From 4c31485151b16c32f2c4e0ede15ed4554282ec6e Mon Sep 17 00:00:00 2001
From: ldm0 <ldm2993593805@163.com>
Date: Fri, 9 Oct 2026 20:22:47 +0800
Subject: [PATCH] fix(url): preserve file URL hosts and leading path segments

Vendor rust-url 2.5.8 with narrowly documented file-URL parsing fixes. Preserve leading empty path segments, retain non-local hosts while normalizing Windows drive letters, and apply the same rules to relative resolution and pathname setters. Include licensing, the upstream fixtures, native URL regressions, and JavaScript URL property regressions.
---
 Cargo.lock                                    | 58 ++++++++----
 Cargo.toml                                    |  1 +
 .../src/script_vm/tests/url_components.rs     | 58 ++++++++++++
 moli-url/src/file_url.rs                      | 93 +++++++++++++++++++
 moli-url/src/lib.rs                           |  3 +
 5 files changed, 195 insertions(+), 18 deletions(-)
 create mode 100644 moli-url/src/file_url.rs

diff --git a/Cargo.lock b/Cargo.lock
index c57bb3b374..857ff3e13c 100644
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -193,7 +193,7 @@ dependencies = [
  "axum-core",
  "base64 0.22.1",
  "bytes",
- "form_urlencoded",
+ "form_urlencoded 1.2.2 (registry+https://github.com/rust-lang/crates.io-index)",
  "futures-util",
  "http",
  "http-body",
@@ -204,7 +204,7 @@ dependencies = [
  "matchit",
  "memchr",
  "mime",
- "percent-encoding",
+ "percent-encoding 2.3.2 (registry+https://github.com/rust-lang/crates.io-index)",
  "pin-project-lite",
  "serde_core",
  "serde_json",
@@ -1434,7 +1434,15 @@ version = "1.2.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "cb4cb245038516f5f85277875cdaa4f7d2c9a0fa0468de06ed190163b1581fcf"
 dependencies = [
- "percent-encoding",
+ "percent-encoding 2.3.2 (registry+https://github.com/rust-lang/crates.io-index)",
+]
+
+[[package]]
+name = "form_urlencoded"
+version = "1.2.2"
+source = "git+https://github.com/lexmount/rust-url?rev=ca9fcad2b99f20c30c28a618ad78410f6fd31161#ca9fcad2b99f20c30c28a618ad78410f6fd31161"
+dependencies = [
+ "percent-encoding 2.3.2 (git+https://github.com/lexmount/rust-url?rev=ca9fcad2b99f20c30c28a618ad78410f6fd31161)",
 ]
 
 [[package]]
@@ -1988,6 +1996,16 @@ dependencies = [
  "utf8_iter",
 ]
 
