rdoc-f95 0.0.1

data/lib/rdoc-f95/markup/to_latex.rb

@@ -0,0 +1,328 @@
+require 'rdoc-f95/markup/formatter'
+require 'rdoc-f95/markup/fragments'
+require 'rdoc-f95/markup/inline'
+
+require 'cgi'
+
+##
+# Convert SimpleMarkup to basic LaTeX report format.
+
+class RDocF95::Markup::ToLaTeX < RDocF95::Markup::Formatter
+
+  BS = "\020"   # \
+  OB = "\021"   # {
+  CB = "\022"   # }
+  DL = "\023"   # Dollar
+
+  BACKSLASH   = "#{BS}symbol#{OB}92#{CB}"
+  HAT         = "#{BS}symbol#{OB}94#{CB}"
+  BACKQUOTE   = "#{BS}symbol#{OB}0#{CB}"
+  TILDE       = "#{DL}#{BS}sim#{DL}"
+  LESSTHAN    = "#{DL}<#{DL}"
+  GREATERTHAN = "#{DL}>#{DL}"
+
+  def self.l(str)
+    str.tr('\\', BS).tr('{', OB).tr('}', CB).tr('$', DL)
+  end
+
+  def l(arg)
+    RDocF95::Markup::ToLaTeX.l(arg)
+  end
+
+  LIST_TYPE_TO_LATEX = {
+    :BULLET =>  [ l("\\begin{itemize}"), l("\\end{itemize}") ],
+    :NUMBER =>  [ l("\\begin{enumerate}"), l("\\end{enumerate}"), "\\arabic" ],
+    :UPPERALPHA =>  [ l("\\begin{enumerate}"), l("\\end{enumerate}"), "\\Alph" ],
+    :LOWERALPHA =>  [ l("\\begin{enumerate}"), l("\\end{enumerate}"), "\\alph" ],
+    :LABELED => [ l("\\begin{description}"), l("\\end{description}") ],
+    :NOTE    => [
+      l("\\begin{tabularx}{\\linewidth}{@{} l X @{}}"),
+      l("\\end{tabularx}") ],
+  }
+
+  InlineTag = Struct.new(:bit, :on, :off)
+
+  def initialize
+    init_tags
+    @list_depth = 0
+    @prev_list_types = []
+  end
+
+  ##
+  # Set up the standard mapping of attributes to LaTeX
+
+  def init_tags
+    @attr_tags = [
+      InlineTag.new(RDocF95::Markup::Attribute.bitmap_for(:BOLD), l("\\textbf{"), l("}")),
+      InlineTag.new(RDocF95::Markup::Attribute.bitmap_for(:TT),   l("\\texttt{"), l("}")),
+      InlineTag.new(RDocF95::Markup::Attribute.bitmap_for(:EM),   l("\\emph{"), l("}")),
+    ]
+  end
+
+  ##
+  # Escape a LaTeX string
+
+  def escape(str)
+    $stderr.print "FE: ", str if $DEBUG_RDOC
+    s = str.
+       sub(/\s+$/, '').
+      gsub(/([_\${}&%#])/, "#{BS}\\1").
+      gsub(/\\/, BACKSLASH).
+      gsub(/\^/, HAT).
+      gsub(/~/,  TILDE).
+      gsub(/</,  LESSTHAN).
+      gsub(/>/,  GREATERTHAN).
+      gsub(/,,/, ",{},").
