1
0
mirror of https://github.com/webrecorder/pywb.git synced 2025-03-15 00:03:28 +01:00

- add referrer self-redirect check and test case

- dispatching: cleanup wbrequestresponse, move tests to a seperate file
- wbrequest: store both rel_prefix and host_prefix, with wb_prefix either full
or rel path as needed, so that full and relative paths are
both available in wbrequest
- create WbUrlHandler to differentiate handlers which
support WbUrl (timestamp[mod]/url) semantic vs other request handlers.
This commit is contained in:
Ilya Kreymer 2014-02-23 23:31:54 -08:00
parent a4f1224d16
commit 9194e867ea
12 changed files with 234 additions and 136 deletions

View File

@ -3,13 +3,13 @@ import re
from wbrequestresponse import WbRequest, WbResponse
from pywb.rewrite.url_rewriter import UrlRewriter
from pywb.rewrite.wburl import WbUrl
#=================================================================
# ArchivalRouter -- route WB requests in archival mode
#=================================================================
class ArchivalRouter:
def __init__(self, routes, hostpaths = None, abs_path = True, home_view = None, error_view = None):
def __init__(self, routes, hostpaths=None, abs_path=True, home_view=None, error_view=None):
self.routes = routes
self.fallback = ReferRedirect(hostpaths)
self.abs_path = abs_path
@ -69,24 +69,25 @@ class Route:
if not matcher:
return None
rel_prefix = matcher.group(0)
matched_str = matcher.group(0)
if rel_prefix:
wb_prefix = env['SCRIPT_NAME'] + '/' + rel_prefix + '/'
wb_url_str = request_uri[len(rel_prefix) + 2:] # remove the '/' + rel_prefix part of uri
if matched_str:
rel_prefix = env['SCRIPT_NAME'] + '/' + matched_str + '/'
wb_url_str = request_uri[len(matched_str) + 2:] # remove the '/' + rel_prefix part of uri
else:
wb_prefix = env['SCRIPT_NAME'] + '/'
rel_prefix = env['SCRIPT_NAME'] + '/'
wb_url_str = request_uri[1:] # the request_uri is the wb_url, since no coll
coll = matcher.group(self.coll_group)
wbrequest = WbRequest(env,
request_uri = request_uri,
wb_url_str = wb_url_str,
wb_prefix = wb_prefix,
coll = coll,
host_prefix = WbRequest.make_host_prefix(env) if use_abs_prefix else '',
wburl_class = self.handler.get_wburl_type())
request_uri=request_uri,
wb_url_str=wb_url_str,
rel_prefix=rel_prefix,
coll=coll,
use_abs_prefix=use_abs_prefix,
wburl_class = self.handler.get_wburl_type(),
urlrewriter_class=UrlRewriter)
# Allow for applying of additional filters

View File

@ -132,8 +132,8 @@ org,iana)/domains/root/db 20140126200928 http://www.iana.org/domains/root/db tex
('filename', 'dupes.warc.gz')]
# NOTE: external dependency -- need self-contained test
>>> x = CDXServer('http://web.archive.org/cdx/search/cdx').load_cdx(url = 'example.com', output = 'raw', limit = '2')
>>> pprint.pprint(x.next().items())
#>>> x = CDXServer('http://web.archive.org/cdx/search/cdx').load_cdx(url = 'example.com', output = 'raw', limit = '2')
#>>> pprint.pprint(x.next().items())
[('urlkey', 'com,example)/'),
('timestamp', '20020120142510'),
('original', 'http://example.com:80/'),