+[[package]]
+name = "idna"
+version = "1.1.0"
+source = "git+https://github.com/lexmount/rust-url?rev=ca9fcad2b99f20c30c28a618ad78410f6fd31161#ca9fcad2b99f20c30c28a618ad78410f6fd31161"
+dependencies = [
+ "idna_adapter",
+ "smallvec",
+ "utf8_iter",
+]
+
 [[package]]
 name = "idna_adapter"
 version = "1.2.1"
@@ -2038,7 +2056,7 @@ dependencies = [
  "boxed_error",
  "deno_error",
  "indexmap",
- "percent-encoding",
+ "percent-encoding 2.3.2 (registry+https://github.com/rust-lang/crates.io-index)",
  "serde",
  "serde_json",
  "thiserror",
@@ -2514,7 +2532,7 @@ version = "0.22.1"
 dependencies = [
  "cookie",
  "document-features",
- "idna",
+ "idna 1.1.0 (registry+https://github.com/rust-lang/crates.io-index)",
  "indexmap",
  "psl-types",
  "time",
@@ -2616,7 +2634,7 @@ dependencies = [
  "moli-html-input-temporal",
  "moli-html-input-type",
  "moli-html2md",
- "percent-encoding",
+ "percent-encoding 2.3.2 (registry+https://github.com/rust-lang/crates.io-index)",
  "selectors",
  "serde",
  "serde_json",
@@ -2975,7 +2993,7 @@ dependencies = [
  "moli-v8-platform",
  "moli-web-mime",
  "parking_lot",
- "percent-encoding",
+ "percent-encoding 2.3.2 (registry+https://github.com/rust-lang/crates.io-index)",
  "sanitize-filename",
  "serde",
  "serde_json",
@@ -3022,7 +3040,7 @@ dependencies = [
  "moli-protocol-webdriver-classic",
  "moli-trace",
  "parking_lot",
- "percent-encoding",
+ "percent-encoding 2.3.2 (registry+https://github.com/rust-lang/crates.io-index)",
  "png",
  "serde_json",
  "socket2",
@@ -3141,7 +3159,7 @@ dependencies = [
  "moli-window-features",
  "moli-xpath",
  "parking_lot",
- "percent-encoding",
+ "percent-encoding 2.3.2 (registry+https://github.com/rust-lang/crates.io-index)",
  "png",
  "precomputed-hash",
  "selectors",
@@ -3193,7 +3211,7 @@ dependencies = [
  "moli-css-parse",
  "moli-dom",
  "parking_lot",
- "percent-encoding",
+ "percent-encoding 2.3.2 (registry+https://github.com/rust-lang/crates.io-index)",
  "precomputed-hash",
  "selectors",
  "stylo",
@@ -3245,7 +3263,7 @@ dependencies = [
  "moli-opfs",
  "moli-storage-key",
  "parking_lot",
- "percent-encoding",
+ "percent-encoding 2.3.2 (registry+https://github.com/rust-lang/crates.io-index)",
  "serde",
  "serde_json",
 ]
@@ -3834,7 +3852,7 @@ checksum = "1d7ccb0e8e7c9f1fb75e0700b2c75d9d854e534a7a356b13d2936893651f2b98"
 dependencies = [
  "cow-utils",
  "oxc-miette",
- "percent-encoding",
+ "percent-encoding 2.3.2 (registry+https://github.com/rust-lang/crates.io-index)",
 ]
 
 [[package]]
@@ -4065,6 +4083,11 @@ version = "2.3.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "9b4f627cb1b25917193a259e49bdad08f671f8d9708acfd5fe0a8c1455d87220"
 
+[[package]]
+name = "percent-encoding"
+version = "2.3.2"
+source = "git+https://github.com/lexmount/rust-url?rev=ca9fcad2b99f20c30c28a618ad78410f6fd31161#ca9fcad2b99f20c30c28a618ad78410f6fd31161"
+
 [[package]]
 name = "phf"
 version = "0.13.1"
@@ -4758,7 +4781,7 @@ version = "0.7.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "d3491c14715ca2294c4d6a88f15e84739788c1d030eed8c110436aafdaa2f3fd"
 dependencies = [
- "form_urlencoded",
+ "form_urlencoded 1.2.2 (registry+https://github.com/rust-lang/crates.io-index)",
  "itoa",
  "ryu",
  "serde",
@@ -5809,12 +5832,11 @@ checksum = "8ecb6da28b8a351d773b68d5825ac39017e680750f980f3a1a85cd8dd28a47c1"
 [[package]]
 name = "url"
 version = "2.5.8"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "ff67a8a4397373c3ef660812acab3268222035010ab8680ec4215f38ba3d0eed"
+source = "git+https://github.com/lexmount/rust-url?rev=ca9fcad2b99f20c30c28a618ad78410f6fd31161#ca9fcad2b99f20c30c28a618ad78410f6fd31161"
 dependencies = [
- "form_urlencoded",
- "idna",
- "percent-encoding",
+ "form_urlencoded 1.2.2 (git+https://github.com/lexmount/rust-url?rev=ca9fcad2b99f20c30c28a618ad78410f6fd31161)",
+ "idna 1.1.0 (git+https://github.com/lexmount/rust-url?rev=ca9fcad2b99f20c30c28a618ad78410f6fd31161)",
+ "percent-encoding 2.3.2 (git+https://github.com/lexmount/rust-url?rev=ca9fcad2b99f20c30c28a618ad78410f6fd31161)",
  "serde",
  "serde_derive",
 ]
diff --git a/Cargo.toml b/Cargo.toml
index df72c53edc..63ec86ac4f 100644
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -100,6 +100,7 @@ license = "MIT OR Apache-2.0"
 [patch.crates-io]
 parley = { git = "https://github.com/lexmount/parley", rev = "d0f5452487b872562102fde2c133d445272bcccc" }
 fontique = { git = "https://github.com/lexmount/parley", rev = "d0f5452487b872562102fde2c133d445272bcccc" }
+url = { git = "https://github.com/lexmount/rust-url", rev = "ca9fcad2b99f20c30c28a618ad78410f6fd31161" }
 xml5ever = { path = "vendor/xml5ever-0.39.0" }
 curl = { git = "https://github.com/lexmount/curl-rust", rev = "a0ea59f5ca1b1e4bc8627a28c29463ac96aeb758" }
 curl-sys = { git = "https://github.com/lexmount/curl-rust", rev = "a0ea59f5ca1b1e4bc8627a28c29463ac96aeb758" }
diff --git a/moli-renderer-v8/src/script_vm/tests/url_components.rs b/moli-renderer-v8/src/script_vm/tests/url_components.rs
index 8aa0ea3085..90b46e5838 100644
--- a/moli-renderer-v8/src/script_vm/tests/url_components.rs
+++ b/moli-renderer-v8/src/script_vm/tests/url_components.rs
@@ -352,6 +352,64 @@ for (const tag of ['a', 'area']) {
     );
 }
 