+      gsub(/\`/,  BACKQUOTE)
+    $stderr.print "-> ", s, "\n" if $DEBUG_RDOC
+    s
+  end
+
+  ##
+  # Add a new set of LaTeX tags for an attribute. We allow
+  # separate start and end tags for flexibility
+
+  def add_tag(name, start, stop)
+    @attr_tags << InlineTag.new(RDocF95::Markup::Attribute.bitmap_for(name), start, stop)
+  end
+
+  ##
+  # Here's the client side of the visitor pattern
+
+  def start_accepting
+    @res = ""
+    @in_list_entry = []
+  end
+
+  def end_accepting
+    @res.tr(BS, '\\').tr(OB, '{').tr(CB, '}').tr(DL, '$')
+  end
+
+  def accept_paragraph(am, fragment)
+    @res << wrap(convert_flow(am.flow(fragment.txt)))
+    @res << "\n"
+  end
+
+  def accept_verbatim(am, fragment)
+    @res << "\n\\begin{code}\n"
+    @res << fragment.txt.sub(/[\n\s]+\Z/, '')
+    @res << "\n\\end{code}\n\n"
+  end
+
+  def accept_rule(am, fragment)
+    size = fragment.param
+    size = 10 if size > 10
+    @res << "\n\n\\rule{\\linewidth}{#{size}pt}\n\n"
+  end
+
+  def accept_list_start(am, fragment)
+    @res << list_name(fragment.type, true) << "\n"
+    @in_list_entry.push false
+  end
+
+  def accept_list_end(am, fragment)
+    if tag = @in_list_entry.pop
+      @res << tag << "\n"
+    end
+    @res << list_name(fragment.type, false) << "\n"
+  end
+
+  def accept_list_item(am, fragment)
+    if tag = @in_list_entry.last
+      @res << tag << "\n"
+    end
+    @res << list_item_start(am, fragment)
+    @res << wrap(convert_flow(am.flow(fragment.txt))) << "\n"
+    @in_list_entry[-1] = list_end_for(fragment.type)
+  end
+
+  def accept_blank_line(am, fragment)
+    # @res << "\n"
+  end
+
+  def accept_heading(am, fragment)
+    @res << convert_heading(fragment.head_level, am.flow(fragment.txt))
+  end
+
+  ##
+  # This is a higher speed (if messier) version of wrap
+
+  def wrap(txt, line_len = 76)
+    res = ""
+    sp = 0
+    ep = txt.length
+    while sp < ep
+      # scan back for a space
+      p = sp + line_len - 1
+      if p >= ep
+        p = ep
+      else
+        while p > sp and txt[p] != ?\s
+          p -= 1
+        end
+        if p <= sp
+          p = sp + line_len
+          while p < ep and txt[p] != ?\s
+            p += 1
+          end
+        end
+      end
+      res << txt[sp...p] << "\n"
+      sp = p
+      sp += 1 while sp < ep and txt[sp] == ?\s
+    end
+    res
+  end
+
+  private
+
+  def on_tags(res, item)
+    attr_mask = item.turn_on
+    return if attr_mask.zero?
+
+    @attr_tags.each do |tag|
+      if attr_mask & tag.bit != 0
+        res << tag.on
+      end
+    end
+  end
+
+  def off_tags(res, item)
+    attr_mask = item.turn_off
+    return if attr_mask.zero?
+
+    @attr_tags.reverse_each do |tag|
+      if attr_mask & tag.bit != 0
+        res << tag.off
+      end
+    end
+  end
+
+  def convert_flow(flow)
+    res = ""
+    flow.each do |item|
+      case item
+      when String
+        $stderr.puts "Converting '#{item}'" if $DEBUG_RDOC
+        res << convert_string(item)
+      when AttrChanger
+        off_tags(res, item)
+        on_tags(res,  item)
+      when Special
+        res << convert_special(item)
+      else
+        raise "Unknown flow element: #{item.inspect}"
+      end
+    end
+    res
+  end
+
+  ##
+  # some of these patterns are taken from SmartyPants...
+
+  def convert_string(item)
+    escape(item).
+
+    # convert ... to elipsis (and make sure .... becomes .<elipsis>)
+      gsub(/\.\.\.\./, '.\ldots{}').gsub(/\.\.\./, '\ldots{}').
+
+    # convert single closing quote
+      gsub(%r{([^ \t\r\n\[\{\(])\'}, '\1\'').
+      gsub(%r{\'(?=\W|s\b)}, "'" ).
