coradoc-docx 0.1.2 → 0.1.3

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: e46f27d58495c1034b6fcb7e18a4b3f7361399bae351f07feecc70bf96998c00
-  data.tar.gz: e635bc232ec683c7ac69d29ca3051cb7ff79858fb6726d97cf57abeef736405a
+  metadata.gz: b2d3e5bd2f76d67000623cf3a2cb3889d5b49105a7420494f6f09622bac8a61b
+  data.tar.gz: 2c611f389f427ac79c73b5ce6e727e602b1d392f97f6090af837d2c6a5d05d9a
 SHA512:
-  metadata.gz: 8f59bc132f05dbc709ed0c7b31e6c633e94efde7a95e4bbf1661c18e3a737ad0ca8e592257a1029c9935e57ce679f771705cef1bbdeea0b4d00c11e4e112e4d9
-  data.tar.gz: 926131183805ae1163e544f464417c21fe38dfb5705cb363427c83e053d7f60c8a71f65cfd0189f9003cc6498f1ec0fa54215ef3d21d87fd721619c844de7b3a
+  metadata.gz: faff02134845b773db02e2e8e8f4a0f99e58548b013db9ad99212fe349f1e8be7088fc153465bb32b54c584c05030d24c8cc5a07dbf0e8365ab2cd3368ed5bc6
+  data.tar.gz: 308659e8aaee319757fb35fd19abf69296aebe00e8d31d4405b9d262454cd1e8e020aed702a6ab043664dd13e64670bf2758bcfda41ef7c80115d6c02d4f513b

data/lib/coradoc/docx/transform/from_core_model.rb

@@ -89,6 +89,8 @@ module Coradoc
           doc = Uniword::Builder::DocumentBuilder.new
           doc.title(element.title) if element.title
 
+          FrontmatterCoreProperties.apply(doc, FrontmatterCoreProperties.extract(element))
+
           transform_children(element.children, doc)
 
           doc.build

data/lib/coradoc/docx/transform/frontmatter_core_properties.rb

@@ -0,0 +1,107 @@
+# frozen_string_literal: true
+
+module Coradoc
+  module Docx
+    module Transform
+      # Single source of truth for FrontmatterBlock -> OOXML core
+      # properties mapping. (MECE: DOCX-specific concerns live in the
+      # DOCX gem; CoreModel has no knowledge of OOXML.)
+      #
+      # Uniword's DocumentBuilder exposes setter methods (title, author,
+      # subject, keywords, description, date_field) that write the
+      # docProps/core.xml Dublin Core elements. This module reads
+      # FrontmatterBlock entries and invokes the right setters, so the
+      # mapping table lives in exactly one place.
+      #
+      # Mapping table (single source of truth):
+      #
+      #   | Frontmatter key       | OOXML element        | Builder method |
+      #   |-----------------------|----------------------|----------------|
+      #   | title                 | dc:title             | title          |
+      #   | author                | dc:creator           | author         |
+      #   | date                  | dc:date              | created        |
+      #   | description, excerpt  | dc:description       | description    |
+      #   | subject               | dc:subject           | subject        |
+      #   | tags, categories      | cp:keywords          | keywords       |
+      module FrontmatterCoreProperties
+        DESCRIPTION_KEYS = %w[description excerpt].freeze
+        KEYWORDS_KEYS = %w[tags categories].freeze
+
+        class << self
+          # Apply FrontmatterBlock data to a Uniword DocumentBuilder.
+          # Skips nil/empty values. Does not overwrite builder values
+          # that the caller has set explicitly (caller invokes this
+          # AFTER setting document.title, for example).
+          #
+          # @param builder [Uniword::Builder::DocumentBuilder]
+          # @param block [Coradoc::CoreModel::FrontmatterBlock, nil]
+          # @return [void]
+          def apply(builder, block)
+            return unless block.is_a?(Coradoc::CoreModel::FrontmatterBlock)
+
+            data = block.data || {}
+
+            if (v = find_first_scalar(data, KEYWORDS_KEYS))
+              builder.keywords(v)
+            end
+            if (v = find_first_scalar(data, %w[author]))
+              builder.author(v)
+            end
+            if (v = find_first_scalar(data, %w[subject]))
+              builder.subject(v)
+            end
+            if (v = find_first_scalar(data, DESCRIPTION_KEYS))
+              builder.description(v)
+            end
+            if (v = find_first_scalar(data, %w[date]))
+              builder.created(v)
+            end
+          end
+
+          # Locate the FrontmatterBlock in a DocumentElement's children,
+          # if any.
+          #
+          # @param document [Coradoc::CoreModel::DocumentElement]
+          # @return [Coradoc::CoreModel::FrontmatterBlock, nil]
+          def extract(document)
+            return nil unless document.is_a?(Coradoc::CoreModel::DocumentElement)
+
+            Array(document.children).find { |c| c.is_a?(Coradoc::CoreModel::FrontmatterBlock) }
+          end
+
+          private
+
+          def find_first_scalar(data, keys)
+            keys.each do |k|
+              value = data[k.to_s]
+              next if value.nil?
+
+              str = scalar_value(value)
+              return str if str && !str.empty?
+            end
+            nil
+          end
+
+          def scalar_value(value)
+            return nil if value.nil?
+
+            case value
+            when String then value
+            when Integer, Float, TrueClass, FalseClass then value.to_s
+            when Date, Time, DateTime then value.iso8601
+            when Symbol then value.to_s
+            when Array then array_to_csv(value)
+            when Hash then nil
+            else value.to_s
+            end
+          end
+
+          def array_to_csv(value)
+            strings = value.map { |i| scalar_value(i) }.compact
+            strings.any? ? strings.join(', ') : nil
+          end
+        end
+      end
+    end
+  end
+end

