Merge pull request #2395 from rdumont/data-subdirectories

features/data.feature

@@ -60,6 +60,35 @@ Feature: Data
     And I should see "Jack" in "_site/index.html"
     And I should see "Leon" in "_site/index.html"
 
+  Scenario: autoload *.yaml files in subdirectories in _data directory
+    Given I have a _data directory
+    And I have a _data/categories directory
+    And I have a "_data/categories/dairy.yaml" file with content:
+      """
+      name: Dairy Products
+      """
+    And I have an "index.html" page that contains "{{ site.data.categories.dairy.name }}"
+    When I run jekyll build
+    Then the "_site/index.html" file should exist
+    And I should see "Dairy Products" in "_site/index.html"
+
+  Scenario: folders should have precedence over files with the same name
+    Given I have a _data directory
+    And I have a _data/categories directory
+    And I have a "_data/categories/dairy.yaml" file with content:
+      """
+      name: Dairy Products
+      """
+    And I have a "_data/categories.yaml" file with content:
+      """
+      dairy:
+        name: Should not display this
+      """
+    And I have an "index.html" page that contains "{{ site.data.categories.dairy.name }}"
+    When I run jekyll build
+    Then the "_site/index.html" file should exist
+    And I should see "Dairy Products" in "_site/index.html"
+
   Scenario: should be backward compatible with site.data in _config.yml
     Given I have a "_config.yml" file with content:
       """

lib/jekyll/site.rb

@@ -196,17 +196,33 @@ module Jekyll
     # Returns nothing
     def read_data(dir)
       base = File.join(source, dir)
-      return unless File.directory?(base) && (!safe || !File.symlink?(base))
+      read_data_to(base, self.data)
+    end
 
-      entries = Dir.chdir(base) { Dir['*.{yaml,yml,json}'] }
+    # Read and parse all yaml files under <dir> and add them to the
-      entries.delete_if { |e| File.directory?(File.join(base, e)) }
+    # <data> variable.
+    #
+    # dir - The string absolute path of the directory to read.
+    # data - The variable to which data will be added.
+    #
+    # Returns nothing
+    def read_data_to(dir, data)
+      return unless File.directory?(dir) && (!safe || !File.symlink?(dir))
+
+      entries = Dir.chdir(dir) do
+        Dir['*.{yaml,yml,json}'] + Dir['*'].select { |fn| File.directory?(fn) }
+      end
 
       entries.each do |entry|
-        path = File.join(source, dir, entry)
+        path = File.join(dir, entry)
         next if File.symlink?(path) && safe
 
         key = sanitize_filename(File.basename(entry, '.*'))
-        self.data[key] = SafeYAML.load_file(path)
+        if File.directory?(path)
+          read_data_to(path, data[key] = {})
+        else
+          data[key] = SafeYAML.load_file(path)
+        end
       end
     end
 

site/docs/datafiles.md

@@ -61,3 +61,47 @@ You can now render the list of members in a template:
 </ul>
 {% endraw %}
 {% endhighlight %}
+
+## Example: Organizations
+
+Data files can also be placed in sub-folders of the `_data` folder. Each folder level will be added to a variable's namespace. The example bellow shows how GitHub organizations could be defined separately in a file under the `orgs` folder:
+
+In `_data/orgs/jekyll.yml`:
+
+{% highlight yaml %}
+username: jekyll
+name: Jekyll
+members:
+  - name: Tom Preston-Werner
+    github: mojombo
+
+  - name: Parker Moore
+    github: parkr
+{% endhighlight %}
+
+In `_data/orgs/doeorg.yml`:
+
+{% highlight yaml %}
+username: doeorg
+name: Doe Org
+members:
+  - name: John Doe
+    github: jdoe
+{% endhighlight %}
+
+The organizations can then be accessed via `site.data.orgs`, followed by the file name:
+
+{% highlight html %}
+{% raw %}
+<ul>
+{% for org in site.data.orgs %}
+  <li>
+    <a href="https://github.com/{{ org.username }}">
+      {{ org.name }}
+    </a>
+    ({{ org.members | size }} members)
+  </li>
+{% endfor %}
+</ul>
+{% endraw %}
+{% endhighlight %}

test/source/_data/categories/dairy.yaml

@@ -0,0 +1,6 @@
+name: Dairy
+products:
+- name: cheese
+  price: 5.3
+- name: milk
+  price: 2.5

test/test_site.rb

@@ -389,6 +389,16 @@ class TestSite < Test::Unit::TestCase
         assert_equal site.site_payload['site']['data']['members'], file_content
       end
 
+      should 'auto load yaml files in subdirectory' do
+        site = Site.new(Jekyll.configuration)
+        site.process
+
+        file_content = SafeYAML.load_file(File.join(source_dir, '_data', 'categories', 'dairy.yaml'))
+
+        assert_equal site.data['categories']['dairy'], file_content
+        assert_equal site.site_payload['site']['data']['categories']['dairy'], file_content
+      end
+
       should "load symlink files in unsafe mode" do
         site = Site.new(Jekyll.configuration.merge({'safe' => false}))
         site.process