docs: link multipage version of HTML spec

6 files changed, 61 insertions, 13 deletions

diff --git a/Cargo.toml b/Cargo.toml
index 943f710..d32b406 100644
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -21,6 +21,7 @@ include = ["src/**/*", "LICENSE", "README.md"]
 codespan-reporting = "0.11.1"
 insta = "1.31.0"
 similar-asserts = { workspace = true }
+walkdir = "2.3.3"
 
 [features]
 # Feature used by integration tests in tests/ to get access to library internals.
diff --git a/README.md b/README.md
index 652edd8..9700826 100644
--- a/README.md
+++ b/README.md
@@ -69,7 +69,7 @@ Licensed under the MIT license, see [`./LICENSE`](./LICENSE).
 
 
 [html5gum]: https://crates.io/crates/html5gum
-[tokenization]: https://html.spec.whatwg.org/#tokenization
+[tokenization]: https://html.spec.whatwg.org/multipage/parsing.html#tokenization
 [html5lib's tokenizer test suite]: https://github.com/html5lib/html5lib-tests/tree/master/tokenizer
-[charset detection]: https://html.spec.whatwg.org/#determining-the-character-encoding
-[misnested tags]: https://html.spec.whatwg.org/#an-introduction-to-error-handling-and-strange-cases-in-the-parser
+[charset detection]: https://html.spec.whatwg.org/multipage/parsing.html#determining-the-character-encoding
+[misnested tags]: https://html.spec.whatwg.org/multipage/parsing.html#an-introduction-to-error-handling-and-strange-cases-in-the-parser
diff --git a/src/emitter.rs b/src/emitter.rs
index 4fc2159..d1180a5 100644
--- a/src/emitter.rs
+++ b/src/emitter.rs
@@ -521,7 +521,7 @@ impl<O: Offset> Comment<O> {
 pub struct Doctype<O> {
     /// The [force-quirks flag].
     ///
-    /// [force-quirks flag]: https://html.spec.whatwg.org/#force-quirks-flag
+    /// [force-quirks flag]: https://html.spec.whatwg.org/multipage/parsing.html#force-quirks-flag
     pub force_quirks: bool,
 
     /// The doctype's name. For HTML documents this is "html".
diff --git a/src/error.rs b/src/error.rs
index 401937b..3ba8f63 100644
--- a/src/error.rs
+++ b/src/error.rs
@@ -4,7 +4,7 @@ macro_rules! impl_error {
     )*) => {
         /// All [parse errors] this tokenizer can emit.
         ///
-        /// [parse errors]: https://html.spec.whatwg.org/#parse-errors
+        /// [parse errors]: https://html.spec.whatwg.org/multipage/parsing.html#parse-errors
         #[derive(Debug, Eq, PartialEq)]
         pub enum Error {
             $(
diff --git a/src/tokenizer.rs b/src/tokenizer.rs
index 3a6fb32..469cbd1 100644
--- a/src/tokenizer.rs
+++ b/src/tokenizer.rs
@@ -76,31 +76,31 @@ impl<R: Reader, O: Offset, E: Emitter<O>> Tokenizer<R, O, E> {
 pub enum State {
     /// The [data state].
     ///
-    /// [data state]: https://html.spec.whatwg.org/#data-state
+    /// [data state]: https://html.spec.whatwg.org/multipage/parsing.html#data-state
     Data,
     /// The [PLAINTEXT state].
     ///
-    /// [PLAINTEXT state]: https://html.spec.whatwg.org/#plaintext-state
+    /// [PLAINTEXT state]: https://html.spec.whatwg.org/multipage/parsing.html#plaintext-state
     PlainText,
     /// The [RCDATA state].
     ///
-    /// [RCDATA state]: https://html.spec.whatwg.org/#rcdata-state
+    /// [RCDATA state]: https://html.spec.whatwg.org/multipage/parsing.html#rcdata-state
     RcData,
     /// The [RAWTEXT state].
     ///
-    /// [RAWTEXT state]: https://html.spec.whatwg.org/#rawtext-state
+    /// [RAWTEXT state]: https://html.spec.whatwg.org/multipage/parsing.html#rawtext-state
     RawText,
     /// The [script data state].
     ///
-    /// [script data state]: https://html.spec.whatwg.org/#script-data-state
+    /// [script data state]: https://html.spec.whatwg.org/multipage/parsing.html#script-data-state
     ScriptData,
     /// The [script data escaped state].
     ///
-    /// [script data escaped state]: https://html.spec.whatwg.org/#script-data-escaped-state
+    /// [script data escaped state]: https://html.spec.whatwg.org/multipage/parsing.html#script-data-escaped-state
     ScriptDataEscaped,
     /// The [script data double escaped state].
     ///
-    /// [script data double escaped state]: https://html.spec.whatwg.org/#script-data-double-escaped-state
+    /// [script data double escaped state]: https://html.spec.whatwg.org/multipage/parsing.html#script-data-double-escaped-state
     ScriptDataDoubleEscaped,
 }
 
@@ -145,7 +145,7 @@ impl<R: Reader + Position<O>, O: Offset, E: Emitter<O>> Tokenizer<R, O, E> {
     ///
     /// See also WHATWG's definition of [appropriate end tag token].
     ///
-    /// [appropriate end tag token]: https://html.spec.whatwg.org/#appropriate-end-tag-token
+    /// [appropriate end tag token]: https://html.spec.whatwg.org/multipage/parsing.html#appropriate-end-tag-token
     #[inline]
     pub(crate) fn current_end_tag_is_appropriate(&mut self) -> bool {
         self.current_tag_name == self.last_start_tag_name
diff --git a/tests/misc.rs b/tests/misc.rs
new file mode 100644
index 0000000..416e506
--- /dev/null
+++ b/tests/misc.rs
@@ -0,0 +1,47 @@
+use similar_asserts::assert_eq;
+use walkdir::{DirEntry, WalkDir};
+
+#[test]
+fn links_to_html_spec_use_multipage_version() {
+    for entry in WalkDir::new(".")
+        .min_depth(1)
+        .into_iter()
+        .filter_entry(is_source_file)
+        .flatten()
+    {
+        if !entry.file_type().is_file() {
+            continue;
+        }
+
+        let actual = match std::fs::read_to_string(entry.path()) {
+            Ok(content) => content,
+            Err(err) => panic!("invalid UTF-8 in file content: {:?}: {}", entry.path(), err),
+        };
+
+        let expected = actual.replace(
+            concat!("://html.spec.whatwg.org/", "#"),
+            concat!("://html.spec.whatwg.org/multipage/???.html#"),
+        );
+
+        assert_eq!(
+            actual,
+            expected,
+            "Found a link to the one-page version of the HTML spec, which is huge and takes long to load. We want to link the multipage version instead."
+        );
+    }
+}
+
+fn is_source_file(entry: &DirEntry) -> bool {
+    let Some(filename) = entry.file_name().to_str() else {
+        panic!("invalid UTF-8 in filename: {:?}", entry.path())
+    };
+
+    if entry.depth() == 1 && filename == "target" || filename == "Cargo.lock" {
+        return false; // cargo files
+    }
+    if filename == "html5lib-tests" {
+        return false; // git submodule
+    }
+
+    !filename.starts_with('.') // .git, etc.
+}