newstile 1.0.1

data/lib/newstile/parser/newstile.rb

@@ -0,0 +1,325 @@
+# -*- coding: utf-8 -*-
+#
+#--
+# Copyright (C) 2009-2010 Thomas Leitner <t_leitner@gmx.at>
+#
+# This file is part of newstile.
+#
+# newstile is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+#++
+#
+
+require 'strscan'
+require 'stringio'
+
+#TODO: use [[:alpha:]] in all regexp to allow parsing of international values in 1.9.1
+#NOTE: use @src.pre_match only before other check/match?/... operations, otherwise the content is changed
+
+module Newstile
+
+  module Parser
+
+    # Used for parsing a document in newstile format.
+    #
+    # If you want to extend the functionality of the parser, you need to the following:
+    #
+    # * Create a new subclass
+    # * add the needed parser methods
+    # * modify the @block_parsers and @span_parsers variables and add the names of your parser
+    #   methods
+    #
+    # Here is a small example for an extended parser class that parses ERB style tags as raw text if
+    # they are used as span level elements (an equivalent block level parser should probably also be
+    # made to handle the block case):
+    #
+    #   require 'newstile/parser/newstile'
+    #
+    #   class Newstile::Parser::ERBNewstile < Kramdown::Parser::Kramdown
+    #
+    #      def initialize(doc)
+    #        super(doc)
+    #        @span_parsers.unshift(:erb_tags)
+    #      end
+    #
+    #      ERB_TAGS_START = /<%.*?%>/
+    #
+    #      def parse_erb_tags
+    #        @src.pos += @src.matched_size
+    #        @tree.children << Element.new(:raw, @src.matched)
+    #      end
+    #      define_parser(:erb_tags, ERB_TAGS_START, '<%')
+    #
+    #   end
+    #
+    # The new parser can be used like this:
+    #
+    #   require 'newstile/document'
+    #   # require the file with the above parser class
+    #
+    #   Newstile::Document.new(input_text, :input => 'ERBNewstile').to_html
+    #
+    class Newstile < Base
+
+      include ::Newstile
+
+      attr_reader :tree
+      attr_reader :doc
+      attr_reader :options
+
+      # Create a new Newstile parser object for the Newstile::Document +doc+.
+      def initialize(doc)
+        super(doc)
+
+        @src = nil
+        @tree = nil
+        @stack = []
+        @text_type = :raw_text
+        @block_ial = nil
+
+        @doc.parse_infos[:ald] = {}
+        @doc.parse_infos[:link_defs] = {}
+        @doc.parse_infos[:abbrev_defs] = {}
+        @doc.parse_infos[:footnotes] = {}
+
+        @block_parsers = [:blank_line, :codeblock, :codeblock_fenced, :blockquote, :summary, :table, :atx_header,
+                          :setext_header, 
+                          :newstile_header, 
+                          :horizontal_rule, :list, :definition_list, :link_definition, :block_html,
+                          :footnote_definition, :abbrev_definition, :ald, :block_math,
+                          :block_extension, :block_ial, :eob_marker, :paragraph]
+        @span_parsers =  [ 
+                          :emphasis, :codespan, :autolink, :span_html, :footnote_marker, :link, 
+                          :newstile_link, 
+                          :smart_quotes, 
+                          :inline_math,
+                          :span_extension, :span_ial, :html_entity, :typographic_syms, :line_break, :escaped_chars]
+
+      end
+      private_class_method(:new, :allocate)
+
+
+      # The source string provided on initialization is parsed and the created +tree+ is returned.
+      def parse(source)
+        configure_parser
+        tree = Element.new(:root)
+        parse_blocks(tree, adapt_source(source))
+        update_tree(tree)
+        replace_abbreviations(tree)
+        @doc.parse_infos[:footnotes].each do |name, data|
+          update_tree(data[:content])
+        end
+        tree
+      end
+
+      #######
+      protected
+      #######
+
+      # Adapt the object to allow parsing like specified in the options.
+      def configure_parser
+        @parsers = {}
+        (@block_parsers + @span_parsers).each do |name|
+          if self.class.has_parser?(name)
+            @parsers[name] = self.class.parser(name)
+          else
+            raise Newstile::Error, "Unknown parser: #{name}"
+          end
+        end
+        @span_start, @span_start_re = span_parser_regexps
+      end
+
+      # Create the needed span parser regexps.