View File

@ -10,19 +10,28 @@ from wbexceptions import WbException, NotFoundException
from views import TextCapturesView
class BaseHandler:
@staticmethod
def get_wburl_type():
return WbUrl
#=================================================================
class BaseHandler(object):
def __call__(self, wbrequest):
return wbrequest
def get_wburl_type(self):
return None
#=================================================================
class WbUrlHandler(BaseHandler):
def get_wburl_type(self):
return WbUrl
#=================================================================
# Standard WB Handler
#=================================================================
class WBHandler(BaseHandler):
def __init__(self, index_reader, replay, html_view = None, search_view = None):
class WBHandler(WbUrlHandler):
def __init__(self, index_reader, replay,
html_view=None, search_view=None):
self.index_reader = index_reader
self.replay = replay
@ -31,7 +40,6 @@ class WBHandler(BaseHandler):
self.html_view = html_view
self.search_view = search_view
def __call__(self, wbrequest):
if wbrequest.wb_url_str == '/':
return self.render_search_page(wbrequest)
@ -61,6 +69,7 @@ class WBHandler(BaseHandler):
def __str__(self):
return 'WBHandler: ' + str(self.index_reader) + ', ' + str(self.replay)
#=================================================================
# CDX-Server Handler -- pass all params to cdx server
#=================================================================
@ -75,11 +84,6 @@ class CDXHandler(BaseHandler):
return self.view.render_response(wbrequest, cdx_lines)
@staticmethod
def get_wburl_type():
return None
def __str__(self):
return 'Index Reader: ' + str(self.index_reader)
@ -115,10 +119,6 @@ class StaticHandler(BaseHandler):
except IOError:
raise NotFoundException('Static File Not Found: ' + wbrequest.wb_url_str)
@staticmethod
def get_wburl_type():
return None
def __str__(self):
return 'Static files from ' + self.static_path
@ -130,6 +130,7 @@ class DebugEchoEnvHandler(BaseHandler):
def __call__(self, wbrequest):
return WbResponse.text_response(str(wbrequest.env))
#=================================================================
class DebugEchoHandler(BaseHandler):
def __call__(self, wbrequest):
@ -150,5 +151,3 @@ class PerfTimer:
self.end = time.clock()
if self.perfdict is not None:
self.perfdict[self.name] = str(self.end - self.start)

View File

@ -37,7 +37,7 @@ class IndexReader(object):
def load_cdx(self, **params):
return self.cdx_server.load_cdx(**params)
def get_query_params(self, wburl, limit = 150000, collapse_time = None, replay_closest = 10):
def get_query_params(self, wburl, limit = 150000, collapse_time = None, replay_closest = 100):
if wburl.type == wburl.URL_QUERY:
raise NotImplementedError('Url Query Not Yet Supported')

View File

@ -45,14 +45,14 @@ class ProxyRouter:
return None
wbrequest = WbRequest(env,
request_uri = url,
wb_url_str = url,
wb_prefix = '',
coll = '',
host_prefix = self.hostpaths[0],
wburl_class = self.handler.get_wburl_type(),
url_rewriter_class = ProxyHttpsUrlRewriter,
is_proxy = True)
request_uri=url,
wb_url_str=url,
#rel_prefix=url,
#host_prefix=self.hostpaths[0],
wburl_class=self.handler.get_wburl_type(),
urlrewriter_class=ProxyHttpsUrlRewriter,
use_abs_prefix=False,
is_proxy=True)
return self.handler(wbrequest)

View File

@ -7,7 +7,6 @@ from wbrequestresponse import WbResponse
from wbexceptions import CaptureException, InternalRedirect
from pywb.warc.recordloader import ArchiveLoadFailed
#=================================================================
class ReplayView:
def __init__(self, content_loader, content_rewriter, head_insert_view = None,
@ -49,6 +48,9 @@ class ReplayView:
# check if redir is needed
self._redirect_if_needed(wbrequest, cdx)
# one more check for referrer-based self-redirect
self._reject_referrer_self_redirect(wbrequest, status_headers)
response = None
if self.content_rewriter and wbrequest.wb_url.mod != 'id_':
@ -148,6 +150,7 @@ class ReplayView:
def _reject_self_redirect(self, wbrequest, cdx, status_headers):
# self-redirect via location
if status_headers.statusline.startswith('3'):
request_url = wbrequest.wb_url.url.lower()
location_url = status_headers.get_header('Location').lower()
@ -156,3 +159,16 @@ class ReplayView:
if (UrlRewriter.strip_protocol(request_url) == UrlRewriter.strip_protocol(location_url)):
raise CaptureException('Self Redirect: ' + str(cdx))
def _reject_referrer_self_redirect(self, wbrequest, status_headers):
# at correct timestamp now, but must check for referrer redirect
# indirect self-redirect, via meta-refresh, if referrer is same as current url
if status_headers.statusline.startswith('2'):
# build full url even if using relative-rewriting
request_url = wbrequest.host_prefix + wbrequest.rel_prefix + str(wbrequest.wb_url)
referrer_url = wbrequest.referrer
if (referrer_url and UrlRewriter.strip_protocol(request_url) == UrlRewriter.strip_protocol(referrer_url)):
raise CaptureException('Self Redirect via Referrer: ' + str(wbrequest.wb_url))