+
+    # convert single opening quote
+      gsub(/'/, '`').
+
+    # convert double closing quote
+      gsub(%r{([^ \t\r\n\[\{\(])\"(?=\W)}, "\\1''").
+
+    # convert double opening quote
+      gsub(/"/, "``").
+
+    # convert copyright
+      gsub(/\(c\)/, '\copyright{}')
+
+  end
+
+  def convert_special(special)
+    handled = false
+    Attribute.each_name_of(special.type) do |name|
+      method_name = "handle_special_#{name}"
+      if self.respond_to? method_name
+        special.text = send(method_name, special)
+        handled = true
+      end
+    end
+    raise "Unhandled special: #{special}" unless handled
+    special.text
+  end
+
+  def convert_heading(level, flow)
+    res =
+      case level
+      when 1 then "\\chapter{"
+      when 2 then "\\section{"
+      when 3 then "\\subsection{"
+      when 4 then "\\subsubsection{"
+      else  "\\paragraph{"
+      end +
+      convert_flow(flow) +
+      "}\n"
+  end
+
+  def list_name(list_type, is_open_tag)
+    tags = LIST_TYPE_TO_LATEX[list_type] || raise("Invalid list type: #{list_type.inspect}")
+    if tags[2] # enumerate
+      if is_open_tag
+        @list_depth += 1
+        if @prev_list_types[@list_depth] != tags[2]
+          case @list_depth
+          when 1
+            roman = "i"
+          when 2
+            roman = "ii"
+          when 3
+            roman = "iii"
+          when 4
+            roman = "iv"
+          else
+            raise("Too deep list: level #{@list_depth}")
+          end
+          @prev_list_types[@list_depth] = tags[2]
+          return l("\\renewcommand{\\labelenum#{roman}}{#{tags[2]}{enum#{roman}}}") + "\n" + tags[0]
+        end
+      else
+        @list_depth -= 1
+      end
+    end
+    tags[ is_open_tag ? 0 : 1]
+  end
+
+  def list_item_start(am, fragment)
+    case fragment.type
+    when :BULLET, :NUMBER, :UPPERALPHA, :LOWERALPHA then
+      "\\item "
+
+    when :LABELED then
+      "\\item[" + convert_flow(am.flow(fragment.param)) + "] "
+
+    when :NOTE then
+        convert_flow(am.flow(fragment.param)) + " & "
+    else
+      raise "Invalid list type"
+    end
+  end
+
+  def list_end_for(fragment_type)
+    case fragment_type
+    when :BULLET, :NUMBER, :UPPERALPHA, :LOWERALPHA, :LABELED then
+      ""
+    when :NOTE
+      "\\\\\n"
+    else
+      raise "Invalid list type"
+    end
+  end
+
+end
+

data/lib/rdoc-f95/markup/to_test.rb

@@ -0,0 +1,50 @@
+require 'rdoc-f95/markup'
+require 'rdoc-f95/markup/formatter'
+
+##
+# This Markup outputter is used for testing purposes.
+
+class RDocF95::Markup::ToTest < RDocF95::Markup::Formatter
+
+  def start_accepting
+    @res = []
+  end
+
+  def end_accepting
+    @res
+  end
+
+  def accept_paragraph(am, fragment)
+    @res << fragment.to_s
+  end
+
+  def accept_verbatim(am, fragment)
+    @res << fragment.to_s
+  end
+
+  def accept_list_start(am, fragment)
+    @res << fragment.to_s
+  end
+
+  def accept_list_end(am, fragment)
+    @res << fragment.to_s
+  end
+
+  def accept_list_item(am, fragment)
+    @res << fragment.to_s
+  end
+
+  def accept_blank_line(am, fragment)
+    @res << fragment.to_s
+  end
+
+  def accept_heading(am, fragment)
+    @res << fragment.to_s
+  end
+
+  def accept_rule(am, fragment)
+    @res << fragment.to_s
+  end
+
+end
+

data/lib/rdoc-f95/markup/to_xhtml_texparser.rb

@@ -0,0 +1,234 @@
+require 'fileutils'
+require 'rdoc-f95/markup/to_html_crossref'
+require 'rdoc-f95/markup/mathml_wrapper'
+
+##
+#<b>Note that Japanese and English are described in parallel.</b>
+#
+#== TeX �ο����� MathML ���Ѵ�
+#
+#TeX �ǵ��Ҥ��줿������ MathML ���Ѵ����ޤ�.
