Class: Softcover::Builders::Epub

Inherits:
Softcover::Builder show all
Includes:
EpubUtils, Output
Defined in:
lib/softcover/builders/epub.rb

Constant Summary

Constants included from Utils

Utils::UNITS

Instance Attribute Summary

Attributes inherited from Softcover::Builder

#built_files, #manifest

Instance Method Summary collapse

Methods included from EpubUtils

#content_opf_template, #cover?, #cover_img, #cover_img_path, #escape, #images_dir, #nav_html_template, #toc_ncx_template

Methods included from Output

should_output?, silence!, silent?, #system, unsilence!

Methods inherited from Softcover::Builder

#clean!, #initialize

Methods included from Utils

#add_highlight_class!, #article?, #as_size, #book_file_lines, #chapter_label, #commands, #current_book, #dependency_filename, #digest, #executable, #execute, #filename_or_default, #get_filename, #in_book_directory?, #language_labels, #linux?, #logged_in?, #master_content, #master_filename, #master_latex_header, #mkdir, #non_comment_lines, #os_x?, #path, #raw_lines, #reset_current_book!, #rm, #rm_r, #silence, #source, #template_dir, #tmpify, #unpublish_slug, #write_master_latex_file, #write_pygments_file

Constructor Details

This class inherits a constructor from Softcover::Builder

Instance Method Details

#build!(options = {}) ⇒ Object 



127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
 
# File 'lib/softcover/builders/epub.rb', line 127

def build!(options={})
  @preview = options[:preview]
  Softcover::Builders::Html.new.build!
  if manifest.markdown?
    opts = options.merge({ source: :polytex, origin: :markdown })
    self.manifest = Softcover::BookManifest.new(opts)
  end
  remove_html
  remove_images
  create_directories
  write_mimetype
  write_container_xml
  write_ibooks_xml
  write_toc
  write_nav
  copy_image_files
  write_html(options)
  write_contents
  create_style_files(options)
  make_epub(options)
  move_epub
end

#chapter_name(n) ⇒ Object 



525
526
527
 
# File 'lib/softcover/builders/epub.rb', line 525

def chapter_name(n)
  n == 0 ? language_labels["frontmatter"] : chapter_label(n)
end

#chapter_template(title, content) ⇒ Object

Returns the HTML template for a chapter.



544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
 
# File 'lib/softcover/builders/epub.rb', line 544

def chapter_template(title, content)
  %(<?xml version="1.0" encoding="utf-8"?>
  <!DOCTYPE html>

  <html xmlns="http://www.w3.org/1999/xhtml">
  <head>
    <title>#{title}</title>
    <link rel="stylesheet" href="styles/pygments.css" type="text/css" />
    <link rel="stylesheet" href="styles/softcover.css" type="text/css" />
    <link rel="stylesheet" href="styles/epub.css" type="text/css" />
    <link rel="stylesheet" href="styles/custom.css" type="text/css"/>
    <link rel="stylesheet" href="styles/custom_epub.css" type="text/css"/>
    <link rel="stylesheet" type="application/vnd.adobe-page-template+xml" href="styles/page-template.xpgt" />
  </head>

  <body>
    #{content}
  </body>
  </html>)
end

#chapters ⇒ Object

Returns the chapters to write.



202
203
204
 
# File 'lib/softcover/builders/epub.rb', line 202

def chapters
  preview? ? manifest.preview_chapters : manifest.chapters
end

#clean_book_id(filename) ⇒ Object

Removes the ‘#book’ CSS id. For some reason, EPUB books hate the #book ids in the stylesheet (i.e., such books fail to validate), so remove them.



378
379
380
 
# File 'lib/softcover/builders/epub.rb', line 378

def clean_book_id(filename)
  File.read(filename).gsub(/#book/, '')
end

#container_xml ⇒ Object 



448
449
450
451
452
453
454
455
 
# File 'lib/softcover/builders/epub.rb', line 448

def container_xml
%(<?xml version="1.0"?>
<container version="1.0" xmlns="urn:oasis:names:tc:opendocument:xmlns:container">
    <rootfiles>
  <rootfile full-path="OEBPS/content.opf" media-type="application/oebps-package+xml"/>
   </rootfiles>
</container>)
end

#content_opf ⇒ Object

Returns the content configuration file.