View File

@ -24,9 +24,9 @@ def test_example_2():
def test_example_3():
status_headers, buff = get_rewritten('http://archive.org/', urlrewriter)
#def test_example_3():
# status_headers, buff = get_rewritten('http://archive.org/', urlrewriter)
assert '/pywb/20131226101010/http://example.com/about/terms.php' in buff, buff
# assert '/pywb/20131226101010/http://example.com/about/terms.php' in buff, buff

View File

@ -103,10 +103,12 @@ class UrlRewriter:
return self.prefix + self.wburl.to_str(timestamp=timestamp, url=url)
def set_base_url(self, newUrl):
self.wburl.url = newUrl
def __repr__(self):
return "UrlRewriter('{0}', '{1}')".format(self.wburl, self.prefix)
@staticmethod
def strip_protocol(url):
for protocol in UrlRewriter.PROTOCOLS:

View File

@ -1,13 +1,19 @@
"""
Test Route
# route with relative path
>>> Route('web', BaseHandler())({'REL_REQUEST_URI': '/web/test.example.com', 'SCRIPT_NAME': ''}, False)
{'wb_url': ('latest_replay', '', '', 'http://test.example.com', 'http://test.example.com'), 'coll': 'web', 'wb_prefix': '/web/', 'request_uri': '/web/test.example.com'}
# Test WbRequest parsed via a Route
# route with relative path, print resulting wbrequest
>>> print_req(Route('web', WbUrlHandler())({'REL_REQUEST_URI': '/web/test.example.com', 'SCRIPT_NAME': ''}, False))
{'coll': 'web',
'request_uri': '/web/test.example.com',
'wb_prefix': '/web/',
'wb_url': ('latest_replay', '', '', 'http://test.example.com', 'http://test.example.com')}
# route with absolute path, running at script /my_pywb
>>> Route('web', BaseHandler())({'REL_REQUEST_URI': '/web/2013im_/test.example.com', 'SCRIPT_NAME': '/my_pywb', 'HTTP_HOST': 'localhost:8081', 'wsgi.url_scheme': 'https'}, True)
{'wb_url': ('replay', '2013', 'im_', 'http://test.example.com', '2013im_/http://test.example.com'), 'coll': 'web', 'wb_prefix': 'https://localhost:8081/my_pywb/web/', 'request_uri': '/web/2013im_/test.example.com'}
# route with absolute path, running at script /my_pywb, print resultingwbrequest
>>> print_req(Route('web', WbUrlHandler())({'REL_REQUEST_URI': '/web/2013im_/test.example.com', 'SCRIPT_NAME': '/my_pywb', 'HTTP_HOST': 'localhost:8081', 'wsgi.url_scheme': 'https'}, True))
{'coll': 'web',
'request_uri': '/web/2013im_/test.example.com',
'wb_prefix': 'https://localhost:8081/my_pywb/web/',
'wb_url': ('replay', '2013', 'im_', 'http://test.example.com', '2013im_/http://test.example.com')}
# not matching route -- skipped
>>> Route('web', BaseHandler())({'REL_REQUEST_URI': '/other/test.example.com', 'SCRIPT_NAME': ''}, False)
@ -65,7 +71,12 @@ False
"""
from pywb.archivalrouter import Route, ReferRedirect
from pywb.handlers import BaseHandler
from pywb.handlers import BaseHandler, WbUrlHandler
import pprint
def print_req(req):
varlist = vars(req)
pprint.pprint({k: varlist[k] for k in ('request_uri', 'wb_prefix', 'wb_url', 'coll')})
def _test_redir(match_host, request_uri, referrer, script_name = '', coll = 'coll', http_host = None):
@ -74,7 +85,7 @@ def _test_redir(match_host, request_uri, referrer, script_name = '', coll = 'col
if http_host:
env['HTTP_HOST'] = http_host
routes = [Route(coll, BaseHandler())]
routes = [Route(coll, WbUrlHandler())]
redir = ReferRedirect(match_host)
#req = WbRequest.from_uri(request_uri, env)
@ -85,4 +96,6 @@ def _test_redir(match_host, request_uri, referrer, script_name = '', coll = 'col
return rep.status_headers.get_header('Location')
if __name__ == "__main__":
import doctest
doctest.testmod()