+#����饤���ɽ�����������, TeX �ο�����ʲ��Τ褦�� $ ... $ �Ǥ����ä�
+#���Ҥ��Ƥ�������. $ ��
+#����ˤ�Ⱦ�Ѷ�����ʸ���ʾ�����Ʋ�����.
+#(�ʤ�, "$ID: ... $" �� "$LOG: ... $
+#�Ȥ��ä�, CVS �Υ�����ɤȤ���
+#�Ѥ����Ƥ�������Ͽ����Ȥ��ư����ޤ���.)
+#
+#  ����饤���ɽ����������� $ f(x) = x^2 + 1 $ �Τ褦�˵��Ҥ��ޤ�.
+#  ($ ������˶����˺��ʤ�).
+#
+#�֥��å���ɽ��������, �ʲ��Τ褦�� \[ ��
+#\] �ȤǤ����äƵ��Ҥ��Ƥ�������. \[, ��ɬ����Ƭ�˵��Ҥ��Ƥ�������.
+#
+#  \[
+#     \sum_{i=1}^nf_n(x)
+#  \]
+#
+#������ʣ���Ԥ�ɽ��������ˤ�, ���Ԥ�����ʬ�� "\] \[" �򵭽Ҥ��Ƥ�������.
+#
+# \[
+#     d\zeta/dt + J(\psi,\zeta) = Ra \; dT/dx + \nabla\zeta, \] \[
+#     dT/dt + J(\psi,T) - d\psi/dx = \nabla T,               \] \[
+#     \nabla\psi = \zeta
+# \]
+#
+#TeX �ο������� MathML �Ѵ��ˤ�
+#<b>Ruby �� MathML �饤�֥��ΥС������ 0.8</b> ����Ѥ��Ƥ��ޤ�.
+#���Υ饤�֥���{�Ҥ餯�ι�˼}[http://www.hinet.mydns.jp/~hiraku/]
+#�ˤƸ�������Ƥ��ޤ�. ���ѤǤ��� TeX ���ޥ�ɤξܺ٤˴ؤ��Ƥ�
+#������Υ����Ȥ򻲾Ȥ��Ƥ�������.
+#
+#�������줿�ɥ�����Ȥ��������ݤˤ� MathML ���б�����
+#�֥饦������Ѥ���ɬ�פ�
+#����ޤ�. {MathML ���ܸ����}[http://washitake.com/MathML/] 
+#�� {MathML Software - Browsers}[http://www.w3.org/Math/Software/mathml_software_cat_browsers.html]
+#�ʤɤ򻲾Ȥ��Ƥ�������.
+#
+#
+#== TeX is converted to MathML
+#
+#TeX formula is converted to MathML.
+#When inline display, TeX formula should be bundled by $ ... $
+#as follows. 
+#One or more normal-width blank is necessary before and behind "$".
+#(The format of CVS keywords, that is "$ID: ... $" or
+#"$LOG: ... $ etc. is ignored.)
+#
+#  Inline formula is $ f(x) = x^2 + 1 $ .
+#
+#When block display, TeX formula should be bundled by \[ ... \]
+#as follows. 
+#Describe \[ at the head of line.
+#
+#  \[
+#     \sum_{i=1}^nf_n(x)
+#  \]
+#
+#To write equations across multiple lines, describe "\] \["
+#as follows.
