kramdown 0.12.0 → 0.13.1

data/doc/syntax.page

@@ -688,12 +688,12 @@ There are four different line types that can be used in a table:
 
   A table row is any line that contains at least one pipe character and is not identified as any
   other type of table line! The table row is divided into individual table cells by pipe characters.
-  An optional trailing pipe character is ignored.
+  An optional trailing pipe character is ignored. Note that literal pipe characters need to be
+  escaped *except* if they occur in code spans or HTML `<code>` elements!
 
   Header rows, footer rows and normal rows are all done using these table rows. Table cells can only
   contain a single line of text, no multi-line text is supported. The text of a table cell is parsed
-  as span-level elements. Note that literal pipe characters need to be escaped except if they occur
-  in code spans!
+  as span-level elements.
 
   Here are some example table rows:
 
@@ -1108,10 +1108,10 @@ Notes:
 
 To create a reference style link, you need to surround the link text with square brackets (as with
 inline links), followed by optional spaces/tabs/line breaks and then optionally followed with
-another set of square brackets with the link identifier in them. A link identifier may only contain
-numbers, letters, spaces (line breaks and tabs are converted to spaces, multiple spaces are
-compressed to one) and punctuation characters (i.e. `_.:,;!?-`) and is not case sensitive. For
-example:
+another set of square brackets with the link identifier in them. A link identifier may not contain a
+closing bracket and, when specified in a link definition, newline characters; it is also not case
+sensitive, line breaks and tabs are converted to spaces and multiple spaces are compressed into one.
+For example:
 
     This is a [reference style link][linkid] to a page. And [this]
     [linkid] is also a link. As is [this][] and [THIS].
@@ -1144,6 +1144,16 @@ The link definition has the following structure:
 > allowed for consistency with the inline title.
 {: .markdown-difference}
 
+If you have some text that looks like a link definition but should really be a link and some text,
+you can escape the colon after the link identifier:
+
+    The next paragraph contains a link and some text.
+
+    [Room 100]\: There you should find everything you need!
+
+    [Room 100]: link_to_room_100.html
+
+
 ### Images
 
 Images can be specified via a syntax that is similar to the one used by links. The difference is

data/doc/tests.page

@@ -45,7 +45,7 @@ different Ruby interpreters:
 ![ruby 1.8.6p399]({relocatable: img/graph-ruby-1.8.6-399.png})
 ![ruby 1.8.7p249]({relocatable: img/graph-ruby-1.8.7-249.png})
 ![ruby 1.8.7p302]({relocatable: img/graph-ruby-1.8.7-302.png})
-![ruby 1.9.2p0]({relocatable: img/graph-ruby-1.9.2p0-0.png})
+![ruby 1.9.2p136]({relocatable: img/graph-ruby-1.9.2p136-136.png})
 ![jruby 1.5.3]({relocatable: img/graph-jruby-1.5.3-249.png})
 
 [Markdown Test Suite]: http://daringfireball.net/projects/downloads/MarkdownTest_1.0.zip

data/lib/kramdown/converter/base.rb

@@ -132,12 +132,7 @@ module Kramdown
       # Return +true+ if the header element +el+ should be used for the table of contents (as
       # specified by the +toc_levels+ option).
       def in_toc?(el)
-        if @options[:toc_depth] >= 0
-          warn('Option toc_depth is deprecated, use the new option toc_levels instead')
-          @options[:toc_depth] == 0 || el.options[:level] <= @options[:toc_depth]
-        else
-          @options[:toc_levels].include?(el.options[:level])
-        end
+        @options[:toc_levels].include?(el.options[:level])
       end
 
       # Generate an unique alpha-numeric ID from the the string +str+ for use as a header ID.
@@ -159,6 +154,14 @@ module Kramdown
         @options[:auto_id_prefix] + gen_id
       end
 