View File

@ -0,0 +1,87 @@
"""
# WbRequest Tests
# =================
>>> print_req_from_uri('/save/_embed/example.com/?a=b')
{'wb_url': ('latest_replay', '', '', 'http://_embed/example.com/?a=b', 'http://_embed/example.com/?a=b'), 'coll': 'save', 'wb_prefix': '/save/', 'request_uri': '/save/_embed/example.com/?a=b'}
>>> print_req_from_uri('/2345/20101024101112im_/example.com/?b=c')
{'wb_url': ('replay', '20101024101112', 'im_', 'http://example.com/?b=c', '20101024101112im_/http://example.com/?b=c'), 'coll': '2345', 'wb_prefix': '/2345/', 'request_uri': '/2345/20101024101112im_/example.com/?b=c'}
>>> print_req_from_uri('/2010/example.com')
{'wb_url': ('latest_replay', '', '', 'http://example.com', 'http://example.com'), 'coll': '2010', 'wb_prefix': '/2010/', 'request_uri': '/2010/example.com'}
>>> print_req_from_uri('../example.com')
{'wb_url': ('latest_replay', '', '', 'http://example.com', 'http://example.com'), 'coll': '', 'wb_prefix': '/', 'request_uri': '../example.com'}
# Abs path
>>> print_req_from_uri('/2010/example.com', {'wsgi.url_scheme': 'https', 'HTTP_HOST': 'localhost:8080'}, use_abs_prefix = True)
{'wb_url': ('latest_replay', '', '', 'http://example.com', 'http://example.com'), 'coll': '2010', 'wb_prefix': 'https://localhost:8080/2010/', 'request_uri': '/2010/example.com'}
# No Scheme, so stick to relative
>>> print_req_from_uri('/2010/example.com', {'HTTP_HOST': 'localhost:8080'}, use_abs_prefix = True)
{'wb_url': ('latest_replay', '', '', 'http://example.com', 'http://example.com'), 'coll': '2010', 'wb_prefix': '/2010/', 'request_uri': '/2010/example.com'}
# WbResponse Tests
# =================
>>> WbResponse.text_response('Test')
{'body': ['Test'], 'status_headers': StatusAndHeaders(protocol = '', statusline = '200 OK', headers = [('Content-Type', 'text/plain')])}
>>> WbResponse.text_stream(['Test', 'Another'], '404')
{'body': ['Test', 'Another'], 'status_headers': StatusAndHeaders(protocol = '', statusline = '404', headers = [('Content-Type', 'text/plain')])}
>>> WbResponse.redir_response('http://example.com/otherfile')
{'body': [], 'status_headers': StatusAndHeaders(protocol = '', statusline = '302 Redirect', headers = [('Location', 'http://example.com/otherfile')])}
"""
from pywb.rewrite.wburl import WbUrl
from pywb.rewrite.url_rewriter import UrlRewriter
from pywb.utils.statusandheaders import StatusAndHeaders
from pywb.wbrequestresponse import WbRequest, WbResponse
def print_req_from_uri(request_uri, env={}, use_abs_prefix=False):
response = req_from_uri(request_uri, env, use_abs_prefix)
varlist = vars(response)
print str({k: varlist[k] for k in ('request_uri', 'wb_prefix', 'wb_url', 'coll')})
def req_from_uri(request_uri, env={}, use_abs_prefix=False):
if not request_uri:
request_uri = env.get('REL_REQUEST_URI')
parts = request_uri.split('/', 2)
# Has coll prefix
if len(parts) == 3:
rel_prefix = '/' + parts[1] + '/'
wb_url_str = parts[2]
coll = parts[1]
# No Coll Prefix
elif len(parts) == 2:
rel_prefix = '/'
wb_url_str = parts[1]
coll = ''
else:
rel_prefix = '/'
wb_url_str = parts[0]
coll = ''
return WbRequest(env,
request_uri=request_uri,
rel_prefix=rel_prefix,
wb_url_str=wb_url_str,
coll=coll,
wburl_class=WbUrl,
urlrewriter_class=UrlRewriter,
use_abs_prefix=use_abs_prefix)
if __name__ == "__main__":
import doctest
doctest.testmod()

