#!/usr/bin/env python """ $Id$ Copyright (c) 2006-2011 sqlmap developers (http://www.sqlmap.org/) See the file 'doc/COPYING' for copying permission """ import httplib import re import socket import time import urllib2 import urlparse import traceback from extra.multipart import multipartpost from lib.core.agent import agent from lib.core.common import average from lib.core.common import calculateDeltaSeconds from lib.core.common import clearConsoleLine from lib.core.common import cpuThrottle from lib.core.common import extractRegexResult from lib.core.common import getCurrentThreadData from lib.core.common import getFilteredPageContent from lib.core.common import getUnicode from lib.core.common import logHTTPTraffic from lib.core.common import parseTargetUrl from lib.core.common import readInput from lib.core.common import removeReflectiveValues from lib.core.common import singleTimeWarnMessage from lib.core.common import stdev from lib.core.common import urlEncodeCookieValues from lib.core.common import wasLastRequestDelayed from lib.core.convert import unicodeencode from lib.core.convert import urlencode from lib.core.data import conf from lib.core.data import kb from lib.core.data import logger from lib.core.enums import HTTPHEADER from lib.core.enums import HTTPMETHOD from lib.core.enums import NULLCONNECTION from lib.core.enums import PAYLOAD from lib.core.enums import PLACE from lib.core.exception import sqlmapConnectionException from lib.core.exception import sqlmapSyntaxException from lib.core.settings import HTTP_ACCEPT_HEADER_VALUE from lib.core.settings import HTTP_SILENT_TIMEOUT from lib.core.settings import META_REFRESH_REGEX from lib.core.settings import IS_WIN from lib.core.settings import MIN_TIME_RESPONSES from lib.core.settings import WARN_TIME_STDEV from lib.core.settings import URI_HTTP_HEADER from lib.core.threads import getCurrentThreadData from lib.request.basic import decodePage from lib.request.basic import forgeHeaders from lib.request.basic import processResponse from lib.request.direct import direct from lib.request.comparison import comparison from lib.request.methodrequest import MethodRequest from lib.utils.checkpayload import checkPayload class Connect: """ This class defines methods used to perform HTTP requests """ @staticmethod def __getPageProxy(**kwargs): return Connect.getPage(**kwargs) @staticmethod def __retryProxy(**kwargs): threadData = getCurrentThreadData() threadData.retriesCount += 1 if kb.testMode and kb.previousMethod == PAYLOAD.METHOD.TIME: # timed based payloads can cause web server unresponsiveness # if the injectable piece of code is some kind of JOIN-like query warnMsg = "most probably web server instance hasn't recovered yet " warnMsg += "from previous timed based payload. if the problem " warnMsg += "persists please wait for few minutes and rerun " warnMsg += "without flag T in --technique option " warnMsg += "(e.g. --flush-session --technique=BEUS) or try to " warnMsg += "lower the --time-sec value (e.g. --time-sec=2)" singleTimeWarnMessage(warnMsg) elif kb.originalPage is None: warnMsg = "if the problem persists please check that the provided " warnMsg += "target url is valid. If it is, you can try to rerun " warnMsg += "with the --random-agent switch turned on " warnMsg += "and/or proxy switches (--ignore-proxy, --proxy,...)" singleTimeWarnMessage(warnMsg) elif conf.threads > 1: warnMsg = "if the problem persists please try to lower " warnMsg += "the number of used threads (--threads)" singleTimeWarnMessage(warnMsg) time.sleep(1) kwargs['retrying'] = True return Connect.