fix language filtering to exclude unconfigured languages

lib/jekyll/polyglot/liquid/tags/i18n_headers.rb

@@ -21,13 +21,17 @@ def render(context)
           i18n = ""
 
           # Find all documents with the same page_id
+          valid_languages = ([site.default_lang] + site.languages).uniq
           docs_with_same_id = site.collections.values
             .flat_map(&:docs)
             .filter { |doc| !doc.data['page_id'].nil? }
             .select { |doc| doc.data['page_id'] == page_id }
 
           # Build a hash of lang => permalink for all matching docs
-          lang_to_permalink = docs_with_same_id.to_h { |doc| [doc.data['lang'], doc.data['permalink']] }
+          # Filter by explicit lang to exclude unconfigured languages even after normalization
+          lang_to_permalink = docs_with_same_id
+            .reject { |doc| doc.data['lang'] && !valid_languages.include?(doc.data['lang']) }
+            .to_h { |doc| [doc.data['lang'] || site.default_lang, doc.data['permalink']] }
 
           # Canonical should always point to the current page's permalink (active_lang)
           current_lang = site.active_lang

lib/jekyll/polyglot/patches/jekyll/site.rb

@@ -148,8 +148,23 @@ def derive_lang_from_path(doc)
     def coordinate_documents(docs)
       regex = document_url_regex
       approved = {}
+      # Build set of valid languages (default + configured)
+      valid_languages = ([@default_lang] + @languages).uniq
+
       docs.each do |doc|
-        lang = doc.data['lang'] || derive_lang_from_path(doc) || @default_lang
+        # Get the explicitly declared language (frontmatter or path-derived)
+        explicit_lang = doc.data['lang'] || derive_lang_from_path(doc)
+        lang = explicit_lang || @default_lang
+
+        # FILTER: Skip documents whose explicit lang is not in configured languages.
+        # Check the explicit value (not the fallback) so that documents with an
+        # unconfigured lang like 'de' are excluded even if normalization would
+        # map them to default_lang.
+        if explicit_lang && !valid_languages.include?(explicit_lang)
+          Jekyll.logger.warn "Polyglot:", "Skipping #{doc.relative_path} - lang '#{explicit_lang}' not in configured languages #{valid_languages.inspect}"
+          next
+        end
+
         lang_exclusive = doc.data['lang-exclusive'] || []
         url = doc.url.gsub(regex, '/')
         page_id = doc.data['page_id'] || url
@@ -220,11 +235,22 @@ def assignPageLanguagePermalinks(doc, docs)
       pageId = doc.data['page_id']
       if !pageId.nil? && !pageId.empty?
         unless doc.data['permalink_lang'] then doc.data['permalink_lang'] = {} end
+
+        # Build set of valid languages
+        valid_languages = ([@default_lang] + @languages).uniq
+
         permalinkDocs = docs.select do |dd|
           dd.data['page_id'] == pageId
         end
         permalinkDocs.each do |dd|
-          doclang = dd.data['lang'] || derive_lang_from_path(dd) || @default_lang
+          explicit_lang = dd.data['lang'] || derive_lang_from_path(dd)
+          doclang = explicit_lang || @default_lang
+
+          # FILTER: Only include permalinks for configured languages.
+          # Check explicit lang so unconfigured languages are excluded
+          # even if normalization would map them to default_lang.
+          next if explicit_lang && !valid_languages.include?(explicit_lang)
+
           doc.data['permalink_lang'][doclang] = dd.data['permalink']
         end
       end