467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
 
# File 'lib/softcover/builders/epub.rb', line 467

def content_opf
  man_ch = chapters.map do |chapter|
             %(<item id="#{chapter.slug}" href="#{chapter.fragment_name}" media-type="application/xhtml+xml"/>)
           end
  toc_ch = chapters.map do |chapter|
             %(<itemref idref="#{chapter.slug}"/>)
           end
  image_files = Dir['epub/OEBPS/images/**/*'].select { |f| File.file?(f) }
  images = image_files.map do |image|
             ext = File.extname(image).sub('.', '')   # e.g., 'png'
             ext = 'jpeg' if ext == 'jpg'
             # Strip off the leading 'epub/OEBPS'.
             sep  = File::SEPARATOR
             href = image.split(sep)[2..-1].join(sep)
             # Define an id based on the filename.
             # Prefix with 'img-' in case the filname starts with an
             # invalid character such as a number.
             label = File.basename(image).gsub('.', '-')
             id = "img-#{label}"
             %(<item id="#{id}" href="#{href}" media-type="image/#{ext}"/>)
           end
  content_opf_template(manifest.title, manifest.copyright,
                       manifest.author, manifest.uuid, cover_id,
                       toc_ch, man_ch, images)
end

#copy_image_files ⇒ Object

Copies the image files from the HTML version of the document.



383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
 
# File 'lib/softcover/builders/epub.rb', line 383

def copy_image_files
  # Copy over all images to guarantee the same directory structure.
  FileUtils.cp_r(File.join('html', 'images'),
                 File.join('epub', 'OEBPS'))
  # Parse the full HTML file with Nokogiri to get images actually used.
  html = File.read(manifest.full_html_file)
  html_image_filenames = Nokogiri::HTML(html).css('img').map do |node|
                           node.attributes['src'].value
                         end
  # Form the corresponding EPUB image paths.
  used_image_filenames = html_image_filenames.map do |filename|
                           "epub/OEBPS/#{filename}"
                         end.to_set
  # Delete unused images.
  Dir.glob("epub/OEBPS/images/**/*").each do |image|
    next if File.directory?(image)
    rm image unless used_image_filenames.include?(image)
  end
end

#cover_id ⇒ Object 



509
510
511
 
# File 'lib/softcover/builders/epub.rb', line 509

def cover_id
  cover? ? "img-#{cover_img.sub('.', '-')}" : nil
end

#cover_page ⇒ Object 



493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
 
# File 'lib/softcover/builders/epub.rb', line 493

def cover_page
%(<?xml version="1.0" encoding="utf-8"?>
<!DOCTYPE html>
<html xmlns="http://www.w3.org/1999/xhtml">
<head>
  <title>Cover</title>
</head>
<body>
  <div id="cover">
     <img width="573" height="800" src="images/#{cover_img}" alt="cover" />
  </div>
</body>
</html>
)
end

#create_directories ⇒ Object 



166
167
168
169
170
171
172
173
 
# File 'lib/softcover/builders/epub.rb', line 166

def create_directories
  mkdir('epub')
  mkdir(path('epub/OEBPS'))
  mkdir(path('epub/OEBPS/styles'))
  mkdir(path('epub/META-INF'))
  mkdir(images_dir)
  mkdir('ebooks')
end

#create_style_files(options) ⇒ Object 



357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
 
# File 'lib/softcover/builders/epub.rb', line 357

def create_style_files(options)
  html_styles = File.join('html', 'stylesheets')
  epub_styles = File.join('epub', 'OEBPS', 'styles')

  FileUtils.cp(File.join(html_styles, 'pygments.css'), epub_styles)
  File.write(File.join(epub_styles, 'softcover.css'),
             clean_book_id(path("#{html_styles}/softcover.css")))

  # Copy over the EPUB-specific CSS.
  template_dir = Softcover::Utils.template_dir(options)
  epub_css     = File.join(template_dir, epub_styles, 'epub.css')
  FileUtils.cp(epub_css, epub_styles)

  # Copy over custom CSS.
  File.write(File.join(epub_styles, 'custom.css'),
             clean_book_id(path("#{html_styles}/custom.css")))
end

#html_with_math(chapter, images_dir, texmath_dir, pngs, options = {}) ⇒ Object

Returns HTML for HTML source that includes math. As a side-effect, html_with_math creates PNGs corresponding to any math in the given source. The technique involves using PhantomJS to hit the HTML source for each page containing math to create SVGs for every math element. Since ereader support for SVGs is spotty, they are then converted to PNGs using Inkscape. The filenames are SHAs of their contents, which arranges both for unique filenames and for automatic disk caching.