__getPageProxy(**kwargs) @staticmethod def getPage(**kwargs): """ This method connects to the target url or proxy and returns the target url page content """ if conf.delay is not None and isinstance(conf.delay, (int, float)) and conf.delay > 0: time.sleep(conf.delay) elif conf.cpuThrottle: cpuThrottle(conf.cpuThrottle) threadData = getCurrentThreadData() threadData.lastRequestUID += 1 url = kwargs.get('url', conf.url) get = kwargs.get('get', None) post = kwargs.get('post', None) method = kwargs.get('method', None) cookie = kwargs.get('cookie', None) ua = kwargs.get('ua', None) referer = kwargs.get('referer', None) direct = kwargs.get('direct', False) multipart = kwargs.get('multipart', False) silent = kwargs.get('silent', False) raise404 = kwargs.get('raise404', True) auxHeaders = kwargs.get('auxHeaders', None) response = kwargs.get('response', False) ignoreTimeout = kwargs.get('ignoreTimeout', kb.ignoreTimeout) refreshing = kwargs.get('refreshing', False) retrying = kwargs.get('retrying', False) redirecting = kwargs.get('redirecting', False) crawling = kwargs.get('crawling', False) if not urlparse.urlsplit(url).netloc: url = urlparse.urljoin(conf.url, url) # flag to know if we are dealing with the same target host target = reduce(lambda x, y: x == y, map(lambda x: urlparse.urlparse(x).netloc.split(':')[0], [url, conf.url])) if not retrying: # Reset the number of connection retries threadData.retriesCount = 0 # fix for known issue when urllib2 just skips the other part of provided # url splitted with space char while urlencoding it in the later phase url = url.replace(" ", "%20") page = "" cookieStr = "" requestMsg = "HTTP request [#%d]:\n%s " % (threadData.lastRequestUID, method or (HTTPMETHOD.POST if post else HTTPMETHOD.GET)) requestMsg += "%s" % urlparse.urlsplit(url)[2] or "/" responseMsg = "HTTP response " requestHeaders = "" responseHeaders = None logHeaders = "" # fix for known issues when using url in unicode format # (e.g. UnicodeDecodeError: "url = url + '?' + query" in redirect case) url = unicodeencode(url) try: if silent: socket.setdefaulttimeout(HTTP_SILENT_TIMEOUT) else: socket.setdefaulttimeout(conf.timeout) if direct: if "?" in url: url, params = url.split("?") params = urlencode(params) url = "%s?%s" % (url, params) requestMsg += "?%s" % params elif multipart: # Needed in this form because of potential circle dependency # problem (option -> update -> connect -> option) from lib.core.option import proxyHandler multipartOpener = urllib2.build_opener(proxyHandler, multipartpost.MultipartPostHandler) conn = multipartOpener.open(unicodeencode(url), multipart) page = conn.read() responseHeaders = conn.info() responseHeaders[URI_HTTP_HEADER] = conn.geturl() page = decodePage(page, responseHeaders.get(HTTPHEADER.CONTENT_ENCODING), responseHeaders.get(HTTPHEADER.CONTENT_TYPE)) return page elif any ([refreshing, crawling]): pass elif target: if conf.parameters.has_key(PLACE.GET) and not get: get = conf.parameters[PLACE.GET] if get: url = "%s?%s" % (url, get) requestMsg += "?%s" % get if conf.method == HTTPMETHOD.POST and not post: for place in (PLACE.POST, PLACE.SOAP): if conf.parameters.has_key(place): post = conf.parameters[place] break elif get: url = "%s?%s" % (url, get) requestMsg += "?%s" % get requestMsg += " %s" % httplib.HTTPConnection._http_vsn_str # Perform HTTP request headers = forgeHeaders(cookie, ua, referer) if conf.realTest: headers[HTTPHEADER.REFERER] = "%s://%s" % (conf.scheme, conf.hostname) if kb.authHeader: headers[HTTPHEADER.AUTHORIZATION] = kb.authHeader if kb.proxyAuthHeader: headers[HTTPHEADER.PROXY_AUTHORIZATION] = kb.proxyAuthHeader headers[HTTPHEADER.ACCEPT] = HTTP_ACCEPT_HEADER_VALUE headers[HTTPHEADER.HOST] = urlparse.urlparse(url).netloc if any(map(lambda x: headers[HTTPHEADER.HOST].endswith(':%d' % x), [80, 443])): headers[HTTPHEADER.HOST] = headers[HTTPHEADER.HOST].split(':')[0] if auxHeaders: for key, item in auxHeaders.items(): headers[key] = item for key, item in headers.items(): del headers[key] headers[unicodeencode(key, kb.pageEncoding)] = unicodeencode(item, kb.pageEncoding) post = unicodeencode(post, kb.pageEncoding) if method: req = MethodRequest(url, post, headers) req.set_method(method) else: req = urllib2.Request(url, post, headers) if not conf.dropSetCookie and conf.cj: for _, cookie in enumerate(conf.cj): if not cookieStr: cookieStr = "Cookie: " cookie = getUnicode(cookie) index = cookie.index(" for ") cookieStr += "%s; " % cookie[8:index] if not req.has_header(HTTPHEADER.ACCEPT_ENCODING): requestHeaders += "%s: identity\n" % HTTPHEADER.ACCEPT_ENCODING requestHeaders += "\n".join(["%s: %s" % (key.capitalize() if isinstance(key, basestring) else