Spec update: treat URLs without a special scheme differently

This matches whatwg/url#185.
jsdom · Dec 29, 2016 · ecb808e · ecb808e
1 parent 9096654
commit ecb808e
Show file tree

Hide file tree

Showing 4 changed files with 38 additions and 12 deletions.
diff --git a/README.md b/README.md
@@ -23,7 +23,8 @@ The following methods are exported for use by places like jsdom that need to imp
 - [Serialize an integer](https://url.spec.whatwg.org/#serialize-an-integer): `serializeInteger(number)`
 - [Origin](https://url.spec.whatwg.org/#concept-url-origin) [Unicode serializer](https://html.spec.whatwg.org/multipage/browsers.html#unicode-serialisation-of-an-origin): `serializeURLToUnicodeOrigin(urlRecord)`
 - [Set the username](https://url.spec.whatwg.org/#set-the-username): `setTheUsername(urlRecord, usernameString)`
-- [Set the password](https://url.spec.whatwg.org/#set-the-password): `setThePassword(urlRecord, passwordString)`.
+- [Set the password](https://url.spec.whatwg.org/#set-the-password): `setThePassword(urlRecord, passwordString)`
+- [Is special](https://url.spec.whatwg.org/#is-special): `isSpecial(urlRecord)`
 
 The `stateOverride` parameter is one of the following strings:
 

diff --git a/lib/URL-impl.js b/lib/URL-impl.js
@@ -140,6 +140,10 @@ exports.implementation = class URLImpl {
       return this._url.path[0];
     }
 
+    if (!usm.isSpecial(this._url) && this._url.path.length === 0) {
+      return "";
+    }
+
     return "/" + this._url.path.join("/");
   }
 

diff --git a/scripts/get-latest-platform-tests.js b/scripts/get-latest-platform-tests.js
@@ -15,7 +15,7 @@ const request = require("request");
 // 1. Go to https://github.com/w3c/web-platform-tests/tree/master/url
 // 2. Press "y" on your keyboard to get a permalink
 // 3. Copy the commit hash
-const commitHash = "5be497b5f5a7036e26dc14739aa8d42f643cf94f";
+const commitHash = "2c87b7d14b07b9383b837278e475d202969f6b64";
 
 const sourceURL = `https://raw.githubusercontent.com/w3c/web-platform-tests/${commitHash}/url/urltestdata.json`;
 const setterSourceURL = `https://raw.githubusercontent.com/w3c/web-platform-tests/${commitHash}/url/setters_tests.json`;

diff --git a/src/url-state-machine.js b/src/url-state-machine.js
@@ -390,6 +390,23 @@ function parseHost(input, isUnicode) {
   return isUnicode ? tr46.toUnicode(asciiDomain, false).domain : asciiDomain;
 }
 
+function parseURLHost(input, isSpecial) {
+  if (isSpecial) {
+    if (input === "") {
+      return failure;
+    }
+
+    return parseHost(input);
+  }
+
+  let output = "";
+  const decoded = punycode.ucs2.decode(input);
+  for (let i = 0; i < decoded.length; ++i) {
+    output += encodeChar(decoded[i], isSimpleEncode);
+  }
+  return output;
+}
+
 function findLongestZeroSequence(arr) {
   let maxIdx = null;
   let maxLen = 1; // only find elements > 1
@@ -750,11 +767,7 @@ URLStateMachine.prototype["parse authority"] = function parseAuthority(c, cStr)
 URLStateMachine.prototype["parse hostname"] =
 URLStateMachine.prototype["parse host"] = function parseHostName(c, cStr) {
   if (c === p(":") && !this.arrFlag) {
-    if (isSpecial(this.url) && this.buffer === "") {
-      return failure;
-    }
-
-    const host = parseHost(this.buffer);
+    const host = parseURLHost(this.buffer, isSpecial(this.url));
     if (host === failure) {
       return failure;
     }
@@ -768,11 +781,8 @@ URLStateMachine.prototype["parse host"] = function parseHostName(c, cStr) {
   } else if (isNaN(c) || c === p("/") || c === p("?") || c === p("#") ||
              (isSpecial(this.url) && c === p("\\"))) {
     --this.pointer;
-    if (isSpecial(this.url) && this.buffer === "") {
-      return failure;
-    }
 
-    const host = parseHost(this.buffer);
+    const host = parseURLHost(this.buffer, isSpecial(this.url));
     if (host === failure) {
       return failure;
     }
@@ -1078,7 +1088,13 @@ function serializeURL(url, excludeFragment) {
     if (url.username !== "" || url.password !== null) {
       output += "@";
     }
-    output += serializeHost(url.host);
+
+    if (isSpecial(url)) {
+      output += serializeHost(url.host);
+    } else {
+      output += url.host;
+    }
+
     if (url.port !== null) {
       output += ":" + url.port;
     }
@@ -1089,6 +1105,9 @@ function serializeURL(url, excludeFragment) {
   if (url.cannotBeABaseURL) {
     output += url.path[0];
   } else {
+    if (!isSpecial(url) && url.path.length === 0) {
+      return output;
+    }
     output += "/" + url.path.join("/");
   }
 
@@ -1196,3 +1215,5 @@ module.exports.parseURL = function (input, options) {
   // We don't handle blobs, so this just delegates:
   return module.exports.basicURLParse(input, { baseURL: options.baseURL, encodingOverride: options.encodingOverride });
 };
+
+module.exports.isSpecial = isSpecial;