def test_seen():
c = {'Robots': {'RobotsCacheSize': 1, 'RobotsCacheTimeout': 1}}
config.set_config(c)
dl = datalayer.Datalayer()
assert not dl.seen_url(URL('http://example.com'))
dl.add_seen_url(URL('http://example.com'))
assert dl.seen_url(URL('http://example.com'))
def create_queue():
queue = asyncio.Queue()
# add a fake domain to make sure the dns doesn't send unknown hosts to a search
# note that mail.foo.com and mx.foo.com don't generally get bogus answers, it's foo.com or www.foo.com that do
for _ in range(2):
r = random.Random()
host = str(r.randrange(1000000000)) + str(r.randrange(1000000000)) + str(r.randrange(1000000000))
queue.put_nowait((URL('http://' + host + '.com'), 'fake'))
# read list of domains to query -- from alexa top million
head, tail = os.path.split(__file__)
alexa = os.path.join(head, os.pardir, 'data', 'top-1k.txt')
alexa_count = 0
try:
with open(alexa, 'r') as f:
print('Using top-1k from Alexa, expect a few failures')
for line in f:
queue.put_nowait((URL('http://'+line.strip()), 'real'))
alexa_count += 1
if alexa_count > args.count:
break
except FileNotFoundError:
# the alexa file wasn't available (it is not in the repo) so just do a few
print('Cannot find top-1k file, so all queries are www.google.com')
for _ in range(args.count):
queue.put_nowait((URL('http://www.google.com'), 'real'))
return queue
def test_facets_from_embeds():
embeds = (URL('http://example.com'), URL('http://cdn.ampproject.org'),
URL('googletagmanager.com?asdf&id=GTM-ZZZXXX&fdsa'),
URL('https://www.facebook.com/tr?foo&id=1234567890123456'))
facets = facet.facets_from_embeds(embeds)
assert facets == [('thing-google amp', True),
('thing-google tag manager', 'GTM-ZZZXXX'),
('thing-facebook events', '1234567890123456')]
def test_summarize(capsys):
c = {'Robots': {'RobotsCacheSize': 1, 'RobotsCacheTimeout': 1}}
config.set_config(c)
dl = datalayer.Datalayer()
dl.add_seen_url(URL('http://example.com'))
dl.add_seen_url(URL('http://example2.com'))
dl.summarize()
out, err = capsys.readouterr()
assert len(err) == 0
assert out.startswith('2 seen_urls')
async def test_cocrawler(capsys):
config.config(None, None)
# we have to get around the useragent checks
config.write('pytest', 'UserAgent', 'MyPrefix')
config.write('http://example.com/pytest-test-cocrawler.py', 'UserAgent',
'URL')
# and configure url_allowed
config.write('AllDomains', 'Plugins', 'url_allowed')
crawler = cocrawler.Crawler()
crawler.add_url(0, {'url': URL('http://example1.com/')})
crawler.add_url(0, {'url': URL('http://example2.com/')})
crawler.add_url(0, {'url': URL('http://example3.com/')})
assert crawler.qsize == 3
f = tempfile.NamedTemporaryFile(delete=False)
name = f.name
with open(name, 'wb') as f:
crawler.save(f)
assert crawler.qsize == 0
crawler.add_url(0, {'url': URL('http://example4.com/')})
assert crawler.qsize == 1
with open(name, 'rb') as f:
crawler.load(f)
assert crawler.qsize == 3
os.unlink(name)
assert not os.path.exists(name)
# clear out the existing capture
out, err = capsys.readouterr()
crawler.summarize()
out, err = capsys.readouterr()
assert err == ''
assert len(
out) >= 200 # not a very good test, but at least it is something
await crawler.close() # needed for smooth shutdown
def test_do_burner_work_html():
urlj = URL('http://example.com')
test_html_bytes = test_html.encode(encoding='utf-8', errors='replace')
headers = {}
links, embeds, sha1, facets = parse.do_burner_work_html(test_html, test_html_bytes, headers, url=urlj)
assert len(links) == 3
assert len(embeds) == 2
linkset = set(u.url for u in links)
embedset = set(u.url for u in embeds)
assert 'http://example.com/foo3.html' in linkset
assert 'http://example.com/foo.gif' in embedset
assert sha1 == 'sha1:3458e0857ec379ec56d4c7fb39d33c90c8b5ae93'