View File

@ -1,99 +1,75 @@
from pywb.rewrite.wburl import WbUrl
from pywb.rewrite.url_rewriter import UrlRewriter
from pywb.utils.statusandheaders import StatusAndHeaders
import pprint
#WB Request and Response
#=================================================================
class WbRequest:
"""
>>> WbRequest.from_uri('/save/_embed/example.com/?a=b')
{'wb_url': ('latest_replay', '', '', 'http://_embed/example.com/?a=b', 'http://_embed/example.com/?a=b'), 'coll': 'save', 'wb_prefix': '/save/', 'request_uri': '/save/_embed/example.com/?a=b'}
Represents the main pywb request object.
>>> WbRequest.from_uri('/2345/20101024101112im_/example.com/?b=c')
{'wb_url': ('replay', '20101024101112', 'im_', 'http://example.com/?b=c', '20101024101112im_/http://example.com/?b=c'), 'coll': '2345', 'wb_prefix': '/2345/', 'request_uri': '/2345/20101024101112im_/example.com/?b=c'}
Contains various info from wsgi env, add additional info
about the request, such as coll, relative prefix,
host prefix, absolute prefix.
>>> WbRequest.from_uri('/2010/example.com')
{'wb_url': ('latest_replay', '', '', 'http://example.com', 'http://example.com'), 'coll': '2010', 'wb_prefix': '/2010/', 'request_uri': '/2010/example.com'}
>>> WbRequest.from_uri('../example.com')
{'wb_url': ('latest_replay', '', '', 'http://example.com', 'http://example.com'), 'coll': '', 'wb_prefix': '/', 'request_uri': '../example.com'}
# Abs path
>>> WbRequest.from_uri('/2010/example.com', {'wsgi.url_scheme': 'https', 'HTTP_HOST': 'localhost:8080'}, use_abs_prefix = True)
{'wb_url': ('latest_replay', '', '', 'http://example.com', 'http://example.com'), 'coll': '2010', 'wb_prefix': 'https://localhost:8080/2010/', 'request_uri': '/2010/example.com'}
# No Scheme, so stick to relative
>>> WbRequest.from_uri('/2010/example.com', {'HTTP_HOST': 'localhost:8080'}, use_abs_prefix = True)
{'wb_url': ('latest_replay', '', '', 'http://example.com', 'http://example.com'), 'coll': '2010', 'wb_prefix': '/2010/', 'request_uri': '/2010/example.com'}
If a wburl and url rewriter classes are specified, the class
also contains the url rewriter.
"""
@staticmethod
def from_uri(request_uri, env = {}, use_abs_prefix = False):
if not request_uri:
request_uri = env.get('REL_REQUEST_URI')
parts = request_uri.split('/', 2)
# Has coll prefix
if len(parts) == 3:
wb_prefix = '/' + parts[1] + '/'
wb_url_str = parts[2]
coll = parts[1]
# No Coll Prefix
elif len(parts) == 2:
wb_prefix = '/'
wb_url_str = parts[1]
coll = ''
else:
wb_prefix = '/'
wb_url_str = parts[0]
coll = ''
host_prefix = WbRequest.make_host_prefix(env) if use_abs_prefix else ''
return WbRequest(env, request_uri, wb_prefix, wb_url_str, coll, host_prefix = host_prefix)
@staticmethod
def make_host_prefix(env):
try:
return env['wsgi.url_scheme'] + '://' + env['HTTP_HOST']
host = env.get('HTTP_HOST')
if not host:
host = env['SERVER_NAME'] + ':' + env['SERVER_PORT']
return env['wsgi.url_scheme'] + '://' + host
except KeyError:
return ''
def __init__(self, env, request_uri, wb_prefix, wb_url_str, coll,
host_prefix = '',
wburl_class = WbUrl,
url_rewriter_class = UrlRewriter,
is_proxy = False):
def __init__(self, env,
