Module: Jekyll::PandocExports

Defined in:
lib/jekyll-pandoc-exports/hooks.rb,
lib/jekyll-pandoc-exports/command.rb,
lib/jekyll-pandoc-exports/version.rb,
lib/jekyll-pandoc-exports/generator.rb,
lib/jekyll-pandoc-exports/statistics.rb

Defined Under Namespace

Classes: Command, ExportRunner, Hooks, Statistics

Constant Summary collapse

VERSION = 
'0.2.0'

Class Method Summary collapse

Class Method Details

.apply_template(html_content, config) ⇒ Object 



195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
 
# File 'lib/jekyll-pandoc-exports/generator.rb', line 195

def self.apply_template(html_content, config)
  template = config['template'] || {}
  header = template['header'].to_s
  footer = template['footer'].to_s
  css = template['css'].to_s
  return html_content if header.empty? && footer.empty? && css.empty?
  
  # Add custom CSS
  unless css.empty?
    css_tag = "<style>#{css}</style>"
    html_content = html_content.sub(/<\/head>/, "#{css_tag}\n</head>")
  end
  
  # Add header after body tag
  unless header.empty?
    html_content = html_content.sub(/<body[^>]*>/, "\&\n#{header}")
  end
  
  # Add footer before closing body tag
  unless footer.empty?
    html_content = html_content.sub(/<\/body>/, "#{footer}\n</body>")
  end
  
  html_content
end

.build_download_html(generated_files, config) ⇒ Object 



309
310
311
312
313
314
315
316
317
318
319
 
# File 'lib/jekyll-pandoc-exports/generator.rb', line 309

def self.build_download_html(generated_files, config)
  download_html = "<div class=\"#{config['download_class']}\" style=\"#{config['download_style']}\">" +
                 "<p><strong>Download Options:</strong></p>" +
                 "<ul style=\"margin: 5px 0; padding-left: 20px;\">"
  
  generated_files.each do |file|
    download_html += "<li><a href=\"#{file[:url]}\" style=\"color: #007bff; text-decoration: none; font-weight: bold;\">#{file[:type]}</a></li>"
  end
  
  download_html += "</ul></div>"
end

.build_download_url(site, config, filename, extension) ⇒ Object 



329
330
331
332
333
334
335
336
 
# File 'lib/jekyll-pandoc-exports/generator.rb', line 329

def self.build_download_url(site, config, filename, extension)
  output_dir = config['output_dir'] || ''
  if output_dir.empty?
    "#{site.baseurl}/#{filename}.#{extension}"
  else
    "#{site.baseurl}/#{output_dir}/#{filename}.#{extension}"
  end
end

.clean_unicode_characters(html) ⇒ Object 



290
291
292
293
 
# File 'lib/jekyll-pandoc-exports/generator.rb', line 290

def self.clean_unicode_characters(html)
  # Remove emoji and symbol ranges that cause LaTeX issues
  html.gsub(/[\u{1F000}-\u{1F9FF}]|[\u{2600}-\u{26FF}]|[\u{2700}-\u{27BF}]/, '')
end

.generate_docx(html_content, filename, output_dir, site, generated_files, config = {}) ⇒ Object 



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
 
# File 'lib/jekyll-pandoc-exports/generator.rb', line 221

def self.generate_docx(html_content, filename, output_dir, site, generated_files, config = {})
  return unless validate_content_size(html_content, config)
  
  begin
    # Run pre-conversion hooks
    processed_html = Hooks.run_pre_conversion_hooks(html_content, config, { format: :docx, filename: filename })
    
    docx_content = PandocRuby.convert(processed_html, from: :html, to: :docx)
    
    # Run post-conversion hooks
    docx_content = Hooks.run_post_conversion_hooks(docx_content, :docx, config, { filename: filename })
    docx_file = File.join(output_dir, "#{filename}.docx")
    
    File.open(docx_file, 'wb') { |file| file.write(docx_content) }
    
    generated_files << { 
      type: 'Word Document (.docx)', 
      url: build_download_url(site, config, filename, 'docx')
    }
    @stats&.record_conversion_success(:docx)
    log_message(config, "Generated #{filename}.docx")
  rescue => e
    @stats&.record_conversion_failure(:docx, e)
    log_error(config, "Failed to generate #{filename}.docx: #{e.message}")
  end
end

.generate_pdf(html_content, filename, output_dir, site, generated_files, page, config) ⇒ Object 



248
249
250
251
252
253
254
255
256
257
258
259
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
 
# File 'lib/jekyll-pandoc-exports/generator.rb', line 248

def self.generate_pdf(html_content, filename, output_dir, site, generated_files, page, config)
  return unless validate_content_size(html_content, config)
  
  begin
    # Run pre-conversion hooks
    processed_html = Hooks.run_pre_conversion_hooks(html_content, config, { format: :pdf, filename: filename })
    
    pdf_html = processed_html.dup
    
    # Apply Unicode cleanup if enabled
    if config['unicode_cleanup']
      pdf_html = clean_unicode_characters(pdf_html)
    end
    