# as a handwave, let's expect these defective pages to also work.
test_html_bytes = test_html_no_body.encode(encoding='utf-8', errors='replace')
links, embeds, sha1, facets = parse.do_burner_work_html(test_html, test_html_bytes, headers, url=urlj)
assert len(links) == 3
assert len(embeds) == 2
test_html_bytes = test_html_no_head.encode(encoding='utf-8', errors='replace')
links, embeds, sha1, facets = parse.do_burner_work_html(test_html, test_html_bytes, headers, url=urlj)
assert len(links) == 3
assert len(embeds) == 2
test_html_bytes = test_html_no_nothing.encode(encoding='utf-8', errors='replace')
links, embeds, sha1, facets = parse.do_burner_work_html(test_html, test_html_bytes, headers, url=urlj)
assert len(links) == 3
assert len(embeds) == 2
def test_do_burner_work_html():
urlj = URL('http://example.com')
test_html_bytes = test_html.encode(encoding='utf-8', errors='replace')
headers = {}
links, embeds, sha1, facets, base = parse.do_burner_work_html(test_html, test_html_bytes, headers, url=urlj)
assert len(links) == 4
assert len(embeds) == 2
linkset = set(u.url for u in links)
embedset = set(e.url for e in embeds)
assert 'http://example.com/foo3.html' in linkset
assert 'http://example.com/foo.gif' in embedset
assert sha1 == 'sha1:cdcb087d39afd827d5d523e165a6566d65a2e9b3'
assert base is None
# as a handwave, let's expect these defective pages to also work.
test_html_bytes = test_html_no_body.encode(encoding='utf-8', errors='replace')
links, embeds, sha1, facets, base = parse.do_burner_work_html(test_html_no_body, test_html_bytes, headers, url=urlj)
assert len(links) == 3
assert len(embeds) == 2
test_html_bytes = test_html_no_head.encode(encoding='utf-8', errors='replace')
links, embeds, sha1, facets, base = parse.do_burner_work_html(test_html_no_head, test_html_bytes, headers, url=urlj)
assert len(links) == 3
assert len(embeds) == 1
test_html_bytes = test_html_no_nothing.encode(encoding='utf-8', errors='replace')
links, embeds, sha1, facets, base = parse.do_burner_work_html(test_html_no_nothing, test_html_bytes, headers, url=urlj)
assert len(links) == 3
assert len(embeds) == 1
async def test_prefetch():
url = URL('http://example.com/')
config.config(None, None)
resolver = dns.get_resolver()
iplist = await dns.prefetch(url, resolver)
assert len(iplist) > 0
iplist2 = await dns.prefetch(url, resolver)
assert iplist == iplist2
def test_cocrawler(capsys):
config = conf.config(None, None, confighome=False)
# ok, we have to get around the useragent checks
config['UserAgent']['MyPrefix'] = 'pytest'
config['UserAgent']['URL'] = 'http://example.com/pytest-test-cocrawler.py'
loop = asyncio.get_event_loop()
crawler = cocrawler.Crawler(loop, config)
crawler.add_url(0, URL('http://example1.com/'), seed=True)
crawler.add_url(0, URL('http://example2.com/'), seed=True)
crawler.add_url(0, URL('http://example3.com/'), seed=True)
assert crawler.qsize == 3
f = tempfile.NamedTemporaryFile(delete=False)
name = f.name
with open(name, 'wb') as f:
crawler.save(f)
assert crawler.qsize == 0
crawler.add_url(0, URL('http://example4.com/'), seed=True)
assert crawler.qsize == 1
with open(name, 'rb') as f:
crawler.load(f)
assert crawler.qsize == 3
os.unlink(name)
assert not os.path.exists(name)
# clear out the existing capture
out, err = capsys.readouterr()
crawler.summarize()
