uniword 1.0.10 → 1.2.4

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: a7feae85851d80946524005c3cc6419419d492c717e559628093494ee7532cd9
-  data.tar.gz: eba8d3d3d433aa158b5a4214659e6afd0e12d74e3fada13997da066021a78f9d
+  metadata.gz: 1517a752442fd3914efbcc8f94130cf6f4983c110807b0790f03bebf9424d875
+  data.tar.gz: a88ecadccf094ffd2c1031d32e13ded47f361b98ca0d594c373bd48cc74f6898
 SHA512:
-  metadata.gz: 940e11bb305505cd3e6acdc6e9cca98e002d151a0f75e7c8872d613353b4c5d85b424062a661572b14474c342b62db3072d2cc081cb9e924f95bb8fdf80e6c39
-  data.tar.gz: 3ca3c9bea8133f7a18e64897b687ef8dd10843c85aa2e4422dabceed5152ae5a5f9cce56fdff8999f03e734f8510b919732f9f5b5b0d5ac0e9470c8dd0db11fb
+  metadata.gz: d728a6dc5a0df3560cc607281fb6ca8dd384c56d75247d7ca4b57a1791a15312b8fa8a2e5a8ae18c4ec60ff0574f76fafde2e68994c4b13b326b329cb5fd134e
+  data.tar.gz: cd7339df777bb848a7ad69c09b3d44fc0478fd9ece6a9f4d9c79ebb3114fa205824487f9bab85753338f24ac8aa22345319a254fc6a841d9f54117d2c0f07945

data/config/font_metadata.yml

@@ -90,6 +90,62 @@ fonts:
   # --- East Asian fonts (per locale) ---
 
   # Chinese Simplified (zh-CN)
+  SimSun:
+    alt_name: "宋体"
+    panose1: "02010600030101010101"
+    charset: "86"
+    family: roman
+    pitch: variable
+    sig:
+      usb0: "00000003"
+      usb1: "080E0000"
+      usb2: "00000010"
+      usb3: "00000000"
+      csb0: "00040000"
+      csb1: "00000000"
+
+  SimHei:
+    alt_name: "黑体"
+    panose1: "02010609060101010101"
+    charset: "86"
+    family: modern
+    pitch: fixed
+    sig:
+      usb0: "800002BF"
+      usb1: "38CF7CFA"
+      usb2: "00000016"
+      usb3: "00000000"
+      csb0: "0004000F"
+      csb1: "00000000"
+
+  MS Mincho:
+    alt_name: "ＭＳ 明朝"
+    panose1: "02020609040205080304"
+    charset: "80"
+    family: roman
+    pitch: fixed
+    sig:
+      usb0: "E00002FF"
+      usb1: "6FFF07FF"
+      usb2: "00000020"
+      usb3: "00000000"
+      csb0: "0000019F"
+      csb1: "00000000"
+
+  MS Gothic:
+    alt_name: "ＭＳ ゴシック"
+    panose1: "020B0609020205080204"
+    charset: "80"
+    family: modern
+    pitch: fixed
+    sig:
+      usb0: "E00002FF"
+      usb1: "6FFF07FF"
+      usb2: "00000020"
+      usb3: "00000000"
+      csb0: "0000019F"
+      csb1: "00000000"
+
   DengXian:
     alt_name: "等线"
     panose1: "02010600030101010101"

data/lib/uniword/accessibility/rules/descriptive_headings_rule.rb

@@ -88,7 +88,7 @@ module Uniword
           if paragraph.is_a?(Lutaml::Model::Serializable) && paragraph.class.attributes.key?(:text)
             paragraph.text
           elsif paragraph.is_a?(Lutaml::Model::Serializable) && paragraph.class.attributes.key?(:runs)
-            paragraph.runs.map(&:text).join
+            paragraph.runs.map(&:text_string).join
           end
         end
       end

data/lib/uniword/assembly/toc.rb

@@ -166,11 +166,11 @@ module Uniword
       # @param paragraph [Wordprocessingml::Paragraph] Paragraph to check
       # @return [Integer, nil] Heading level or nil
       def heading_level(paragraph)
-        style_name = paragraph.properties&.style
-        return nil unless style_name
+        style = paragraph.style
+        return nil unless style
 
         # Check for heading styles (Heading 1, Heading 2, etc.)
-        match = style_name.match(/^Heading\s*(\d+)$/i)
+        match = style.to_s.match(/^Heading\s*(\d+)$/i)
         return match[1].to_i if match
 
