Class: Html2Doc

Inherits:
Object
  • Object
show all
Defined in:
lib/html2doc/base.rb,
lib/html2doc/math.rb,
lib/html2doc/mime.rb,
lib/html2doc/lists.rb,
lib/html2doc/notes.rb,
lib/html2doc/version.rb

Constant Summary collapse

NOKOHEAD = 
<<~HERE.freeze
  <!DOCTYPE html SYSTEM
  "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
  <html xmlns="http://www.w3.org/1999/xhtml">
  <head> <title></title> <meta charset="UTF-8" /> </head>
  <body> </body> </html>
HERE
DOCTYPE = 
<<~"DOCTYPE".freeze
  <!DOCTYPE html SYSTEM "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
DOCTYPE
<<~XML.freeze

  <xml>
  <w:WordDocument>
  <w:View>Print</w:View>
  <w:Zoom>100</w:Zoom>
  <w:DoNotOptimizeForBrowser/>
  </w:WordDocument>
  </xml>
  <meta http-equiv='Content-Type' content="text/html; charset=utf-8"/>
XML
HTML_NS = 
'xmlns="http://www.w3.org/1999/xhtml"'.freeze
IMAGE_PATH = 
"//*[local-name() = 'img' or local-name() = 'imagedata']".freeze
TOPLIST = 
"[not(ancestor::ul) and not(ancestor::ol)]".freeze
FN = 
"<span class='MsoFootnoteReference'>"\
"<span style='mso-special-character:footnote'/></span>".freeze
VERSION = 
"1.5.4".freeze

Instance Method Summary collapse

Constructor Details

#initialize(hash) ⇒ Html2Doc

Returns a new instance of Html2Doc.



7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
 
# File 'lib/html2doc/base.rb', line 7

def initialize(hash)
  @filename = hash[:filename]
  @dir = hash[:dir]
  @dir1 = create_dir(@filename, @dir)
  @header_file = hash[:header_file]
  @asciimathdelims = hash[:asciimathdelims]
  @imagedir = hash[:imagedir]
  @debug = hash[:debug]
  @liststyles = hash[:liststyles]
  @stylesheet = hash[:stylesheet]
  @c = HTMLEntities.new
  @xsltemplate =
    Nokogiri::XSLT(File.read(File.join(File.dirname(__FILE__), "mml2omml.xsl"),
                             encoding: "utf-8"))
end

Instance Method Details

#accent_tr(xml) ⇒ Object 



159
160
161
162
163
164
165
 
# File 'lib/html2doc/math.rb', line 159

def accent_tr(xml)
  xml.xpath(".//*[local-name()='accPr']/*[local-name()='chr']").each do |x|
    x["m:val"] &&= accent_tr1(x["m:val"])
    x["val"] &&= accent_tr1(x["val"])
  end
  xml
end

#accent_tr1(accent) ⇒ Object 



167
168
169
170
171
172
173
174
 
# File 'lib/html2doc/math.rb', line 167

def accent_tr1(accent)
  case accent
  when "\u2192" then "\u20D7"
  when "^" then "\u0302"
  when "~" then "\u0303"
  else accent
  end
end

#add_stylesheet(head, title, css) ⇒ Object 



197
198
199
200
201
202
203
204
205
 
# File 'lib/html2doc/base.rb', line 197

def add_stylesheet(head, title, css)
  if head.children.empty?
    head.add_child css
  elsif title.nil?
    head.children.first.add_previous_sibling css
  else
    title.add_next_sibling css
  end
end

#asciimath_to_mathml(doc, delims, retain_asciimath: false) ⇒ Object 



20
21
22
23
24
25
26
27
28
29
 
# File 'lib/html2doc/math.rb', line 20

def asciimath_to_mathml(doc, delims, retain_asciimath: false)
  return doc if delims.nil? || delims.size < 2

  m = doc.split(/(#{Regexp.escape(delims[0])}|#{Regexp.escape(delims[1])})/)
  m.each_slice(4).map.with_index do |(*a), i|
    progress_conv(i, 500, (m.size / 4).floor, 1000, "AsciiMath")
    a[2].nil? or a[2] = asciimath_to_mathml1(a[2], retain_asciimath)
    a.size > 1 ? a[0] + a[2] : a[0]
  end.join
end

#asciimath_to_mathml1(expr, retain_asciimath) ⇒ Object 



8
9
10
11
12
13
14
15
16
17
18
 
# File 'lib/html2doc/math.rb', line 8

def asciimath_to_mathml1(expr, retain_asciimath)
  ret = Plurimath::Math.parse(HTMLEntities.new.decode(expr), "asciimath").to_mathml
    .gsub(/<math>/, "<math xmlns='http://www.w3.org/1998/Math/MathML'>").strip
  retain_asciimath and
    ret += "<asciimath>#{@c.encode(@c.decode(expr), :basic)}</asciimath>"
  ret
rescue StandardError => e
  puts "parsing: #{expr}"
  puts e.message
  raise e
end

#bookmarks(docxml) ⇒ Object 



220
221
222
223
224
225
226
227
228
229
230
231
 
# File 'lib/html2doc/base.rb', line 220

def bookmarks(docxml)
  docxml.xpath("//*[@id][not(@name)][not(@style = 'mso-element:footnote')]")
    .each do |x|
    next if x["id"].empty? ||
      %w(shapetype v:shapetype shape v:shape).include?(x.name)

    if x.children.empty? then x.add_child("<a name='#{x['id']}'></a>")
    else x.children.first.previous = "<a name='#{x['id']}'></a>"
    end
    x.delete("id")
  end
end