260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
 
# File 'lib/softcover/builders/epub.rb', line 260

def html_with_math(chapter, images_dir, texmath_dir, pngs, options={})
  content = File.read(File.join("html", "#{chapter.slug}.html"))
  pagejs = "#{File.dirname(__FILE__)}/utils/page.js"
  url = "file://#{Dir.pwd}/html/#{chapter.slug}.html"
  cmd = "#{phantomjs} #{pagejs} #{url}"
  silence { silence_stream(STDERR) { system cmd } }
  # Sometimes in tests the phantomjs_source.html file is missing.
  # It shouldn't ever happen, but it does no harm to skip it.
  return nil unless File.exist?('phantomjs_source.html')
  raw_source = File.read('phantomjs_source.html')
  source = strip_attributes(Nokogiri::HTML(raw_source))
  rm 'phantomjs_source.html'
  # Remove the first body div, which is the hidden MathJax SVGs.
  if (mathjax_svgs = source.at_css('body div'))
    mathjax_svgs.remove
  else
    # There's not actually any math, so return nil.
    return nil
  end
  # Remove all the unneeded raw TeX displays.
  source.css('script').each(&:remove)
  # Remove all the MathJax preview spans.
  source.css('MathJax_Preview').each(&:remove)

  # Suck out all the SVGs
  svgs   = source.css('div#book svg')
  frames = source.css('span.MathJax_SVG')
  svgs.zip(frames).each do |svg, frame|
    # Save the SVG file.
    svg['viewBox'] = svg['viewbox']
    svg.remove_attribute('viewbox')
    first_child = frame.children.first
    first_child.replace(svg) unless svg == first_child
    output = svg.to_xhtml
    svg_filename = File.join(texmath_dir, "#{digest(output)}.svg")
    File.write(svg_filename, output)
    # Convert to PNG.
    png_filename = svg_filename.sub('.svg', '.png')
    pngs << png_filename
    unless File.exist?(png_filename)
      unless options[:silent] || options[:quiet]
        puts "Creating #{png_filename}"
      end
      svg_height = svg['style'].scan(/height: (.*?);/).flatten.first
      scale_factor = 8   # This scale factor turns out to look good.
      h = scale_factor * svg_height.to_f
      cmd = "#{inkscape} -f #{svg_filename} -e #{png_filename} -h #{h}pt"
      if options[:silent]
        silence { silence_stream(STDERR) { system cmd } }
      else
        silence_stream(STDERR) { system cmd }
      end
    end
    rm svg_filename
    png = Nokogiri::XML::Node.new('img', source)
    png['src'] = File.join('images', 'texmath',
                           File.basename(png_filename))
    png['alt'] = png_filename.sub('.png', '')
    svg.replace(png)
  end
  # Make references relative.
  source.css('a.hyperref').each do |ref_node|
    ref_node['href'] = ref_node['href'].sub('.html', '_fragment.html')
  end
  source.at_css('div#book').children.to_xhtml
end

#ibooks_xml ⇒ Object 



457
458
459
460
461
462
463
464
 
# File 'lib/softcover/builders/epub.rb', line 457

def ibooks_xml
%(<?xml version="1.0" encoding="UTF-8"?>
<display_options>
  <platform name="*">
    <option name="specified-fonts">true</option>
  </platform>
</display_options>)
end

#inkscape ⇒ Object

Returns the Inkscape executable (if available).



333
334
335
 
# File 'lib/softcover/builders/epub.rb', line 333

def inkscape
  @inkscape ||= executable(dependency_filename(:inkscape))
end

#make_epub(options = {}) ⇒ Object

Make the EPUB, which is basically just a zipped HTML file.



