Properly handle quotes in query selector

src/queryselector/parser.rs

@@ -25,12 +25,17 @@ impl<'a> Parser<'a> {
         has_whitespace
     }
 
+    #[inline]
     fn read_identifier(&mut self) -> &'a [u8] {
+        self.read_while(util::is_ident)
+    }
+
+    fn read_while(&mut self, mut predicate: impl FnMut(u8) -> bool) -> &'a [u8] {
         let start = self.stream.idx;
 
         while !self.stream.is_eof() {
-            let is_ident = self.stream.current().copied().map_or(false, util::is_ident);
-            if !is_ident {
+            let matched = self.stream.current().copied().map_or(false, &mut predicate);
+            if !matched {
                 break;
             } else {
                 self.stream.advance();
@@ -84,23 +89,27 @@ impl<'a> Parser<'a> {
             Some(b'=') => {
                 self.stream.advance();
                 let quote = self.stream.expect_oneof_and_skip(&[b'"', b'\'']);
-                let value = self.read_identifier();
-                if let Some(quote) = quote {
-                    // Only require the given quote if the value starts with a quote
+                let value = if let Some(quote) = quote {
+                    let value = self.read_while(|c| c != quote);
                     self.stream.expect_and_skip(quote)?;
-                }
+                    value
+                } else {
+                    self.read_identifier()
+                };
                 self.stream.expect_and_skip(b']')?;
                 Selector::AttributeValue(attribute, value)
             }
             Some(c @ b'~' | c @ b'^' | c @ b'$' | c @ b'*') => {
                 self.stream.advance();
                 self.stream.expect_and_skip(b'=')?;
                 let quote = self.stream.expect_oneof_and_skip(&[b'"', b'\'']);
-                let value = self.read_identifier();
-                if let Some(quote) = quote {
-                    // Only require the given quote if the value starts with a quote
+                let value = if let Some(quote) = quote {
+                    let value = self.read_while(|c| c != quote);
                     self.stream.expect_and_skip(quote)?;
-                }
+                    value
+                } else {
+                    self.read_identifier()
+                };
                 self.stream.expect_and_skip(b']')?;
                 match c {
                     b'~' => Selector::AttributeValueWhitespacedContains(attribute, value),

src/tests.rs

@@ -585,11 +585,11 @@ mod query_selector {
 
     #[test]
     fn query_selector_with_quote() {
-        let input = r#"<div><meta property="og:title" content="hello" /></div>"#;
+        let input = r#"<div><meta property="og:'title'" content="hello" /></div>"#;
         let dom = parse(input, ParserOptions::default()).unwrap();
         let parser = dom.parser();
         let node_option = dom
-            .query_selector(r#"meta[property="og:title"]"#)
+            .query_selector(r#"meta[property="og:'title'"]"#)
             .and_then(|mut iter| iter.next());
         let value = if let Some(node) = node_option {
             Some(

src/util.rs

@@ -5,9 +5,6 @@ pub fn is_ident(c: u8) -> bool {
         || (b'a'..=b'z').contains(&c)
         || c == b'-'
         || c == b'_'
-        || c == b':'
-        || c == b'+'
-        || c == b'/'
 }
 
 #[inline(always)]