#cleanup(docxml) ⇒ Object 



68
69
70
71
72
73
74
75
76
77
 
# File 'lib/html2doc/base.rb', line 68

def cleanup(docxml)
  namespace(docxml.root)
  image_cleanup(docxml, @dir1, @imagedir)
  mathml_to_ooml(docxml)
  lists(docxml, @liststyles)
  footnotes(docxml)
  bookmarks(docxml)
  msonormal(docxml)
  docxml
end

#clear_dir(dir) ⇒ Object 



41
42
43
44
45
46
47
 
# File 'lib/html2doc/base.rb', line 41

def clear_dir(dir)
  Dir.foreach(dir) do |f|
    fn = File.join(dir, f)
    File.delete(fn) if f != "." && f != ".."
  end
  dir
end

#contentid(mhtml) ⇒ Object 



69
70
71
72
73
74
75
76
77
 
# File 'lib/html2doc/mime.rb', line 69

def contentid(mhtml)
  mhtml.gsub %r{(<img[^>]*?src=")([^"']+)(['"])}m do |m|
    repl = "#{$1}cid:#{File.basename($2)}#{$3}"
    /^data:|^https?:/ =~ $2 ? m : repl
  end.gsub %r{(<v:imagedata[^>]*?src=")([^"']+)(['"])}m do |m|
    repl = "#{$1}cid:#{File.basename($2)}#{$3}"
    /^data:|^https?:/ =~ $2 ? m : repl
  end
end

#create_dir(filename, dir) ⇒ Object 



49
50
51
52
53
54
 
# File 'lib/html2doc/base.rb', line 49

def create_dir(filename, dir)
  dir and return clear_dir(dir)
  dir = "#{filename}_files"
  FileUtils.mkdir_p(dir)
  clear_dir(dir)
end

#define_head(docxml) ⇒ Object 



187
188
189
190
191
192
193
194
195
 
# File 'lib/html2doc/base.rb', line 187

def define_head(docxml)
  title = docxml.at("//*[local-name() = 'head']/*[local-name() = 'title']")
  head = docxml.at("//*[local-name() = 'head']")
  css = stylesheet(@filename, @header_file, @stylesheet)
  add_stylesheet(head, title, css)
  filename_substitute(head, @header_file)
  define_head1(docxml, @dir1)
  rootnamespace(docxml.root)
end

#define_head1(docxml, _dir) ⇒ Object 



150
151
152
153
154
155
156
157
 
# File 'lib/html2doc/base.rb', line 150

def define_head1(docxml, _dir)
  docxml.xpath("//*[local-name() = 'head']").each do |h|
    h.children.first.add_previous_sibling <<~XML
      #{PRINT_VIEW}
        <link rel="File-List" href="cid:filelist.xml"/>
    XML
  end
end

#esc_space(xml) ⇒ Object

escape space as &#x32;; we are removing any spaces generated by XML indentation



178
179
180
181
182
183
184
185
 
# File 'lib/html2doc/math.rb', line 178

def esc_space(xml)
  xml.traverse do |n|
    next unless n.text?

    n = n.text.gsub(/ /, "&#x32;")
  end
  xml
end

#filename_substitute(head, header_filename) ⇒ Object 



159
160
161
162
163
164
165
166
167
168
 
# File 'lib/html2doc/base.rb', line 159

def filename_substitute(head, header_filename)
  return if header_filename.nil?

  head.xpath(".//*[local-name() = 'style']").each do |s|
    s1 = s.to_xml.gsub(/url\("[^"]+"\)/) do |m|
      /FILENAME/.match?(m) ? "url(cid:header.html)" : m
    end
    s.replace(s1)
  end
end

#find_page_size(stylesheet, klass, in_xml) ⇒ Object

if in_xml, CSS is embedded in XML <style> tag



170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
 
# File 'lib/html2doc/mime.rb', line 170

def find_page_size(stylesheet, klass, in_xml)
  xml_found = false
  found = false
  ret = ""
  stylesheet&.lines&.each do |l|
    in_xml && l.include?("<style") and xml_found = true and found = false
    in_xml && l.include?("</style>") and xml_found = false
    /^\s*@page\s+#{klass}/.match?(l) and found = true
    found && /^\s*\{?size:/.match?(l) and ret += l
    found && /^\s*\{?margin:/.match?(l) and ret += l
    if found && /}/.match?(l)
      !ret.blank? && (!in_xml || xml_found) and return ret
      ret = ""
      found = false
    end
  end
  nil
end

#find_page_size_in_doc(stylesheet, doc) ⇒ Object 



161
162
163
164
165
166
167
 
# File 'lib/html2doc/mime.rb', line 161

def find_page_size_in_doc(stylesheet, doc)
  find_page_size(stylesheet, "WordSection2", false) ||
    find_page_size(stylesheet, "WordSection3", false) ||
    find_page_size(doc, "WordSection2", true) ||
    find_page_size(doc, "WordSection3", true) ||
    find_page_size(stylesheet, "", false) || find_page_size(doc, "", true)
end

#footnote?(elem) ⇒ Boolean

Returns:

  • (Boolean) 


85
86
87
88
 
# File 'lib/html2doc/notes.rb', line 85

def footnote?(elem)
  elem["epub:type"]&.casecmp("footnote")&.zero? ||
    elem["class"]&.casecmp("footnote")&.zero?
end

#footnote_cleanup(docxml) ⇒ Object

We expect that the content of the footnote text received is one or more text containers, p or aside or div (which we have already converted to p). We do not expect any <a name> or links back to text; if they are present in the HTML, they need to have been cleaned out before passing to this gem



