Class: Uniword::Transformation::MhtmlMetadataBuilder Private

Inherits:
Object
  • Object
show all
Defined in:
lib/uniword/transformation/mhtml_metadata_builder.rb

Overview

This class is part of a private API. You should avoid using this class if possible, as it may be removed or be changed in the future.

Builds metadata, document properties, and file parts for MHTML output.

Extracted from OoxmlToMhtmlConverter for separation of responsibilities.

Instance Method Summary collapse

Constructor Details

#initialize(document, core_properties, relationships, document_name) ⇒ MhtmlMetadataBuilder

This method is part of a private API. You should avoid using this method if possible, as it may be removed or be changed in the future.

Returns a new instance of MhtmlMetadataBuilder.



11
12
13
14
15
16
 
# File 'lib/uniword/transformation/mhtml_metadata_builder.rb', line 11

def initialize(document, core_properties, relationships, document_name)
  @document = document
  @core_properties = core_properties
  @relationships = relationships
  @provided_document_name = document_name
end

Instance Method Details

#build_custom_document_properties ⇒ Object

This method is part of a private API. You should avoid using this method if possible, as it may be removed or be changed in the future.

Build o:CustomDocumentProperties for AssetID and other custom fields



136
137
138
139
140
141
142
143
144
145
 
# File 'lib/uniword/transformation/mhtml_metadata_builder.rb', line 136

def build_custom_document_properties
  asset_id = custom_property("AssetID")
  return "" unless asset_id

  <<~XML
    <o:CustomDocumentProperties xmlns:o="urn:schemas-microsoft-com:office:office">
     <o:AssetID dt:dt="string">#{escape_xml(asset_id)}</o:AssetID>
    </o:CustomDocumentProperties>
  XML
end

#build_document_properties ⇒ Object

This method is part of a private API. You should avoid using this method if possible, as it may be removed or be changed in the future.

Build document properties from core properties



48
49
50
51
52
53
54
55
56
57
58
59
60
61
 
# File 'lib/uniword/transformation/mhtml_metadata_builder.rb', line 48

def build_document_properties
  props = Uniword::Mhtml::Metadata::DocumentProperties.new
  cp = core_properties

  if cp.is_a?(Uniword::Ooxml::CoreProperties)
    props.author = cp.creator.to_s if cp.creator
    props.created = cp.created.value.to_s if cp.created
    props.last_author = cp.last_modified_by.to_s if cp.last_modified_by
    props.last_saved = cp.modified.value.to_s if cp.modified
    props.application = "Microsoft Word"
  end

  props
end

#build_filelist_part ⇒ Object

This method is part of a private API. You should avoid using this method if possible, as it may be removed or be changed in the future.

Build filelist.xml part



64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
 
# File 'lib/uniword/transformation/mhtml_metadata_builder.rb', line 64

def build_filelist_part
  image_entries = if @document.image_parts && !@document.image_parts.empty?
                    @document.image_parts.map do |_r_id, image_data|
                      %(<o:File HRef="#{image_data[:target]}"/>)
                    end.join("\n            ")
                  else
                    ""
                  end

  filelist_xml = <<~XML
    <xml xmlns:o="urn:schemas-microsoft-com:office:office">
      <o:MainFile HRef="../#{document_name}.htm"/>
      <o:File HRef="filelist.xml"/>
      #{image_entries}
    </xml>
  XML

  part = Uniword::Mhtml::XmlPart.new
  part.content_type = 'text/xml; charset="utf-8"'
  part.content_transfer_encoding = "quoted-printable"
  part.raw_content = filelist_xml
  part.content_location = "file:///C:/D057922B/#{document_name}.fld/filelist.xml"

  part
end

#build_image_parts ⇒ Object

This method is part of a private API. You should avoid using this method if possible, as it may be removed or be changed in the future.

Build image parts from document.image_parts



91
92
93
94
95
96
97
98
99
100
101
102
 
# File 'lib/uniword/transformation/mhtml_metadata_builder.rb', line 91

def build_image_parts
  return [] unless @document.image_parts && !@document.image_parts.empty?

  @document.image_parts.map do |_r_id, image_data|
    part = Uniword::Mhtml::ImagePart.new
    part.content_type = image_data[:content_type] || "image/png"
    part.content_transfer_encoding = "base64"
    part.raw_content = image_data[:data]
    part.content_location = "file:///C:/D057922B/#{document_name}.fld/#{image_data[:target]}"
    part
  end
end

#build_link_tags ⇒ Object

This method is part of a private API. You should avoid using this method if possible, as it may be removed or be changed in the future.