+      def span_parser_regexps(parsers = @span_parsers)
+        span_start = /#{parsers.map {|name| @parsers[name].span_start}.join('|')}/
+        [span_start, /(?=#{span_start})/]
+      end
+
+      # Parse all block level elements in +text+ into the element +el+.
+      def parse_blocks(el, text = nil)
+        @stack.push([@tree, @src])
+        @tree, @src = el, (text.nil? ? @src : StringScanner.new(text))
+
+        status = catch(:stop_block_parsing) do
+          while !@src.eos?
+            block_ial_set = @block_ial
+            @block_parsers.any? do |name|
+              if @src.check(@parsers[name].start_re)
+                send(@parsers[name].method)
+              else
+                false
+              end
+            end || begin
+              warning('Warning: this should not occur - no block parser handled the line')
+              add_text(@src.scan(/.*\n/))
+            end
+            @block_ial = nil if block_ial_set
+          end
+        end
+
+        @tree, @src = *@stack.pop
+        status
+      end
+
+      # Update the tree by parsing all <tt>:raw_text</tt> elements with the span level parser
+      # (resets +@tree+, +@src+ and the +@stack+) and by updating the attributes from the IALs.
+      def update_tree(element)
+        last_blank = nil
+        element.children.map! do |child|
+          if child.type == :raw_text
+            last_blank = nil
+            @stack, @tree, @text_type = [], nil, :text
+            @src = StringScanner.new(child.value)
+            parse_spans(child)
+            child.children
+          elsif child.type == :eob
+            []
+          elsif child.type == :blank
+            if last_blank
+              last_blank.value += child.value
+              []
+            else
+              last_blank = child
+              child
+            end
+          else
+            last_blank = nil
+            update_tree(child)
+            update_attr_with_ial(child.attr, child.options[:ial]) if child.options[:ial]
+            child
+          end
+        end.flatten!
+      end
+
+      # Parse all span level elements in the source string.
+      def parse_spans(el, stop_re = nil, parsers = nil, text_type = @text_type)
+        @stack.push([@tree, @text_type]) unless @tree.nil?
+        @tree, @text_type = el, text_type
+
+        span_start = @span_start
+        span_start_re = @span_start_re
+        span_start, span_start_re = span_parser_regexps(parsers) if parsers
+        parsers = parsers || @span_parsers
+
+        used_re = (stop_re.nil? ? span_start_re : /(?=#{Regexp.union(stop_re, span_start)})/)
+        stop_re_found = false
+        while !@src.eos? && !stop_re_found
+          if result = @src.scan_until(used_re)
+            add_text(result)
+            if stop_re && (stop_re_matched = @src.check(stop_re))
+              stop_re_found = (block_given? ? yield : true)
+            end
+            processed = parsers.any? do |name|
+              if @src.check(@parsers[name].start_re)
+                send(@parsers[name].method)
+                true
+              else
+                false
+              end
+            end unless stop_re_found
+            add_text(@src.scan(/./)) if !processed && !stop_re_found
+          else
+            add_text(@src.scan(/.*/m)) unless stop_re
+            break
+          end
+        end
+
+        @tree, @text_type = @stack.pop
+
+        stop_re_found
+      end
+
+      # Update the attributes with the information from the inline attribute list and all referenced ALDs.
+      def update_attr_with_ial(attr, ial)
+        ial[:refs].each do |ref|
+          update_attr_with_ial(attr, ref) if ref = @doc.parse_infos[:ald][ref]
+        end if ial[:refs]
+        ial.each do |k,v|
+          if k == 'class'
+            attr[k] = ((attr[k] || '') + " #{v}").lstrip
+          elsif k.kind_of?(String)
+            attr[k] = v
+          end
+        end
+      end
+
+      # Create a new block level element, taking care of applying a preceding block IAL if it exists.
+      def new_block_el(*args)
+        el = Element.new(*args)
+        el.options[:category] ||= :block
+        el.options[:ial] = @block_ial if @block_ial && el.type != :blank && el.type != :eob
+        el
+      end
+
+      @@parsers = {}
+
+      # Holds all the needed data for one block/span level parser.
+      Data = Struct.new(:name, :start_re, :span_start, :method)
+
+      # Add a parser method
+      #
+      # * with the given +name+,
+      # * using +start_re+ as start regexp
+      # * and, for span parsers, +span_start+ as a String that can be used in a regexp and
+      #   which identifies the starting character(s)
+      #
+      # to the registry. The method name is automatically derived from the +name+ or can explicitly
+      # be set by using the +meth_name+ parameter.
