jekyll-import 0.1.0.beta3 → 0.1.0.beta4

data/lib/jekyll-import/importers/drupal7.rb

@@ -0,0 +1,102 @@
+# NOTE: This converter requires Sequel and the MySQL gems.
+# The MySQL gem can be difficult to install on OS X. Once you have MySQL
+# installed, running the following commands should work:
+# $ sudo gem install sequel
+# $ sudo gem install mysql -- --with-mysql-config=/usr/local/mysql/bin/mysql_config
+
+module JekyllImport
+  module Importers
+    class Drupal7 < Importer
+      # Reads a MySQL database via Sequel and creates a post file for each story
+      # and blog node.
+      QUERY = "SELECT n.nid, \
+                      n.title, \
+                      fdb.body_value, \
+                      n.created, \
+                      n.status \
+               FROM node AS n, \
+                    field_data_body AS fdb \
+               WHERE (n.type = 'blog' OR n.type = 'story' OR n.type = 'article') \
+               AND n.nid = fdb.entity_id \
+               AND n.vid = fdb.revision_id"
+
+      def self.validate(options)
+        %w[dbname user].each do |option|
+          if options[option].nil?
+            abort "Missing mandatory option --#{option}."
+          end
+        end
+      end
+
+      def self.specify_options(c)
+        c.option 'dbname', '--dbname DB', 'Database name'
+        c.option 'user', '--user USER', 'Database user name'
+        c.option 'password', '--password PW', 'Database user\'s password (default: "")'
+        c.option 'host', '--host HOST', 'Database host name (default: "localhost")'
+        c.option 'prefix', '--prefix PREFIX', 'Table prefix name'
+      end
+
+      def self.require_deps
+        JekyllImport.require_with_fallback(%w[
+          rubygems
+          sequel
+          fileutils
+          safe_yaml
+        ])
+      end
+
+      def self.process(options)
+        dbname = options.fetch('dbname')
+        user   = options.fetch('user')
+        pass   = options.fetch('password', "")
+        host   = options.fetch('host', "localhost")
+        prefix = options.fetch('prefix', "")
+
+        db = Sequel.mysql(dbname, :user => user, :password => pass, :host => host, :encoding => 'utf8')
+
+        unless prefix.empty?
+          QUERY[" node "] = " " + prefix + "node "
+          QUERY[" field_data_body "] = " " + prefix + "field_data_body "
+        end
+
+        FileUtils.mkdir_p "_posts"
+        FileUtils.mkdir_p "_drafts"
+        FileUtils.mkdir_p "_layouts"
+
+        db[QUERY].each do |post|
+          # Get required fields and construct Jekyll compatible name
+          node_id = post[:nid]
+          title = post[:title]
+          content = post[:body_value]
+          created = post[:created]
+          time = Time.at(created)
+          is_published = post[:status] == 1
+          dir = is_published ? "_posts" : "_drafts"
+          slug = title.strip.downcase.gsub(/(&|&amp;)/, ' and ').gsub(/[\s\.\/\\]/, '-').gsub(/[^\w-]/, '').gsub(/[-_]{2,}/, '-').gsub(/^[-_]/, '').gsub(/[-_]$/, '')
+          name = time.strftime("%Y-%m-%d-") + slug + '.md'
+
+          # Get the relevant fields as a hash, delete empty fields and convert
+          # to YAML for the header
+          data = {
+             'layout' => 'default',
+             'title' => title.to_s,
+             'created' => created,
+           }.delete_if { |k,v| v.nil? || v == ''}.to_yaml
+
+          # Write out the data and content to file
+          File.open("#{dir}/#{name}", "w") do |f|
+            f.puts data
+            f.puts "---"
+            f.puts content
+          end
+
+        end
+
+        # TODO: Make dirs & files for nodes of type 'page'
+          # Make refresh pages for these as well
+
+        # TODO: Make refresh dirs & files according to entries in url_alias table
+      end
+    end
+  end
+end