+      SMART_QUOTE_INDICES = {:lsquo => 0, :rsquo => 1, :ldquo => 2, :rdquo => 3} # :nodoc:
+
+      # Return the entity that represents the given smart_quote element.
+      def smart_quote_entity(el)
+        res = @options[:smart_quotes][SMART_QUOTE_INDICES[el.value]]
+        ::Kramdown::Utils::Entities.entity(res)
+      end
+
     end
 
   end

data/lib/kramdown/converter/html.rb

@@ -178,7 +178,7 @@ module Kramdown
       end
 
       # A list of all HTML tags that need to have a body (even if the body is empty).
-      HTML_TAGS_WITH_BODY=['div', 'script', 'iframe', 'textarea'] # :nodoc:
+      HTML_TAGS_WITH_BODY=['div', 'script', 'iframe', 'textarea', 'a'] # :nodoc:
 
       def convert_html_element(el, indent)
         res = inner(el, indent)
@@ -311,7 +311,7 @@ module Kramdown
       end
 
       def convert_smart_quote(el, indent)
-        entity_to_str(::Kramdown::Utils::Entities.entity(el.value.to_s))
+        entity_to_str(smart_quote_entity(el))
       end
 
       def convert_math(el, indent)

data/lib/kramdown/converter/latex.rb

@@ -84,7 +84,7 @@ module Kramdown
         if el.children.size == 1 && el.children.first.type == :img && !(img = convert_img(el.children.first, opts)).empty?
           convert_standalone_image(el, opts, img)
         else
-          "#{inner(el, opts)}\n\n"
+          "#{latex_link_target(el)}#{inner(el, opts)}\n\n"
         end
       end
 
@@ -92,20 +92,23 @@ module Kramdown
       # <tt>:img</tt> element.
       def convert_standalone_image(el, opts, img)
         attrs = attribute_list(el)
-        "\\begin{figure}#{attrs}\n\\begin{center}\n#{img}\n\\end{center}\n\\caption{#{escape(el.children.first.attr['alt'])}}\n\\end{figure}#{attrs}\n"
+        "\\begin{figure}#{attrs}\n\\begin{center}\n#{img}\n\\end{center}\n\\caption{#{escape(el.children.first.attr['alt'])}}\n#{latex_link_target(el, true)}\n\\end{figure}#{attrs}\n"
       end
 
       def convert_codeblock(el, opts)
         show_whitespace = el.attr['class'].to_s =~ /\bshow-whitespaces\b/
         lang = el.attr['lang']
         if show_whitespace || lang
-          result = "\\lstset{showspaces=%s,showtabs=%s}\n" % (show_whitespace ? ['true', 'true'] : ['false', 'false'])
-          result << "\\lstset{language=#{lang}}\n" if lang
-          result << "\\lstset{basicstyle=\\ttfamily\\footnotesize}\\lstset{columns=fixed,frame=tlbr}\n"
+          options = []
+          options << "showspaces=%s,showtabs=%s" % (show_whitespace ? ['true', 'true'] : ['false', 'false'])
+          options << "language=#{lang}" if lang
+          options << "basicstyle=\\ttfamily\\footnotesize,columns=fixed,frame=tlbr"
+          id = el.attr['id']
+          options << "label=#{id}" if id
           attrs = attribute_list(el)
-          result << "\\begin{lstlisting}#{attrs}\n#{el.value}\n\\end{lstlisting}#{attrs}\n"
+          "#{latex_link_target(el)}\\begin{lstlisting}[#{options.join(',')}]\n#{el.value}\n\\end{lstlisting}#{attrs}\n"
         else
-          "\\begin{verbatim}#{el.value}\\end{verbatim}\n"
+          "#{latex_link_target(el)}\\begin{verbatim}#{el.value}\\end{verbatim}\n"
         end
       end
 
@@ -117,7 +120,7 @@ module Kramdown
         type = @options[:latex_headers][el.options[:level] - 1]
         if ((id = el.attr['id']) ||
             (@options[:auto_ids] && (id = generate_id(el.options[:raw_text])))) && in_toc?(el)
-          "\\hypertarget{#{id}}{}\\#{type}{#{inner(el, opts)}}\\label{#{id}}\n\n"
+          "\\#{type}{#{inner(el, opts)}}\\hypertarget{#{id}}{}\\label{#{id}}\n\n"
         else
           "\\#{type}*{#{inner(el, opts)}}\n\n"
         end
@@ -125,7 +128,7 @@ module Kramdown
 
       def convert_hr(el, opts)
         attrs = attribute_list(el)
-        "\\begin{center}#{attrs}\n\\rule{3in}{0.4pt}\n\\end{center}#{attrs}\n"
+        "#{latex_link_target(el)}\\begin{center}#{attrs}\n\\rule{3in}{0.4pt}\n\\end{center}#{attrs}\n"
       end
 
