Sitemap hreflang syntax invalid for regional language variants fix (#…

…5638)

Sitemap hreflang syntax invalid for regional language variants fix

readthedocs/core/views/serve.py

@@ -354,6 +354,17 @@ def priorities_generator():
         priorities = [1, 0.9, 0.8, 0.7, 0.6, 0.5, 0.4, 0.3, 0.2]
         yield from itertools.chain(priorities, itertools.repeat(0.1))
 
+    def hreflang_formatter(lang):
+        """
+        sitemap hreflang should follow correct format.
+
+        Use hyphen instead of underscore in language and country value.
+        ref: https://en.wikipedia.org/wiki/Hreflang#Common_Mistakes
+        """
+        if '_' in lang:
+            return lang.replace("_", "-")
+        return lang
+
     def changefreqs_generator():
         """
         Generator returning ``changefreq`` needed by sitemap.xml.
@@ -407,7 +418,7 @@ def changefreqs_generator():
                         private=False,
                     )
                     element['languages'].append({
-                        'hreflang': translation.language,
+                        'hreflang': hreflang_formatter(translation.language),
                         'href': href,
                     })
 

readthedocs/rtd_tests/tests/test_doc_serving.py

@@ -246,6 +246,13 @@ def test_sitemap_xml(self):
             main_language_project=self.public,
             language='translation-es'
         )
+        # sitemap hreflang should follow correct format.
+        # ref: https://en.wikipedia.org/wiki/Hreflang#Common_Mistakes
+        hreflang_test_translation_project = fixture.get(
+            Project,
+            main_language_project=self.public,
+            language='zh_CN'
+        )
         response = self.client.get(
             reverse('sitemap_xml'),
             HTTP_HOST='public.readthedocs.io',
@@ -283,3 +290,6 @@ def test_sitemap_xml(self):
                 private=False,
             ),
         )
+        # hreflang should use hyphen instead of underscore
+        # in language and country value. (zh_CN should be zh-CN)
+        self.assertContains(response, 'zh-CN')