data/lib/coradoc/docx/transform/style_resolver.rb

@@ -94,7 +94,12 @@ module Coradoc
           return false unless paragraph.properties
 
           num_id = paragraph.properties.num_id
-          num_id.to_i.positive?
+          # Uniword (via lutaml-model) returns an UninitializedClass
+          # sentinel for attributes that weren't deserialized from source.
+          # Treat anything that isn't a real Integer as "no list id".
+          return false unless num_id.is_a?(Integer)
+
+          num_id.positive?
         end
 
         # Check if paragraph has a specific role based on style name

data/lib/coradoc/docx/transform.rb

@@ -11,6 +11,7 @@ module Coradoc
       autoload :StyleResolver, 'coradoc/docx/transform/style_resolver'
       autoload :NumberingResolver, 'coradoc/docx/transform/numbering_resolver'
       autoload :OrderedContent, 'coradoc/docx/transform/ordered_content'
+      autoload :FrontmatterCoreProperties, 'coradoc/docx/transform/frontmatter_core_properties'
 
       # Element transform rules
       module Rules

data/lib/coradoc/docx/version.rb

@@ -2,6 +2,6 @@
 
 module Coradoc
   module Docx
-    VERSION = '0.1.2'
+    VERSION = '0.1.3'
   end
 end

data/lib/coradoc/docx.rb

@@ -1,7 +1,6 @@
 # frozen_string_literal: true
 
 require 'coradoc'
-require 'coradoc/core_model'
 require 'uniword'
 
 # Coradoc::Docx provides DOCX (OOXML) format support for Coradoc.

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: coradoc-docx
 version: !ruby/object:Gem::Version
-  version: 0.1.2
+  version: 0.1.3
 platform: ruby
 authors:
 - Ribose Inc.
@@ -108,6 +108,7 @@ files:
 - lib/coradoc/docx/transform.rb
 - lib/coradoc/docx/transform/context.rb
 - lib/coradoc/docx/transform/from_core_model.rb
+- lib/coradoc/docx/transform/frontmatter_core_properties.rb
 - lib/coradoc/docx/transform/numbering_resolver.rb
 - lib/coradoc/docx/transform/ordered_content.rb
 - lib/coradoc/docx/transform/rule.rb
@@ -136,6 +137,7 @@ licenses:
 metadata:
   homepage_uri: https://github.com/lutaml/coradoc
   source_code_uri: https://github.com/lutaml/coradoc
+  rubygems_mfa_required: 'true'
 rdoc_options: []
 require_paths:
 - lib