    # Apply title cleanup patterns from config
    # (Note: also applied in process_html_content for both formats)
    
    # Get PDF options from config or page front matter
    pdf_options = page.data['pdf_options'] || config['pdf_options']
    
    # Merge custom pandoc options
    all_options = pdf_options.merge(config['pandoc_options'])
    pdf_content = PandocRuby.new(pdf_html, from: :html, to: :pdf).convert(all_options)
    
    # Run post-conversion hooks
    pdf_content = Hooks.run_post_conversion_hooks(pdf_content, :pdf, config, { filename: filename })
    pdf_file = File.join(output_dir, "#{filename}.pdf")
    
    File.open(pdf_file, 'wb') { |file| file.write(pdf_content) }
    
    generated_files << { 
      type: 'PDF Document (.pdf)', 
      url: build_download_url(site, config, filename, 'pdf')
    }
    @stats&.record_conversion_success(:pdf)
    log_message(config, "Generated #{filename}.pdf")
  rescue => e
    @stats&.record_conversion_failure(:pdf, e)
    log_error(config, "Failed to generate #{filename}.pdf: #{e.message}")
  end
end

.get_html_file_path(site, page) ⇒ Object 



167
168
169
170
171
172
173
174
 
# File 'lib/jekyll-pandoc-exports/generator.rb', line 167

def self.get_html_file_path(site, page)
  # Handle different Jekyll URL structures
  if page.url.end_with?('/')
    File.join(site.dest, page.url, 'index.html')
  else
    File.join(site.dest, "#{page.url.gsub('/', '')}.html")
  end
end

.get_output_directory(site, config) ⇒ Object 



134
135
136
137
138
139
140
141
142
 
# File 'lib/jekyll-pandoc-exports/generator.rb', line 134

def self.get_output_directory(site, config)
  if config['output_dir'].empty?
    site.dest
  else
    output_path = File.join(site.dest, config['output_dir'])
    FileUtils.mkdir_p(output_path) unless Dir.exist?(output_path)
    output_path
  end
end

.get_output_filename(item) ⇒ Object 



121
122
123
124
125
126
127
128
129
130
131
132
 
# File 'lib/jekyll-pandoc-exports/generator.rb', line 121

def self.get_output_filename(item)
  # Support custom export filenames via front matter
  if item.respond_to?(:data) && item.data['export_filename']
    return item.data['export_filename']
  end

  if item.respond_to?(:basename)
    File.basename(item.basename, '.md')
  else
    File.basename(item.path, '.md')
  end
end

.inject_download_links(html_content, generated_files, html_file, config) ⇒ Object 



295
296
297
298
299
300
301
302
303
304
305
306
307
 
# File 'lib/jekyll-pandoc-exports/generator.rb', line 295

def self.inject_download_links(html_content, generated_files, html_file, config)
  download_html = build_download_html(generated_files, config)
  
  # Insert after first heading or at beginning of body
  if html_content.match(/<h[1-6][^>]*>/)
    html_content.sub!(/<\/h[1-6]>/, "\\&\n#{download_html}")
  else
    html_content.sub!(/<body[^>]*>/, "\\&\n#{download_html}")
  end
  
  File.write(html_file, html_content)
  html_content
end

.log_error(config, message) ⇒ Object 



338
339
340
 
# File 'lib/jekyll-pandoc-exports/generator.rb', line 338

def self.log_error(config, message)
  Jekyll.logger.error "Pandoc Exports:", message
end

.log_message(config, message) ⇒ Object 



321
322
323
324
325
326
327
 
# File 'lib/jekyll-pandoc-exports/generator.rb', line 321

def self.log_message(config, message)
  if config['debug']
    Jekyll.logger.info "Pandoc Exports [DEBUG]:", message
  else
    Jekyll.logger.info "Pandoc Exports:", message
  end
end

.process_collections(site, config) ⇒ Object 



63
64
65
66
67
68
69
70
71
72
73
74
75
 
# File 'lib/jekyll-pandoc-exports/generator.rb', line 63

def self.process_collections(site, config)
  config['collections'].each do |collection_name|
    case collection_name
    when 'pages'
      site.pages.each { |item| process_item(site, item, config) }
    when 'posts'
      site.posts.docs.each { |item| process_item(site, item, config) }
    else
      collection = site.collections[collection_name]
      collection&.docs&.each { |item| process_item(site, item, config) }
    end
  end
end

.process_html_content(html_content, site, config) ⇒ Object 



176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
 
# File 'lib/jekyll-pandoc-exports/generator.rb', line 176

def self.process_html_content(html_content, site, config)
  processed = html_content.dup
  
  # Apply template customizations
  processed = apply_template(processed, config)
  
  # Apply HTML cleanup patterns (removes unwanted elements before conversion)
  (config['title_cleanup'] || []).each do |pattern|
    processed.gsub!(Regexp.new(pattern, Regexp::MULTILINE), '')
  end
  