102
103
104
105
106
107
108
109
 
# File 'lib/html2doc/notes.rb', line 102

def footnote_cleanup(docxml)
  docxml.xpath('//div[@style="mso-element:footnote"]/a')
    .each do |x|
    n = x.next_element
    n&.children&.first&.add_previous_sibling(x.remove)
  end
  docxml
end

#footnote_container(docxml, idx) ⇒ Object 



40
41
42
43
44
45
46
47
48
 
# File 'lib/html2doc/notes.rb', line 40

def footnote_container(docxml, idx)
  ref = docxml&.at("//a[@href='#_ftn#{idx}']")&.children&.to_xml(indent: 0)
    &.gsub(/>\n</, "><") || FN
  <<~DIV
    <div style='mso-element:footnote' id='ftn#{idx}'>
      <a style='mso-footnote-id:ftn#{idx}' href='#_ftn#{idx}'
         name='_ftnref#{idx}' title='' id='_ftnref#{idx}'>#{ref.strip}</a></div>
  DIV
end

#footnote_div_to_p(elem) ⇒ Object 



26
27
28
29
30
31
32
33
34
35
 
# File 'lib/html2doc/notes.rb', line 26

def footnote_div_to_p(elem)
  if %w{div aside}.include? elem.name
    if elem.at(".//p")
      elem.replace(elem.children)
    else
      elem.name = "p"
      elem["class"] = "MsoFootnoteText"
    end
  end
end

#footnotes(docxml) ⇒ Object 



4
5
6
7
8
9
10
11
12
13
 
# File 'lib/html2doc/notes.rb', line 4

def footnotes(docxml)
  i = 1
  fn = []
  docxml.xpath("//a").each do |a|
    next unless process_footnote_link(docxml, a, i, fn)

    i += 1
  end
  process_footnote_texts(docxml, fn)
end

#from_xhtml(xml) ⇒ Object 



102
103
104
105
106
107
108
 
# File 'lib/html2doc/base.rb', line 102

def from_xhtml(xml)
  xml.to_xml.sub(%{ xmlns="http://www.w3.org/1999/xhtml"}, "")
    .sub(DOCTYPE, "").gsub(%{ />}, "/>")
    .gsub(/<!-- MSWORD-COMMENT (.+?) -->/, "<!--[\\1]>")
    .gsub(/<!-- MSWORD-COMMENT-END -->/, "<![endif]-->")
    .gsub("\n--&gt;\n", "\n-->\n")
end

#generate_filelist(filename, dir) ⇒ Object 



220
221
222
223
224
225
226
227
228
229
230
231
 
# File 'lib/html2doc/mime.rb', line 220

def generate_filelist(filename, dir)
  File.open(File.join(dir, "filelist.xml"), "w") do |f|
    f.write %{<xml xmlns:o="urn:schemas-microsoft-com:office:office">
      <o:MainFile HRef="../#{filename}.htm"/>}
    Dir.entries(dir).sort.each do |item|
      next if item == "." || item == ".." || /^\./.match(item)

      f.write %{  <o:File HRef="#{item}"/>\n}
    end
    f.write("</xml>\n")
  end
end

#get_image_size(img, path) ⇒ Object 



93
94
95
96
97
98
99
 
# File 'lib/html2doc/mime.rb', line 93

def get_image_size(img, path)
  realsize = ImageSize.path(path).size
  s = [img["width"].to_i, img["height"].to_i]
  s = realsize if s[0].zero? && s[1].zero?
  s = [nil, nil]  if realsize.nil? || realsize[0].nil? || realsize[1].nil?
  [s, realsize]
end

#header_image_cleanup(doc, dir, filename, localdir) ⇒ Object

do not parse the header through Nokogiri, since it will contain non-XML like <![if !supportFootnotes]>



202
203
204
205
206
 
# File 'lib/html2doc/mime.rb', line 202

def header_image_cleanup(doc, dir, filename, localdir)
  doc.split(%r{(<img [^>]*>|<v:imagedata [^>]*>)}).each_slice(2).map do |a|
    header_image_cleanup1(a, dir, filename, localdir)
  end.join
end

#header_image_cleanup1(a, dir, _filename, localdir) ⇒ Object 



208
209
210
211
212
213
214
215
216
217
218
 
# File 'lib/html2doc/mime.rb', line 208

def header_image_cleanup1(a, dir, _filename, localdir)
  if a.size == 2 && !(/ src="https?:/.match a[1]) &&
      !(%r{ src="data:(image|application)/[^;]+;base64}.match a[1])
    m = / src=['"](?<src>[^"']+)['"]/.match a[1]
    m2 = /\.(?<suffix>[a-zA-Z_0-9]+)$/.match m[:src]
    new_filename = "#{mkuuid}.#{m2[:suffix]}"
    FileUtils.cp localname(m[:src], localdir), File.join(dir, new_filename)
    a[1].sub!(%r{ src=['"](?<src>[^"']+)['"]}, " src='cid:#{new_filename}'")
  end
  a.join
end

#image_cleanup(docxml, dir, localdir) ⇒ Object

only processes locally stored images



116
117
118
119
120
121
122
123
124
125
 
# File 'lib/html2doc/mime.rb', line 116

def image_cleanup(docxml, dir, localdir)
  maxheight, maxwidth = page_dimensions(docxml)
  docxml.traverse do |i|
    skip_image_cleanup?(i) and next
    local_filename = rename_image(i, dir, localdir)
    i["width"], i["height"] = image_resize(i, local_filename, maxheight,
                                           maxwidth)
  end
  docxml
end

