pandoku 0.1

data/lib/pandoku.rb

@@ -0,0 +1,9 @@
+require File.dirname(__FILE__) + '/pandoku/document'
+require File.dirname(__FILE__) + '/pandoku/format'
+
+# %w<markdown rst html latex
+# context man mediawiki texinfo docbook opendocument odt s5 rtf>.each do |f|
+%w<markdown html>.each do |f|
+  require "#{File.dirname(__FILE__)}/pandoku/formats/#{f}"
+end
+

data/lib/pandoku/document.rb

@@ -0,0 +1,37 @@
+require File.dirname(__FILE__) + '/format'
+
+module Pandoku
+  PANDOC_PATH = 'pandoc'
+
+  class Document
+    attr_reader :format, :text
+
+    def initialize(format, text)
+      unless format.is_a?(InputFormat)
+        raise TypeError, 'format must be InputFormat'
+      end
+      @format = format
+      @text = text
+    end
+
+    # Makes a command string to execute Pandoc.
+    def command_for(format)
+      <<-CMD
+        #{PANDOC_PATH} -f #{self.format.class.name}
+                       -t #{format.class.name}
+                       #{self.format.cliopts}
+                       #{format.cliopts}
+      CMD
+    end
+
+    # Compiles the document to given +format+.
+    # If a second argument +io+ is true, returns +IO+ instead of +String+.
+    def compile(format, io = false)
+      unless format.is_a?(OutputFormat)
+        raise TypeError, 'format must be OutputFormat'
+      end
+      format.compile(self, io)
+    end
+  end
+end
+

data/lib/pandoku/format.rb

@@ -0,0 +1,54 @@
+require 'open3'
+require File.dirname(__FILE__) + '/document'
+
+module Pandoku
+  class Format
+    attr_reader :options
+
+    def self.name
+      raise NotImplementedError,
+            'Format.name should be overrided to return Symbol'
+    end
+
+    def self.default_options
+      {}
+    end
+
+    def initialize(options = {})
+      @options = self.class.default_options
+      keys = @options.keys
+      @options.merge!(Hash[options.select {|k, v| keys.include?(k) }])
+    end
+
+    def cliopts
+      escapeshellarg = lambda do |arg|
+        "'" + arg.gsub(/[^\\]'/) {|s| %<#{s.chars.first}\\'> } + "'"
+      end
+      @options.select {|k, v| v } \
+              .collect {|p| %<--#{p[0].gsub('_', '-')}> +
+                            (p[1] != true ? %<=#{p[1]}> : '') } \
+              .join(' ')
+    end
+  end
+
+  module InputFormat
+    def parse(text)
+      Document.new(self, text)
+    end
+  end
+
+  module OutputFormat
+    # Compiles the given +document+ to the format.
+    # If a second argument +io+ is true, returns +IO+ instead of +String+.
+    def compile(document, io = false)
+      cin, cout, cerr = Open3.popen3(document.command_for(self))
+      cin.print(document.text)
+      cin.close
+      return cout if io
+      result = cout.read
+      cout.close
+      return result
+    end
+  end
+end
+

data/lib/pandoku/formats/html.rb

@@ -0,0 +1,66 @@
+require File.dirname(File.dirname(__FILE__)) + '/format'
+
+module Pandoku::Formats
+  # HTML format.
+  #
+  # == Available Options
+  #
+  # * <tt>:standalone</tt> - Produce output with an appropriate header
+  #   and footer.
+  # * <tt>:parse_raw</tt> - Parse untranslatable HTML codes as raw HTML,
+  #   instead of ignoring them.
+  # * <tt>latexmathml</tt> - Use LaTeXMathML to display embedded TeX math in
+  #   HTML output. To insert a link to a local copy of the <b>LaTeXMathML.js</b>
+  #   script, provide a URL string. If just <tt>true</tt> is provided, the
+  #   contents of the script will be inserted directly into the HTML header.
+  # * <tt>:jsmath</tt> - Use jsMath to display embedded TeX math in HTML output.
+  #   The URL should point to the jsMath load script; if provided,
+  #   it will be linked to in the header of standalone HTML documents.
+  # * <tt>:gladtex</tt> - Enclose TeX math in +<eq>+ tags in HTML output.
+  #   These can then be processed by gladTeX to produce links
+  #   to images of the typeset formulas.
+  # * <tt>:mimetex</tt> - Render TeX math using the mimeTeX CGI script.
+  #   If URL is not specified, it is assumed that the script is at
+  #   <b>/cgi-bin/mimetex.cgi</b>.
+  # * <tt>:no_wrap</tt> - Disable text wrapping in output.
+  #   (Default is to wrap text.)
+  # * <tt>:sanitize_html</tt> - Sanitizes HTML using a whitelist.
+  #   Unsafe tags are replaced by HTML comments; unsafe attributes are omitted.
+  #   URIs in links and images are also checked against a whitelist of URI
+  #   schemes.
+  # * <tt>:email_obfuscation</tt> -
+  #   (<tt>nil</tt>|<tt>:javascript</tt>|<tt>:references</tt>)
+  #   Specify a method for obfuscating <tt>mailto:</tt> links in HTML documents.
+  #   * <tt>:nil</tt> leaves <tt>mailto:</tt> links as they are.
+  #   * <tt>:javascript</tt> obfuscates them using javascript.
+  #   * <tt>:references</tt> obfuscates them by printing their letters
+  #     as decimal or hexadecimal character references.
+  #     If <tt>:strict => true</tt> is specified, references is used regardless
+  #     of the presence of this option.
+  # * <tt>:toc</tt> - Include an automatically generated table of contents.
+  # * <tt>:css</tt> -  Link to a CSS style sheet.
+  #   The pathname of the style sheet goes value.
+  # * <tt>:include_in_header</tt> - Include contents of the given filename
+  #   at the end of the header. Implies <tt>:standalone</tt>.
+  # * <tt>:include_before_body</tt> - Include contents of the given filename
+  #   at the beginning of the document body.
+  # * <tt>:include_after_body</tt> - Include contents of the given filename
+  #   at the end of the document body.
+  # * <tt>:custom_header</tt> - Use contents of the given filename as the
+  #   document header. Implies <tt>:standalone</tt>.
+  class HTML < Pandoku::Format
+    include Pandoku::InputFormat, Pandoku::OutputFormat
+
+    def self.name
+      :html
+    end
+
+    def self.default_options
+      { :standalone => false, :parse_raw => false, :latexmathml => nil,
+        :jsmath => nil, :gladtex => false, :mimetex => nil, :no_wrap => false,
+        :sanitize_html => false, :email_obfuscation => false, :toc => false,
+        :css => nil, :include_in_header => nil, :include_before_body => nil,
+        :include_after_body => nil, :custom_header => nil }
+    end
+  end
+end