key, getUnicode(value)) for (key, value) in req.header_items()]) if not req.has_header(HTTPHEADER.COOKIE) and cookieStr: requestHeaders += "\n%s" % cookieStr[:-2] if not req.has_header(HTTPHEADER.CONNECTION): requestHeaders += "\n%s: close" % HTTPHEADER.CONNECTION requestMsg += "\n%s" % requestHeaders if post: requestMsg += "\n\n%s" % post requestMsg += "\n" logger.log(8, requestMsg) conn = urllib2.urlopen(req) if not kb.authHeader and req.has_header(HTTPHEADER.AUTHORIZATION): kb.authHeader = req.get_header(HTTPHEADER.AUTHORIZATION) if not kb.proxyAuthHeader and req.has_header(HTTPHEADER.PROXY_AUTHORIZATION): kb.proxyAuthHeader = req.get_header(HTTPHEADER.PROXY_AUTHORIZATION) if hasattr(conn, "setcookie"): kb.redirectSetCookie = conn.setcookie if hasattr(conn, "redurl") and hasattr(conn, "redcode") and target\ and not redirecting and not conf.realTest: if kb.alwaysRedirect is None: msg = "sqlmap got a %d redirect to " % conn.redcode msg += "'%s'. Do you want to follow redirects " % conn.redurl msg += "from now on (or stay on the original page)? [Y/n]" choice = readInput(msg, default="Y") kb.alwaysRedirect = choice not in ("n", "N") kwargs['url'] = conn.redurl if kb.alwaysRedirect else conf.url kwargs['redirecting'] = True return Connect.__getPageProxy(**kwargs) # Return response object if response: return conn, None # Get HTTP response page = conn.read() code = conn.code responseHeaders = conn.info() responseHeaders[URI_HTTP_HEADER] = conn.geturl() page = decodePage(page, responseHeaders.get(HTTPHEADER.CONTENT_ENCODING), responseHeaders.get(HTTPHEADER.CONTENT_TYPE)) status = getUnicode(conn.msg) if extractRegexResult(META_REFRESH_REGEX, page, re.DOTALL | re.IGNORECASE) and not refreshing: url = extractRegexResult(META_REFRESH_REGEX, page, re.DOTALL | re.IGNORECASE) debugMsg = "got HTML meta refresh header" logger.debug(debugMsg) if kb.alwaysRefresh is None: msg = "sqlmap got a refresh request " msg += "(redirect like response common to login pages). " msg += "Do you want to apply the refresh " msg += "from now on (or stay on the original page)? [Y/n]" choice = readInput(msg, default="Y") kb.alwaysRefresh = choice not in ("n", "N") if kb.alwaysRefresh: if url.lower().startswith('http://'): kwargs['url'] = url else: kwargs['url'] = conf.url[:conf.url.rfind('/')+1] + url threadData.lastRedirectMsg = (threadData.lastRequestUID, page) kwargs['refreshing'] = True kwargs['get'] = None kwargs['post'] = None try: return Connect.__getPageProxy(**kwargs) except sqlmapSyntaxException: pass # Explicit closing of connection object if not conf.keepAlive: try: conn.fp._sock.close() conn.close() except Exception, msg: warnMsg = "problem occured during connection closing ('%s')" % msg logger.warn(warnMsg) except urllib2.HTTPError, e: page = None responseHeaders = None try: page = e.read() responseHeaders = e.info() responseHeaders[URI_HTTP_HEADER] = e.geturl() page = decodePage(page, responseHeaders.get(HTTPHEADER.CONTENT_ENCODING), responseHeaders.get(HTTPHEADER.CONTENT_TYPE)) except socket.timeout: warnMsg = "connection timed out while trying " warnMsg += "to get error page information (%d)" % e.code logger.warn(warnMsg) return None, None except: pass code = e.code threadData.lastHTTPError = (threadData.lastRequestUID, code) if code not in kb.httpErrorCodes: kb.httpErrorCodes[code] = 0 kb.httpErrorCodes[code] += 1 status = getUnicode(e.msg) responseMsg += "[#%d] (%d %s):\n" % (threadData.lastRequestUID, code, status) if responseHeaders: logHeaders = "\n".join(["%s: %s" % (key.capitalize() if isinstance(key, basestring) else key, getUnicode(value)) for (key, value) in responseHeaders.items()]) logHTTPTraffic(requestMsg, "%s%s\n\n%s" % (responseMsg, logHeaders, page if isinstance(page, unicode) else getUnicode(page))) if conf.verbose <= 5: responseMsg += getUnicode(logHeaders) elif conf.verbose > 5: responseMsg += "%s\n\n%s\n" % (logHeaders, page) logger.log(7, responseMsg) if e.code == 401: errMsg = "not authorized, try to provide right HTTP " errMsg += "authentication type and valid credentials (%d)" % code raise sqlmapConnectionException, errMsg elif e.code == 404: if raise404: errMsg = "page not found (%d)" % code raise sqlmapConnectionException, errMsg else: debugMsg = "page not found (%d)" % code logger.debug(debugMsg) processResponse(page, responseHeaders) elif e.code == 504: if ignoreTimeout: return None, None else: warnMsg = "unable to connect to the target url (%d - %s)" % (e.code, httplib.responses[e.code]) if threadData.retriesCount < conf.retries and not kb.threadException and not conf.realTest: warnMsg += ", sqlmap is going to retry the request" logger.critical(warnMsg) return Connect.