data/lib/jekyll-import/importers/enki.rb

@@ -0,0 +1,76 @@
+# Adapted by Rodrigo Pinto <rodrigopqn@gmail.com>
+# Based on typo.rb by Toby DiPasquale
+
+module JekyllImport
+    module Importers
+    class Enki < Importer
+      SQL = <<-EOS
+        SELECT p.id,
+               p.title,
+               p.slug,
+               p.body,
+               p.published_at as date,
+               p.cached_tag_list as tags
+        FROM posts p
+EOS
+
+      def self.validate(options)
+        %w[dbname user].each do |option|
+          if options[option].nil?
+            abort "Missing mandatory option --#{option}."
+          end
+        end
+      end
+
+      def self.specify_options(c)
+        c.option 'dbname', '--dbname', 'Database name'
+        c.option 'user', '--user', 'Database name'
+        c.option 'password', '--password', 'Database name (default: "")'
+        c.option 'host', '--host', 'Database name'
+      end
+
+      def self.require_deps
+        JekyllImport.require_with_fallback(%w[
+          rubygems
+          sequel
+          fileutils
+          pg
+        ])
+      end
+
+      # Just working with postgres, but can be easily adapted
+      # to work with both mysql and postgres.
+      def self.process(options)
+        dbname = options.fetch('dbname')
+        user   = options.fetch('user')
+        pass   = options.fetch('password', "")
+        host   = options.fetch('host', "localhost")
+
+        FileUtils.mkdir_p('_posts')
+        db = Sequel.postgres(:database => dbname,
+                             :user => user,
+                             :password => pass,
+                             :host => host,
+                             :encoding => 'utf8')
+
+        db[SQL].each do |post|
+          name = [ sprintf("%.04d", post[:date].year),
+                   sprintf("%.02d", post[:date].month),
+                   sprintf("%.02d", post[:date].day),
+                   post[:slug].strip ].join('-')
+          name += '.textile'
+
+          File.open("_posts/#{name}", 'w') do |f|
+            f.puts({ 'layout'   => 'post',
+                     'title'    => post[:title].to_s,
+                     'enki_id'  => post[:id],
+                     'categories'  => post[:tags]
+                   }.delete_if { |k, v| v.nil? || v == '' }.to_yaml)
+            f.puts '---'
+            f.puts post[:body].delete("\r")
+          end
+        end
+      end
+    end
+  end
+end

data/lib/jekyll-import/importers/google_reader.rb

@@ -0,0 +1,68 @@
+# Usage:
+#   (Local file)
+#   ruby -r 'jekyll/jekyll-import/rss' -e "JekyllImport::GoogleReader.process(:source => './somefile/on/your/computer.xml')"
+
+module JekyllImport
+  module Importers
+    class GoogleReader < Importer
+      def self.validate(options)
+        if options['source'].nil?
+          abort "Missing mandatory option --source."
+        end
+      end
+
+      def self.specify_options(c)
+        c.option 'source', '--source', 'Source XML file of Google Reader export'
+      end
+
+      def self.require_deps
+        JekyllImport.require_with_fallback(%w[
+          rubygems
+          rss
+          fileutils
+          safe_yaml
+          open-url
+          rexml/document
+          date
+        ])
+      end
+
+      # Process the import.
+      #
+      # source - a URL or a local file String.
+      #
+      # Returns nothing.
+      def self.process(options)
+        source = options.fetch('source')
+
+        open(source) do |content|
+          feed = RSS::Parser.parse(content)
+
+          raise "There doesn't appear to be any RSS items at the source (#{source}) provided." unless feed
+
+          feed.items.each do |item|
+            title = item.title.content.to_s
+            formatted_date = Date.parse(item.published.to_s)
+            post_name = title.split(%r{ |!|/|:|&|-|$|,}).map do |i|
+              i.downcase if i != ''
+            end.compact.join('-')
+            name = "#{formatted_date}-#{post_name}"
+
+            header = {
+              'layout' => 'post',
+              'title' => title
+            }
+
+            FileUtils.mkdir_p("_posts")
+
+            File.open("_posts/#{name}.html", "w") do |f|
+              f.puts header.to_yaml
+              f.puts "---\n\n"
+              f.puts item.content.content.to_s
+            end
+          end
+        end
+      end
+    end
+  end
+end