data/lib/pandoku/formats/markdown.rb

@@ -0,0 +1,31 @@
+require File.dirname(File.dirname(__FILE__)) + '/format'
+
+module Pandoku::Formats
+  # Markdown format. http://daringfireball.net/projects/markdown/
+  #
+  # == Available Options
+  #
+  # * <tt>:strict</tt> - Use strict syntax, with no extensions or variants.
+  # * <tt>:reference_links</tt> - Use reference-style links,
+  #   rather than inline links, in writing.
+  # * <tt>:smart</tt> - Use smart quotes, dashes, and ellipses.
+  # * <tt>:no_wrap</tt> - Disable text wrapping in output.
+  #   (Default is to wrap text.)
+  # * <tt>:sanitize_html</tt> - Sanitizes HTML using a whitelist.
+  #   Unsafe tags are replaced by HTML comments; unsafe attributes are omitted.
+  #   URIs in links and images are also checked against a whitelist of URI
+  #   schemes.
+  # * <tt>:toc</tt> - Include an automatically generated table of contents.
+  class Markdown < Pandoku::Format
+    include Pandoku::InputFormat, Pandoku::OutputFormat
+
+    def self.name
+      :markdown
+    end
+
+    def self.default_options
+      { :strict => false, :reference_links => false, :smart => false,
+        :no_wrap => false, :sanitize_html => false, :toc => false }
+    end
+  end
+end

data/lib/pandoku/formats/odt.rb

@@ -0,0 +1,27 @@
+require File.dirname(File.dirname(__FILE__)) + '/format'
+
+module Pandoku::Formats
+  # ODT (OpenOffice text document) format.
+  class ODT
+    include OutputFormat
+
+    def self.name
+      :odt
+    end
+
+    def cliopts
+      @tmpfile = @tmpfile || self._tmpfile
+      "--output=#{@tmpfile} #{super.cliopts}"
+    end
+
+    def compile(document)
+      @tmpfile = self._tmpfile
+    end
+
+    private
+
+    def _tmpfile
+      Dir.tmpdir
+    end
+  end
+end

metadata

@@ -0,0 +1,60 @@
+--- !ruby/object:Gem::Specification 
+name: pandoku
+version: !ruby/object:Gem::Version 
+  version: "0.1"
+platform: ruby
+authors: 
+- Hong, MinHee
+autorequire: 
+bindir: bin
+cert_chain: []
+
+date: 2009-10-08 00:00:00 +09:00
+default_executable: 
+dependencies: []
+
+description: Loose Ruby interface for Pandoc, the most powerful markup processor.
+email: minhee@dahlia.kr
+executables: []
+
+extensions: []
+
+extra_rdoc_files: []
+
+files: 
+- lib/pandoku.rb
+- lib/pandoku/formats/odt.rb
+- lib/pandoku/formats/html.rb
+- lib/pandoku/formats/markdown.rb
+- lib/pandoku/document.rb
+- lib/pandoku/format.rb
+has_rdoc: true
+homepage: http://github.com/dahlia/pandoku
+licenses: []
+
+post_install_message: 
+rdoc_options: []
+
+require_paths: 
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement 
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      version: "0"
+  version: 
+required_rubygems_version: !ruby/object:Gem::Requirement 
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      version: "0"
+  version: 
+requirements: []
+
+rubyforge_project: 
+rubygems_version: 1.3.5
+signing_key: 
+specification_version: 3
+summary: Loose Ruby interface for Pandoc, the most powerful markup processor.
+test_files: []
+