         # Check for outline level in properties
@@ -187,17 +187,7 @@ module Uniword
       # @param paragraph [Wordprocessingml::Paragraph] Source paragraph
       # @return [String] Extracted text
       def extract_text(paragraph)
-        text_parts = []
-
-        paragraph.runs.each do |run|
-          run_text = run.text
-          next unless run_text
-
-          # Handle Text objects by converting to string
-          text_parts << run_text.to_s
-        end
-
-        text_parts.join
+        paragraph.runs.map(&:text_string).join
       end
 
       # Create title paragraph.

data/lib/uniword/assembly/variable_substitutor.rb

@@ -186,10 +186,9 @@ module Uniword
       # @param run [Run] Run to process
       # @return [void]
       def substitute_run(run)
-        return unless run.text
-
-        content = run.text.to_s
-        run.text = substitute(content)
+        run.substitute_with_block(VARIABLE_PATTERN) do |match|
+          resolve_variable(match[1])
+        end
       end
 
       # Collect all keys from nested hash.

data/lib/uniword/batch/stages/normalize_styles_stage.rb

@@ -76,8 +76,10 @@ module Uniword
 
           # Detect style based on properties
           detected_style = detect_paragraph_style(paragraph)
+          target = detected_style && @target_styles[detected_style]
+          next unless target
 
-          paragraph.properties.style = @target_styles[detected_style] if detected_style && @target_styles[detected_style]
+          paragraph.properties.style = target
         end
       end
 
@@ -102,7 +104,7 @@ module Uniword
 
         # Check style name
         if paragraph.style
-          style_name = paragraph.style.downcase
+          style_name = paragraph.style.to_s.downcase
           return :heading1 if style_name.include?("heading 1")
           return :heading2 if style_name.include?("heading 2")
           return :heading3 if style_name.include?("heading 3")

data/lib/uniword/builder/base_builder.rb

@@ -39,6 +39,31 @@ module Uniword
         raise NotImplementedError,
               "#{name} must implement .default_model_class"
       end
+
+      private
+
+      # Tag name for the properties element (override in subclasses).
+      # e.g., "pPr" for ParagraphBuilder, "tcPr" for TableCellBuilder.
+      def properties_tag
+        raise NotImplementedError,
+              "#{self.class.name} must implement #properties_tag"
+      end
+
+      def track_element_order(tag)
+        @model.element_order ||= []
+        ensure_properties_in_order
+        @model.element_order << Lutaml::Xml::Element.new("Element", tag)
+      end
+
+      def ensure_properties_in_order
+        return unless @model.element_order
+        return if @model.element_order.any? { |e| e.element? && e.name == properties_tag }
+        return unless @model.properties
+
+        @model.element_order.unshift(
+          Lutaml::Xml::Element.new("Element", properties_tag)
+        )
+      end
     end
   end
 end

data/lib/uniword/builder/chart_builder.rb

@@ -1,7 +1,7 @@
 # frozen_string_literal: true
 
 require "nokogiri"
-require "securerandom"
+require "digest"
 
 module Uniword
   module Builder
@@ -134,6 +134,8 @@ module Uniword
 
       private
 
+      include DeterministicId
+
       # Build the inner chart XML
       def build_chart_xml(xml)
         xml["c"].chart do
@@ -265,19 +267,36 @@ module Uniword
 
       # Build a chart axis
       def build_axis(xml, tag, id, position, cross_id)
-        xml["c"].public_send(tag) do
-          xml["c"].axId("val" => id)
-          xml["c"].scaling do
-            xml["c"].orientation("val" => "minMax")
+        case tag
+        when "catAx"
+          xml["c"].catAx do
+            write_axis_content(xml, id, position, cross_id)
+          end
+        when "valAx"
+          xml["c"].valAx do
+            write_axis_content(xml, id, position, cross_id)
           end
-          xml["c"].delete("val" => "0")
-          xml["c"].axPos("val" => position)
-          xml["c"].majorGridlines
-          xml["c"].numFmt("formatCode" => "General",
-                          "sourceLinked" => "1")
-          xml["c"].tickLblPos("val" => "nextTo")
-          xml["c"].crossAx("val" => cross_id)
+        when "serAx"
+          xml["c"].serAx do
+            write_axis_content(xml, id, position, cross_id)
+          end
+        else
+          raise ArgumentError, "Unknown axis type: #{tag}"
+        end
+      end
+
+      def write_axis_content(xml, id, position, cross_id)
+        xml["c"].axId("val" => id)
+        xml["c"].scaling do
+          xml["c"].orientation("val" => "minMax")
         end
+        xml["c"].delete("val" => "0")
+        xml["c"].axPos("val" => position)
+        xml["c"].majorGridlines
+        xml["c"].numFmt("formatCode" => "General",
+                        "sourceLinked" => "1")
+        xml["c"].tickLblPos("val" => "nextTo")
+        xml["c"].crossAx("val" => cross_id)
       end
 