+      def self.define_parser(name, start_re, span_start = nil, meth_name = "parse_#{name}")
+        raise "A parser with the name #{name} already exists!" if @@parsers.has_key?(name)
+        @@parsers[name] = Data.new(name, start_re, span_start, meth_name)
+      end
+
+      # Return the Data structure for the parser +name+.
+      def self.parser(name = nil)
+        @@parsers[name]
+      end
+
+      # Return +true+ if there is a parser called +name+.
+      def self.has_parser?(name)
+        @@parsers.has_key?(name)
+      end
+
+      INDENT = /^(?:\t| {4})/
+      OPT_SPACE = / {0,3}/
+
+      require 'newstile/parser/newstile/blank_line'
+      require 'newstile/parser/newstile/eob'
+      require 'newstile/parser/newstile/paragraph'
+      require 'newstile/parser/newstile/header'
+      require 'newstile/parser/newstile/blockquote'
+      require 'newstile/parser/newstile/table'
+      require 'newstile/parser/newstile/codeblock'
+      require 'newstile/parser/newstile/horizontal_rule'
+      require 'newstile/parser/newstile/list'
+      require 'newstile/parser/newstile/link'
+      require 'newstile/parser/newstile/attribute_list'
+      require 'newstile/parser/newstile/extension'
+      require 'newstile/parser/newstile/footnote'
+      require 'newstile/parser/newstile/html'
+      require 'newstile/parser/newstile/escaped_chars'
+      require 'newstile/parser/newstile/html_entity'
+      require 'newstile/parser/newstile/line_break'
+      require 'newstile/parser/newstile/typographic_symbol'
+      require 'newstile/parser/newstile/autolink'
+      require 'newstile/parser/newstile/codespan'
+      require 'newstile/parser/newstile/emphasis'
+      require 'newstile/parser/newstile/smart_quotes'
+      require 'newstile/parser/newstile/math'
+      require 'newstile/parser/newstile/abbreviation'
+
+    end
+
+  end
+
+end

data/lib/newstile/parser/newstile/abbreviation.rb

@@ -0,0 +1,66 @@
+# -*- coding: utf-8 -*-
+#
+#--
+# Copyright (C) 2009-2010 Thomas Leitner <t_leitner@gmx.at>
+#
+# This file is part of newstile.
+#
+# newstile is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+#++
+#
+
+module Newstile
+  module Parser
+    class Newstile
+
+      ABBREV_DEFINITION_START = /^#{OPT_SPACE}\*\[(.+?)\]:(.*?)\n/
+
+      # Parse the link definition at the current location.
+      def parse_abbrev_definition
+        @src.pos += @src.matched_size
+        abbrev_id, abbrev_text = @src[1], @src[2].strip
+        warning("Duplicate abbreviation ID '#{abbrev_id}' - overwriting") if @doc.parse_infos[:abbrev_defs][abbrev_id]
+        @doc.parse_infos[:abbrev_defs][abbrev_id] = abbrev_text
+        @tree.children << Element.new(:eob, :abbrev_def)
+        true
+      end
+      define_parser(:abbrev_definition, ABBREV_DEFINITION_START)
+
+      # Replace the abbreviation text with elements.
+      def replace_abbreviations(el, regexps = nil)
+        return if @doc.parse_infos[:abbrev_defs].empty?
+        if !regexps
+          regexps = [Regexp.union(*@doc.parse_infos[:abbrev_defs].keys.map {|k| /#{Regexp.escape(k)}/})]
+          regexps << /(?=(?:\W|^)#{regexps.first}(?!\w))/ # regexp should only match on word boundaries
+        end
+        el.children.map! do |child|
+          if child.type == :text
+            result = []
+            strscan = StringScanner.new(child.value)
+            while temp = strscan.scan_until(regexps.last)
+              temp += strscan.scan(/\W|^/)
+              abbr = strscan.scan(regexps.first)
+              result += [Element.new(:text, temp), Element.new(:abbreviation, abbr)]
+            end
+            result + [Element.new(:text, extract_string(strscan.pos..-1, strscan))]
+          else
+            replace_abbreviations(child, regexps)
+            child
+          end
+        end.flatten!
+      end
+
+    end
+  end
+end

data/lib/newstile/parser/newstile/attribute_list.rb

@@ -0,0 +1,111 @@
+# -*- coding: utf-8 -*-
+#
+#--
+# Copyright (C) 2009-2010 Thomas Leitner <t_leitner@gmx.at>
+#
+# This file is part of newstile.
