Merge pull request #2922 from ehuss/header-id-lowercase

Lowercase heading IDs

Author: ehuss

CHANGELOG.md

@@ -58,8 +58,11 @@ The following is a summary of the changes that may require your attention when u
   [#2847](https://github.com/rust-lang/mdBook/pull/2847)
 - Added support for admonitions. These are enabled by default, with the option `output.html.admonitions` to disable it.
   [#2851](https://github.com/rust-lang/mdBook/pull/2851)
-- Headers that start or end with HTML characters like `<`, `&`, or `>` now replace those characters in the link ID with `-` instead of being stripped. This brings the header ID generation closer to other tools and sites.
-  [#2844](https://github.com/rust-lang/mdBook/pull/2844)
+- Header ID generation has some minor changes to bring the ID generation closer to other tools and sites:
+  - IDs now use Unicode lowercase instead of ASCII lowercase.
+    [#2922](https://github.com/rust-lang/mdBook/pull/2922)
+  - Headers that start or end with HTML characters like `<`, `&`, or `>` now replace those characters in the link ID with `-` instead of being stripped.
+    [#2844](https://github.com/rust-lang/mdBook/pull/2844)
 
 ### CLI changes
 

crates/mdbook-html/src/utils.rs

@@ -74,12 +74,22 @@ pub(crate) fn unique_id(id: &str, used: &mut HashSet<String>) -> String {
 
 /// Generates an HTML id from the given text.
 pub(crate) fn id_from_content(content: &str) -> String {
+    // This is intended to be close to how header ID generation is done in
+    // other sites and tools, but is not 100% the same. Not all sites and
+    // tools use the same algorithm. See these for more information:
+    //
+    // - https://docs.github.com/en/get-started/writing-on-github/getting-started-with-writing-and-formatting-on-github/basic-writing-and-formatting-syntax#section-links
+    // - https://docs.gitlab.com/user/markdown/#heading-ids-and-links
+    // - https://pandoc.org/MANUAL.html#extension-auto_identifiers
+    // - https://kramdown.gettalong.org/converter/html#auto-ids
+    // - https://docs.rs/comrak/latest/comrak/options/struct.Extension.html#structfield.header_ids
     content
         .trim()
+        .to_lowercase()
         .chars()
         .filter_map(|ch| {
             if ch.is_alphanumeric() || ch == '_' || ch == '-' {
-                Some(ch.to_ascii_lowercase())
+                Some(ch)
             } else if ch.is_whitespace() {
                 Some('-')
             } else {
@@ -120,6 +130,6 @@ mod tests {
         assert_eq!(id_from_content("한국어"), "한국어");
         assert_eq!(id_from_content(""), "");
         assert_eq!(id_from_content("中文標題 CJK title"), "中文標題-cjk-title");
-        assert_eq!(id_from_content("Über"), "Über");
+        assert_eq!(id_from_content("Über"), "über");
     }
 }