       # Create a Drawing element with chart reference
@@ -288,8 +307,8 @@ module Uniword
         inline.extent = WpDrawing::Extent.new(cx: @width, cy: @height)
         inline.effect_extent = WpDrawing::EffectExtent.new
         inline.doc_properties = WpDrawing::DocProperties.new(
-          id: SecureRandom.random_number(1_000_000_000),
-          name: "Chart #{SecureRandom.random_number(1_000_000)}",
+          id: deterministic_id("chart", @chart_type, r_id),
+          name: "Chart #{@chart_type}",
         )
 
         graphic = Drawingml::Graphic.new

data/lib/uniword/builder/deterministic_id.rb

@@ -0,0 +1,17 @@
+# frozen_string_literal: true
+
+require "digest"
+
+module Uniword
+  module Builder
+    # Shared deterministic ID generation for builders.
+    #
+    # Produces stable integer IDs from seed values using SHA256,
+    # ensuring reproducible output across runs.
+    module DeterministicId
+      def deterministic_id(*seeds)
+        Digest::SHA256.hexdigest(seeds.join(":")).to_i(16) % 1_000_000_000
+      end
+    end
+  end
+end

data/lib/uniword/builder/document_builder.rb

@@ -29,15 +29,17 @@ module Uniword
     #   doc.footer { |f| f << Builder.page_number_field }
     #   doc.save('report.docx')
     class DocumentBuilder < BaseBuilder
+      attr_reader :allocator
+
       def self.default_model_class
         Wordprocessingml::DocumentRoot
       end
 
-      def initialize(model = nil)
-        super
-        @bookmark_counter = 0
-        @footnote_builder = FootnoteBuilder.new(self)
-        @comment_counter = 0
+      def initialize(model = nil, allocator: nil)
+        super(model)
+        @allocator = allocator
+        @model.allocator = allocator if allocator
+        @footnote_builder = FootnoteBuilder.new(self, allocator: @allocator)
       end
 
       # Load a document from file for manipulation
@@ -48,6 +50,90 @@ module Uniword
         new(Uniword.load(path))
       end
 
