fix handling of script content, bump html5gum (#181)

* scriptdata bad testcase

* bump html5gum

Cargo.toml

@@ -21,7 +21,7 @@ rayon = "1.5.1"
 pulldown-cmark = "0.12.2"
 blake3 = "1.0.0"
 
-html5gum = "0.6.0"
+html5gum = "0.6.1"
 jwalk = "0.8.1"
 bumpalo = { version = "3.11.1", features = ["collections"] }
 percent-encoding = "2.1.0"

src/html/mod.rs

@@ -570,3 +570,26 @@ fn test_document_join_bare_html() {
         Href("locations/oslo#grünerløkka")
     );
 }
+
+#[test]
+fn test_json_script() {
+    use crate::paragraph::ParagraphHasher;
+
+    let doc = Document::new(
+        Path::new("/"),
+        Path::new("/html5gum/struct.Tokenizer.html"),
+    );
+
+    let html = r#"<script type="text/json" id="notable-traits-data">{"InfallibleTokenizer<R, E>":"<h3>Notable traits for <code><a class=\"struct\" href=\"struct.InfallibleTokenizer.html\" title=\"struct html5gum::InfallibleTokenizer\">InfallibleTokenizer</a>&lt;R, E&gt;</code></h3><pre><code><div class=\"where\">impl&lt;R: <a class=\"trait\" href=\"trait.Reader.html\" title=\"trait html5gum::Reader\">Reader</a>&lt;Error = <a class=\"enum\" href=\"https://doc.rust-lang.org/1.82.0/core/convert/enum.Infallible.html\" title=\"enum core::convert::Infallible\">Infallible</a>&gt;, E: <a class=\"trait\" href=\"emitters/trait.Emitter.html\" title=\"trait html5gum::emitters::Emitter\">Emitter</a>&gt; <a class=\"trait\" href=\"https://doc.rust-lang.org/1.82.0/core/iter/traits/iterator/trait.Iterator.html\" title=\"trait core::iter::traits::iterator::Iterator\">Iterator</a> for <a class=\"struct\" href=\"struct.InfallibleTokenizer.html\" title=\"struct html5gum::InfallibleTokenizer\">InfallibleTokenizer</a>&lt;R, E&gt;</div><div class=\"where\">    type <a href=\"https://doc.rust-lang.org/1.82.0/core/iter/traits/iterator/trait.Iterator.html#associatedtype.Item\" class=\"associatedtype\">Item</a> = E::<a class=\"associatedtype\" href=\"emitters/trait.Emitter.html#associatedtype.Token\" title=\"type html5gum::emitters::Emitter::Token\">Token</a>;</div>"}</script>"#;
+
+    let mut doc_buf = DocumentBuffers::default();
+
+    let links = doc
+        .links_from_read::<_, ParagraphHasher>(&mut doc_buf, html.as_bytes(), false)
+        .unwrap();
+
+    assert_eq!(
+        links.collect::<Vec<_>>(),
+        &[]
+    );
+}