Merge pull request #92 from metanorma/features/list-start

allow lists with arbitrary starting number: https://github.com/metano…

Author: opoudjis

lib/html2doc/base.rb

@@ -13,7 +13,7 @@ def initialize(hash)
     @imagedir = hash[:imagedir]
     @debug = hash[:debug]
     @liststyles = hash[:liststyles]
-    @stylesheet = hash[:stylesheet]
+    @stylesheet = read_stylesheet(hash[:stylesheet])
     @c = HTMLEntities.new
   end
 
@@ -74,8 +74,7 @@ def cleanup(docxml)
   end
 
   def locate_landscape(_docxml)
-    css = read_stylesheet(@stylesheet)
-    @landscape = css.scan(/div\.\S+\s+\{\s*page:\s*[^;]+L;\s*\}/m)
+    @landscape = @stylesheet.scan(/div\.\S+\s+\{\s*page:\s*[^;]+L;\s*\}/m)
       .map { |e| e.sub(/^div\.(\S+).*$/m, "\\1") }
   end
 
@@ -99,11 +98,9 @@ def filename_substitute(head, header_filename)
     end
   end
 
-  def stylesheet(_filename, _header_filename, cssname)
-    stylesheet = read_stylesheet(cssname)
+  def stylesheet(_filename, _header_filename, _cssname)
+    stylesheet = "#{@stylesheet}\n#{@newliststyledefs}"
     xml = Nokogiri::XML("<style/>")
-    # s = Nokogiri::XML::CDATA.new(xml, "\n#{stylesheet}\n")
-    # xml.children.first << Nokogiri::XML::Comment.new(xml, s)
     xml.children.first << Nokogiri::XML::CDATA
       .new(xml, "\n<!--\n#{stylesheet}\n-->\n")
     xml.root.to_s

lib/html2doc/lists.rb

@@ -4,8 +4,7 @@
 
 class Html2Doc
   def style_list(elem, level, liststyle, listnumber)
-    return unless liststyle
-
+    liststyle or return
     if elem["style"]
       elem["style"] += ";"
     else
@@ -30,16 +29,37 @@ def list_add1(elem, liststyles, listtype, level)
 
   def list_add(xpath, liststyles, listtype, level)
     xpath.each do |l|
-      level == 1 and l["seen"] = true and @listnumber += 1
+      level == 1 && l["seen"] = true and @listnumber += 1
       l["id"] ||= UUIDTools::UUID.random_create
+      liststyle = derive_liststyle(l, liststyles[listtype], level)
       (l.xpath(".//li") - l.xpath(".//ol//li | .//ul//li")).each do |li|
-        style_list(li, level, liststyles[listtype], @listnumber)
+        style_list(li, level, liststyle, @listnumber)
         list_add1(li, liststyles, listtype, level)
       end
       list_add_tail(l, liststyles, listtype, level)
     end
   end
 
+  def derive_liststyle(list, liststyle, level)
+    list["start"] && list["start"] != "1" or return liststyle
+    @liststyledefsidx += 1
+    ret = "l#{@liststyledefsidx}"
+    @newliststyledefs += newliststyle(list["start"], liststyle, ret, level)
+    ret
+  end
+
+  def newliststyle(start, liststyle, newstylename, level)
+    s = @liststyledefs[liststyle]
+      .gsub(/@list\s+#{liststyle}/, "@list #{newstylename}")
+      .sub(/@list\s+#{newstylename}\s+\{[^}]*\}/m, <<~LISTSTYLE)
+        @list #{newstylename}\n{mso-list-id:#{rand(100_000_000..999_999_999)};
+        mso-list-template-ids:#{rand(100_000_000..999_999_999)};}
+      LISTSTYLE
+      .sub(/@list\s+#{newstylename}:level#{level}\s+\{/m,
+           "\\0mso-level-start-at:#{start};\n")
+    "#{s}\n"
+  end
+
   def list_add_tail(list, liststyles, listtype, level)
     list.xpath(".//ul[not(ancestor::li/ancestor::*/@id = '#{list['id']}')] | "\
                ".//ol[not(ancestor::li/ancestor::*/@id = '#{list['id']}')]")
@@ -49,16 +69,15 @@ def list_add_tail(list, liststyles, listtype, level)
   end
 