out, err = capsys.readouterr()
assert err == ''
assert len(out) >= 242 # not a very good test, but at least it is something
def test_setup_seeds_prefix():
seeds = {
'http://example.com/asdf', 'http://example.com/a',
'http://example.com/a', 'http://example.com/b',
'http://example.com/asdff', 'http://example2.com/a'
}
url_allowed.setup(policy='SeedsPrefix')
url_allowed.setup_seeds([URL(s) for s in seeds])
SEEDS = {'example.com': {'/a', '/b'}, 'example2.com': {'/a'}}
assert SEEDS == url_allowed.SEEDS
def test_saveload():
tf = tempfile.NamedTemporaryFile(delete=False)
name = tf.name
c = {'Robots': {'RobotsCacheSize': 1, 'RobotsCacheTimeout': 1}}
config.set_config(c)
dl = datalayer.Datalayer()
dl.add_seen_url(URL('http://example.com'))
assert dl.seen_url(URL('http://example.com'))
with open(name, 'wb') as f:
dl.save(f)
dl.add_seen_url(URL('http://example2.com'))
with open(name, 'rb') as f:
dl.load(f)
assert dl.seen_url(URL('http://example.com'))
assert not dl.seen_url(URL('http://example2.com'))
os.unlink(name)
assert not os.path.exists(name)
async def test_prefetch_dns():
url = URL('http://google.com/')
mock_url = None
resolver = aiohttp.resolver.AsyncResolver(nameservers=ns)
connector = aiohttp.connector.TCPConnector(resolver=resolver,
family=socket.AF_INET)
session = aiohttp.ClientSession(connector=connector)
# whew
iplist = await dns.prefetch_dns(url, mock_url, session)
assert len(iplist) > 0
def test_extension_allowed():
assert url_allowed.extension_allowed(URL('https://example.com/'))
assert url_allowed.extension_allowed(
URL('https://example.com/thing.with.dots/'))
assert url_allowed.extension_allowed(
URL('https://example.com/thing.with.dots'))
assert url_allowed.extension_allowed(URL('https://example.com/index.html'))
assert not url_allowed.extension_allowed(
URL('https://example.com/foo.jpg'))
assert not url_allowed.extension_allowed(
URL('https://example.com/foo.tar.gz'))
'127.0.0.1',
description='desc',
creator='test',
operator='alice')
sub.create_default_info('1.0', '0.99', '127.0.0.1')
fake_dns_result = [{
'host': '172.217.6.78'
}, {
'host': '172.217.6.78'
}, {
'host': '172.217.6.78'
}]
with mock.patch('cocrawler.warc.timestamp_now', return_value='20190215073137'):
main.write_dns(fake_dns_result, 10, URL('http://google.com'))
fake_url = 'https://www.google.com/'
fake_req_headers = [('Host', 'www.google.com')]
fake_resp_headers = [(b'Content-Type', b'text/html; charset=UTF-8')]
fake_payload = b'<html><body>Hello, world!</body></html>'
# to make sure that warcio is actually using our digest, multilate it
# this means we can't use a warc checker!
fake_digest = 'sha1:FAKE_DIGEST'
main.write_request_response_pair(fake_url,
'1.2.3.4',
fake_req_headers,
fake_resp_headers,
False,
def test_URL():
url = URL('http://www.example.com/')
assert url.url == 'http://www.example.com/'
assert list(url.urlparse) == ['http', 'www.example.com', '/', '', '', '']
assert url.netloc == 'www.example.com'
assert url.hostname == 'www.example.com'
assert url.hostname_without_www == 'example.com'
assert url.registered_domain == 'example.com'
assert url.original_frag is None
url = URL('http://www.example.com/#foo#foo')
assert url.original_frag == 'foo#foo'
url = URL('http://www.example.com/#')
assert url.original_frag is None
# canonicalization
url = URL('http://www.example.com/?')
