/** * Tests the canonicalizing URL's */ public function testCanonicalize() { $this->assertEquals(GSB_UrlUtil::getCanonicalizedUrl("http://host/%25%32%35"), "http://host/%25"); $this->assertEquals(GSB_UrlUtil::getCanonicalizedUrl("http://host/%25%32%35%25%32%35"), "http://host/%25%25"); $this->assertEquals(GSB_UrlUtil::getCanonicalizedUrl("http://host/%2525252525252525"), "http://host/%25"); $this->assertEquals(GSB_UrlUtil::getCanonicalizedUrl("http://host/asdf%25%32%35asd"), "http://host/asdf%25asd"); $this->assertEquals(GSB_UrlUtil::getCanonicalizedUrl("http://host/%%%25%32%35asd%%"), "http://host/%25%25%25asd%25%25"); $this->assertEquals(GSB_UrlUtil::getCanonicalizedUrl("http://www.google.com/"), "http://www.google.com/"); $this->assertEquals(GSB_UrlUtil::getCanonicalizedUrl("http://%31%36%38%2e%31%38%38%2e%39%39%2e%32%36/%2E%73%65%63%75%72%65/%77%77%77%2E%65%62%61%79%2E%63%6F%6D/"), "http://168.188.99.26/.secure/www.ebay.com/"); $this->assertEquals(GSB_UrlUtil::getCanonicalizedUrl("http://195.127.0.11/uploads/%20%20%20%20/.verify/.eBaysecure=updateuserdataxplimnbqmn-xplmvalidateinfoswqpcmlx=hgplmcx/"), "http://195.127.0.11/uploads/%20%20%20%20/.verify/.eBaysecure=updateuserdataxplimnbqmn-xplmvalidateinfoswqpcmlx=hgplmcx/"); $this->assertEquals(GSB_UrlUtil::getCanonicalizedUrl("http://host%23.com/%257Ea%2521b%2540c%2523d%2524e%25f%255E00%252611%252A22%252833%252944_55%252B"), 'http://host%23.com/~a!b@c%23d$e%25f^00&11*22(33)44_55+'); $this->assertEquals(GSB_UrlUtil::getCanonicalizedUrl("http://3279880203/blah"), "http://195.127.0.11/blah"); $this->assertEquals(GSB_UrlUtil::getCanonicalizedUrl("http://www.google.com/blah/.."), "http://www.google.com/"); $this->assertEquals(GSB_UrlUtil::getCanonicalizedUrl("www.google.com/"), "http://www.google.com/"); $this->assertEquals(GSB_UrlUtil::getCanonicalizedUrl("www.google.com"), "http://www.google.com/"); $this->assertEquals(GSB_UrlUtil::getCanonicalizedUrl("http://www.evil.com/blah#frag"), "http://www.evil.com/blah"); $this->assertEquals(GSB_UrlUtil::getCanonicalizedUrl("http://www.GOOgle.com/"), "http://www.google.com/"); $this->assertEquals(GSB_UrlUtil::getCanonicalizedUrl("http://www.google.com.../"), "http://www.google.com/"); $this->assertEquals(GSB_UrlUtil::getCanonicalizedUrl("http://www.google.com/foo\tbar\rbaz\n2"), "http://www.google.com/foobarbaz2"); $this->assertEquals(GSB_UrlUtil::getCanonicalizedUrl("http://www.google.com/q?"), "http://www.google.com/q?"); $this->assertEquals(GSB_UrlUtil::getCanonicalizedUrl("http://www.google.com/q?r?"), "http://www.google.com/q?r?"); $this->assertEquals(GSB_UrlUtil::getCanonicalizedUrl("http://www.google.com/q?r?s"), "http://www.google.com/q?r?s"); $this->assertEquals(GSB_UrlUtil::getCanonicalizedUrl("http://evil.com/foo#bar#baz"), "http://evil.com/foo"); $this->assertEquals(GSB_UrlUtil::getCanonicalizedUrl("http://evil.com/foo;"), "http://evil.com/foo;"); $this->assertEquals(GSB_UrlUtil::getCanonicalizedUrl("http://evil.com/foo?bar;"), "http://evil.com/foo?bar;"); $this->assertEquals(GSB_UrlUtil::getCanonicalizedUrl("http://€.com/"), "http://%01%80.com/"); $this->assertEquals(GSB_UrlUtil::getCanonicalizedUrl("http://notrailingslash.com"), "http://notrailingslash.com/"); $this->assertEquals(GSB_UrlUtil::getCanonicalizedUrl("http://www.gotaport.com:1234/"), "http://www.gotaport.com:1234/"); $this->assertEquals(GSB_UrlUtil::getCanonicalizedUrl(" http://www.google.com/ "), "http://www.google.com/"); $this->assertEquals(GSB_UrlUtil::getCanonicalizedUrl("http:// leadingspace.com/"), "http://%20leadingspace.com/"); $this->assertEquals(GSB_UrlUtil::getCanonicalizedUrl("http://%20leadingspace.com/"), "http://%20leadingspace.com/"); $this->assertEquals(GSB_UrlUtil::getCanonicalizedUrl("%20leadingspace.com/"), "http://%20leadingspace.com/"); $this->assertEquals(GSB_UrlUtil::getCanonicalizedUrl("https://www.securesite.com/"), "https://www.securesite.com/"); $this->assertEquals(GSB_UrlUtil::getCanonicalizedUrl("http://host.com/ab%23cd"), "http://host.com/ab%23cd"); $this->assertEquals(GSB_UrlUtil::getCanonicalizedUrl("http://host.com//twoslashes?more//slashes"), "http://host.com/twoslashes?more//slashes"); $this->assertEquals(GSB_UrlUtil::getCanonicalizedUrl("http://host.com/foo/"), "http://host.com/foo/"); $this->assertEquals(GSB_UrlUtil::getCanonicalizedUrl("http://host.com/foo?"), "http://host.com/foo?"); $this->assertEquals("http://host.com/", GSB_UrlUtil::getCanonicalizedUrl("http://host.com:")); $this->assertEquals("http://host.com/", GSB_UrlUtil::getCanonicalizedUrl("http://host.com:80")); $this->assertEquals("http://host.com/", GSB_UrlUtil::getCanonicalizedUrl("http://host.com:80/")); $this->assertEquals("https://host.com/", GSB_UrlUtil::getCanonicalizedUrl("https://host.com:443/")); /* $url = 'http://host.com/foo%3Fbar'; $expected = array( 'canonical' => $url, 'original'=> $url, 'host' => 'host.com', 'path' => '/foo', 'query' => 'bar', 'is_ip' => FALSE ); $this->assertEquals($expected, GSB_UrlUtil::canonicalize($url)); */ }
/** * Organizes an array of host keys to a map which is indexed by the host_key * value. * * NOTE: this function is made non-private so we can unit test it. * * @param array[int]array $urls * @return array[string]array */ function hostkeysToHostKeyMap($urls) { $map = array(); foreach ((array) $urls as $url) { // first canonicalize the URLs $canurl = GSB_UrlUtil::canonicalize($url); $phashes = GSB_UrlUtil::makePrefixesHashes($canurl['host'], $canurl['path'], $canurl['query'], $canurl['is_ip']); // make hostkeys $hostkeys = GSB_UrlUtil::makeHostKeyList($canurl['host'], $canurl['is_ip']); foreach ((array) $hostkeys as $hostkey) { $hostkey['url'] = $url; $hostkey['phashes'] = $phashes; $map[$hostkey['host_key']] = $hostkey; } } return $map; }