request_uri=None,
rel_prefix='',
wb_url_str='/',
coll='',
host_prefix='',
use_abs_prefix=False,
wburl_class=None,
urlrewriter_class=None,
is_proxy=False):
self.env = env
self.request_uri = request_uri if request_uri else env.get('REL_REQUEST_URI')
self.host_prefix = host_prefix
self.coll = coll
if not host_prefix:
host_prefix = self.make_host_prefix(env)
self.host_prefix = host_prefix
self.rel_prefix = rel_prefix
if use_abs_prefix:
self.wb_prefix = host_prefix + rel_prefix
else:
self.wb_prefix = rel_prefix
self.wb_prefix = host_prefix + wb_prefix
if not wb_url_str:
wb_url_str = '/'
self.wb_url_str = wb_url_str
# wb_url present and not root page
if wb_url_str != '/' and wburl_class:
self.wb_url_str = wb_url_str
self.wb_url = wburl_class(wb_url_str)
self.urlrewriter = url_rewriter_class(self.wb_url, self.wb_prefix)
self.urlrewriter = urlrewriter_class(self.wb_url, self.wb_prefix)
else:
# no wb_url, just store blank wb_url
self.wb_url_str = wb_url_str
self.wb_url = None
self.urlrewriter = None
self.coll = coll
self.referrer = env.get('HTTP_REFERER')
self.is_ajax = self._is_ajax()
@ -122,24 +98,19 @@ class WbRequest:
def __repr__(self):
#return "WbRequest(env, '" + (self.wb_url) + "', '" + (self.coll) + "')"
#return str(vars(self))
varlist = vars(self)
return str({k: varlist[k] for k in ('request_uri', 'wb_prefix', 'wb_url', 'coll')})
varstr = pprint.pformat(varlist)
return varstr
#=================================================================
class WbResponse:
"""
>>> WbResponse.text_response('Test')
{'body': ['Test'], 'status_headers': StatusAndHeaders(protocol = '', statusline = '200 OK', headers = [('Content-Type', 'text/plain')])}
Represnts a pywb wsgi response object.
>>> WbResponse.text_stream(['Test', 'Another'], '404')
{'body': ['Test', 'Another'], 'status_headers': StatusAndHeaders(protocol = '', statusline = '404', headers = [('Content-Type', 'text/plain')])}
>>> WbResponse.redir_response('http://example.com/otherfile')
{'body': [], 'status_headers': StatusAndHeaders(protocol = '', statusline = '302 Redirect', headers = [('Location', 'http://example.com/otherfile')])}
Holds a status_headers object and a response iter, to be
returned to wsgi container.
"""
def __init__(self, status_headers, value = []):
self.status_headers = status_headers
self.body = value
@ -180,8 +151,3 @@ class WbResponse:
def __repr__(self):
return str(vars(self))
if __name__ == "__main__":
import doctest
doctest.testmod()

View File

@ -124,6 +124,20 @@ class TestWb:
assert resp.content_type == 'text/css'
def test_referrer_self_redirect(self):
uri = '/pywb/20140127171239/http://www.iana.org/_css/2013.1/screen.css'
host = 'somehost:8082'
referrer = 'http://' + host + uri
# capture is normally a 200
resp = self.testapp.get(uri)
assert resp.status_int == 200
# redirect causes skip of this capture, redirect to next
resp = self.testapp.get(uri, headers = [('Referer', referrer), ('Host', host)], status = 302)
assert resp.status_int == 302
def test_excluded_content(self):
resp = self.testapp.get('/pywb/http://www.iana.org/_img/bookmark_icon.ico', status = 403)
assert resp.status_int == 403