       def convert_ul(el, opts)
@@ -143,7 +146,7 @@ module Kramdown
       end
 
       def convert_li(el, opts)
-        "\\item #{inner(el, opts).sub(/\n+\Z/, '')}\n"
+        "\\item #{latex_link_target(el, true)}#{inner(el, opts).sub(/\n+\Z/, '')}\n"
       end
 
       def convert_dt(el, opts)
@@ -151,7 +154,7 @@ module Kramdown
       end
 
       def convert_dd(el, opts)
-        "#{inner(el, opts)}\n\n"
+        "#{latex_link_target(el)}#{inner(el, opts)}\n\n"
       end
 
       def convert_html_element(el, opts)
@@ -179,7 +182,7 @@ module Kramdown
       def convert_table(el, opts)
         align = el.options[:alignment].map {|a| TABLE_ALIGNMENT_CHAR[a]}.join('|')
         attrs = attribute_list(el)
-        "\\begin{tabular}{|#{align}|}#{attrs}\n\\hline\n#{inner(el, opts)}\\hline\n\\end{tabular}#{attrs}\n\n"
+        "#{latex_link_target(el)}\\begin{tabular}{|#{align}|}#{attrs}\n\\hline\n#{inner(el, opts)}\\hline\n\\end{tabular}#{attrs}\n\n"
       end
 
       def convert_thead(el, opts)
@@ -227,7 +230,7 @@ module Kramdown
           ''
         elsif !el.options.attr['src'].empty?
           @data[:packages] << 'graphicx'
-          "\\includegraphics{#{el.attr['src']}}"
+          "#{latex_link_target(el)}\\includegraphics{#{el.attr['src']}}"
         else
           warning("Cannot include image with empty path")
           ''
@@ -235,7 +238,7 @@ module Kramdown
       end
 
       def convert_codespan(el, opts)
-        "{\\tt #{escape(el.value)}}"
+        "{\\tt #{latex_link_target(el)}#{escape(el.value)}}"
       end
 
       def convert_footnote(el, opts)
@@ -252,11 +255,11 @@ module Kramdown
       end
 
       def convert_em(el, opts)
-        "\\emph{#{inner(el, opts)}}"
+        "\\emph{#{latex_link_target(el)}#{inner(el, opts)}}"
       end
 
       def convert_strong(el, opts)
-        "\\textbf{#{inner(el, opts)}}"
+        "\\textbf{#{latex_link_target(el)}#{inner(el, opts)}}"
       end
 
       # Inspired by Maruku: entity conversion table based on the one from htmltolatex
@@ -387,8 +390,8 @@ module Kramdown
         34 => ['"'],
         39 => ['\''],
         169 => ['\copyright'],
-        60 => ['\textless{}'],
-        62 => ['\textgreater{}'],
+        60 => ['\textless'],
+        62 => ['\textgreater'],
         338 => ['\OE'],
         339 => ['\oe'],
         352 => ['\v{S}'],
@@ -503,20 +506,27 @@ module Kramdown
         253 => ['\\\'y'],
         254 => ['\thorn', 'wasysym'],
         255 => ['\"y'],
+        8201 => ['\thinspace'],
+        8194 => ['\hskip .5em\relax'],
+        8195 => ['\quad'],
       } # :nodoc:
-      ENTITY_CONV_TABLE.each {|k,v| ENTITY_CONV_TABLE[k] = v.unshift(v.shift + '{}')}
+      ENTITY_CONV_TABLE.each {|k,v| ENTITY_CONV_TABLE[k][0].insert(-1, '{}')}
 
-      def convert_entity(el, opts)
-        text, package = ENTITY_CONV_TABLE[el.value.code_point]
+      def entity_to_latex(entity)
+        text, package = ENTITY_CONV_TABLE[entity.code_point]
         if text
           @data[:packages] << package if package
           text
         else
-          warning("Couldn't find entity in substitution table!")
+          warning("Couldn't find entity with code #{entity.code_point} in substitution table!")
           ''
         end
       end
 
+      def convert_entity(el, opts)
+        entity_to_latex(el.value)
+      end
+
       TYPOGRAPHIC_SYMS = {
         :mdash => '---', :ndash => '--', :hellip => '\ldots{}',
         :laquo_space => '\guillemotleft{}~', :raquo_space => '~\guillemotright{}',
@@ -526,10 +536,9 @@ module Kramdown
         TYPOGRAPHIC_SYMS[el.value]
       end
 
