Question

I am beginner in Ruby and having trouble understanding this code

    require_relative 'custom_page'

module Jekyll   
  class Tag < CustomPage
    def initialize(site, base, dir, tag)
      super site, base, dir, 'tag'

      self.data['tag'] = tag
      self.data['title'] = "#{site.config['tag_title_prefix'] || 'Tag: '}#{tag}"
      self.data['description'] = "#{site.config['tag_meta_description_prefix'] || 'Tag: '}#{tag}"
    end
  end

  class Tags < CustomPage
    def initialize(site, base, dir)
      super site, base, dir, 'tags'
      self.data['tags'] = site.categories.keys.sort
      #1# puts self.data['tags']
    end
  end

  class Site
    # generate_tags_categories is called by the custom process function in site_process.rb

    def generate_tags_categories            
      dir = self.config['tag_dir'] || 'tags'
      write_page Tags.new(self, self.source, dir) if self.layouts.key? 'tags'        

      self.categories.keys.each do |tag|
      puts "dd"
      #2# puts tag    
      write_page Tag.new(self, self.source, File.join(dir, tag.slugize), tag)
      end
    end
  end
end

In the above code, the statement puts self.data['tags'] (marked 1) outputs more than 10 values as expected. However, the line puts tag (marked 2) outputs only one value implying that array contains only one value. Isn’t self.categories.keys.each expected to loop through all the values which itself is assigned to self.data['tags']

Answer 1

You may start with ensuring that the ‘categories’ still contain many values before entering the loop:

puts "categories.keys: #{self.categories.keys.inspect}" # <<- here
self.categories.keys.each do |tag|
  puts "dd"
  #2# puts tag

If the first puts shows you more than one value, and the loop is called only once, then you may want to investigate what possibly makes the loop break. Maybe write_page throws an exception, catch somewhere up the calling stack?

When you debug some values, it’s better to use inspect method instead of (automagically called by puts) method to_s. The output of inspect is more debug-friendly.

It is possible, (however unlikely) that the ‘tag’ contains some control characters which clear the previous output, or something like that. The inspect is always safe. (ok, not always but in most of cases 😉