+      # Load a DOCX as a template, reset its body to a clean state, and
+      # seed the IdAllocator from the template's existing relationships.
+      #
+      # Use this when you want the template's OOXML scaffolding (settings,
+      # fonts, styles, theme, content_types, namespace declarations) but
+      # will replace the document body with new content. This is the
+      # recommended path to producing Word-valid output: starting from a
+      # known-good DOCX preserves mc:Ignorable prefixes, rsid values, and
+      # other structural guarantees that Word expects.
+      #
+      # Resets performed:
+      # - Body content cleared (paragraphs, tables, SDTs, bookmarks,
+      #   element_order, section_properties)
+      # - User footnotes/endnotes cleared (separator/continuation kept)
+      # - custom_properties and custom_xml_items cleared
+      # - Stale image and customXml relationships removed
+      # - IdAllocator seeded from document_rels and package_rels
+      #
+      # @param path [String] Path to .docx template file
+      # @return [DocumentBuilder]
+      # @raise [ArgumentError] if path does not exist
+      def self.from_template(path)
+        raise ArgumentError, "Template not found: #{path}" unless File.exist?(path)
+
+        root = Uniword.load(path)
+        reset_template_body(root)
+        clear_user_notes(root)
+        root.custom_properties = nil
+        root.custom_xml_items = nil
+        remove_stale_relationships(root)
+        seed_allocator(root)
+        new(root, allocator: root.allocator)
+      end
+
+      def self.reset_template_body(root)
+        return unless root.body
+
+        root.body.paragraphs.clear
+        root.body.tables.clear
+        root.body.structured_document_tags.clear
+        root.body.bookmark_starts.clear
+        root.body.bookmark_ends.clear
+        root.body.element_order = [] if root.body.element_order
+        root.body.section_properties = nil
+      end
+      private_class_method :reset_template_body
+
+      def self.clear_user_notes(root)
+        clear_user_note_entries(root.footnotes, :footnote) if root.footnotes
+        clear_user_note_entries(root.endnotes, :endnote) if root.endnotes
+      end
+      private_class_method :clear_user_notes
+
+      def self.clear_user_note_entries(notes, kind)
+        entries = kind == :footnote ? notes.footnote_entries : notes.endnote_entries
+        entries.reject! { |e| e.type != "separator" && e.type != "continuationSeparator" }
+        notes.element_order = [] if notes.element_order
+      end
+      private_class_method :clear_user_note_entries
+
+      def self.remove_stale_relationships(root)
+        remove_rels_by_type_fragment(root.document_rels, "/image")
+        remove_rels_by_type_fragment(root.document_rels, "/customXml")
+        remove_rels_by_type_fragment(root.package_rels, "custom-properties")
+        root.content_types&.overrides&.reject! do |o|
+          part = o.part_name.to_s
+          part == "/docProps/custom.xml" || part.include?("customXml/")
+        end
+        root.image_parts = nil
+      end
+      private_class_method :remove_stale_relationships
+
+      def self.remove_rels_by_type_fragment(rels, fragment)
+        return unless rels&.relationships
+
+        rels.relationships.reject! { |r| r.type.to_s.include?(fragment) }
+      end
+      private_class_method :remove_rels_by_type_fragment
+
+      def self.seed_allocator(root)
+        root.allocator = Docx::IdAllocator.populate_from_package(root)
+      end
+      private_class_method :seed_allocator
+
       # Append a top-level element (paragraph or table)
       #
       # @param element [Paragraph, Table, ParagraphBuilder, TableBuilder]
@@ -56,12 +142,16 @@ module Uniword
         case element
         when Wordprocessingml::Paragraph
           @model.body.paragraphs << element
+          @model.body.append_to_element_order("p")
         when Wordprocessingml::Table
           @model.body.tables << element
+          @model.body.append_to_element_order("tbl")
         when ParagraphBuilder
           @model.body.paragraphs << element.build
+          @model.body.append_to_element_order("p")
         when TableBuilder
           @model.body.tables << element.build
+          @model.body.append_to_element_order("tbl")
         else
           raise ArgumentError, "Cannot add #{element.class} to document"
         end
@@ -74,10 +164,10 @@ module Uniword
       # @yield [ParagraphBuilder] Builder for configuration
       # @return [ParagraphBuilder] The paragraph builder
       def paragraph(text = nil, &block)
-        para = ParagraphBuilder.new
+        para = ParagraphBuilder.new(allocator: @allocator)
         para << text if text
         block.call(para) if block_given?
-        @model.body.paragraphs << para.build
+        self << para
         para
       end
 
@@ -88,10 +178,10 @@ module Uniword
       # @yield [ParagraphBuilder] Builder for additional configuration
       # @return [ParagraphBuilder] The paragraph builder
       def heading(text, level: 1)
-        para = ParagraphBuilder.new
+        para = ParagraphBuilder.new(allocator: @allocator)
         para.style = "Heading#{level}"
         para << text
-        @model.body.paragraphs << para.build
+        self << para
         para
       end
 
@@ -99,7 +189,7 @@ module Uniword
       #
       # @return [self]
       def page_break
-        @model.body.paragraphs << Wordprocessingml::Paragraph.new(
+        self << Wordprocessingml::Paragraph.new(
           runs: [Builder.page_break]
         )
         self
@@ -112,7 +202,7 @@ module Uniword
       def table(&block)
         tbl = TableBuilder.new
         block.call(tbl) if block_given?
-        @model.body.tables << tbl.build
+        self << tbl
         tbl
       end
 
@@ -135,7 +225,7 @@ module Uniword
       # @yield [HeaderFooterBuilder] Builder for header content
       # @return [HeaderFooterBuilder] The header/footer builder
       def header(type: "default", &block)
-        hf = HeaderFooterBuilder.new(:header, type: type)
+        hf = HeaderFooterBuilder.new(:header, type: type, allocator: @allocator)
         block.call(hf) if block_given?
         (@model.headers ||= {})[type] = hf.build
         hf
@@ -147,7 +237,7 @@ module Uniword
       # @yield [HeaderFooterBuilder] Builder for footer content
       # @return [HeaderFooterBuilder] The header/footer builder
       def footer(type: "default", &block)
-        hf = HeaderFooterBuilder.new(:footer, type: type)
+        hf = HeaderFooterBuilder.new(:footer, type: type, allocator: @allocator)
         block.call(hf) if block_given?
         (@model.footers ||= {})[type] = hf.build
         hf
@@ -160,7 +250,7 @@ module Uniword
       # @return [self]
       def toc(title: "Table of Contents", styles: nil)
         TocBuilder.build(title: title, styles: styles).each do |para|
-          @model.body.paragraphs << para
+          self << para
         end
         self
       end
@@ -198,16 +288,17 @@ module Uniword
       # @yield [ParagraphBuilder] Builder for bookmark content
       # @return [ParagraphBuilder] The paragraph builder
       def bookmark(name, &block)
-        @bookmark_counter += 1
-        id = @bookmark_counter.to_s
+        id = @allocator ? @allocator.alloc_bookmark_id : begin
+          @bookmark_counter ||= 0
+          @bookmark_counter += 1
+          @bookmark_counter.to_s
+        end
 