  # Apply image path fixes from config
  config['image_path_fixes'].each do |fix|
    processed.gsub!(Regexp.new(fix['pattern']), fix['replacement'].gsub('{{site.dest}}', site.dest))
  end
  
  processed
end

.process_item(site, item, config) ⇒ Object 



77
78
79
80
81
82
83
84
 
# File 'lib/jekyll-pandoc-exports/generator.rb', line 77

def self.process_item(site, item, config)
  return unless item.data['docx'] || item.data['pdf']
  
  # Check if file was modified (incremental build)
  return if skip_unchanged_file?(site, item, config)
  
  process_page(site, item, config)
end

.process_page(site, page, config) ⇒ Object 



144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
 
# File 'lib/jekyll-pandoc-exports/generator.rb', line 144

def self.process_page(site, page, config)
  @stats&.record_processing_start
  @stats&.record_file_processed
  
  html_file = get_html_file_path(site, page)
  return unless File.exist?(html_file)
  
  html_content = File.read(html_file)
  processed_html = process_html_content(html_content, site, config)
  filename = get_output_filename(page)
  output_dir = get_output_directory(site, config)
  generated_files = []
  
  generate_docx(processed_html, filename, output_dir, site, generated_files, config) if page.data['docx']
  generate_pdf(processed_html, filename, output_dir, site, generated_files, page, config) if page.data['pdf']
  
  if config['inject_downloads'] && generated_files.any?
    inject_download_links(html_content, generated_files, html_file, config)
  end
  
  @stats&.record_processing_end
end

.setup_configuration(site) ⇒ Object 



22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
 
# File 'lib/jekyll-pandoc-exports/generator.rb', line 22

def self.setup_configuration(site)
  config = site.config['pandoc_exports'] || {}
  {
    'enabled' => true,
    'output_dir' => '',
    'collections' => ['pages', 'posts'],
    'pdf_options' => { 'variable' => 'geometry:margin=1in' },
    'unicode_cleanup' => true,
    'inject_downloads' => true,
    'download_class' => 'pandoc-downloads no-print',
    'download_style' => 'margin: 20px 0; padding: 15px; background-color: #f8f9fa; border: 1px solid #dee2e6; border-radius: 5px;',
    'title_cleanup' => [],
    'image_path_fixes' => [],
    'debug' => false,
    'max_file_size' => 10_000_000,
    'strict_size_limit' => false,
    'performance_monitoring' => false,
    'template' => {
      'header' => '',
      'footer' => '',
      'css' => ''
    },
    'pandoc_options' => {}
  }.merge(config)
end

.skip_unchanged_file?(site, item, config) ⇒ Boolean

Returns:

  • (Boolean) 


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
 
# File 'lib/jekyll-pandoc-exports/generator.rb', line 86

def self.skip_unchanged_file?(site, item, config)
  return false unless config['incremental']
  
  source_file = item.respond_to?(:path) ? item.path : item.relative_path
  return false unless File.exist?(source_file)
  
  filename = get_output_filename(item)
  output_dir = get_output_directory(site, config)
  
  docx_file = File.join(output_dir, "#{filename}.docx")
  pdf_file = File.join(output_dir, "#{filename}.pdf")
  
  source_mtime = File.mtime(source_file)
  
  # Check if any required output files are missing
  if item.data['docx'] && !File.exist?(docx_file)
    return false
  end
  
  if item.data['pdf'] && !File.exist?(pdf_file)
    return false
  end
  
  # Check if any existing output files are older than source
  if item.data['docx'] && File.exist?(docx_file)
    return false if File.mtime(docx_file) < source_mtime
  end
  
  if item.data['pdf'] && File.exist?(pdf_file)
    return false if File.mtime(pdf_file) < source_mtime
  end
  
  true
end

.validate_content_size(html_content, config) ⇒ Object 



342
343
344
345
346
347
348
349
350
351
 
# File 'lib/jekyll-pandoc-exports/generator.rb', line 342

def self.validate_content_size(html_content, config)
  max_size = config['max_file_size'] || 10_000_000 # 10MB default
  
  if html_content.bytesize > max_size
    Jekyll.logger.warn "Pandoc Exports:", "Content size (#{html_content.bytesize} bytes) exceeds recommended limit (#{max_size} bytes)"
    return false if config['strict_size_limit']
  end
  
  true
end

.validate_dependencies ⇒ Object 



48
49
50
51
52
53
54
55
56
57
58
59
60
61
 
# File 'lib/jekyll-pandoc-exports/generator.rb', line 48

def self.validate_dependencies
  pandoc_available = system('pandoc --version > /dev/null 2>&1')
  latex_available = system('pdflatex --version > /dev/null 2>&1')
  
  unless pandoc_available
    Jekyll.logger.warn "Pandoc Exports:", "Pandoc not found. Install with: brew install pandoc (macOS) or apt-get install pandoc (Ubuntu)"
  end
  
  unless latex_available
    Jekyll.logger.warn "Pandoc Exports:", "LaTeX not found. Install with: brew install --cask mactex (macOS) or apt-get install texlive-latex-base (Ubuntu)"
  end
  
  pandoc_available
end