   def list2para(list)
-    return if list.xpath("./li").empty?
-
+    list.xpath("./li").empty? and return
     list2para_position(list)
     list.xpath("./li").each do |l|
       l.name = "p"
       l["class"] ||= "MsoListParagraphCxSpMiddle"
-      next unless l.first_element_child&.name == "p"
-
+      l.first_element_child&.name == "p" or next
       l["style"] ||= ""
-      l["style"] += (l.first_element_child["style"]&.sub(/mso-list[^;]+;/, "") || "")
+      l["style"] += l.first_element_child["style"]
+        &.sub(/mso-list[^;]+;/, "") || ""
       l.first_element_child.replace(l.first_element_child.children)
     end
     list.replace(list.children)
@@ -100,12 +119,82 @@ def lists_unstyled(docxml, liststyles)
   end
 
   def lists(docxml, liststyles)
-    return if liststyles.nil?
-
-    @listnumber = 0
+    liststyles.nil? and return
+    parse_stylesheet_line_styles
     liststyles.each_key { |k| lists1(docxml, liststyles, k) }
     lists_unstyled(docxml, liststyles)
     liststyles.has_key?(:ul) and docxml.xpath("//ul").each { |u| list2para(u) }
     liststyles.has_key?(:ol) and docxml.xpath("//ol").each { |u| list2para(u) }
   end
+
+  def parse_stylesheet_line_styles
+    @listnumber = 0
+    result = process_stylesheet_lines(@stylesheet.split("\n"))
+    @liststyledefs = clean_result_content(result)
+    @newliststyledefs = ""
+    @liststyledefsidx = @liststyledefs.keys.map do |k|
+      k.sub(/^.*(\d+)$/, "\\1").to_i
+    end.max
+  end
+
+  private
+
+  def extract_list_name(line)
+    match = line.match(/^\s*@list\s+([^:\s]+)(?::.*)?/)
+    match ? match[1] : nil
+  end
+
+  def list_declaration?(line)
+    !extract_list_name(line).nil?
+  end
+
+  def save_current_list(result, current_base, current_content)
+    current_base.nil? || current_content.empty? and return result
+    if result[current_base]
+      result[current_base] += current_content
+    else
+      result[current_base] = current_content
+    end
+    result
+  end
+
+  def process_stylesheet_lines(lines)
+    result = {}
+    current_base = nil
+    current_content = ""
+    parsing_active = false
+
+    lines.each do |line|
+      if list_declaration?(line)
+        base_name = extract_list_name(line)
+        if current_base == base_name
+          current_content += "#{line}\n"
+        else
+          # save accumulated list style definition, new list style
+          save_current_list(result, current_base, current_content)
+          current_base = base_name
+          current_content = "#{line}\n"
+        end
+        parsing_active = true
+
+      elsif parsing_active && line.include?("}")
+        # End of current block - add this line and stop parsing
+        current_content += "#{line}\n"
+        parsing_active = false
+
+      elsif parsing_active
+        # Continue adding content while parsing is active
+        current_content += "#{line}\n"
+      end
+      # If parsing_active is false and no @list declaration, skip the line
+    end
+    # Save the last list if we were still parsing
+    save_current_list(result, current_base, current_content)
+    result
+  end
+
+  def clean_result_content(result)
+    result.each { |k, v| result[k] = v.rstrip }
+    result
+  end
 end