-      SMART_QUOTE_SYMS = {:lsquo => '`', :rsquo => '\'', :ldquo => '``', :rdquo => '\'\''} # :nodoc:
       def convert_smart_quote(el, opts)
-        res = SMART_QUOTE_SYMS[el.value]
-        res << "{}" if (nel = opts[:parent].children[opts[:index]+1]) && nel.type == :smart_quote
+        res = entity_to_latex(smart_quote_entity(el)).chomp('{}')
+        res << "{}" if ((nel = opts[:parent].children[opts[:index]+1]) && nel.type == :smart_quote) || res =~ /\w$/
         res
       end
 
@@ -556,7 +565,18 @@ module Kramdown
       # environments.
       def latex_environment(type, el, text)
         attrs = attribute_list(el)
-        "\\begin{#{type}}#{attrs}\n#{text.rstrip}\n\\end{#{type}}#{attrs}\n"
+        "\\begin{#{type}}#{latex_link_target(el)}#{attrs}\n#{text.rstrip}\n\\end{#{type}}#{attrs}\n"
+      end
+
+      # Return a string containing a valid <tt>\hypertarget</tt> command if the element has an ID
+      # defined, or +nil+ otherwise. If the parameter +add_label+ is +true+, a <tt>\label</tt>
+      # command will also be used additionally to the <tt>\hypertarget</tt> command.
+      def latex_link_target(el, add_label = false)
+        if (id = el.attr['id'])
+          "\\hypertarget{#{id}}{}" << (add_label ? "\\label{#{id}}" : '')
+        else
+          nil
+        end
       end
 
       # Return a LaTeX comment containing all attributes as <tt>key="value"</tt> pairs.

data/lib/kramdown/options.rb

@@ -119,6 +119,31 @@ module Kramdown
       data
     end
 
+    # ----------------------------
+    # :section: Option Validators
+    #
+    # This sections contains all pre-defined option validators.
+    # ----------------------------
+
+    # Ensures that the option value +val+ for the option called +name+ is a valid array. The
+    # parameter +val+ can be
+    #
+    # - a comma separated string which is split into an array of values
+    # - or an array.
+    #
+    # Additionally, the array is checked for the correct size.
+    def self.simple_array_validator(val, name, size)
+      if String === val
+        val = val.split(/,/)
+      elsif !(Array === val)
+        raise Kramdown::Error, "Invalid type #{val.class} for option #{name}"
+      end
+      if val.size != size
+        raise Kramdown::Error, "Option #{name} needs exactly #{size} values"
+      end
+      val
+    end
+
     # ----------------------------
     # :section: Option Definitions
     #
@@ -141,7 +166,7 @@ file still cannot be found, the templates name is interpreted as a
 template name that is provided by kramdown (without the converter
 extension).
 
-kramdown provides a default template named 'default' for each converter.
+kramdown provides a default template named 'document' for each converter.
 
 Default: ''
 Used by: all converters
@@ -277,17 +302,6 @@ on Ruby 1.9).
 
 Default: :as_char
 Used by: HTML converter, kramdown converter
-EOF
-
-    define(:toc_depth, Integer, -1, <<EOF)
-DEPRECATED: Defines the maximum level of headers which will be used to generate the table of
-contents. For instance, with a value of 2, toc entries will be generated for h1
-and h2 headers but not for h3, h4, etc. A value of 0 uses all header levels.
-
-Use option toc_levels instead!
-
-Default: -1
-Used by: HTML/Latex converter
 EOF
 
     define(:toc_levels, Object, (1..6).to_a, <<EOF) do |val|
@@ -335,14 +349,22 @@ separating them with commas.
 Default: section,subsection,subsubsection,paragraph,subparagraph,subsubparagraph
 Used by: Latex converter
 EOF
-      if String === val
-        val = val.split(/,/)
-      elsif !(Array === val)
-        raise Kramdown::Error, "Invalid type #{val.class} for option latex_headers"
-      end
-      if val.size != 6
-        raise Kramdown::Error, "Option latex_headers needs exactly six LaTeX commands"
-      end
+      simple_array_validator(val, :latex_headers, 6)
+    end
+
+    define(:smart_quotes, Object, %w{lsquo rsquo ldquo rdquo}, <<EOF) do |val|
+Defines the HTML entity names or code points for smart quote output
+
+The entities identified by entity name or code point that should be
+used for, in order, a left single quote, a right single quote, a left
+double and a right double quote are specified by separating them with
+commas.
+
+Default: lsquo,rsquo,ldquo,rdquo
+Used by: HTML/Latex converter
+EOF
+      val = simple_array_validator(val, :smart_quotes, 4)
+      val.map! {|v| Integer(v) rescue v}
       val
     end
 

