Use document encoding when parsing url in href setter (#33825)

* Implement encoding-parse-a-url algorithm

Signed-off-by: Simon Wülker <simon.wuelker@arcor.de>

* Parse url with encoding when setting href on HtmlAnchorElement

Signed-off-by: Simon Wülker <simon.wuelker@arcor.de>

* Update WPT expectations

Signed-off-by: Simon Wülker <simon.wuelker@arcor.de>

---------

Signed-off-by: Simon Wülker <simon.wuelker@arcor.de>

components/script/dom/document.rs

@@ -766,7 +766,7 @@ impl Document {
         &self.origin
     }
 
-    // https://dom.spec.whatwg.org/#concept-document-url
+    /// <https://dom.spec.whatwg.org/#concept-document-url>
     pub fn url(&self) -> ServoUrl {
         self.url.borrow().clone()
     }
@@ -775,7 +775,7 @@ impl Document {
         *self.url.borrow_mut() = url;
     }
 
-    // https://html.spec.whatwg.org/multipage/#fallback-base-url
+    /// <https://html.spec.whatwg.org/multipage/#fallback-base-url>
     pub fn fallback_base_url(&self) -> ServoUrl {
         let document_url = self.url();
         if let Some(browsing_context) = self.browsing_context() {
@@ -800,7 +800,7 @@ impl Document {
         document_url
     }
 
-    // https://html.spec.whatwg.org/multipage/#document-base-url
+    /// <https://html.spec.whatwg.org/multipage/#document-base-url>
     pub fn base_url(&self) -> ServoUrl {
         match self.base_element() {
             // Step 1.
@@ -3019,6 +3019,30 @@ impl Document {
     pub(crate) fn status_code(&self) -> Option<u16> {
         self.status_code
     }
+
+    /// <https://html.spec.whatwg.org/multipage/#encoding-parsing-a-url>
+    pub fn encoding_parse_a_url(&self, url: &str) -> Result<ServoUrl, url::ParseError> {
+        // NOTE: This algorithm is defined for both Document and environment settings objects.
+        // This implementation is only for documents.
+
+        // Step 1. Let encoding be UTF-8.
+        // Step 2. If environment is a Document object, then set encoding to environment's character encoding.
+        let encoding = self.encoding.get();
+
+        // Step 3. Otherwise, if environment's relevant global object is a Window object, set encoding to environment's
+        // relevant global object's associated Document's character encoding.
+
+        // Step 4. Let baseURL be environment's base URL, if environment is a Document object;
+        // otherwise environment's API base URL.
+        let base_url = self.base_url();
+
+        // Step 5. Return the result of applying the URL parser to url, with baseURL and encoding.
+        url::Url::options()
+            .base_url(Some(base_url.as_url()))
+            .encoding_override(Some(&|s| encoding.encode(s).0))
+            .parse(url)
+            .map(ServoUrl::from)
+    }
 }
 
 fn is_character_value_key(key: &Key) -> bool {

components/script/dom/htmlanchorelement.rs

@@ -76,15 +76,29 @@ impl HTMLAnchorElement {
         )
     }
 
-    // https://html.spec.whatwg.org/multipage/#concept-hyperlink-url-set
+    /// <https://html.spec.whatwg.org/multipage/#concept-hyperlink-url-set>
     fn set_url(&self) {
+        // Step 1. Set this element's url to null.
+        *self.url.borrow_mut() = None;
+
+        // Step 2. If this element's href content attribute is absent, then return.
         let attribute = self
             .upcast::<Element>()
             .get_attribute(&ns!(), &local_name!("href"));
-        *self.url.borrow_mut() = attribute.and_then(|attribute| {
+
+        let Some(attribute) = attribute else {
+            return;
+        };
+
+        // Step 3. Let url be the result of encoding-parsing a URL given this element's
+        // href content attribute's value, relative to this element's node document.
         let document = document_from_node(self);
-            document.base_url().join(&attribute.value()).ok()
+        let url = document.encoding_parse_a_url(&attribute.value());
-        });
+
+        // Step 4. If url is not failure, then set this element's url to url.
+        if let Ok(url) = url {
+            *self.url.borrow_mut() = Some(url);
+        }
     }
 
     // https://html.spec.whatwg.org/multipage/#reinitialise-url

tests/wpt/meta/url/percent-encoding.window.js.ini

@@ -1,10 +1,4 @@
 [percent-encoding.window.html]
-  [Input † with encoding windows-1252]
-    expected: FAIL
-
-  [Input − with encoding shift_jis]
-    expected: FAIL
-
   [Input \x0eA with encoding iso-2022-jp]
     expected: FAIL
 
@@ -14,11 +8,5 @@
   [Input † with encoding big5]
     expected: FAIL
 
-  [Input † with encoding euc-kr]
-    expected: FAIL
-
-  [Input ‾\\ with encoding iso-2022-jp]
-    expected: FAIL
-
   [Input U+d800 with encoding windows-1252]
     expected: FAIL