Build link tags for the head



158
159
160
161
162
 
# File 'lib/uniword/transformation/mhtml_metadata_builder.rb', line 158

def build_link_tags
  <<~LINK
    <link rel=File-List href="#{document_name}.fld/filelist.xml">
  LINK
end

#build_meta_tags ⇒ Object

This method is part of a private API. You should avoid using this method if possible, as it may be removed or be changed in the future.

Build meta tags for the head



148
149
150
151
152
153
154
155
 
# File 'lib/uniword/transformation/mhtml_metadata_builder.rb', line 148

def build_meta_tags
  <<~META
    <meta http-equiv=Content-Type content="text/html; charset=utf-8">
    <meta name=ProgId content=Word.Document>
    <meta name=Generator content="Microsoft Word 15">
    <meta name=Originator content="Microsoft Word 15">
  META
end

#build_metadata_comments ⇒ Object

This method is part of a private API. You should avoid using this method if possible, as it may be removed or be changed in the future.

Build the o:DocumentProperties content lines



105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
 
# File 'lib/uniword/transformation/mhtml_metadata_builder.rb', line 105

def build_metadata_comments
  props = core_properties
  return "" unless props

  author = props.is_a?(Uniword::Ooxml::CoreProperties) ? props.creator : nil
  last_author = props.is_a?(Uniword::Ooxml::CoreProperties) ? props.last_modified_by : nil
  revision = props.is_a?(Uniword::Ooxml::CoreProperties) ? props.revision : nil
  created = props.is_a?(Uniword::Ooxml::CoreProperties) && props.created ? props.created.value : nil
  last_saved = props.is_a?(Uniword::Ooxml::CoreProperties) && props.modified ? props.modified.value : nil

  stats = calculate_document_stats

  doc_props = []
  doc_props << " <o:Author>#{escape_xml(author)}</o:Author>" if author
  doc_props << " <o:LastAuthor>#{escape_xml(last_author)}</o:LastAuthor>" if last_author
  doc_props << " <o:Revision>#{escape_xml(revision)}</o:Revision>" if revision
  doc_props << " <o:TotalTime>0</o:TotalTime>"
  doc_props << " <o:Created>#{escape_xml(created)}</o:Created>" if created
  doc_props << " <o:LastSaved>#{escape_xml(last_saved)}</o:LastSaved>" if last_saved
  doc_props << " <o:Pages>#{stats[:pages]}</o:Pages>"
  doc_props << " <o:Words>#{stats[:words]}</o:Words>"
  doc_props << " <o:Characters>#{stats[:characters]}</o:Characters>"
  doc_props << " <o:Lines>#{stats[:lines]}</o:Lines>"
  doc_props << " <o:Paragraphs>#{stats[:paragraphs]}</o:Paragraphs>"
  doc_props << " <o:CharactersWithSpaces>#{stats[:characters_with_spaces]}</o:CharactersWithSpaces>"
  doc_props << " <o:Version>16.00</o:Version>"

  doc_props.join("\n")
end

#core_properties ⇒ Object

This method is part of a private API. You should avoid using this method if possible, as it may be removed or be changed in the future.

Get the core properties to use (provided or from document)



19
20
21
 
# File 'lib/uniword/transformation/mhtml_metadata_builder.rb', line 19

def core_properties
  @core_properties || @document.core_properties
end

#document_name ⇒ Object

This method is part of a private API. You should avoid using this method if possible, as it may be removed or be changed in the future.

Get document name - use provided name, or extract from relationships, or default



25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
 
# File 'lib/uniword/transformation/mhtml_metadata_builder.rb', line 25

def document_name
  @document_name ||= @provided_document_name || begin
    if @relationships
      rel = @relationships.relationships.find do |r|
        r.target.to_s.include?("document.xml")
      end
      if rel
        target = rel.target
        parts = target.split("/")
        return parts.last.sub(/\.xml$/, "") if parts.last
      end
    end
    location = @document.document_rels&.relationships&.first&.target || "document"
    File.basename(location, ".*")
  rescue StandardError => e
    Uniword.logger&.debug do
      "Falling back to 'document' name: #{e.message}"
    end
    "document"
  end
end