lib/html2doc/mime.rb

@@ -135,8 +135,7 @@ def skip_image_cleanup?(img)
   # Scan both @stylesheet and docxml.to_xml (where @standardstylesheet has ended up)
   # Allow 0.9 * height to fit caption
   def page_dimensions(docxml)
-    stylesheet = read_stylesheet(@stylesheet)
-    page_size = find_page_size_in_doc(stylesheet, docxml.to_xml) or
+    page_size = find_page_size_in_doc(@stylesheet, docxml.to_xml) or
       return [680, 400]
     m_size = /size:\s*(\S+)\s+(\S+)\s*;/.match(page_size) or return [680, 400]
     m_marg = /margin:\s*(\S+)\s+(\S+)\s*(\S+)\s*(\S+)\s*;/.match(page_size) or

lib/html2doc/version.rb

@@ -1,3 +1,3 @@
 class Html2Doc
-  VERSION = "1.9.2".freeze
+  VERSION = "1.10.0".freeze
 end

spec/html2doc_spec.rb

@@ -263,6 +263,9 @@ def word_body(xml, footnote)
 DEFAULT_STYLESHEET = File.read("lib/html2doc/wordstyle.css",
                                encoding: "utf-8").freeze
 
+NEW_LIST_STYLES = File.read("spec/wordstyle-custom-lists.css",
+                               encoding: "utf-8")
+
 def guid_clean(xml)
   xml.gsub(/NextPart_[0-9a-f.]+/, "NextPart_")
 end
@@ -965,6 +968,32 @@ def image_clean(xml)
       OUTPUT
   end
 
+  it "sets custom start numbering of lists" do
+    simple_body = <<~BODY
+      <div>
+      <ol id="1"><li><div><p><ol id="2" start="3"><li><ul id="3"><li><p><ol id="4" start="5"><li><ol id="5" start="7"><li>A</li></ol></li></ol></p></li></ul></li></ol></p></div></li></ol>
+      <ol id="6" start="2"><li><div><p><ol id="7" start="1"><li><ul id="8"><li><p><ol id="9"><li><ol id="10"><li>A</li></ol></li></ol></p></li></ul></li></ol></p></div></li></ol></div>
+    BODY
+    Html2Doc.new(filename: "test", liststyles: { ul: "l0", ol: "l1" })
+      .process(html_input(simple_body))
+    style = "#{DEFAULT_STYLESHEET} #{NEW_LIST_STYLES}"
+      .gsub(/mso-list-id:\d+/, "mso-list-id:_")
+  .gsub(/mso-list-template-ids:\d+/, "mso-list-template-ids:_")
+    doc = File.read("test.doc", encoding: "utf-8")
+      .gsub(/mso-list-id:\d+/, "mso-list-id:_")
+  .gsub(/mso-list-template-ids:\d+/, "mso-list-template-ids:_")
+
+    expect(guid_clean(doc))
+      .to match_fuzzy(<<~OUTPUT)
+        #{WORD_HDR} #{style} #{WORD_HDR_END}
+        #{word_body('<div>
+        <p style="mso-list:l1 level1 lfo1;" class="MsoListParagraphCxSpFirst"><div><p class="MsoNormal"><p style="mso-list:l2 level2 lfo1;" class="MsoListParagraphCxSpFirst"><p style="mso-list:l3 level4 lfo1;" class="MsoListParagraphCxSpFirst"><p style="mso-list:l4 level5 lfo1;" class="MsoListParagraphCxSpFirst">A</p></p></p></p></div></p>
+       <p style="mso-list:l5 level1 lfo2;" class="MsoListParagraphCxSpFirst"><div><p class="MsoNormal"><p style="mso-list:l1 level2 lfo2;" class="MsoListParagraphCxSpFirst"><p style="mso-list:l1 level4 lfo2;" class="MsoListParagraphCxSpFirst"><p style="mso-list:l1 level5 lfo2;" class="MsoListParagraphCxSpFirst">A</p></p></p></p></div></p></div>',
+                    '<div style="mso-element:footnote-list"/>')}
+        #{WORD_FTR1}
+      OUTPUT
+  end
+
   it "labels lists with multiple list styles" do
     simple_body = <<~BODY
       <div><ul class="steps" id="0">