data/lib/kramdown/parser/html.rb

@@ -260,7 +260,12 @@ module Kramdown
                         elsif %w{mdash ndash hellip laquo raquo}.include?(val)
                           Element.new(:typographic_sym, val.intern)
                         else
-                          Element.new(:entity, entity(val), nil, :original => src.matched)
+                          begin
+                            Element.new(:entity, entity(val), nil, :original => src.matched)
+                          rescue ::Kramdown::Error
+                            src.pos -= src.matched_size - 1
+                            Element.new(:entity, ::Kramdown::Utils::Entities.entity('amp'))
+                          end
                         end
             else
               result << Element.new(:text, src.rest)
@@ -339,14 +344,19 @@ module Kramdown
           end
         end
 
-        def convert_b(el)
-          set_basics(el, :strong)
-          process_children(el)
+        EMPHASIS_TYPE_MAP = {'em' => :em, 'i' => :em, 'strong' => :strong, 'b' => :strong}
+        def convert_em(el)
+          text = ''
+          extract_text(el, text)
+          if text =~ /\A\s/ || text =~ /\s\z/
+            process_html_element(el, false)
+          else
+            set_basics(el, EMPHASIS_TYPE_MAP[el.value])
+            process_children(el)
+          end
         end
-
-        def convert_i(el)
-          set_basics(el, :em)
-          process_children(el)
+        %w{b strong i}.each do |i|
+          alias_method("convert_#{i}".to_sym, :convert_em)
         end
 
         def convert_h1(el)
@@ -405,15 +415,25 @@ module Kramdown
           set_basics(el, :table)
           el.options[:alignment] = []
 
+          nr_cols = 0
           calc_alignment = lambda do |c|
-            if c.type == :tr && el.options[:alignment].empty?
-              el.options[:alignment] = [:default] * c.children.length
+            align = c.attr['align']
+            if c.type == :html_element && c.value == 'col' && (align.nil? || %w{left right center}.include?(align))
+              el.options[:alignment] << (align.nil? ? :default : align.to_sym)
+            elsif c.type == :tr
+              nr_cols = c.children.length
               break
             else
               c.children.each {|cc| calc_alignment.call(cc)}
             end
           end
           calc_alignment.call(el)
+          if el.options[:alignment].length > nr_cols
+            el.options[:alignment][nr_cols..-1] = []
+          else
+            el.options[:alignment] += [:default] * (nr_cols - el.options[:alignment].length)
+          end
+          el.children.delete_if {|c| c.type == :html_element}
 
           change_th_type = lambda do |c|
             if c.type == :th
@@ -446,12 +466,31 @@ module Kramdown
           end
           check_cells.call(el)
 
+          nr_cells = 0
+          check_nr_cells = lambda do |t|
+            if t.value == 'tr'
+              count = t.children.select {|cc| cc.value == 'th' || cc.value == 'td'}.length
+              if count != nr_cells
+                if nr_cells == 0
+                  nr_cells = count
+                else
+                  nr_cells = -1
+                  break
+                end
+              end
+            else
+              t.children.each {|cc| check_nr_cells.call(cc)}
+            end
+          end
+          check_nr_cells.call(el)
+          return false if nr_cells == -1
+
           check_rows = lambda do |t, type|
             t.children.all? {|r| (r.value == 'tr' || r.type == :text) && r.children.all? {|c| c.value == type || c.type == :text}}
           end
           check_rows.call(el, 'td') ||
             (el.children.all? do |t|
-               t.type == :text || (t.value == 'thead' && check_rows.call(t, 'th')) ||
+               t.type == :text || t.value == 'col' || (t.value == 'thead' && check_rows.call(t, 'th')) ||
                  ((t.value == 'tfoot' || t.value == 'tbody') && check_rows.call(t, 'td'))
              end && el.children.any? {|t| t.value == 'tbody'})
         end