assert url.url == 'http://www.example.com/'
url = URL('http://www.example.com')
assert url.url == 'http://www.example.com/'
url = URL('http://www.example.com/?#')
assert url.url == 'http://www.example.com/'
url = URL('http://www.example.com/foo')
assert url.url == 'http://www.example.com/foo'
url = URL('http://www.example.com/foo/')
assert url.url == 'http://www.example.com/foo/'
# urljoin
urlj1 = URL('http://www.example.com/foo/')
urlj2 = 'http://www.example.com/foo/'
url = URL('foo', urljoin=urlj1)
assert url.url == 'http://www.example.com/foo/foo'
url = URL('foo', urljoin=urlj1)
assert url.url == 'http://www.example.com/foo/foo'
url = URL('/bar', urljoin=urlj1)
assert url.url == 'http://www.example.com/bar'
url = URL('/bar', urljoin=urlj2)
assert url.url == 'http://www.example.com/bar'
url = URL('http://sub.example.com/', urljoin=urlj1)
assert url.url == 'http://sub.example.com/'
url = URL('http://sub.example.com/', urljoin=urlj2)
assert url.url == 'http://sub.example.com/'
# read-only
with pytest.raises(AttributeError):
url.url = 'foo'
def test_facets_from_embeds():
embeds = set((URL('http://example.com'), URL('http://cdn.ampproject.org')))
facets = facet.facets_from_embeds(embeds)
assert facets == [('google amp', True)]
def test_URL():
url = URL('http://www.example.com/')
assert url.url == 'http://www.example.com/'
assert list(url.urlsplit) == ['http', 'www.example.com', '/', '', '']
assert url.netloc == 'www.example.com'
assert url.hostname == 'www.example.com'
assert url.hostname_without_www == 'example.com'
assert url.registered_domain == 'example.com'
assert url.original_frag is None
url = URL('http://www.example.com/#foo#foo')
assert url.original_frag == '#foo#foo'
url = URL('http://www.example.com/#')
assert url.original_frag is None
# canonicalization
url = URL('http://www.example.com/?')
assert url.url == 'http://www.example.com/'
url = URL('http://www.example.com')
assert url.url == 'http://www.example.com/'
url = URL('http://www.example.com/?#')
assert url.url == 'http://www.example.com/'
url = URL('http://www.example.com/foo')
assert url.url == 'http://www.example.com/foo'
url = URL('http://www.example.com/foo/')
assert url.url == 'http://www.example.com/foo/'
# urljoin
urlj1 = URL('http://www.example.com/foo/')
urlj2 = 'http://www.example.com/foo/'
url = URL('foo', urljoin=urlj1)
assert url.url == 'http://www.example.com/foo/foo'
url = URL('foo', urljoin=urlj1)
assert url.url == 'http://www.example.com/foo/foo'
url = URL('/bar', urljoin=urlj1)
assert url.url == 'http://www.example.com/bar'
url = URL('/bar', urljoin=urlj2)
assert url.url == 'http://www.example.com/bar'
url = URL('http://sub.example.com/', urljoin=urlj1)
assert url.url == 'http://sub.example.com/'
url = URL('http://sub.example.com/', urljoin=urlj2)
assert url.url == 'http://sub.example.com/'
url = URL('foo', urljoin='http://example.com/subdir/') # base can cause this
assert url.url == 'http://example.com/subdir/foo'
# read-only
with pytest.raises(AttributeError):
url.url = 'foo'
# urljoin examples from RFC 3986 -- python takes care of . and ..
urlj = URL('http://a/b/c/d;p?q')
# assert URL('g:h', urljoin=urlj).url == 'g:h' # absolute url missing hostname
assert URL('g', urljoin=urlj).url == 'http://a/b/c/g'
assert URL('./g', urljoin=urlj).url == 'http://a/b/c/g'
assert URL('g/', urljoin=urlj).url == 'http://a/b/c/g/'
assert URL('/g', urljoin=urlj).url == 'http://a/g'
assert URL('//g', urljoin=urlj).url == 'http://g/' # altered because I insist on the trailing /
assert URL('?y', urljoin=urlj).url == 'http://a/b/c/d;p?y'
assert URL('g?y', urljoin=urlj).url == 'http://a/b/c/g?y'
assert URL('#s', urljoin=urlj).url == 'http://a/b/c/d;p?q' # I drop the frag
assert URL('g#s', urljoin=urlj).url == 'http://a/b/c/g' # I drop the frag
assert URL('g?y#s', urljoin=urlj).url == 'http://a/b/c/g?y' # I drop the frag
assert URL(';x', urljoin=urlj).url == 'http://a/b/c/;x'
assert URL('g;x', urljoin=urlj).url == 'http://a/b/c/g;x'
assert URL('g;x?y#s', urljoin=urlj).url == 'http://a/b/c/g;x?y' # I drop the frag
assert URL('', urljoin=urlj).url == 'http://a/b/c/d;p?q'
assert URL('.', urljoin=urlj).url == 'http://a/b/c/'
assert URL('./', urljoin=urlj).url == 'http://a/b/c/'
assert URL('..', urljoin=urlj).url == 'http://a/b/'
assert URL('../', urljoin=urlj).url == 'http://a/b/'
assert URL('../g', urljoin=urlj).url == 'http://a/b/g'
assert URL('../..', urljoin=urlj).url == 'http://a/'
assert URL('../../', urljoin=urlj).url == 'http://a/'
assert URL('../../g', urljoin=urlj).url == 'http://a/g'
def test_URL():
url = URL('http://www.example.com/')
assert url.url == 'http://www.example.com/'
assert list(url.urlsplit) == ['http', 'www.example.com', '/', '', '']
assert url.netloc == 'www.example.com'
assert url.hostname == 'www.example.com'
assert url.hostname_without_www == 'example.com'
assert url.registered_domain == 'example.com'
assert url.original_frag is None
url = URL('http://www.example.com/#foo#foo')
assert url.original_frag == '#foo#foo'
url = URL('http://www.example.com/#')
assert url.original_frag is None
# canonicalization
url = URL('http://www.example.com/?')
