Add sorting by slug

components/content/src/sorting.rs

@@ -16,6 +16,7 @@ pub fn sort_pages(pages: &[&Page], sort_by: SortBy) -> (Vec<PathBuf>, Vec<PathBu
             }
             SortBy::Title | SortBy::TitleBytes => page.meta.title.is_some(),
             SortBy::Weight => page.meta.weight.is_some(),
+            SortBy::Slug => true,
             SortBy::None => unreachable!(),
         });
 
@@ -32,6 +33,7 @@ pub fn sort_pages(pages: &[&Page], sort_by: SortBy) -> (Vec<PathBuf>, Vec<PathBu
                 a.meta.title.as_ref().unwrap().cmp(b.meta.title.as_ref().unwrap())
             }
             SortBy::Weight => a.meta.weight.unwrap().cmp(&b.meta.weight.unwrap()),
+            SortBy::Slug => natural_lexical_cmp(&a.slug, &b.slug),
             SortBy::None => unreachable!(),
         };
 
@@ -73,6 +75,16 @@ mod tests {
         Page::new(format!("content/hello-{}.md", weight), front_matter, &PathBuf::new())
     }
 
+    fn create_page_with_slug(slug: &str) -> Page {
+        let front_matter = PageFrontMatter { slug: Some(slug.to_owned()), ..Default::default() };
+        let mut page =
+            Page::new(format!("content/hello-{}.md", slug), front_matter, &PathBuf::new());
+        // Normally, the slug field is populated when a page is parsed, but
+        // since we're creating one manually, we have to set it explicitly
+        page.slug = slug.to_owned();
+        page
+    }
+
     #[test]
     fn can_sort_by_dates() {
         let page1 = create_page_with_date("2018-01-01", None);
@@ -185,6 +197,28 @@ mod tests {
         );
     }
 
+    #[test]
+    fn can_sort_by_slug() {
+        let page1 = create_page_with_slug("2");
+        let page2 = create_page_with_slug("3");
+        let page3 = create_page_with_slug("1");
+        let (pages, ignored_pages) = sort_pages(&[&page1, &page2, &page3], SortBy::Slug);
+        assert_eq!(pages[0], page3.file.path);
+        assert_eq!(pages[1], page1.file.path);
+        assert_eq!(pages[2], page2.file.path);
+        assert_eq!(ignored_pages.len(), 0);
+
+        // 10 should come after 2
+        let page1 = create_page_with_slug("1");
+        let page2 = create_page_with_slug("10");
+        let page3 = create_page_with_slug("2");
+        let (pages, ignored_pages) = sort_pages(&[&page1, &page2, &page3], SortBy::Slug);
+        assert_eq!(pages[0], page1.file.path);
+        assert_eq!(pages[1], page3.file.path);
+        assert_eq!(pages[2], page2.file.path);
+        assert_eq!(ignored_pages.len(), 0);
+    }
+
     #[test]
     fn can_find_ignored_pages() {
         let page1 = create_page_with_date("2018-01-01", None);

components/content/src/types.rs

@@ -15,6 +15,8 @@ pub enum SortBy {
     TitleBytes,
     /// Lower weight comes first
     Weight,
+    /// Sort by slug
+    Slug,
     /// No sorting
     None,
 }

docs/content/documentation/content/section.md

@@ -187,11 +187,14 @@ of the Swedish alphabet, åäö, for example would be considered by the natural
 sort as aao. In that case the standard byte-order sort may be more suitable.
 
 ### `weight`
-This will be sort all pages by their `weight` field, from lightest weight
-(at the top of the list) to heaviest (at the bottom of the list).  Each
+This will sort all pages by their `weight` field, from the lightest weight
+(at the top of the list) to the heaviest (at the bottom of the list). Each
 page gets `page.lower` and `page.higher` variables that contain the
 pages with lighter and heavier weights, respectively.
 
+### `slug`
+This will sort pages or sections by their slug in natural lexical order.
+
 ### Reversed sorting
 When iterating through pages, you may wish to use the Tera `reverse` filter,
 which reverses the order of the pages.  For example, after using the `reverse` filter,