#image_resize(img, path, maxheight, maxwidth) ⇒ Object 



79
80
81
82
83
84
85
86
87
88
89
90
91
 
# File 'lib/html2doc/mime.rb', line 79

def image_resize(img, path, maxheight, maxwidth)
  s, realsize = get_image_size(img, path)
  return s if s[0] == nil && s[1] == nil

  if img.name == "svg" && !img["viewBox"]
    img["viewBox"] = "0 0 #{s[0]} #{s[1]}"
  end
  s[1] = s[0] * realsize[1] / realsize[0] if s[1].zero? && !s[0].zero?
  s[0] = s[1] * realsize[0] / realsize[1] if s[0].zero? && !s[1].zero?
  s = [(s[0] * maxheight / s[1]).ceil, maxheight] if s[1] > maxheight
  s = [maxwidth, (s[1] * maxwidth / s[0]).ceil] if s[0] > maxwidth
  s
end

#list2para(list) ⇒ Object 



51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
 
# File 'lib/html2doc/lists.rb', line 51

def list2para(list)
  return if list.xpath("./li").empty?

  list2para_position(list)
  list.xpath("./li").each do |l|
    l.name = "p"
    l["class"] ||= "MsoListParagraphCxSpMiddle"
    next unless l.first_element_child&.name == "p"

    l["style"] ||= ""
    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)
end

#list2para_position(list) ⇒ Object 



67
68
69
70
71
72
73
 
# File 'lib/html2doc/lists.rb', line 67

def list2para_position(list)
  list.xpath("./li").first["class"] ||= "MsoListParagraphCxSpFirst"
  list.xpath("./li").last["class"] ||= "MsoListParagraphCxSpLast"
  list.xpath("./li/p").each do |p|
    p["class"] ||= "MsoListParagraphCxSpMiddle"
  end
end

#list_add(xpath, liststyles, listtype, level) ⇒ Object 



31
32
33
34
35
36
37
38
39
40
41
 
# File 'lib/html2doc/lists.rb', line 31

def list_add(xpath, liststyles, listtype, level)
  xpath.each do |l|
    level == 1 and l["seen"] = true and @listnumber += 1
    l["id"] ||= UUIDTools::UUID.random_create
    (l.xpath(".//li") - l.xpath(".//ol//li | .//ul//li")).each do |li|
      style_list(li, level, liststyles[listtype], @listnumber)
      list_add1(li, liststyles, listtype, level)
    end
    list_add_tail(l, liststyles, listtype, level)
  end
end

#list_add1(elem, liststyles, listtype, level) ⇒ Object 



17
18
19
20
21
22
23
24
25
26
27
28
29
 
# File 'lib/html2doc/lists.rb', line 17

def list_add1(elem, liststyles, listtype, level)
  if %i[ul ol].include? listtype
    list_add(elem.xpath(".//ul") - elem.xpath(".//ul//ul | .//ol//ul"),
             liststyles, :ul, level + 1)
    list_add(elem.xpath(".//ol") - elem.xpath(".//ul//ol | .//ol//ol"),
             liststyles, :ol, level + 1)
  else
    list_add(elem.xpath(".//ul") - elem.xpath(".//ul//ul | .//ol//ul"),
             liststyles, listtype, level + 1)
    list_add(elem.xpath(".//ol") - elem.xpath(".//ul//ol | .//ol//ol"),
             liststyles, listtype, level + 1)
  end
end

#list_add_tail(list, liststyles, listtype, level) ⇒ Object 



43
44
45
46
47
48
49
 
# File 'lib/html2doc/lists.rb', line 43

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']}')]")
    .each do |li|
    list_add1(li.parent, liststyles, listtype, level - 1)
  end
end

#lists(docxml, liststyles) ⇒ Object 



102
103
104
105
106
107
108
109
110
 
# File 'lib/html2doc/lists.rb', line 102

def lists(docxml, liststyles)
  return if liststyles.nil?

  @listnumber = 0
  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

#lists1(docxml, liststyles, style) ⇒ Object 



77
78
79
80
81
82
83
84
85
86
87
88
 
# File 'lib/html2doc/lists.rb', line 77

def lists1(docxml, liststyles, style)
  case style
  when :ul then list_add(docxml.xpath("//ul[not(@class)]#{TOPLIST}"),
                         liststyles, :ul, 1)
  when :ol then list_add(docxml.xpath("//ol[not(@class)]#{TOPLIST}"),
                         liststyles, :ol, 1)
  else
    list_add(docxml.xpath("//ol[@class = '#{style}']#{TOPLIST} | "\
                          "//ul[@class = '#{style}']#{TOPLIST}"),
             liststyles, style, 1)
  end
end

#lists_unstyled(docxml, liststyles) ⇒ Object 



90
91
92
93
94
95
96
97
98
99
100
 
# File 'lib/html2doc/lists.rb', line 90

def lists_unstyled(docxml, liststyles)
  liststyles.has_key?(:ul) and
    list_add(docxml.xpath("//ul#{TOPLIST}[not(@seen)]"),
             liststyles, :ul, 1)
  liststyles.has_key?(:ol) and
    list_add(docxml.xpath("//ol#{TOPLIST}[not(@seen)]"),
             liststyles, :ul, 1)
  docxml.xpath("//ul[@seen] | //ol[@seen]").each do |l|
    l.delete("seen")
  end
end

#localname(src, localdir) ⇒ Object 



111
112
113
 
# File 'lib/html2doc/mime.rb', line 111

def localname(src, localdir)
  %r{^([A-Z]:)?/}.match?(src) ? src : File.join(localdir, src)
end

