createPortsList()
		self.es = Elasticsearch([{'host':es_ip,'port':es_port}])
		self.index_name = index_name

	def displayInputFileName(self):
		print self.input_file

	def __importXML(self):
		#Parse XML directly from the file path
                self.displayInputFileName()
		return xml.parse(self.input_file)
                self.displayInputFileName()

	def __createIssuesList(self):
		"Returns a list of dictionaries for each issue in the report"
		issuesList = [];
		for result in self.root.iter('result'):
			issueDict = {};
			issueDict['scanner'] = 'openvas'
			if result.find('host') is not None:
				issueDict['ip'] = unicode(result.find('host').text)
				#print issueDict['host']
			for nvt in result.iter('nvt'):
				issueDict['oid'] = unicode(nvt.attrib['oid'])
				for child in nvt:
					issueDict[child.tag] = unicode(child.text)

			if result.find('description') is not None:
				issueDict['description'] = unicode(result.find('description').text)
			if result.find('port') is not None:
				issueDict['port'] = unicode(result.find('port').text)
			if result.find('threat') is not None:
				issueDict['threat'] = unicode(result.find('threat').text)
			if result.find('severity') is not None:
				issueDict['severity'] = unicode(result.find('severity').text)
			if result.find('scan_nvt_version') is not None:
                            issueDict['scan_nvt_version'] = unicode(result.find('scan_nvt_version').text)
			if issueDict:
				issuesList.append(issueDict)

		#for x in issuesList:
		#	print x['description']
		return issuesList



	def __createPortsList(self):
		"Returns a list of dictionaries for each ports in the report"
		portsList = [];
		for p in self.root.iter('ports'):
			for port in p:
				portDict = {};
				portDict['scanner'] = 'openvas'
				if port.text != 'general/tcp':
					d = self.parsePort(port.text)
					#print d['service']
					if port.find('host').text is not None: portDict['ip'] = port.find('host').text
					if d != None:
						portDict['service'] = d['service']
						portDict['port'] = d['port']
						portDict['protocol'] = d['protocol']
						portsList.append(portDict)



		return portsList

	def parsePort(self,string):
		fieldsDict={};
		portsParsed = re.search(r'(\S*\b)\s\((\d+)\/(\w+)',string)
		#portsParsed = re.search('(\S*)\s\((\d+)\/(\w+)',string)
		#print string
		if portsParsed:
			fieldsDict['service'] = unicode(portsParsed.group(1))
			fieldsDict['port'] = unicode(portsParsed.group(2))
			fieldsDict['protocol'] = unicode(portsParsed.group(3))
			#print fieldsDict
			return fieldsDict
		return None


	def toES(self):
		for item in self.issueList:
			self.es.index(index=self.index_name,doc_type="vuln", body=json.dumps(item))
		for port in self.portList:
			self.es.index(index=self.index_name,doc_type="vuln", body=json.dumps(port))


def usage():
		print "Usage: VulntoES.py [-i input_file | input_file=input_file] [-e elasticsearch_ip | es_ip=es_ip_address] [-p elasticsearch_port | es_port=es_server_port] [-I index_name] [-r report_type | --report_type=type] [-s name=value] [-h | --help]"
def main():

        letters = 'i:I:e:p:r:s:h' #input_file, index_name es_ip_address, report_type, create_sql, create_xml, help
	keywords = ['input-file=', 'index_name=', 'es_ip=','es_port=','report_type=', 'static=', 'help' ]
	try:
		opts, extraparams = getopt.getopt(sys.argv[1:], letters, keywords)
	except getopt.GetoptError, err:
		print str(err)
		usage()
		sys.exit()
	in_file = ''
	es_ip = ''
	es_port = 9200
	report_type = ''
	index_name = ''
        static_fields = dict()

	for o,p in opts:
	  if o in ['-i','--input-file=']:
		in_file = p
	  elif o in ['-r', '--report_type=']:
	  	report_type = p
	  elif o in ['-e', '--es_ip=']:
	  	es_ip=p
	  elif o in ['-p', '--es_port=']:
		es_port=p
	  elif o in ['-I', '--index_name=']:
		index_name=p
          elif o in ['-s', '--static']:
                name, value = p.split("=", 1)
                static_fields[name] = value
	  elif o in ['-h', '--help']:
		 usage()
		 sys.exit()


	if (len(sys.argv) < 1):
		usage()
		sys.exit()

	try:
		with open(in_file) as f: pass
	except IOError as e:
		print "Input file does not exist. Exiting."
		sys.exit()

	if report_type.lower() == 'nessus':
		print "Sending Nessus data to Elasticsearch"
		np = NessusES(in_file,es_ip,es_port,index_name, static_fields)
		np.toES()
	elif report_type.lower() == 'nikto':
		print "Sending Nikto data to Elasticsearch"
		np = NiktoES(in_file,es_ip,es_port,index_name)
		np.toES()
	elif report_type.lower() == 'nmap':
		print "Sending Nmap data to Elasticsearch"
		np = NmapES(in_file,es_ip,es_port,index_name)
		np.toES()
	elif report_type.lower() == 'openvas':
		np = OpenVasES(in_file,es_ip,es_port,index_name)
		np.toES()
	else:
		print "Error: Invalid report type specified. Available options: nessus, nikto, nmap, openvas"
		sys.exit()

if __name__ == "__main__":
	main()