Merge branch 'master' into addl-percent-encode-sets

Author: cryptoquick

data-url/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "data-url"
-version = "0.2.0"
+version = "0.3.0"
 authors = ["Simon Sapin <simon.sapin@exyr.org>"]
 description = "Processing of data: URL according to WHATWG’s Fetch Standard"
 categories = ["no_std"]

form_urlencoded/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "form_urlencoded"
-version = "1.1.0"
+version = "1.2.0"
 authors = ["The rust-url developers"]
 description = "Parser and serializer for the application/x-www-form-urlencoded syntax, as used by HTML forms."
 categories = ["no_std"]
@@ -18,4 +18,4 @@ std = ["alloc", "percent-encoding/std"]
 alloc = ["percent-encoding/alloc"]
 
 [dependencies]
-percent-encoding = { version = "2.2.0", default-features = false, path = "../percent_encoding" }
+percent-encoding = { version = "2.3.0", default-features = false, path = "../percent_encoding" }

idna/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "idna"
-version = "0.3.0"
+version = "0.4.0"
 authors = ["The rust-url developers"]
 description = "IDNA (Internationalizing Domain Names in Applications) and Punycode."
 categories = ["no_std"]

percent_encoding/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "percent-encoding"
-version = "2.2.0"
+version = "2.3.0"
 authors = ["The rust-url developers"]
 description = "Percent encoding and decoding"
 categories = ["no_std"]

url/Cargo.toml

@@ -2,7 +2,7 @@
 
 name = "url"
 # When updating version, also modify html_root_url in the lib.rs
-version = "2.3.1"
+version = "2.4.0"
 authors = ["The rust-url developers"]
 
 description = "URL library for Rust, based on the WHATWG URL Standard"
@@ -28,9 +28,9 @@ debugger_test = "0.1"
 debugger_test_parser = "0.1"
 
 [dependencies]
-form_urlencoded = { version = "1.1.0", path = "../form_urlencoded" }
-idna = { version = "0.3.0", path = "../idna" }
-percent-encoding = { version = "2.2.0", path = "../percent_encoding" }
+form_urlencoded = { version = "1.2.0", path = "../form_urlencoded" }
+idna = { version = "0.4.0", path = "../idna" }
+percent-encoding = { version = "2.3.0", path = "../percent_encoding" }
 serde = {version = "1.0", optional = true, features = ["derive"]}
 
 [features]

url/src/lib.rs

@@ -121,7 +121,7 @@ url = { version = "2", features = ["serde"] }
 
 */
 
-#![doc(html_root_url = "https://docs.rs/url/2.3.1")]
+#![doc(html_root_url = "https://docs.rs/url/2.4.0")]
 #![cfg_attr(
     feature = "debugger_visualizer",
     feature(debugger_visualizer),

url/src/parser.rs

@@ -178,7 +178,7 @@ pub fn default_port(scheme: &str) -> Option<u16> {
     }
 }
 
-#[derive(Clone)]
+#[derive(Clone, Debug)]
 pub struct Input<'i> {
     chars: str::Chars<'i>,
 }
@@ -1173,7 +1173,7 @@ impl<'a> Parser<'a> {
     ) -> Input<'i> {
         // Relative path state
         loop {
-            let segment_start = self.serialization.len();
+            let mut segment_start = self.serialization.len();
             let mut ends_with_slash = false;
             loop {
                 let input_before_c = input.clone();
@@ -1202,6 +1202,14 @@ impl<'a> Parser<'a> {
                     }
                     _ => {
                         self.check_url_code_point(c, &input);
+                        if scheme_type.is_file()
+                            && is_normalized_windows_drive_letter(
+                                &self.serialization[path_start + 1..],
+                            )
+                        {
+                            self.serialization.push('/');
+                            segment_start += 1;
+                        }
                         if self.context == Context::PathSegmentSetter {
                             if scheme_type.is_special() {
                                 self.serialization
@@ -1249,7 +1257,10 @@ impl<'a> Parser<'a> {
                 }
                 _ => {
                     // If url’s scheme is "file", url’s path is empty, and buffer is a Windows drive letter, then
-                    if scheme_type.is_file() && is_windows_drive_letter(segment_before_slash) {
+                    if scheme_type.is_file()
+                        && segment_start == path_start + 1
+                        && is_windows_drive_letter(segment_before_slash)
+                    {
                         // Replace the second code point in buffer with U+003A (:).
                         if let Some(c) = segment_before_slash.chars().next() {
                             self.serialization.truncate(segment_start);

url/tests/unit.rs

@@ -1262,3 +1262,39 @@ fn test_authority() {
         "%C3%A0lex:%C3%A0lex@xn--lex-8ka.xn--p1ai.example.com"
     );
 }
+
+#[test]
+/// https://github.com/servo/rust-url/issues/838
+fn test_file_with_drive() {
+    let s1 = "fIlE:p:?../";
+    let url = url::Url::parse(s1).unwrap();
+    assert_eq!(url.to_string(), "file:///p:?../");
+    assert_eq!(url.path(), "/p:");
+
+    let testcases = [
+        ("a", "file:///p:/a"),
+        ("", "file:///p:?../"),
+        ("?x", "file:///p:?x"),
+        (".", "file:///p:/"),
+        ("..", "file:///p:/"),
+        ("../", "file:///p:/"),
+    ];
+
+    for case in &testcases {
+        let url2 = url::Url::join(&url, case.0).unwrap();
+        assert_eq!(url2.to_string(), case.1);
+    }
+}
+
+#[test]
+/// Similar to test_file_with_drive, but with a path
+/// that could be confused for a drive.
+fn test_file_with_drive_and_path() {
+    let s1 = "fIlE:p:/x|?../";
+    let url = url::Url::parse(s1).unwrap();
+    assert_eq!(url.to_string(), "file:///p:/x|?../");
+    assert_eq!(url.path(), "/p:/x|");
+    let s2 = "a";
+    let url2 = url::Url::join(&url, s2).unwrap();
+    assert_eq!(url2.to_string(), "file:///p:/a");
+}