#mathml_insert_rows(math, docnamespaces) ⇒ Object 



57
58
59
60
61
62
63
64
65
 
# File 'lib/html2doc/math.rb', line 57

def mathml_insert_rows(math, docnamespaces)
  math.xpath(%w(msup msub msubsup munder mover munderover)
          .map { |m| ".//xmlns:#{m}" }.join(" | "), docnamespaces).each do |x|
    next unless x.next_element && x.next_element != "mrow"

    x.next_element.wrap("<mrow/>")
  end
  math
end

#mathml_preserve_space(math, docnamespaces) ⇒ Object 



67
68
69
70
71
72
 
# File 'lib/html2doc/math.rb', line 67

def mathml_preserve_space(math, docnamespaces)
  math.xpath(".//xmlns:mtext", docnamespaces).each do |x|
    x.children = x.children.to_xml.gsub(/^\s/, "&#xA0;").gsub(/\s$/, "&#xA0;")
  end
  math
end

#mathml_to_ooml(docxml) ⇒ Object 



131
132
133
134
135
136
137
138
 
# File 'lib/html2doc/math.rb', line 131

def mathml_to_ooml(docxml)
  docnamespaces = docxml.collect_namespaces
  m = docxml.xpath("//*[local-name() = 'math']")
  m.each_with_index do |x, i|
    progress_conv(i, 100, m.size, 500, "Math OOXML")
    mathml_to_ooml1(x, docnamespaces)
  end
end

#mathml_to_ooml1(xml, docnamespaces) ⇒ Object 



151
152
153
154
155
156
157
 
# File 'lib/html2doc/math.rb', line 151

def mathml_to_ooml1(xml, docnamespaces)
  doc = Nokogiri::XML::Document::new
  doc.root = ooxml_cleanup(xml, docnamespaces)
  ooxml = ooml_clean(unitalic(esc_space(accent_tr(@xsltemplate.transform(doc)))))
  ooxml = uncenter(xml, ooxml)
  xml.swap(ooxml)
end

#mime_attachment(boundary, _filename, item, dir) ⇒ Object 



23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
 
# File 'lib/html2doc/mime.rb', line 23

def mime_attachment(boundary, _filename, item, dir)
  content_type = mime_type(item)
  text_mode = %w[text application].any? { |p| content_type.start_with? p }

  path = File.join(dir, item)
  content = text_mode ? File.read(path, encoding: "utf-8") : IO.binread(path)

  encoded_file = Base64.strict_encode64(content).gsub(/(.{76})/, "\\1\n")
  <<~"FILE"
    --#{boundary}
    Content-ID: <#{File.basename(item)}>
    Content-Disposition: inline; filename="#{File.basename(item)}"
    Content-Transfer-Encoding: base64
    Content-Type: #{content_type}

    #{encoded_file}

  FILE
end

#mime_boundary ⇒ Object 



50
51
52
53
 
# File 'lib/html2doc/mime.rb', line 50

def mime_boundary
  salt = UUIDTools::UUID.random_create.to_s.gsub(/-/, ".")[0..17]
  "----=_NextPart_#{salt}"
end

#mime_package(result, filename, dir) ⇒ Object 



55
56
57
58
59
60
61
62
63
64
65
66
67
 
# File 'lib/html2doc/mime.rb', line 55

def mime_package(result, filename, dir)
  boundary = mime_boundary
  mhtml = mime_preamble(boundary, "#{filename}.htm", result)
  mhtml += mime_attachment(boundary, "#{filename}.htm", "filelist.xml", dir)
  Dir.foreach(dir) do |item|
    next if item == "." || item == ".." || /^\./.match(item) ||
      item == "filelist.xml"

    mhtml += mime_attachment(boundary, "#{filename}.htm", item, dir)
  end
  mhtml += "--#{boundary}--"
  File.open("#{filename}.doc", "w:UTF-8") { |f| f.write contentid(mhtml) }
end

#mime_preamble(boundary, filename, result) ⇒ Object 



8
9
10
11
12
13
14
15
16
17
18
19
20
21
 
# File 'lib/html2doc/mime.rb', line 8

def mime_preamble(boundary, filename, result)
  <<~"PREAMBLE"
    MIME-Version: 1.0
    Content-Type: multipart/related; boundary="#{boundary}"

    --#{boundary}
    Content-ID: <#{File.basename(filename)}>
    Content-Disposition: inline; filename="#{File.basename(filename)}"
    Content-Type: text/html; charset="utf-8"

    #{result}

  PREAMBLE
end

#mime_type(item) ⇒ Object 



43
44
45
46
47
48
 
# File 'lib/html2doc/mime.rb', line 43

def mime_type(item)
  types = MIME::Types.type_for(item)
  type = types ? types.first.to_s : 'text/plain; charset="utf-8"'
  type = %(#{type} charset="utf-8") if /^text/.match(type) && types
  type
end

#mkuuid ⇒ Object 



103
104
105
 
# File 'lib/html2doc/mime.rb', line 103

def mkuuid
  UUIDTools::UUID.random_create.to_s
end

#msonormal(docxml) ⇒ Object 



233
234
235
236
237
238
239
240
 
# File 'lib/html2doc/base.rb', line 233

def msonormal(docxml)
  docxml.xpath("//*[local-name() = 'p'][not(self::*[@class])]").each do |p|
    p["class"] = "MsoNormal"
  end
  docxml.xpath("//*[local-name() = 'li'][not(self::*[@class])]").each do |p|
    p["class"] = "MsoNormal"
  end
end