assert url.url == 'http://www.example.com/'
url = URL('http://www.example.com')
assert url.url == 'http://www.example.com/'
url = URL('http://www.example.com/?#')
assert url.url == 'http://www.example.com/'
url = URL('http://www.example.com/foo')
assert url.url == 'http://www.example.com/foo'
url = URL('http://www.example.com/foo/')
assert url.url == 'http://www.example.com/foo/'
# urljoin
urlj1 = URL('http://www.example.com/foo/')
urlj2 = 'http://www.example.com/foo/'
url = URL('foo', urljoin=urlj1)
assert url.url == 'http://www.example.com/foo/foo'
url = URL('foo', urljoin=urlj1)
assert url.url == 'http://www.example.com/foo/foo'
url = URL('/bar', urljoin=urlj1)
assert url.url == 'http://www.example.com/bar'
url = URL('/bar', urljoin=urlj2)
assert url.url == 'http://www.example.com/bar'
url = URL('http://sub.example.com/', urljoin=urlj1)
assert url.url == 'http://sub.example.com/'
url = URL('http://sub.example.com/', urljoin=urlj2)
assert url.url == 'http://sub.example.com/'
url = URL('foo',
urljoin='http://example.com/subdir/') # base can cause this
assert url.url == 'http://example.com/subdir/foo'
# read-only
with pytest.raises(AttributeError):
url.url = 'foo'
# urljoin examples from RFC 3986 -- python takes care of . and ..
urlj = URL('http://a/b/c/d;p?q')
# assert URL('g:h', urljoin=urlj).url == 'g:h' # absolute url missing hostname
assert URL('g', urljoin=urlj).url == 'http://a/b/c/g'
assert URL('./g', urljoin=urlj).url == 'http://a/b/c/g'
assert URL('g/', urljoin=urlj).url == 'http://a/b/c/g/'
assert URL('/g', urljoin=urlj).url == 'http://a/g'
assert URL(
'//g', urljoin=urlj
).url == 'http://g/' # altered because I insist on the trailing /
assert URL('?y', urljoin=urlj).url == 'http://a/b/c/d;p?y'
assert URL('g?y', urljoin=urlj).url == 'http://a/b/c/g?y'
assert URL('#s',
urljoin=urlj).url == 'http://a/b/c/d;p?q' # I drop the frag
assert URL('g#s', urljoin=urlj).url == 'http://a/b/c/g' # I drop the frag
assert URL('g?y#s',
urljoin=urlj).url == 'http://a/b/c/g?y' # I drop the frag
assert URL(';x', urljoin=urlj).url == 'http://a/b/c/;x'
assert URL('g;x', urljoin=urlj).url == 'http://a/b/c/g;x'
assert URL('g;x?y#s',
urljoin=urlj).url == 'http://a/b/c/g;x?y' # I drop the frag
assert URL('', urljoin=urlj).url == 'http://a/b/c/d;p?q'
assert URL('.', urljoin=urlj).url == 'http://a/b/c/'
assert URL('./', urljoin=urlj).url == 'http://a/b/c/'
assert URL('..', urljoin=urlj).url == 'http://a/b/'
assert URL('../', urljoin=urlj).url == 'http://a/b/'
assert URL('../g', urljoin=urlj).url == 'http://a/b/g'
assert URL('../..', urljoin=urlj).url == 'http://a/'
assert URL('../../', urljoin=urlj).url == 'http://a/'
assert URL('../../g', urljoin=urlj).url == 'http://a/g'
def test_special_redirect():
sr = urls.special_redirect
assert sr(URL('http://example.com/'),
URL('http://example.com/foo')) is None
assert sr(URL('http://example.com/'),
URL('https://example.com/foo')) is None
assert sr(URL('http://example.com/'),
URL('https://www.example.com/foo')) is None
assert sr(URL('http://example.com/'),
URL('http://example.com/?foo=1')) is None
assert sr(URL('http://example.com/'),
URL('http://example.com/bar?foo=1')) is None