-        para = ParagraphBuilder.new
-        para.model.bookmark_starts <<
-          Wordprocessingml::BookmarkStart.new(id: id, name: name)
+        para = ParagraphBuilder.new(allocator: @allocator)
+        para << Wordprocessingml::BookmarkStart.new(id: id, name: name)
         block.call(para) if block_given?
-        para.model.bookmark_ends <<
-          Wordprocessingml::BookmarkEnd.new(id: id)
-        @model.body.paragraphs << para.build
+        para << Wordprocessingml::BookmarkEnd.new(id: id)
+        self << para
         para
       end
 
@@ -306,12 +397,12 @@ module Uniword
       # @param size [Integer] Border size in eighths of a point (default 6)
       # @return [self]
       def horizontal_rule(style: "single", color: "auto", size: 6)
-        para = ParagraphBuilder.new
+        para = ParagraphBuilder.new(allocator: @allocator)
         para.borders(
           bottom: { style: style, color: color, size: size }
         )
         para.spacing(after: 0)
-        @model.body.paragraphs << para.build
+        self << para
         self
       end
 
@@ -327,7 +418,7 @@ module Uniword
         para.runs << ImageBuilder.create_run(
           self, path, width: width, height: height, alt_text: alt_text
         )
-        @model.body.paragraphs << para
+        self << para
         self
       end
 
@@ -350,7 +441,7 @@ module Uniword
           self, path, width: width, height: height, alt_text: alt_text,
                       align: align, wrap: wrap, behind_text: behind_text
         )
-        @model.body.paragraphs << para
+        self << para
         self
       end
 
@@ -389,10 +480,14 @@ module Uniword
       # @yield [CommentBuilder] Builder for rich comment content
       # @return [Comment] The created Comment model
       def comment(author:, text: nil, initials: nil, &block)
-        @comment_counter += 1
+        comment_id = @allocator ? @allocator.alloc_comment_id : begin
+          @comment_counter ||= 0
+          @comment_counter += 1
+          @comment_counter.to_s
+        end
         cb = CommentBuilder.new(
           author: author,
-          comment_id: @comment_counter.to_s,
+          comment_id: comment_id,
           initials: initials
         )
         cb << text if text
@@ -437,7 +532,7 @@ module Uniword
         sdt = SdtBuilder.bibliography.build
         para = Wordprocessingml::Paragraph.new
         para.sdts << sdt
-        @model.body.paragraphs << para
+        self << para
         self
       end
 
@@ -458,7 +553,7 @@ module Uniword
         run.drawings << drawing
         para = Wordprocessingml::Paragraph.new
         para.runs << run
-        @model.body.paragraphs << para
+        self << para
         cb
       end
 
@@ -466,7 +561,7 @@ module Uniword
       #
       # @return [self]
       def page_number
-        @model.body.paragraphs << Builder.page_number_field
+        self << Builder.page_number_field
         self
       end
 
@@ -474,7 +569,7 @@ module Uniword
       #
       # @return [self]
       def total_pages
-        @model.body.paragraphs << Builder.total_pages_field
+        self << Builder.total_pages_field
         self
       end
 
@@ -483,7 +578,7 @@ module Uniword
       # @param format [String] Date format (default 'M/d/yyyy')
       # @return [self]
       def date_field(format: "M/d/yyyy")
-        @model.body.paragraphs << Builder.date_field(format: format)
+        self << Builder.date_field(format: format)
         self
       end
 
@@ -492,7 +587,7 @@ module Uniword
       # @param format [String] Time format (default 'h:mm:ss am/pm')
       # @return [self]
       def time_field(format: "h:mm:ss am/pm")
-        @model.body.paragraphs << Builder.time_field(format: format)
+        self << Builder.time_field(format: format)
         self
       end