spec/jekyll/polyglot/patches/jekyll/site_spec.rb

@@ -1021,5 +1021,213 @@
         expect(doc.data['rendered_lang']).to eq('es')
       end
     end
+
+    describe 'coordinate_documents with unconfigured languages' do
+      before do
+        @collection = Jekyll::Collection.new(@site, 'test')
+      end
+
+      it 'should exclude documents with unconfigured lang in frontmatter' do
+        # Configure site with only en and es
+        @site.config['languages'] = ['en', 'es']
+        @site.config['default_lang'] = 'en'
+        @site.prepare
+
+        docs = [
+          # lang: en - should be included
+          Jekyll::Document.new('test-en.md', site: @site, collection: @collection).tap do |doc|
+            doc.data['lang'] = 'en'
+            doc.data['title'] = 'English Page'
+          end,
+          # lang: de - should be excluded (not in config)
+          Jekyll::Document.new('test-de.md', site: @site, collection: @collection).tap do |doc|
+            doc.data['lang'] = 'de'
+            doc.data['title'] = 'German Page'
+          end,
+          # lang: es - should be included
+          Jekyll::Document.new('test-es.md', site: @site, collection: @collection).tap do |doc|
+            doc.data['lang'] = 'es'
+            doc.data['title'] = 'Spanish Page'
+          end
+        ]
+
+        coordinated = @site.coordinate_documents(docs)
+        coordinated_langs = coordinated.map { |d| d.data['lang'] }
+
+        expect(coordinated_langs).to include('en')
+        expect(coordinated_langs).to include('es')
+        expect(coordinated_langs).not_to include('de')
+      end
+
+      it 'should include documents with default_lang even if not in languages array' do
+        # Configure with es and fr, but default_lang is en
+        @site.config['languages'] = ['es', 'fr']
+        @site.config['default_lang'] = 'en'
+        @site.prepare
+
+        docs = [
+          # lang: en (default_lang) - should be included
+          Jekyll::Document.new('test-en.md', site: @site, collection: @collection).tap do |doc|
+            doc.data['lang'] = 'en'
+            doc.data['title'] = 'English Page'
+          end,
+          # lang: es - should be included
+          Jekyll::Document.new('test-es.md', site: @site, collection: @collection).tap do |doc|
+            doc.data['lang'] = 'es'
+            doc.data['title'] = 'Spanish Page'
+          end,
+          # lang: de - should be excluded
+          Jekyll::Document.new('test-de.md', site: @site, collection: @collection).tap do |doc|
+            doc.data['lang'] = 'de'
+            doc.data['title'] = 'German Page'
+          end
+        ]
+
+        coordinated = @site.coordinate_documents(docs)
+        coordinated_langs = coordinated.map { |d| d.data['lang'] }
+
+        expect(coordinated_langs).to include('en')  # default_lang always included
+        expect(coordinated_langs).to include('es')
+        expect(coordinated_langs).not_to include('de')
+      end
+
+      it 'should log warning when excluding documents with unconfigured language' do
+        @site.config['languages'] = ['en', 'es']
+        @site.config['default_lang'] = 'en'
+        @site.prepare
+
+        doc = Jekyll::Document.new('test-de.md', site: @site, collection: @collection)
+        doc.data['lang'] = 'de'
+        doc.data['title'] = 'German Page'
+
+        expect(Jekyll.logger).to receive(:warn).with("Polyglot:", /lang 'de' not in configured/)
+
+        @site.coordinate_documents([doc])
+      end
+    end
+
+      it 'should not serve unconfigured language pages as default language content' do
+        # Real-world scenario: site configured with limited languages for dev builds,
+        # but pages exist for many languages from production config.
+        # German pages should not appear when only en and pt-BR are configured.
+        @site.config['languages'] = ['en', 'pt-BR']
+        @site.config['default_lang'] = 'en'
+        @site.prepare
+
+        # Simulate Jekyll::Page objects using OpenStruct (like site.pages)
+        pages = [
+          OpenStruct.new(
+            data: { 'lang' => 'en', 'page_id' => 'home', 'permalink' => '/',
+                    'title' => 'The SQL Editor You Love' },
+            url: '/'
+          ),
+          OpenStruct.new(
+            data: { 'lang' => 'de', 'page_id' => 'home', 'permalink' => '/',
+                    'title' => 'Der SQL-Editor Ihrer Träume' },
+            url: '/'