__retryProxy(**kwargs) elif kb.testMode: logger.critical(warnMsg) return None, None else: raise sqlmapConnectionException, warnMsg else: debugMsg = "got HTTP error code: %d (%s)" % (code, status) logger.debug(debugMsg) processResponse(page, responseHeaders) return page, responseHeaders except (urllib2.URLError, socket.error, socket.timeout, httplib.BadStatusLine, httplib.IncompleteRead), e: tbMsg = traceback.format_exc() if "no host given" in tbMsg: warnMsg = "invalid url address used (%s)" % repr(url) raise sqlmapSyntaxException, warnMsg elif "forcibly closed" in tbMsg: warnMsg = "connection was forcibly closed by the target url" elif "timed out" in tbMsg: warnMsg = "connection timed out to the target url" elif "URLError" in tbMsg or "error" in tbMsg: warnMsg = "unable to connect to the target url" elif "BadStatusLine" in tbMsg: warnMsg = "the target url responded with an unknown HTTP " warnMsg += "status code, try to force the HTTP User-Agent " warnMsg += "header with option --user-agent or --random-agent" elif "IncompleteRead" in tbMsg: warnMsg = "there was an incomplete read error while retrieving data " warnMsg += "from the target url" else: warnMsg = "unable to connect to the target url" if "BadStatusLine" not in tbMsg: warnMsg += " or proxy" if "forcibly closed" in tbMsg: logger.critical(warnMsg) return None, None elif silent or (ignoreTimeout and any(map(lambda x: x in tbMsg, ["timed out", "IncompleteRead"]))): return None, None elif threadData.retriesCount < conf.retries and not kb.threadException and not conf.realTest: warnMsg += ", sqlmap is going to retry the request" logger.critical(warnMsg) return Connect.__retryProxy(**kwargs) elif kb.testMode: logger.critical(warnMsg) return None, None else: raise sqlmapConnectionException, warnMsg finally: socket.setdefaulttimeout(conf.timeout) processResponse(page, responseHeaders) responseMsg += "[#%d] (%d %s):\n" % (threadData.lastRequestUID, code, status) if responseHeaders: logHeaders = "\n".join(["%s: %s" % (key.capitalize() if isinstance(key, basestring) else key, getUnicode(value)) for (key, value) in responseHeaders.items()]) logHTTPTraffic(requestMsg, "%s%s\n\n%s" % (responseMsg, logHeaders, page if isinstance(page, unicode) else getUnicode(page))) if conf.verbose <= 5: responseMsg += getUnicode(logHeaders) elif conf.verbose > 5: responseMsg += "%s\n\n%s\n" % (logHeaders, page) logger.log(7, responseMsg) return page, responseHeaders @staticmethod def queryPage(value=None, place=None, content=False, getRatioValue=False, silent=False, method=None, timeBasedCompare=False, noteResponseTime=True, auxHeaders=None, response=False, raise404=None): """ This method calls a function to get the target url page content and returns its page MD5 hash or a boolean value in case of string match check ('--string' command line parameter) """ if conf.direct: return direct(value, content) get = None post = None cookie = None ua = None referer = None page = None pageLength = None uri = None if not place: place = kb.injection.place or PLACE.GET raise404 = place != PLACE.URI if raise404 is None else raise404 payload = agent.extractPayload(value) threadData = getCurrentThreadData() if payload: if kb.tamperFunctions: for function in kb.tamperFunctions: payload = function(payload) value = agent.replacePayload(value, payload) logger.log(9, payload) if place == PLACE.COOKIE and conf.cookieUrlencode: value = agent.removePayloadDelimiters(value) value = urlEncodeCookieValues(value) elif place: if place in (PLACE.GET, PLACE.POST): # payloads in GET and/or POST need to be urlencoded # throughly without safe chars (especially & and =) # addendum: as we support url encoding in tampering # functions therefore we need to use % as a safe char payload = urlencode(payload, "%", False, True) value = agent.replacePayload(value, payload) elif place == PLACE.SOAP: # payloads in SOAP should have chars > and < replaced # with their HTML encoded counterparts payload = payload.replace('>', '>').replace('<', '<') value = agent.replacePayload(value, payload) value = agent.removePayloadDelimiters(value) if conf.checkPayload: checkPayload(value) if PLACE.GET in conf.parameters: get = urlencode(conf.parameters[PLACE.GET] if place != PLACE.GET or not value else value, limit=True) if PLACE.POST in conf.parameters: post = urlencode(conf.parameters[PLACE.POST] if place != PLACE.POST or not value else value) if PLACE.SOAP in conf.parameters: post = conf.parameters[PLACE.SOAP] if place != PLACE.SOAP or not value else value if PLACE.COOKIE in conf.parameters: cookie = conf.parameters[PLACE.COOKIE] if place != PLACE.COOKIE or not value else value if PLACE.UA in conf.parameters: ua = conf.parameters[PLACE.UA] if place != PLACE.UA or not value else value if PLACE.REFERER in conf.parameters: referer = conf.parameters[PLACE.REFERER] if place != PLACE.REFERER or not value else value if PLACE.URI in conf.parameters: uri = conf.url if place != PLACE.URI or not value else value else: uri = conf.url if timeBasedCompare: if len(kb.responseTimes) < MIN_TIME_RESPONSES: clearConsoleLine() warnMsg = "time-based comparison needs larger statistical " warnMsg += "model. Making a few dummy requests, please wait.." singleTimeWarnMessage(warnMsg) while len(kb.responseTimes) < MIN_TIME_RESPONSES: Connect.queryPage(content=True) deviation = stdev(kb.responseTimes) if deviation > WARN_TIME_STDEV: kb.adjustTimeDelay = False warnMsg = "there is considerable lagging (standard deviation: " warnMsg += "%f sec%s) " % (deviation, "s" if deviation > 1 else "") warnMsg += "in connection response(s). Please use as high " warnMsg += "value for --time-sec option as possible (e.g. " warnMsg += "%d or more)" % (conf.timeSec * 2) logger.critical(warnMsg) elif not kb.testMode: warnMsg = "it is very important not to stress the network adapter's " warnMsg += "bandwidth during usage of time-based queries" singleTimeWarnMessage(warnMsg) if conf.safUrl and conf.saFreq > 0: kb.queryCounter += 1 if kb.queryCounter % conf.saFreq == 0: Connect.getPage(url=conf.safUrl, cookie=cookie, direct=True, silent=True, ua=ua, referer=referer) start = time.time() if kb.nullConnection and not content and not response and not timeBasedCompare: if kb.nullConnection == NULLCONNECTION.HEAD: method = HTTPMETHOD.HEAD elif kb.nullConnection == NULLCONNECTION.RANGE: if not auxHeaders: auxHeaders = {} auxHeaders[HTTPHEADER.RANGE] = "bytes=-1" _, headers = Connect.getPage(url=uri, get=get, post=post, cookie=cookie, ua=ua, referer=referer, silent=silent, method=method, auxHeaders=auxHeaders, raise404=raise404) if headers: if kb.nullConnection == NULLCONNECTION.HEAD and HTTPHEADER.CONTENT_LENGTH in headers: pageLength = int(headers[HTTPHEADER.CONTENT_LENGTH]) elif kb.nullConnection == NULLCONNECTION.RANGE and HTTPHEADER.CONTENT_RANGE in headers: pageLength = int(headers[HTTPHEADER.CONTENT_RANGE][headers[HTTPHEADER.CONTENT_RANGE].find('/') + 1:]) if not pageLength: page, headers = Connect.getPage(url=uri, get=get, post=post, cookie=cookie, ua=ua, referer=referer, silent=silent, method=method, auxHeaders=auxHeaders, response=response, raise404=raise404, ignoreTimeout=timeBasedCompare) threadData.lastQueryDuration = calculateDeltaSeconds(start) if kb.testMode: kb.testQueryCount += 1 if conf.cj: conf.cj.clear() if timeBasedCompare: return wasLastRequestDelayed() elif noteResponseTime: kb.responseTimes.append(threadData.lastQueryDuration) if content or response: return page, headers page = removeReflectiveValues(page, payload) if getRatioValue: return comparison(page, headers, getRatioValue=False, pageLength=pageLength), comparison(page, headers, getRatioValue=True, pageLength=pageLength) elif pageLength or page: return comparison(page, headers, getRatioValue, pageLength) else: return False