404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
 
# File 'lib/softcover/builders/epub.rb', line 404

def make_epub(options={})
  filename = manifest.filename
  zfname = filename + '.zip'
  base_file = "#{zip} -X0 #{zfname} mimetype"
  fullzip = "#{zip} -rDXg9"
  meta_info = "#{fullzip} #{zfname} META-INF -x \*.DS_Store -x mimetype"
  main_info = "#{fullzip} #{zfname} OEBPS    -x \*.DS_Store \*.gitkeep"
  rename = "mv #{zfname} #{filename}.epub"
  commands = [base_file, meta_info, main_info, rename]
  command = commands.join(' && ')
  Dir.chdir('epub') do
    if Softcover.test? || options[:quiet] || options[:silent]
      silence { system(command) }
    else
      system(command)
    end
  end
end

#math?(string) ⇒ Boolean

Returns true if a string appears to have LaTeX math. We detect math via opening math commands: (, [, and beginequation This gives a false positive when math is included in verbatim environments and nowhere else, but it does little harm (requiring only an unnecessary call to page.js).

Returns:

  • (Boolean) 


353
354
355
 
# File 'lib/softcover/builders/epub.rb', line 353

def math?(string)
  !!string.match(/(?:\\\(|\\\[|\\begin{equation})/)
end

#move_epub ⇒ Object

Move the completed EPUB book to the ‘ebooks` directory. Note that we handle the case of a preview book as well.



429
430
431
432
433
434
 
# File 'lib/softcover/builders/epub.rb', line 429

def move_epub
  origin = manifest.filename
  target = preview? ? origin + '-preview' : origin
  FileUtils.mv(File.join('epub',   "#{origin}.epub"),
               File.join('ebooks', "#{target}.epub"))
end

#nav_html ⇒ Object

Returns the nav HTML content.



530
531
532
533
534
535
536
 
# File 'lib/softcover/builders/epub.rb', line 530

def nav_html
  nav_list = manifest.chapters.map do |chapter|
               element = preview? ? chapter.title : nav_link(chapter)
               %(<li>#{element}</li>)
             end
  nav_html_template(manifest.title, nav_list)
end

#nav_link(chapter) ⇒ Object

Returns a navigation link for the chapter.



539
540
541
 
# File 'lib/softcover/builders/epub.rb', line 539

def nav_link(chapter)
  %(<a href="#{chapter.fragment_name}">#{chapter.html_title}</a>)
end

#phantomjs ⇒ Object

Returns the PhantomJS executable (if available).



328
329
330
 
# File 'lib/softcover/builders/epub.rb', line 328

def phantomjs
  @phantomjs ||= executable(dependency_filename(:phantomjs))
end

#preview? ⇒ Boolean

Returns true if generating a book preview.

Returns:

  • (Boolean) 


151
152
153
 
# File 'lib/softcover/builders/epub.rb', line 151

def preview?
  !!@preview
end

#remove_html ⇒ Object

Removes HTML. All the HTML is generated, so this clears out any unused files.



157
158
159
 
# File 'lib/softcover/builders/epub.rb', line 157

def remove_html
  FileUtils.rm(Dir.glob(path('epub/OEBPS/*.html')))
end

#remove_images ⇒ Object

Removes images in case they are stale.



162
163
164
 
# File 'lib/softcover/builders/epub.rb', line 162

def remove_images
  rm_r images_dir
end

#strip_attributes(doc) ⇒ Object

Strip attributes that are invalid in EPUB documents.



338
339
340
341
342
343
344
345
346
 
# File 'lib/softcover/builders/epub.rb', line 338

def strip_attributes(doc)
  attrs = %w[data-tralics-id data-label data-number data-chapter
             role aria-readonly target]
  doc.tap do
    attrs.each do |attr|
      doc.xpath("//@#{attr}").remove
    end
  end
end

#toc_ncx ⇒ Object

Returns the Table of Contents for the spine.