url1 = URL('http://example.com/')
assert sr(url1, url1) == 'same'
assert sr(url1, URL('https://example.com/')) == 'tohttps'
assert sr(url1, URL('http://www.example.com/')) == 'towww'
assert sr(url1, URL('https://www.example.com/')) == 'towww+tohttps'
url2str = 'http://www.example.com/'
url2 = URL(url2str)
assert sr(url2, URL('https://www.example.com/')) == 'tohttps'
assert sr(url2, URL('http://example.com/')) == 'tononwww'
assert sr(url2, URL('https://example.com/')) == 'tononwww+tohttps'
assert sr(url2str, 'https://www.example.com/') == 'tohttps'
assert sr(url2str, 'http://example.com/') == 'tononwww'
assert sr(url2str, 'https://example.com/') == 'tononwww+tohttps'
url3 = URL('https://www.example.com/')
assert sr(url3, URL('http://www.example.com/')) == 'tohttp'
assert sr(url3, URL('https://example.com/')) == 'tononwww'
assert sr(url3, URL('http://example.com/')) == 'tononwww+tohttp'
url4 = URL('https://example.com/')
assert sr(url4, URL('http://www.example.com/')) == 'towww+tohttp'
url5 = URL('https://example.com/foo')
url6 = URL('https://example.com/foo/')
assert sr(url5, url6) == 'addslash'
assert sr(url6, url5) == 'removeslash'
def test_scheme_allowed():
assert url_allowed.scheme_allowed(URL('http://example.com'))
assert url_allowed.scheme_allowed(URL('https://example.com'))
assert not url_allowed.scheme_allowed(URL('ftp://example.com'))
def test_url_allowed():
assert not url_allowed.url_allowed(URL('ftp://example.com'))
url_allowed.setup(policy='SeedsDomain')
url_allowed.setup_seeds([URL('http://example.com')])
assert url_allowed.url_allowed(URL('http://example.com'))
assert url_allowed.url_allowed(URL('http://www.example.com'))
assert url_allowed.url_allowed(URL('http://sub.example.com'))
url_allowed.setup(policy='SeedsHostname')
url_allowed.setup_seeds([URL('http://example.com')])
assert url_allowed.url_allowed(URL('http://example.com'))
assert url_allowed.url_allowed(URL('http://www.example.com'))
assert not url_allowed.url_allowed(URL('http://sub.example.com'))
url_allowed.setup(policy='SeedsPrefix')
url_allowed.setup_seeds([URL('http://example.com/prefix1')])
url_allowed.setup_seeds([URL('http://example2.com/prefix2/')])
assert not url_allowed.url_allowed(URL('http://example.com'))
assert url_allowed.url_allowed(URL('http://www.example.com/prefix11'))
assert not url_allowed.url_allowed(URL('http://example2.com'))
assert not url_allowed.url_allowed(URL('http://www.example2.com/prefix21'))
assert not url_allowed.url_allowed(URL('http://www.example2.com/prefix2'))
assert url_allowed.url_allowed(URL('http://www.example2.com/prefix2/'))
assert url_allowed.url_allowed(URL('http://www.example2.com/prefix2/foo'))
url_allowed.setup(policy='OnlySeeds')
url_allowed.setup_seeds([URL('http://example.com')])
assert url_allowed.url_allowed(URL('http://example.com'))
assert not url_allowed.url_allowed(URL('http://example.com/foo'))
url_allowed.setup(policy='AllDomains')
url_allowed.setup_seeds([URL('http://example.com')])
assert url_allowed.url_allowed(URL('http://example.com'))
assert url_allowed.url_allowed(URL('http://exa2mple.com'))
assert url_allowed.url_allowed(URL('http://exa3mple.com'))