#msword_fix(doc) ⇒ Object 



110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
 
# File 'lib/html2doc/base.rb', line 110

def msword_fix(doc)
  # brain damage in MSWord parser
  doc.gsub!(%r{<w:DoNotOptimizeForBrowser></w:DoNotOptimizeForBrowser>},
            "<w:DoNotOptimizeForBrowser/>")
  doc.gsub!(%r{<span style="mso-special-character:footnote"/>},
            '<span style="mso-special-character:footnote"></span>')
  doc.gsub!(%r{<div style="mso-element:footnote-list"></div>},
            '<div style="mso-element:footnote-list"/>')
  doc.gsub!(%r{(<a style="mso-comment-reference:[^>/]+)/>}, "\\1></a>")
  doc.gsub!(%r{<link rel="File-List"}, "<link rel=File-List")
  doc.gsub!(%r{<meta http-equiv="Content-Type"},
            "<meta http-equiv=Content-Type")
  doc.gsub!(%r{></m:jc>}, "/>")
  doc.gsub!(%r{></v:stroke>}, "/>")
  doc.gsub!(%r{></v:f>}, "/>")
  doc.gsub!(%r{></v:path>}, "/>")
  doc.gsub!(%r{></o:lock>}, "/>")
  doc.gsub!(%r{></v:imagedata>}, "/>")
  doc.gsub!(%r{></w:wrap>}, "/>")
  doc.gsub!(%r{<(/)?m:(span|em)\b}, "<\\1\\2")
  doc.gsub!(%r{&tab;|&amp;tab;},
            '<span style="mso-tab-count:1">&#xA0; </span>')
  doc.split(%r{(<m:oMath>|</m:oMath>)}).each_slice(4).map do |a|
    a.size > 2 and a[2] = a[2].gsub(/>\s+</, "><")
    a
  end.join
end

#namespace(root) ⇒ Object 



207
208
209
210
211
212
213
214
 
# File 'lib/html2doc/base.rb', line 207

def namespace(root)
  {
    o: "urn:schemas-microsoft-com:office:office",
    w: "urn:schemas-microsoft-com:office:word",
    v: "urn:schemas-microsoft-com:vml",
    m: "http://schemas.microsoft.com/office/2004/12/omml",
  }.each { |k, v| root.add_namespace_definition(k.to_s, v) }
end

#ooml_clean(xml) ⇒ Object

We need span and em not to be namespaced. Word can’t deal with explicit namespaces. We will end up stripping them out again under Nokogiri 1.11, which correctly insists on inheriting namespace from parent.



144
145
146
147
148
149
 
# File 'lib/html2doc/math.rb', line 144

def ooml_clean(xml)
  xml.to_s
    .gsub(/<\?[^>]+>\s*/, "")
    .gsub(/ xmlns(:[^=]+)?="[^"]+"/, "")
    .gsub(%r{<(/)?(?!span)(?!em)([a-z])}, "<\\1m:\\2")
end

#ooxml_cleanup(math, docnamespaces) ⇒ Object

random fixes to MathML input that OOXML needs to render properly



47
48
49
50
51
52
53
54
55
 
# File 'lib/html2doc/math.rb', line 47

def ooxml_cleanup(math, docnamespaces)
  math = unwrap_accents(
    mathml_preserve_space(
      mathml_insert_rows(math, docnamespaces), docnamespaces
    ),
  )
  math.add_namespace(nil, "http://www.w3.org/1998/Math/MathML")
  math
end

#page_dimensions(docxml) ⇒ Object

we are going to use the 2nd instance of @page in the Word CSS, skipping the cover page. Currently doesn’t deal with Landscape. Scan both @stylesheet and docxml.to_xml (where @standardstylesheet has ended up) Allow 0.9 * height to fit caption



148
149
150
151
152
153
154
155
156
157
158
159
 
# File 'lib/html2doc/mime.rb', line 148

def page_dimensions(docxml)
  stylesheet = read_stylesheet(@stylesheet)
  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
    return [680, 400]
  [0.9 * (units_to_px(m_size[2]) - units_to_px(m_marg[1]) - units_to_px(m_marg[3])),
   units_to_px(m_size[1]) - units_to_px(m_marg[2]) - units_to_px(m_marg[4])]
rescue StandardError
  [680, 400]
end

#process(result) ⇒ Object 



23
24
25
26
27
28
29
30
 
# File 'lib/html2doc/base.rb', line 23

def process(result)
  result = process_html(result)
  process_header(@header_file)
  generate_filelist(@filename, @dir1)
  File.open("#{@filename}.htm", "w:UTF-8") { |f| f.write(result) }
  mime_package result, @filename, @dir1
  rm_temp_files(@filename, @dir, @dir1) unless @debug
end

#process_footnote_link(docxml, elem, idx, footnote) ⇒ Object 



50
51
52
53
54
55
56
57
58
59
60
61
62
63
 
# File 'lib/html2doc/notes.rb', line 50

def process_footnote_link(docxml, elem, idx, footnote)
  return false unless footnote?(elem)

  href = elem["href"].gsub(/^#/, "")
  note = docxml.at("//*[@name = '#{href}' or @id = '#{href}']")
  return false if note.nil?

  set_footnote_link_attrs(elem, idx)
  if elem.at("./span[@class = 'MsoFootnoteReference']")
    process_footnote_link1(elem)
  else elem.children = FN
  end
  footnote << transform_footnote_text(note)
end

#process_footnote_link1(elem) ⇒ Object 



65
66
67
68
69
70
71
72
73
 
# File 'lib/html2doc/notes.rb', line 65

def process_footnote_link1(elem)
  elem.children.each do |c|
    if c.name == "span" && c["class"] == "MsoFootnoteReference"
      c.replace(FN)
    else
      c.wrap("<span class='MsoFootnoteReference'></span>")
    end
  end
end

#process_footnote_texts(docxml, footnotes) ⇒ Object 



15
16
17
18
19
20
21
22
23
24
 
# File 'lib/html2doc/notes.rb', line 15

def process_footnote_texts(docxml, footnotes)
  body = docxml.at("//body")
  list = body.add_child("<div style='mso-element:footnote-list'/>")
  footnotes.each_with_index do |f, i|
    fn = list.first.add_child(footnote_container(docxml, i + 1))
    f.parent = fn.first
    footnote_div_to_p(f)
  end
  footnote_cleanup(docxml)
end

