Revert r239503.
[WebKit-https.git] / LayoutTests / fast / url / host.html
index 32082c2..bb81371 100644 (file)
-<!DOCTYPE HTML PUBLIC "-//IETF//DTD HTML//EN">
+<!DOCTYPE html>
 <html>
 <head>
-<link rel="stylesheet" href="../js/resources/js-test-style.css">
-<script src="../js/resources/js-test-pre.js"></script>
+<script src="../../resources/js-test-pre.js"></script>
 <script src="resources/utilities.js"></script>
 </head>
 <body>
-<p id="description"></p>
-<div id="console"></div>
-<script src="script-tests/host.js"></script>
-<script src="../js/resources/js-test-post.js"></script>
+<script>
+description("Canonicalization of host names.");
+
+cases = [ 
+  // Basic canonicalization, uppercase should be converted to lowercase.
+  ["GoOgLe.CoM", "google.com"],
+  // Spaces and some other characters should be escaped.
+  ["Goo%20 goo%7C|.com", "goo%20%20goo%7C%7C.com"],
+  // Exciting different types of spaces!
+  ["GOO\u00a0\u3000goo.com", "goo%20%20goo.com"],
+  // Other types of space (no-break, zero-width, zero-width-no-break) are
+  // name-prepped away to nothing.
+  ["GOO\u200b\u2060\ufeffgoo.com", "googoo.com"],
+  // Ideographic full stop (full-width period for Chinese, etc.) should be
+  // treated as a dot.
+  ["www.foo\u3002" + "bar.com", "www.foo.bar.com"],
+  // Invalid unicode characters should fail...
+  // ...In wide input, ICU will barf and we'll end up with the input as
+  //    escaped UTF-8 (the invalid character should be replaced with the
+  //    replacement character).
+  ["\ufdd0zyx.com", "%EF%BF%BDzyx.com"],
+  // ...This is the same as previous but with with escaped.
+  ["%ef%b7%90zyx.com", "%EF%BF%BDzyx.com"],
+  // Test name prepping, fullwidth input should be converted to ASCII and NOT
+  // IDN-ized. This is "Go" in fullwidth UTF-8/UTF-16.
+  ["\uff27\uff4f.com", "go.com"],
+  // Test that fullwidth escaped values are properly name-prepped,
+  // then converted or rejected.
+  // ...%41 in fullwidth = 'A' (also as escaped UTF-8 input)
+  ["\uff05\uff14\uff11.com", "a.com"],
+  ["%ef%bc%85%ef%bc%94%ef%bc%91.com", "a.com"],
+  // ...%00 in fullwidth should fail (also as escaped UTF-8 input)
+  ["\uff05\uff10\uff10.com", "%00.com"],
+  ["%ef%bc%85%ef%bc%90%ef%bc%90.com", "%00.com"],
+  // Basic IDN support, UTF-8 and UTF-16 input should be converted to IDN
+  ["\u4f60\u597d\u4f60\u597d", "xn--6qqa088eba"],
+  // Mixed UTF-8 and escaped UTF-8 (narrow case) and UTF-16 and escaped
+  // UTF-8 (wide case). The output should be equivalent to the true wide
+  // character input above).
+  ["%E4%BD%A0%E5%A5%BD\u4f60\u597d", "xn--6qqa088eba"],
+  // Invalid escaped characters should fail and the percents should be
+  // escaped.
+  ["%zz%66%a", "%25zzf%25a"],
+  // If we get an invalid character that has been escaped.
+  ["%25", "%25"],
+  ["hello%00", "hello%00"],
+  // Escaped numbers should be treated like IP addresses if they are.
+  ["%30%78%63%30%2e%30%32%35%30.01", "192.168.0.1"],
+  ["%30%78%63%30%2e%30%32%35%30.01%2e", "192.168.0.1"],
+  // Invalid escaping should trigger the regular host error handling.
+  ["%3g%78%63%30%2e%30%32%35%30%2E.01", "%253gxc0.0250..01"],
+  // Something that isn't exactly an IP should get treated as a host and
+  // spaces escaped.
+  ["192.168.0.1 hello", "192.168.0.1%20hello"],
+  // Fullwidth and escaped UTF-8 fullwidth should still be treated as IP.
+  // These are "0Xc0.0250.01" in fullwidth.
+  ["\uff10\uff38\uff43\uff10\uff0e\uff10\uff12\uff15\uff10\uff0e\uff10\uff11", "192.168.0.1"],
+  // Broken IP addresses get marked as such.
+  ["192.168.0.257", "192.168.0.257"],
+  ["[google.com]", "[google.com]"],
+  // Cyrillic letter followed buy ( should return punicode for ( escaped before punicode string was created. I.e.
+  // if ( is escaped after punicode is created we would get xn--%28-8tb (incorrect).
+  ["\u0442(", "xn--%28-7ed"],
+  ["go\\\\@ogle.com","go/@ogle.com"],
+  ["go/@ogle.com","go/@ogle.com"],
+  ["www.lookout.net::==80::==443::","www.lookout.net::%3D%3D80::%3D%3D443:"],
+  ["www.lookout.net::80::443","www.lookout.net::80::443"],
+  // From http://eaea.sirdarckcat.net/uritest.html
+  ["\\./","./"],
+  ["//:@/","/"],
+  ["\\google.com/foo","google.com/foo"],
+  ["\\\\google.com/foo","google.com/foo"],
+  ["//asdf@/","asdf@/"],
+  ["//:81",":81"],
+  ["://","://"],
+  ["c:","c"],
+  ["xxxx:","xxxx"],
+  [".:.",".:."],
+  ["////@google.com/","google.com/"],
+  ["@google.com","google.com"],
+  ["quip\u2010apple.com", "xn--quipapple-y79d.com"],
+  ["quip\u2011apple.com", "xn--quipapple-y79d.com"],
+  ["quip\u2212apple.com", "xn--quipapple-tf4e.com"],
+  ["iclou\uA771.com", "xn--iclou-rl3s.com"],
+  ["g\u1ECD\u1ECDgle.com", "xn--ggle-gx5aa.com"]
+];
+
+for (var i = 0; i < cases.length; ++i) {
+  test_vector = cases[i][0];
+  expected_result = cases[i][1];
+  shouldBe("canonicalize('http://" + test_vector + "/')",
+           "'http://" + expected_result + "/'");
+}
+</script>
+<script src="../../resources/js-test-post.js"></script>
 </body>
 </html>