data/lib/jekyll-import/importers/joomla.rb

@@ -0,0 +1,83 @@
+# NOTE: This migrator is made for Joomla 1.5 databases.
+# NOTE: This converter requires Sequel and the MySQL gems.
+# The MySQL gem can be difficult to install on OS X. Once you have MySQL
+# installed, running the following commands should work:
+# $ sudo gem install sequel
+# $ sudo gem install mysql -- --with-mysql-config=/usr/local/mysql/bin/mysql_config
+
+module JekyllImport
+  module Importers
+    class Joomla < Importer
+      def self.validate(options)
+        %w[dbname user].each do |option|
+          if options[option].nil?
+            abort "Missing mandatory option --#{option}."
+          end
+        end
+      end
+
+      def self.specify_options(c)
+        c.option 'dbname', '--dbname', 'Database name'
+        c.option 'user', '--user', 'Database user name'
+        c.option 'password', '--password', "Database user's password (default: '')"
+        c.option 'host', '--host', 'Database host name'
+        c.option 'section', '--section', 'Table prefix name'
+        c.option 'prefix', '--prefix', 'Table prefix name'
+      end
+
+      def self.require_deps
+        JekyllImport.require_with_fallback(%w[
+          rubygems
+          sequel
+          fileutils
+          safe_yaml
+        ])
+      end
+
+      def self.process(options)
+        dbname  = options.fetch('dbname')
+        user    = options.fetch('user')
+        pass    = options.fetch('password', '')
+        host    = options.fetch('host', "localhost")
+        section = options.fetch('section', '1')
+        table_prefix = options.fetch('prefix', "jos_")
+
+        db = Sequel.mysql(dbname, :user => user, :password => pass, :host => host, :encoding => 'utf8')
+
+        FileUtils.mkdir_p("_posts")
+
+        # Reads a MySQL database via Sequel and creates a post file for each
+        # post in wp_posts that has post_status = 'publish'. This restriction is
+        # made because 'draft' posts are not guaranteed to have valid dates.
+        query = "SELECT `title`, `alias`, CONCAT(`introtext`,`fulltext`) as content, `created`, `id` FROM #{table_prefix}content WHERE state = '0' OR state = '1' AND sectionid = '#{section}'"
+
+        db[query].each do |post|
+          # Get required fields and construct Jekyll compatible name.
+          title = post[:title]
+          slug = post[:alias]
+          date = post[:created]
+          content = post[:content]
+          name = "%02d-%02d-%02d-%s.markdown" % [date.year, date.month, date.day,
+                                                 slug]
+
+          # Get the relevant fields as a hash, delete empty fields and convert
+          # to YAML for the header.
+          data = {
+             'layout' => 'post',
+             'title' => title.to_s,
+             'joomla_id' => post[:id],
+             'joomla_url' => post[:alias],
+             'date' => date
+           }.delete_if { |k,v| v.nil? || v == '' }.to_yaml
+
+          # Write out the data and content to file
+          File.open("_posts/#{name}", "w") do |f|
+            f.puts data
+            f.puts "---"
+            f.puts content
+          end
+        end
+      end
+    end
+  end
+end