+#
+# newstile is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+#++
+#
+
+module Newstile
+  module Parser
+    class Newstile
+
+      # Parse the string +str+ and extract all attributes and add all found attributes to the hash
+      # +opts+.
+      def parse_attribute_list(str, opts)
+        str.scan(ALD_TYPE_ANY).each do |key, sep, val, id_attr, class_attr, ref|
+          if ref
+            (opts[:refs] ||= []) << ref
+          elsif class_attr
+            opts['class'] = ((opts['class'] || '') + " #{class_attr}").lstrip
+          elsif id_attr
+            opts['id'] = id_attr
+          else
+            opts[key] = val.gsub(/\\(\}|#{sep})/, "\\1")
+          end
+        end
+      end
+
+      # Update the +ial+ with the information from the inline attribute list +opts+.
+      def update_ial_with_ial(ial, opts)
+        (ial[:refs] ||= []) << opts[:refs]
+        opts.each do |k,v|
+          if k == 'class'
+            ial[k] = ((ial[k] || '') + " #{v}").lstrip
+          elsif k.kind_of?(String)
+            ial[k] = v
+          end
+        end
+      end
+
+
+      ALD_ID_CHARS = /[\w-]/
+      ALD_ANY_CHARS = /\\\}|[^\}]/
+      ALD_ID_NAME = /\w#{ALD_ID_CHARS}*/
+      ALD_TYPE_KEY_VALUE_PAIR = /(#{ALD_ID_NAME})=("|')((?:\\\}|\\\2|[^\}\2])*?)\2/
+      ALD_TYPE_CLASS_NAME = /\.(#{ALD_ID_NAME})/
+      ALD_TYPE_ID_NAME = /#(\w[\w:-]*)/
+      ALD_TYPE_REF = /(#{ALD_ID_NAME})/
+      ALD_TYPE_ANY = /(?:\A|\s)(?:#{ALD_TYPE_KEY_VALUE_PAIR}|#{ALD_TYPE_ID_NAME}|#{ALD_TYPE_CLASS_NAME}|#{ALD_TYPE_REF})(?=\s|\Z)/
+      ALD_START = /^#{OPT_SPACE}\{:(#{ALD_ID_NAME}):(#{ALD_ANY_CHARS}+)\}\s*?\n/
+
+      # Parse the attribute list definition at the current location.
+      def parse_ald
+        @src.pos += @src.matched_size
+        parse_attribute_list(@src[2], @doc.parse_infos[:ald][@src[1]] ||= Utils::OrderedHash.new)
+        @tree.children << Element.new(:eob, :ald)
+        true
+      end
+      define_parser(:ald, ALD_START)
+
+
+      IAL_BLOCK = /\{:(?!:|\/)(#{ALD_ANY_CHARS}+)\}\s*?\n/
+      IAL_BLOCK_START = /^#{OPT_SPACE}#{IAL_BLOCK}/
+
+      # Parse the inline attribute list at the current location.
+      def parse_block_ial
+        @src.pos += @src.matched_size
+        if @tree.children.last && @tree.children.last.type != :blank && @tree.children.last.type != :eob
+          parse_attribute_list(@src[1], @tree.children.last.options[:ial] ||= Utils::OrderedHash.new)
+          @tree.children << Element.new(:eob, :ial) unless @src.check(IAL_BLOCK_START)
+        else
+          parse_attribute_list(@src[1], @block_ial = Utils::OrderedHash.new)
+        end
+        true
+      end
+      define_parser(:block_ial, IAL_BLOCK_START)
+
+
+      IAL_SPAN_START = /\{:(#{ALD_ANY_CHARS}+)\}/
+
+      # Parse the inline attribute list at the current location.
+      def parse_span_ial
+        @src.pos += @src.matched_size
+        if @tree.children.last && @tree.children.last.type != :text
+          attr = Utils::OrderedHash.new
+          parse_attribute_list(@src[1], attr)
+          update_ial_with_ial(@tree.children.last.options[:ial] ||= Utils::OrderedHash.new, attr)
+          update_attr_with_ial(@tree.children.last.attr, attr)
+        else
+          warning("Ignoring span IAL because preceding element is just text")
+        end
+      end
+      define_parser(:span_ial, IAL_SPAN_START, '\{:')
+
+    end
+  end
+end