+          ),
+          OpenStruct.new(
+            data: { 'lang' => 'pt-BR', 'page_id' => 'home', 'permalink' => '/',
+                    'title' => 'O Editor SQL dos Seus Sonhos' },
+            url: '/'
+          ),
+          OpenStruct.new(
+            data: { 'lang' => 'fr', 'page_id' => 'home', 'permalink' => '/',
+                    'title' => "L'éditeur SQL de vos rêves" },
+            url: '/'
+          )
+        ]
+
+        # Building for default language (en)
+        @site.active_lang = 'en'
+        coordinated = @site.coordinate_documents(pages)
+
+        # Only configured languages should appear
+        coordinated_langs = coordinated.map { |p| p.data['lang'] }
+        expect(coordinated_langs).not_to include('de')
+        expect(coordinated_langs).not_to include('fr')
+
+        # The winning page for the default language build should be English
+        expect(coordinated.length).to eq(1)
+        expect(coordinated[0].data['title']).to eq('The SQL Editor You Love')
+      end
+
+      it 'should filter by explicit frontmatter lang even if lang resolves to default' do
+        # This test catches a bug where normalize_lang (on other branches) converts
+        # unknown lang codes to nil, which then falls back to default_lang,
+        # bypassing the valid_languages filter.
+        # The filter must check the ORIGINAL frontmatter lang value, not the resolved one.
+        @site.config['languages'] = ['en', 'pt-BR']
+        @site.config['default_lang'] = 'en'
+        @site.prepare
+
+        collection = Jekyll::Collection.new(@site, 'test')
+        docs = [
+          Jekyll::Document.new('index.html', site: @site, collection: collection).tap do |doc|
+            doc.data['lang'] = 'en'
+            doc.data['page_id'] = 'home'
+            doc.data['title'] = 'English Home'
+          end,
+          # This doc has explicit lang: de in frontmatter.
+          # Even if a normalize step maps 'de' -> nil -> default_lang,
+          # it should still be excluded because 'de' is not configured.
+          Jekyll::Document.new('index-de.html', site: @site, collection: collection).tap do |doc|
+            doc.data['lang'] = 'de'
+            doc.data['page_id'] = 'home'
+            doc.data['title'] = 'German Home'
+          end
+        ]
+
+        @site.active_lang = 'en'
+        coordinated = @site.coordinate_documents(docs)
+
+        # The German doc must NOT be in the output
+        titles = coordinated.map { |d| d.data['title'] }
+        expect(titles).to include('English Home')
+        expect(titles).not_to include('German Home')
+
+        # Verify the German doc's explicit lang was the reason for exclusion
+        coordinated_langs = coordinated.map { |d| d.data['lang'] }
+        expect(coordinated_langs).not_to include('de')
+      end
+
+    describe 'assignPageLanguagePermalinks with unconfigured languages' do
+      before do
+        @collection = Jekyll::Collection.new(@site, 'test')
+      end
+
+      it 'should only include configured languages in permalink_lang' do
+        @site.config['languages'] = ['en', 'es']
+        @site.config['default_lang'] = 'en'
+        @site.prepare
+
+        # Create docs with en, es, de (de not configured)
+        docs = [
+          Jekyll::Document.new('test-en.md', site: @site, collection: @collection).tap do |doc|
+            doc.data['lang'] = 'en'
+            doc.data['page_id'] = 'test-page'
+            doc.data['permalink'] = '/test/'
+          end,
+          Jekyll::Document.new('test-es.md', site: @site, collection: @collection).tap do |doc|
+            doc.data['lang'] = 'es'
+            doc.data['page_id'] = 'test-page'
+            doc.data['permalink'] = '/es/test/'
+          end,
+          Jekyll::Document.new('test-de.md', site: @site, collection: @collection).tap do |doc|
+            doc.data['lang'] = 'de'
+            doc.data['page_id'] = 'test-page'
+            doc.data['permalink'] = '/de/test/'
+          end
+        ]
+
+        # Call assignPageLanguagePermalinks on the English doc
+        @site.assignPageLanguagePermalinks(docs[0], docs)
+
+        # Verify permalink_lang only has keys for en and es
+        expect(docs[0].data['permalink_lang'].keys).to include('en')
+        expect(docs[0].data['permalink_lang'].keys).to include('es')
+        expect(docs[0].data['permalink_lang'].keys).not_to include('de')
+      end
+    end
   end
 end