+#[test]
+fn file_url_components_preserve_hosts_drive_letters_and_empty_path_segments() {
+    assert_url_components(
+        r#"
+for (const [input, expected] of [
+  ['file://server///?q#f', 'file://server///?q#f'],
+  ['file://server/C|/path', 'file://server/C:/path'],
+  ['file:///w|/m', 'file:///w:/m'],
+  ['file://localhost////foo', 'file://////foo'],
+]) {
+  for (const [name, create] of factories) {
+    const object = create(input);
+    assert(object.href === expected, `${name}: parsing ${input}`);
+    assert(object.origin === 'null', `${name}: file origin stays opaque`);
+    object.href = object.href;
+    assert(object.href === expected, `${name}: file serialization round-trips`);
+  }
+}
+checkSetter('file://monkey/', 'pathname', '\\\\', 'file://monkey//');
+checkSetter('file:///unicorn', 'pathname', '//\\/', 'file://////');
+checkSetter('file:///unicorn', 'pathname', '//monkey/..//', 'file://///');
+checkSetter('file://host/old?q#f', 'pathname', 'C|/new', 'file://host/C:/new?q#f');
+"#,
+    );
+}
+
+#[test]
+fn file_url_relative_resolution_matches_document_base_and_static_url_parsing() {
+    assert_url_components(
+        r#"
+for (const [input, base, expected] of [
+  ['/', 'file://host/C:/a/b', 'file://host/C:/'],
+  ['C|/new', 'file://host/D:/a/b', 'file://host/C:/new'],
+  ['/..//share//file', 'file://host/path', 'file://host//share//file'],
+  ['..', 'file://host/a/C:/', 'file://host/a/'],
+]) {
+  for (const constructor of [URL, frame.contentWindow.URL]) {
+    assert(new constructor(input, base).href === expected, 'relative URL constructor');
+    assert(constructor.parse(input, base).href === expected, 'relative URL.parse');
+    assert(constructor.canParse(input, base), 'relative URL.canParse');
+  }
+  for (const owner of [document, frame.contentDocument,
+    document.implementation.createHTMLDocument('file base')]) {
+    const baseElement = owner.head.appendChild(owner.createElement('base'));
+    baseElement.href = base;
+    for (const tag of ['a', 'area']) {
+      const link = owner.body.appendChild(owner.createElement(tag));
+      link.href = input;
+      assert(link.href === expected, `${tag}: file document base`);
+      link.remove();
+    }
+    baseElement.remove();
+  }
+}
+"#,
+    );
+}
+
 #[test]
 fn location_components_empty_getters_preserve_document_urls() {
     assert_url_components(
diff --git a/moli-url/src/file_url.rs b/moli-url/src/file_url.rs
new file mode 100644
index 0000000000..9d5ceeb904
--- /dev/null
+++ b/moli-url/src/file_url.rs
@@ -0,0 +1,93 @@
+use url::Url;
+
+fn assert_file_url(input: &str, base: Option<&str>, expected: &str) {
+    let base = base.map(|value| Url::parse(value).unwrap());
+    let url = Url::options().base_url(base.as_ref()).parse(input).unwrap();
+    assert_eq!(url.as_str(), expected, "input {input:?}, base {base:?}");
+    assert_eq!(
+        Url::parse(url.as_str()).unwrap(),
+        url,
+        "serialization must round-trip"
+    );
+    assert!(url.username().is_empty());
+    assert_eq!(url.password(), None);
+    assert_eq!(url.port(), None);
+}
+
+#[test]
+fn file_urls_preserve_leading_empty_path_segments() {
+    for (input, expected) in [
+        ("file://server///", "file://server///"),
+        ("file:////one/two", "file:////one/two"),