#process_header(headerfile) ⇒ Object 



32
33
34
35
36
37
38
39
 
# File 'lib/html2doc/base.rb', line 32

def process_header(headerfile)
  return if headerfile.nil?

  doc = File.read(headerfile, encoding: "utf-8")
  doc = header_image_cleanup(doc, @dir1, @filename,
                             File.dirname(@filename))
  File.open("#{@dir1}/header.html", "w:UTF-8") { |f| f.write(doc) }
end

#process_html(result) ⇒ Object 



56
57
58
59
60
 
# File 'lib/html2doc/base.rb', line 56

def process_html(result)
  docxml = to_xhtml(asciimath_to_mathml(result, @asciimathdelims))
  define_head(cleanup(docxml))
  msword_fix(from_xhtml(docxml))
end

#progress_conv(idx, step, total, threshold, msg) ⇒ Object 



31
32
33
34
35
 
# File 'lib/html2doc/math.rb', line 31

def progress_conv(idx, step, total, threshold, msg)
  return unless (idx % step).zero? && total > threshold && idx.positive?

  warn "#{msg} #{idx} of #{total}"
end

#read_stylesheet(cssname) ⇒ Object 



181
182
183
184
185
 
# File 'lib/html2doc/base.rb', line 181

def read_stylesheet(cssname)
  (cssname.nil? || cssname.empty?) and
    cssname = File.join(File.dirname(__FILE__), "wordstyle.css")
  File.read(cssname, encoding: "UTF-8")
end

#rename_image(img, dir, localdir) ⇒ Object 



127
128
129
130
131
132
133
 
# File 'lib/html2doc/mime.rb', line 127

def rename_image(img, dir, localdir)
  local_filename = localname(img["src"], localdir)
  new_filename = "#{mkuuid}#{File.extname(img['src'])}"
  FileUtils.cp local_filename, File.join(dir, new_filename)
  img["src"] = File.join(File.basename(dir), new_filename)
  local_filename
end

#rm_temp_files(filename, dir, dir1) ⇒ Object 



62
63
64
65
66
 
# File 'lib/html2doc/base.rb', line 62

def rm_temp_files(filename, dir, dir1)
  FileUtils.rm "#{filename}.htm"
  FileUtils.rm_f "#{dir1}/header.html"
  FileUtils.rm_r dir1 unless dir
end

#rootnamespace(root) ⇒ Object 



216
217
218
 
# File 'lib/html2doc/base.rb', line 216

def rootnamespace(root)
  root.add_namespace(nil, "http://www.w3.org/TR/REC-html40")
end

#set_footnote_link_attrs(elem, idx) ⇒ Object 



90
91
92
93
94
95
 
# File 'lib/html2doc/notes.rb', line 90

def set_footnote_link_attrs(elem, idx)
  elem["style"] = "mso-footnote-id:ftn#{idx}"
  elem["href"] = "#_ftn#{idx}"
  elem["name"] = "_ftnref#{idx}"
  elem["title"] = ""
end

#skip_image_cleanup?(img) ⇒ Boolean

Returns:

  • (Boolean) 


135
136
137
138
139
140
141
142
 
# File 'lib/html2doc/mime.rb', line 135

def skip_image_cleanup?(img)
  src = img["src"]
  return true unless img.element? && %w(img v:imagedata).include?(img.name)
  return true if src.nil? || src.empty? || /^http/.match?(src) ||
    %r{^data:(image|application)/[^;]+;base64}.match?(src)

  false
end

#style_list(elem, level, liststyle, listnumber) ⇒ Object 



6
7
8
9
10
11
12
13
14
15
 
# File 'lib/html2doc/lists.rb', line 6

def style_list(elem, level, liststyle, listnumber)
  return unless liststyle

  if elem["style"]
    elem["style"] += ";"
  else
    elem["style"] = ""
  end
  elem["style"] += "mso-list:#{liststyle} level#{level} lfo#{listnumber};"
end

#stylesheet(_filename, _header_filename, cssname) ⇒ Object 



170
171
172
173
174
175
176
177
178
179
 
# File 'lib/html2doc/base.rb', line 170

def stylesheet(_filename, _header_filename, cssname)
  stylesheet = read_stylesheet(cssname)
  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
end

#to_plane1(xml, font) ⇒ Object 



122
123
124
125
126
127
128
129
 
# File 'lib/html2doc/math.rb', line 122

def to_plane1(xml, font)
  xml.traverse do |n|
    next unless n.text?

    n.replace(Plane1Converter.conv(@c.decode(n.text), font))
  end
  xml
end

#to_xhtml(xml) ⇒ Object 



87
88
89
90
91
92
93
94
95
96
 
# File 'lib/html2doc/base.rb', line 87

def to_xhtml(xml)
  xml.gsub!(/<\?xml[^>]*>/, "")
  unless /<!DOCTYPE /.match? xml
    xml = '<!DOCTYPE html SYSTEM
        "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">' + xml
  end
  xml = xml.gsub(/<!--\s*\[([^\]]+)\]>/, "<!-- MSWORD-COMMENT \\1 -->")
    .gsub(/<!\s*\[endif\]\s*-->/, "<!-- MSWORD-COMMENT-END -->")
  Nokogiri::XML.parse(xml)
end