data/lib/jekyll-import/importers/jrnl.rb

@@ -0,0 +1,127 @@
+# Author: Aniket Pant <me@aniketpant.com>
+
+module JekyllImport
+  module Importers
+    class Jrnl < Importer
+
+      def self.require_deps
+        JekyllImport.require_with_fallback(%w[
+          time
+          rubygems
+          safe_yaml
+        ])
+      end
+
+      def self.specify_options(c)
+        c.option 'file', '--file FILENAME', 'Journal file (default: "~/journal.txt")'
+        c.option 'time_format', '--time_format FORMAT', 'Time format of your journal (default: "%Y-%m-%d %H:%M")'
+        c.option 'extension', '--extension EXT', 'Output extension (default: "md")'
+        c.option 'layout', '--layout NAME', 'Output post layout (default: "post")'
+      end
+
+      # Reads a jrnl file and creates a new post for each entry
+      # The following overrides are available:
+      # :file         path to input file
+      # :time_format  the format used by the jrnl configuration
+      # :extension    the extension format of the output files
+      # :layout       explicitly set the layout of the output
+      def self.process(options)
+        file        = options.fetch('file', "~/journal.txt")
+        time_format = options.fetch('time_format', "%Y-%m-%d %H:%M")
+        extension   = options.fetch('extension', "md")
+        layout      = options.fetch('layout', "post")
+
+        date_length = Time.now.strftime(time_format).length
+
+        # convert relative to absolute if needed
+        file = File.expand_path(file)
+
+        abort "The jrnl file was not found. Please make sure '#{file}' exists. You can specify a different file using the --file switch." unless File.file?(file)
+
+        input = File.read(file)
+        entries = input.split("\n\n");
+
+        entries.each do |entry|
+          # split dateline and body
+          # content[0] has the date and title
+          # content[1] has the post body
+          content = entry.split("\n")
+
+          body = get_post_content(content)
+          date = get_date(content[0], date_length)
+          title = get_title(content[0], date_length)
+          slug = create_slug(title)
+          filename = create_filename(date, slug, extension)
+          meta = create_meta(layout, title, date) # prepare YAML meta data
+
+          write_file(filename, meta, body) # write to file
+        end
+      end
+
+      # strip body from jrnl entry
+      def self.get_post_content(content)
+        return content[1]
+      end
+
+      # strip timestamp from the dateline
+      def self.get_date(content, offset)
+        return content[0, offset]
+      end
+
+      # strip title from the dateline
+      def self.get_title(content, offset)
+        return content[offset + 1, content.length]
+      end
+
+      # generate slug
+      def self.create_slug(title)
+        return title.downcase.strip.gsub(' ', '-').gsub(/[^\w-]/, '')
+      end
+
+      # generate filename
+      def self.create_filename(date, slug, extension)
+        return "#{Time.parse(date).strftime("%Y-%m-%d")}-#{slug}.#{extension}"
+      end
+
+      # Prepare YAML meta data
+      #
+      # layout  - name of the layout
+      # title   - title of the entry
+      # date    - date of entry creation
+      #
+      # Examples
+      #
+      #   create_meta("post", "Entry 1", "2013-01-01 13:00")
+      #   # => "---\nlayout: post\ntitle: Entry 1\ndate: 2013-01-01 13:00\n"
+      #
+      # Returns array converted to YAML
+      def self.create_meta(layout, title, date)
+        data = {
+          'layout'        => layout,
+          'title'         => title,
+          'date'          => Time.parse(date).strftime("%Y-%m-%d %H:%M %z")
+        }.to_yaml
+        return data;
+      end
+
+      # Writes given data to file
+      #
+      # filename    - name of the output file
+      # meta        - YAML header data
+      # body        - jrnl entry content
+      #
+      # Examples
+      #
+      #   write_file("2013-01-01-entry-1.md", "---\nlayout: post\ntitle: Entry 1\ndate: 2013-01-01 13:00\n", "This is the first entry for my new journal")
+      #
+      # Writes file to _posts/filename
+      def self.write_file(filename, meta, body)
+        File.open("_posts/#{filename}", "w") do |f|
+          f.puts meta
+          f.puts "---\n\n"
+          f.puts body
+        end
+      end
+    end
+  end
+end