Add canonical-site-url setting

Based on the conversation in #1238,
this implements the suggestion by markhildreth to implement such a setting.

I've additionally infixed `-site-` to highlight the relationship with the
`site-url`, and to distinguish it from the canonical URL as it occurs in a
page.

Author: roberth

CHANGELOG.md

@@ -1,5 +1,10 @@
 # Changelog
 
+## Upcoming release
+
+### Added
+- Added [`canonical-site-url`](https://rust-lang.github.io/mdBook/format/configuration/renderers.html?highlight=canonical-site-url#html-renderer-options) setting, to set `<link rel="canonical">` in the HTML output of each page.
+
 ## mdBook 0.4.52
 [v0.4.51...v0.4.52](https://github.com/rust-lang/mdBook/compare/v0.4.51...v0.4.52)
 

crates/mdbook-core/src/config.rs

@@ -580,6 +580,8 @@ pub struct HtmlConfig {
     pub input_404: Option<String>,
     /// Absolute url to site, used to emit correct paths for the 404 page, which might be accessed in a deeply nested directory
     pub site_url: Option<String>,
+    /// Canonical site url, used to emit <link rel="canonical"> tags in the HTML.
+    pub canonical_site_url: Option<String>,
     /// The DNS subdomain or apex domain at which your book will be hosted. This
     /// string will be written to a file named CNAME in the root of your site,
     /// as required by GitHub Pages (see [*Managing a custom domain for your
@@ -630,6 +632,7 @@ impl Default for HtmlConfig {
             edit_url_template: None,
             input_404: None,
             site_url: None,
+            canonical_site_url: None,
             cname: None,
             live_reload_endpoint: None,
             redirect: HashMap::new(),

crates/mdbook-html/front-end/templates/index.hbs

@@ -10,6 +10,9 @@
         {{#if base_url}}
         <base href="{{ base_url }}">
         {{/if}}
+        {{#if canonical_url}}
+        <link rel="canonical" href="{{ canonical_url }}">
+        {{/if}}
 
 
         <!-- Custom HTML head -->

crates/mdbook-html/src/html_handlebars/hbs_renderer.rs

@@ -74,6 +74,13 @@ impl HtmlHandlebars {
             .to_str()
             .with_context(|| "Could not convert path to str")?;
         let filepath = Path::new(&ctx_path).with_extension("html");
+        let filepath_str = filepath
+            .to_str()
+            .with_context(|| format!("Could not convert path to str: {}", filepath.display()))?;
+        let canonical_url = ctx.html_config.canonical_site_url.map(|canon_url| {
+            let canon_url = canon_url.as_str().trim_end_matches('/');
+            format!("{}/{}", canon_url, self.clean_path(filepath_str))
+        });
 
         // "print.html" is used for the print page.
         if path == Path::new("print.md") {
@@ -95,6 +102,8 @@ impl HtmlHandlebars {
         };
 
         ctx.data.insert("path".to_owned(), json!(path));
+        ctx.data
+            .insert("canonical_url".to_owned(), json!(canonical_url));
         ctx.data.insert("content".to_owned(), json!(content));
         ctx.data.insert("chapter_title".to_owned(), json!(ch.name));
         ctx.data.insert("title".to_owned(), json!(title));
@@ -325,6 +334,15 @@ impl HtmlHandlebars {
 
         Ok(())
     }
+
+    /// Strips `index.html` from the end of a path, if it exists.
+    fn clean_path(&self, path: &str) -> String {
+        if path == "index.html" || path.ends_with("/index.html") {
+            path[..path.len() - 10].to_string()
+        } else {
+            path.to_string()
+        }
+    }
 }
 
 impl Renderer for HtmlHandlebars {

guide/src/format/configuration/renderers.md

@@ -164,6 +164,12 @@ The following configuration options are available:
   navigation links and script/css imports in the 404 file work correctly, even when accessing
   urls in subdirectories. Defaults to `/`. If `site-url` is set,
   make sure to use document relative links for your assets, meaning they should not start with `/`.
+- **canonical-site-url:** Set the canonical URL for the book, which is used by
+  search engines to determine the primary URL for the content. Use this when
+  your site is deployed at multiple URLs. For example, when you have site
+  deployments for a range of versions, you can point all of them to the URL for
+  the latest version. Without this, your content may be penalized for
+  duplication, and visitors may be directed to an outdated version of the book.
 - **cname:** The DNS subdomain or apex domain at which your book will be hosted.
   This string will be written to a file named CNAME in the root of your site, as
   required by GitHub Pages (see [*Managing a custom domain for your GitHub Pages

tests/testsuite/rendering.rs

@@ -41,3 +41,25 @@ fn first_chapter_is_copied_as_index_even_if_not_first_elem() {
             ]],
         );
 }
+
+// Checks that a canonical URL is generated correctly.
+#[test]
+fn canonical_url() {
+    BookTest::from_dir("rendering/canonical_url")
+        .check_file_contains(
+            "book/index.html",
+            "<link rel=\"canonical\" href=\"https://example.com/test/\">",
+        )
+        .check_file_contains(
+            "book/canonical_url.html",
+            "<link rel=\"canonical\" href=\"https://example.com/test/canonical_url.html\">",
+        )
+        .check_file_contains(
+            "book/nested/page.html",
+            "<link rel=\"canonical\" href=\"https://example.com/test/nested/page.html\">",
+        )
+        .check_file_contains(
+            "book/nested/index.html",
+            "<link rel=\"canonical\" href=\"https://example.com/test/nested/\">",
+        );
+}

tests/testsuite/rendering/canonical_url/book.toml

@@ -0,0 +1,6 @@
+[book]
+title = "canonical_url test"
+
+[output.html]
+# trailing slash is not necessary or recommended, but tested here
+canonical-site-url = "https://example.com/test/"

tests/testsuite/rendering/canonical_url/src/SUMMARY.md

@@ -0,0 +1,4 @@
+- [Intro](README.md)
+- [Canonical URL](canonical_url.md)
+- [Nested Page](nested/page.md)
+- [Nested Index](nested/index.md)