#transform_footnote_text(note) ⇒ Object 



75
76
77
78
79
80
81
82
83
 
# File 'lib/html2doc/notes.rb', line 75

def transform_footnote_text(note)
  note["id"] = ""
  note.xpath(".//div").each { |div| div.replace(div.children) }
  note.xpath(".//aside | .//p").each do |p|
    p.name = "p"
    p["class"] = "MsoFootnoteText"
  end
  note.remove
end

#uncenter(math, ooxml) ⇒ Object

if oomml has no siblings, by default it is centered; override this with left/right if parent is so tagged



189
190
191
192
193
194
195
196
197
198
199
200
201
 
# File 'lib/html2doc/math.rb', line 189

def uncenter(math, ooxml)
  alignnode = math.at(".//ancestor::*[@style][local-name() = 'p' or "\
                      "local-name() = 'div' or local-name() = 'td']/@style")
  return ooxml unless alignnode && (math.next == nil && math.previous == nil)

  %w(left right).each do |dir|
    if alignnode.text.include? ("text-align:#{dir}")
      ooxml = "<m:oMathPara><m:oMathParaPr><m:jc "\
              "m:val='#{dir}'/></m:oMathParaPr>#{ooxml}</m:oMathPara>"
    end
  end
  ooxml
end

#unitalic(math) ⇒ Object 



76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
 
# File 'lib/html2doc/math.rb', line 76

def unitalic(math)
  math.xpath(".//xmlns:r[xmlns:rPr[not(xmlns:scr)]/xmlns:sty[@m:val = 'p']]").each do |x|
    x.wrap("<span #{HTML_NS} style='font-style:normal;'></span>")
  end
  math.xpath(".//xmlns:r[xmlns:rPr[not(xmlns:scr)]/xmlns:sty[@m:val = 'bi']]").each do |x|
    x.wrap("<span #{HTML_NS} class='nostem' style='font-weight:bold;'><em></em></span>")
  end
  math.xpath(".//xmlns:r[xmlns:rPr[not(xmlns:scr)]/xmlns:sty[@m:val = 'i']]").each do |x|
    x.wrap("<span #{HTML_NS} class='nostem'><em></em></span>")
  end
  math.xpath(".//xmlns:r[xmlns:rPr[not(xmlns:scr)]/xmlns:sty[@m:val = 'b']]").each do |x|
    x.wrap("<span #{HTML_NS} style='font-style:normal;font-weight:bold;'></span>")
  end
  math.xpath(".//xmlns:r[xmlns:rPr/xmlns:scr[@m:val = 'monospace']]").each do |x|
    to_plane1(x, :monospace)
  end
  math.xpath(".//xmlns:r[xmlns:rPr/xmlns:scr[@m:val = 'double-struck']]").each do |x|
    to_plane1(x, :doublestruck)
  end
  math.xpath(".//xmlns:r[xmlns:rPr[not(xmlns:sty) or xmlns:sty/@m:val = 'p']/xmlns:scr[@m:val = 'script']]").each do |x|
    to_plane1(x, :script)
  end
  math.xpath(".//xmlns:r[xmlns:rPr[xmlns:sty/@m:val = 'b']/xmlns:scr[@m:val = 'script']]").each do |x|
    to_plane1(x, :scriptbold)
  end
  math.xpath(".//xmlns:r[xmlns:rPr[not(xmlns:sty) or xmlns:sty/@m:val = 'p']/xmlns:scr[@m:val = 'fraktur']]").each do |x|
    to_plane1(x, :fraktur)
  end
  math.xpath(".//xmlns:r[xmlns:rPr[xmlns:sty/@m:val = 'b']/xmlns:scr[@m:val = 'fraktur']]").each do |x|
    to_plane1(x, :frakturbold)
  end
  math.xpath(".//xmlns:r[xmlns:rPr[not(xmlns:sty) or xmlns:sty/@m:val = 'p']/xmlns:scr[@m:val = 'sans-serif']]").each do |x|
    to_plane1(x, :sans)
  end
  math.xpath(".//xmlns:r[xmlns:rPr[xmlns:sty/@m:val = 'b']/xmlns:scr[@m:val = 'sans-serif']]").each do |x|
    to_plane1(x, :sansbold)
  end
  math.xpath(".//xmlns:r[xmlns:rPr[xmlns:sty/@m:val = 'i']/xmlns:scr[@m:val = 'sans-serif']]").each do |x|
    to_plane1(x, :sansitalic)
  end
  math.xpath(".//xmlns:r[xmlns:rPr[xmlns:sty/@m:val = 'bi']/xmlns:scr[@m:val = 'sans-serif']]").each do |x|
    to_plane1(x, :sansbolditalic)
  end
  math
end

#units_to_px(measure) ⇒ Object 



189
190
191
192
193
194
195
196
197
198
 
# File 'lib/html2doc/mime.rb', line 189

def units_to_px(measure)
  m = /^(\S+)(pt|cm)/.match(measure)
  ret = case m[2]
        when "px" then (m[1].to_f * 0.75)
        when "pt" then m[1].to_f
        when "cm" then (m[1].to_f * 28.346456693)
        when "in" then (m[1].to_f * 72)
        end
  ret.to_i
end

#unwrap_accents(doc) ⇒ Object 



37
38
39
40
41
42
43
44
 
# File 'lib/html2doc/math.rb', line 37

def unwrap_accents(doc)
  doc.xpath("//*[@accent = 'true']").each do |x|
    x.elements.length > 1 or next
    x.elements[1].name == "mrow" and
      x.elements[1].replace(x.elements[1].children)
  end
  doc
end

#warnsvg(src) ⇒ Object 



107
108
109
 
# File 'lib/html2doc/mime.rb', line 107

def warnsvg(src)
  warn "#{src}: SVG not supported" if /\.svg$/i.match?(src)
end