+#
+# \[
+#     d\zeta/dt + J(\psi,\zeta) = Ra \; dT/dx + \nabla\zeta, \] \[
+#     dT/dt + J(\psi,T) - d\psi/dx = \nabla T,               \] \[
+#     \nabla\psi = \zeta
+# \]
+#
+#<b>MathML library for Ruby version 0.8</b> is needed to
+#convert TeX formula to MathML.
+#This library is available from {Bottega of Hiraku (JAPANESE only)}[http://www.hinet.mydns.jp/~hiraku/].
+#See this site about available TeX commands.
+#
+#When you browse generated documents, you need to use 
+#browers that can handle MathML.
+#See {MathML Software - Browsers}[http://www.w3.org/Math/Software/mathml_software_cat_browsers.html], etc.
+#
+#
+#== \newcommand, \newenvironment �λ�����ˡ
+#
+#\newcommand �� \newenvironment ̿��ˤ��ޥ�������Ѥ���ˤ�,
+#Ruby �Υ����ɥѥ��ʲ��� '<b><tt>mathml/macro</tt></b>' �ǥ��쥯�ȥ��
+#������, ���Υǥ��쥯�ȥ�ʲ��˥ޥ���̿��򵭤����ե�������֤��Ƥ�������.
+#�ե�����̾���䤤�ޤ���.
+#
+#�㤨��, '<tt>/usr/lib/ruby/1.8</tt>' �� Ruby �Υ����ɥѥ��Ǥ�����,
+#'<tt>/usr/lib/ruby/1.8/mathml</tt>' �����
+#'<tt>/usr/lib/ruby/1.8/mathml/macro</tt>' �ǥ��쥯�ȥ�������,
+#���Υǥ��쥯�ȥ�ʲ���, 'D6math.sty' �Ȥ����ե������������ޤ�
+#(���Ҥ����褦�ˤ��Υե�����̾�ϲ��Ǥ⹽���ޤ���). ���Υե��������
+#�ʲ��Τ褦�� \newcommand ̿��򵭽Ҥ��Ƥ������Ȥ�, '<b>\DP{}{}</b>'
+#(����ʬ���ñ�˵��Ҥ��뤿��Υޥ���) ���ޥ�ɤ����Ѳ�ǽ�ˤʤ�ޤ�.
+#
+#    \newcommand{\DP}[2]{\frac{\partial #1}{\partial #2}}
+#
+#�ϵ�ή����Ǿ��������󶡤��Ƥ��� {TeX �ޥ��� (�̾�: ��Ǿ��������)}[http://www.gfd-dennou.org/library/cc-env/TeXmacro/dennou/SIGEN.htm]
+#�˴ޤޤ�� 'D6math.sty' �� Ruby �� Mathml �饤�֥��ǻ��ѤǤ���褦��
+#���������ѥå������� libmathml-macro-dennou-ruby[http://www.gfd-dennou.org/library/cc-env/libmathml-macro-dennou-ruby/debian/stable/] �Ȥ���
+#�󶡤��Ƥ��ޤ�. ����ץ�Ȥ������Ѥ��ƤߤƤ�������.
+#
+#
+#== Usage of \newcommand and \newenvironment
+#
+#If you want to use macros defined by \newcommand and 
+#\newenvironment commands, make '<b><tt>mathml/macro</tt></b>' directory
+#under load paths of Ruby, and prepare a file that macro commands are 
+#described in the directory. A name of the file is free.
+#
+#For example, if your load path of Ruby is '<tt>/usr/lib/ruby/1.8</tt>',
+#you should make '<tt>/usr/lib/ruby/1.8/mathml</tt>' and
+#'<tt>/usr/lib/ruby/1.8/mathml/macro</tt>' directories,
+#and make 'D6math.sty' file (already mentioned, the file name is free).
+#When you describe a following \newcommand command, you can use
+#'<b>\DP{}{}</b>' (a macro for partial differentiations) command.