514
515
516
517
518
519
520
521
522
523
 
# File 'lib/softcover/builders/epub.rb', line 514

def toc_ncx
  chapter_nav = []
  chapters.each_with_index do |chapter, n|
    chapter_nav << %(<navPoint id="#{chapter.slug}" playOrder="#{n+1}">)
    chapter_nav << %(    <navLabel><text>#{chapter_name(n)}</text></navLabel>)
    chapter_nav << %(    <content src="#{chapter.fragment_name}"/>)
    chapter_nav << %(</navPoint>)
  end
  toc_ncx_template(manifest.title, manifest.uuid, chapter_nav)
end

#write_container_xml ⇒ Object

Writes the container XML file. This is required by the EPUB standard.



183
184
185
 
# File 'lib/softcover/builders/epub.rb', line 183

def write_container_xml
  File.write(path('epub/META-INF/container.xml'), container_xml)
end

#write_contents ⇒ Object

Writes the content.opf file. This is required by the EPUB standard.



197
198
199
 
# File 'lib/softcover/builders/epub.rb', line 197

def write_contents
  File.write(path('epub/OEBPS/content.opf'), content_opf)
end

#write_html(options = {}) ⇒ Object

Writes the HTML for the EPUB. Included is a math detector that processes the page with MathJax (via page.js) so that math can be included in EPUB (and thence MOBI).



209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
 
# File 'lib/softcover/builders/epub.rb', line 209

def write_html(options={})
  texmath_dir = File.join(images_dir, 'texmath')
  mkdir images_dir
  mkdir texmath_dir

  File.write(path('epub/OEBPS/cover.html'), cover_page) if cover?

  pngs = []
  chapters.each_with_index do |chapter, i|
    target_filename = path("epub/OEBPS/#{chapter.fragment_name}")
    File.open(target_filename, 'w') do |f|
      content = File.read(path("html/#{chapter.fragment_name}"))
      doc = strip_attributes(Nokogiri::HTML(content))
      body = doc.at_css('body')
      if body.nil?
        $stderr.puts "\nError: Document not built due to empty chapter"
        $stderr.puts "Chapters must include a title using the Markdown"
        $stderr.puts "    # This is a chapter"
        $stderr.puts "or the LaTeX"
        $stderr.puts "    \\chapter{This is a chapter}"
        exit(1)
      end
      inner_html = body.children.to_xhtml
      if math?(inner_html)
        html = html_with_math(chapter, images_dir, texmath_dir, pngs,
                              options)
        html ||= inner_html # handle case of spurious math detection
      else
        html = inner_html
      end
      f.write(chapter_template("Chapter #{i}", html))
    end
  end
  # Clean up unused PNGs.
  png_files = Dir[path("#{texmath_dir}/*.png")]
  (png_files - pngs).each do |f|
    if File.exist?(f)
      puts "Removing unused PNG #{f}" unless options[:silent]
      FileUtils.rm(f)
    end
  end
end

#write_ibooks_xml ⇒ Object

Writes iBooks-specific XML. This allows proper display of monospace fonts in code samples, among other things.



190
191
192
193
 
# File 'lib/softcover/builders/epub.rb', line 190

def write_ibooks_xml
  xml_filename = 'com.apple.ibooks.display-options.xml'
  File.write(path("epub/META-INF/#{xml_filename}"), ibooks_xml)
end

#write_mimetype ⇒ Object

Writes the mimetype file. This is required by the EPUB standard.



177
178
179
 
# File 'lib/softcover/builders/epub.rb', line 177

def write_mimetype
  File.write(path('epub/mimetype'), 'application/epub+zip')
end

#write_nav ⇒ Object

Writes the navigation file. This is required by the EPUB standard.



444
445
446
 
# File 'lib/softcover/builders/epub.rb', line 444

def write_nav
  File.write('epub/OEBPS/nav.html', nav_html)
end

#write_toc ⇒ Object

Writes the Table of Contents. This is required by the EPUB standard.



438
439
440
 
# File 'lib/softcover/builders/epub.rb', line 438

def write_toc
  File.write('epub/OEBPS/toc.ncx', toc_ncx)
end

#zip ⇒ Object 



423
424
425
 
# File 'lib/softcover/builders/epub.rb', line 423

def zip
  @zip ||= executable(dependency_filename(:zip))
end