def create_request(self, operation, request, data):
""" Reimplemented to intercept requests. Stops blacklisted requests
CFG02
"""
request.setAttribute(
QNetworkRequest.HttpPipeliningAllowedAttribute, True)
request.setAttribute(QNetworkRequest.CacheLoadControlAttribute,
QNetworkRequest.PreferCache)
qurl = request.url()
url = qurl.toString()
info = {'notify': self.show_detail,
'message': None,
'detail': None,
'qurl': qurl,
'color': Fore.GREEN}
# if keeping a log of the POST data, do it before anything else
if operation == QNetworkAccessManager.PostOperation:
notify_post(data, url)
# stop here if the request is local enough as for not
# requiring further scrutiny
if is_local(qurl) and not is_font(qurl):
info['message'] = "LOCAL"
return self.__pass(operation, request, data, info=info)
# make optional the loading of webfonts
if is_font(qurl):
if self.load_webfonts:
return self.__pass(operation, request, data)
else:
info['message'] = "TRIMMED WEBFONT"
return self.__block("about:blank", info=info)
# If the request is going to be intercepted a custom request is
# built and returned after optionally reporting the reason
whitelist = get_options()['sites'][self.__instance]['host_whitelist']
info['color'] = Fore.RED
for (stop_case, description, show) in [
# it may be an un-dns'ed request; careful here
(is_numerical(qurl.host()), "NUMERICAL", True),
# whitelist exists, and the requested URL is not in it
(non_whitelisted(whitelist, qurl),
"NON WHITELISTED", self.show_detail)
]:
if stop_case:
info['notify'] = show
info['message'] = description
return self.__block("about:blank", info=info)
# It's not a local request; it should have a proper URL structure
# then. 'domain' and 'suffix' must be non-None (and non-empty).
(subdomain, domain, suffix) = tldextract.extract(url)
subdomain = subdomain if subdomain != '' else None
# FIXME this entire block (and maybe the one above) feels half-baked
for (stop_case, description, detail, show) in [
# if 'domain' or 'suffix' are not valid, stop;
# should never happen (even though it does - some providers
# don't have a proper 'domain' according to tldextract
(domain == '' or suffix == '',
"SHOULD NOT HAPPEN", " {}|{}|{} ".format(subdomain,
domain,
suffix), True),
# site is on a whitelist, but this instance does not use
# whitelists; this means it belong to another instance,
# assuming exclusivity
# TODO 'domain + suffix' is a maybe bad generalization
(whitelist is None and # CFG02
domain + '.' + suffix in get_options()['all_whitelists'],
"WHITELIST^C FILTER", "{} || {} || {} ".format(
subdomain,
domain,
suffix), self.show_detail)
]:
if stop_case:
info['message'] = description
info['detail'] = detail
return self.__block(encode_blocked(description, url),
info=info)
return self.__pass(operation, request, data)
def reply_complete(reply):
""" Prints when a request completes, handles the filter that
chooses between successful and filtered requests
Replies only - if the request doesn't have to go
through the network, it will not be reported here
AB01
"""
qurl = reply.url()
s_url = qurl.toString()
status = reply.attribute(
QNetworkRequest.HttpStatusCodeAttribute)
if is_local(qurl) or status is None:
return
# 'status' will never be None from here on
# used only once
color_status = Fore.GREEN if reply.attribute(
QNetworkRequest.SourceIsFromCacheAttribute) else Fore.RED
# was it a filtered reply?
# used only once
color = Fore.BLUE if 200 <= status < 400 else Fore.RED
origin = reply.request().originatingObject()
if origin:
# used only once
not_same = not compare_host(
origin.requestedUrl().host(), qurl.host())
else:
print("NO ORIGINATING OBJECT", s_url)
return
in_iframe = origin != origin.page().mainFrame()
# this is the response to the original .load(), not a
# resource loaded by the page itself
if origin.requestedUrl() == qurl and not in_iframe:
host = sub("^www.", "", qurl.host())
path = qurl.path().rstrip("/ ")
if ( # AB01
(host not in DO_NOT_STORE) and
(not qurl.hasQuery()) and
len(path.split('/')) < 4 and
200 <= status < 400):
database().store_navigation(host, path, self.prefix)
if not_same or in_iframe:
info = {
'color_status': color_status,
'status': status,
'resource': qurl,
'source': origin.requestedUrl(),
'main_color': color,
'iframe': in_iframe}
show_info(info)
