Merge pull request #17 from niklak/feature/selector-only-text

added `:only-text` pseudo-class support
niklak · Nov 1, 2024 · 50566cf · 50566cf
2 parents 478c8ba + 56fe743
commit 50566cf
Show file tree

Hide file tree

Showing 6 changed files with 70 additions and 1 deletion.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -16,6 +16,7 @@ All notable changes to the `dom_query` crate will be documented in this file.
 - Added `NodeRef::replace_with_html` method that allows to replace a node with a new node created from the given HTML.
 - Added `NodeIdProver` trait and implementations for `NodeRef` and `Node`. Which allows to call some node functions with a `&NodeRef` and `&NodeId`. 
 Previously these functions required `NodeId` as a parameter.
+- Added a new pseudo-class `:only-text` that allows selecting a node with no child elements except a single **text** child node.
 
 ### Fixed
 - Fixed `Tree::append_prev_siblings_from_another_tree` method. It didn't assign `TreeNode.prev_sibling` properly.

diff --git a/src/matcher.rs b/src/matcher.rs
@@ -173,6 +173,9 @@ impl<'i> parser::Parser<'i> for InnerSelectorParser {
             Ok(Checked)
         } else if name.eq_ignore_ascii_case("indeterminate") {
             Ok(Indeterminate)
+
+        } else if name.eq_ignore_ascii_case("only-text") {
+            Ok(OnlyText)
         } else {
             Err(
                 location.new_custom_error(SelectorParseErrorKind::UnsupportedPseudoClassOrElement(
@@ -239,6 +242,8 @@ pub enum NonTSPseudoClass {
     Disabled,
     Checked,
     Indeterminate,
+    /// `:only-text` pseudo-class allows selecting a node with no child elements except a single **text** child node.
+    OnlyText,
     /// `:has-text` pseudo-class represents a selection for the element or one of its descendant element that contains the specified text.
     HasText(CssString),
     /// `:contains` pseudo-class represents a selection for the element that contains the specified text (it's own text and text of all his descendant elements).
@@ -261,6 +266,7 @@ impl ToCss for NonTSPseudoClass {
             NonTSPseudoClass::Disabled => dest.write_str(":disabled"),
             NonTSPseudoClass::Checked => dest.write_str(":checked"),
             NonTSPseudoClass::Indeterminate => dest.write_str(":indeterminate"),
+            NonTSPseudoClass::OnlyText => dest.write_str(":only-text"),
             NonTSPseudoClass::HasText(s) => {
                 dest.write_str(":has-text(")?;
                 s.to_css(dest)?;

diff --git a/src/node/node_ref.rs b/src/node/node_ref.rs
@@ -501,4 +501,12 @@ impl<'a> NodeRef<'a> {
         }
         false
     }
+
+    pub fn has_only_text(&self) -> bool {
+        if self.children_it().count() == 1 {
+            self.first_child().unwrap().is_text()
+        }else {
+            false
+        }
+    }
 }
diff --git a/src/node/selector.rs b/src/node/selector.rs
@@ -146,7 +146,7 @@ impl<'a> selectors::Element for NodeRef<'a> {
                 }
                 None => false,
             },
-
+            OnlyText => self.has_only_text(),
             HasText(s) => self.has_text(s.as_str()),
             Contains(s) => self.text().contains(s.as_str()),
         }

diff --git a/tests/node-manipulation.rs b/tests/node-manipulation.rs
@@ -185,3 +185,30 @@ fn test_node_replace_with_reparent() {
     // #inline is a child of #outline now
     assert!(doc.select("#outline > #inline").exists());
 }
+
+
+#[cfg_attr(not(target_arch = "wasm32"), test)]
+#[cfg_attr(target_arch = "wasm32", wasm_bindgen_test)]
+fn test_node_replace_text_node() {
+
+    let content = r#"<!DOCTYPE html>
+    <html lang="en">
+        <head></head>
+        <body>
+            <div id="main">
+                <p><a href="javascript:void(0)">Some text</a></p>
+            </div>
+        </body>
+    </html>"#;
+    let doc = Document::from(content);
+    // :only-text pseudo-class allows to select nodes that contain only one text node
+    let a_sel = doc.select_single(r#"a[href^="javascript:"]:only-text"#);
+    assert!(a_sel.exists());
+    let a_node = a_sel.nodes().first().unwrap();
+    let text_node =a_node.first_child().unwrap();
+    assert!(text_node.is_text());
+    a_node.replace_with(&text_node);
+
+    assert_eq!(doc.select("#main > p").inner_html(), "Some text".into());
+
+}
diff --git a/tests/pseudo-classes.rs b/tests/pseudo-classes.rs
@@ -145,6 +145,33 @@ fn pseudo_class_contains() {
     assert_eq!(text, "It is not how it works");
 }
 
+
+#[cfg_attr(not(target_arch = "wasm32"), test)]
+#[cfg_attr(target_arch = "wasm32", wasm_bindgen_test)]
+fn pseudo_class_only_text() {
+    //! :only-text pseudo-class allows selecting a node with no child elements except a single **text** child node.
+
+    let html = r#"<!DOCTYPE html>
+    <html>
+        <head>
+            <title>Test</title>
+        </head>
+        <body>
+            <div>
+                <a href="/1">One</a>
+                <a href="/2">Two</a>
+                <a href="/3">Three</a>
+            </div>
+            <div>Only text</div>
+        </body>
+    </html>
+    "#;
+    let document = Document::from(html);
+    let sel = document.select("body div:only-text");
+    assert_eq!(sel.inner_html(), "Only text".into());
+
+}
+
 #[cfg_attr(not(target_arch = "wasm32"), test)]
 #[cfg_attr(target_arch = "wasm32", wasm_bindgen_test)]
 fn pseudo_class_not() {