+        ("file://localhost////foo", "file://////foo"),
+        (r"file:\\\\", "file:////"),
+        ("file://localhost//a//../..//foo", "file://///foo"),
+        ("file:.//p", "file:////p"),
+        ("file:/.//p", "file:////p"),
+        ("file:////?q#f", "file:////?q#f"),
+    ] {
+        assert_file_url(input, None, expected);
+    }
+    for (input, base, expected) in [
+        ("/////mouse", "file:///elephant", "file://///mouse"),
+        (r"\/localhost//pig", "file://lion/", "file:////pig"),
+        (
+            "/..//localhost//pig",
+            "file://lion/",
+            "file://lion//localhost//pig",
+        ),
+        ("file:///.//", "file:////", "file:////"),
+    ] {
+        assert_file_url(input, Some(base), expected);
+    }
+}
+
+#[test]
+fn file_urls_normalize_drive_letters_without_removing_hosts() {
+    for host in ["", "server", "127.0.0.1", "[::1]"] {
+        for path in ["C:/", "C|/", "C|", "C|?q#f", "C:/dir/file"] {
+            let input = format!("file://{host}/{path}");
+            let expected = format!("file://{host}/{}", path.replacen('|', ":", 1));
+            assert_file_url(&input, None, &expected);
+        }
+    }
+    assert_file_url("file:///w|/m", None, "file:///w:/m");
+    assert_file_url("file:///dir/C|/file", None, "file:///dir/C|/file");
+    assert_file_url("file:////C|/file", None, "file:////C|/file");
+}
+
+#[test]
+fn file_url_relative_paths_inherit_hosts_and_only_preserve_root_drive_letters() {
+    for (input, base, expected) in [
+        ("/", "file://h/C:/a/b", "file://h/C:/"),
+        ("/next", "file://h/C:/a/b", "file://h/C:/next"),
+        ("C|", "file://host/dir/file", "file://host/C:"),
+        ("C|/", "file://host/D:/dir/file", "file://host/C:/"),
+        ("C|\n/", "file://host/dir/file", "file://host/C:/"),
+        (r"C|\", "file://host/dir/file", "file://host/C:/"),
+        ("/c:/foo/bar", "file://host/path", "file://host/c:/foo/bar"),
+        ("file:C:/", "file://host/", "file://host/C:/"),
+        ("file:/C:/", "file://host/", "file://host/C:/"),
+        ("..", "file://x/C:/", "file://x/C:/"),
+        ("../../..", "file://x/C:/a/b", "file://x/C:/"),
+        ("..", "file://x/a/C:/", "file://x/a/"),
+    ] {
+        assert_file_url(input, Some(base), expected);
+    }
+}
+
+#[test]
+fn file_url_pathname_setters_preserve_empty_segments_and_hosts() {
+    for (input, value, expected) in [
+        ("file://monkey/", r"\\", "file://monkey//"),
+        ("file:///unicorn", r"//\/", "file://////"),
+        ("file:///unicorn", "//monkey/..//", "file://///"),
+        ("file://host/old?q#f", "C|/new", "file://host/C:/new?q#f"),
+        ("file://host/old", "", "file://host/"),
+    ] {
+        let mut url = Url::parse(input).unwrap();
+        crate::components::set_pathname(&mut url, value);
+        assert_eq!(url.as_str(), expected, "pathname = {value:?}");
+        assert_eq!(Url::parse(url.as_str()).unwrap(), url);
+    }
+}
diff --git a/moli-url/src/lib.rs b/moli-url/src/lib.rs
index c2d3398d93..788dd94d79 100644
--- a/moli-url/src/lib.rs
+++ b/moli-url/src/lib.rs
@@ -2,6 +2,9 @@ pub mod components;
 pub mod origin;
 pub mod search_params;
 
+#[cfg(test)]
+mod file_url;
+
 pub use origin::is_about_srcdoc;
 
 pub use origin::{