+#
+#    \newcommand{\DP}[2]{\frac{\partial #1}{\partial #2}}
+#
+#As a sample, please use libmathml-macro-dennou-ruby[http://www.gfd-dennou.org/library/cc-env/libmathml-macro-dennou-ruby/debian/stable/].
+#The original style file is 'D6math.sty' in {TeX macro (Dennou style)}[http://www.gfd-dennou.org/library/cc-env/TeXmacro/dennou/SIGEN.htm].
+#libmathml-macro-dennou-ruby is a reconfigured package for Mathml library for Ruby.
+#
+#
+class RDocF95::Markup::ToXHtmlTexParser < RDocF95::Markup::ToHtmlCrossref
+
+  attr_accessor :context
+  attr_reader   :block_exceptions
+
+  ##
+  # We need to record the html path of our caller so we can generate
+  # correct relative paths for any hyperlinks that we find
+
+  def initialize(from_path, context, show_hash, mathml=nil)
+    super(from_path, context, show_hash)
+    @mathml = mathml
+
+    if @mathml
+      # TeX inline form
+      @markup.add_special(/(\$(.*?)[^\\]\$)/im, :TEXINLINE)
+
+      # TeX inline delimiter
+      @markup.add_special(/(\\\$)/im, :TEXINLINEDELIMITER)
+
+      # TeX block form
+      @markup.add_special(/(\\\[(.+?)\\\])/im, :TEXBLOCK)
+    end
+
+    @block_exceptions = []
+    if @markup
+      @block_exceptions << {
+        'name'     => :texblockform,
+        'start'    => Regexp.new("^\\\\\\["),
+        'end'      => Regexp.new("\\\\\\]$"),
+        'replaces' => []
+      }
+      @block_exceptions[0]['replaces'] << {
+        'from' => Regexp.new("\\\\\\\\"),
+        'to'   => "\\\\\\\\\\\\\\\\",
+      }
+    end
+
+  end
+
+  def file_location
+    if @context.context.parent
+      class_or_method = @context.context.name
+    end
+    context = @context.context
+    while context.parent
+      context = context.parent
+    end
+    location = context.file_relative_name
+    if class_or_method
+      location += "#"+class_or_method
+    end
+    return location
+  end
+
+  ##
+  # TEXINLINE pattern $...$ is converted to MathML format
+  # when --mathml option is given.
+  #
+  def handle_special_TEXINLINE(special)
+    text = special.text
+    return text unless @mathml
+    raw_text = text.scan(/^.*?\$/).to_s.sub(/\$$/, '')
+    return text if text =~ /^.*?\$[A-Z]\w+:/  # CVS keywords are skipped
+    text.sub!(/^.*?\$/, '')
+    text.sub!(/\$$/, '')
+    tex = MathMLWrapper.new
+    mathml, stat = tex.parse(text)
+    if !(stat == 0)
+      $stderr.puts "Warning: in #{file_location}, following TeX commands can not be converted to MathML\n\n",
+      "    #{text}\n\n"
+    end
+    return raw_text + mathml
+  end
+
+  ##
+  # TEXINLINEDELIMITER pattern "\$" is converted to single dollar "$"
+  # when --mathml option is given.
+  #
+  def handle_special_TEXINLINEDELIMITER(special)
+    text = special.text
+    return text unless @mathml
+    return text.gsub(/\\\$/, '$')
+  end
+
+  ##
+  # TEXBLOCK pattern \[...\] is converted to MathML format
+  # when --mathml option is given.
+  #
+  def handle_special_TEXBLOCK(special)
+    text = special.text
+    return text unless @mathml
+    text.sub!(/^\\\[/, '')
+    text.sub!(/\\\]$/, '')
+    tex = MathMLWrapper.new
+    mathml, stat = tex.parse(text, true)
+    if !(stat == 0)
+      $stderr.puts "Warning: in #{file_location}, following TeX commands can not be converted to MathML\n\n",
+      "    #{text}\n\n"
+    end
+    return mathml
+  end
+
+end