Class: Kramdown::Parser::Html::ElementConverter

Inherits:
Object
  • Object
show all
Includes:
Constants, Utils::Entities
Defined in:
lib/kramdown/parser/html.rb

Overview

Converts HTML elements to native elements if possible.

Constant Summary collapse

REMOVE_TEXT_CHILDREN = 
%w{html head hgroup ol ul dl table colgroup tbody thead tfoot tr select optgroup}
WRAP_TEXT_CHILDREN = 
%w{body section nav article aside header footer address div li dd blockquote figure
figcaption fieldset form}
REMOVE_WHITESPACE_CHILDREN = 
%w{body section nav article aside header footer address
div li dd blockquote figure figcaption td th fieldset form}
STRIP_WHITESPACE = 
%w{address article aside blockquote body caption dd div dl dt fieldset figcaption form footer
header h1 h2 h3 h4 h5 h6 legend li nav p section td th}
SIMPLE_ELEMENTS = 
%w{em strong blockquote hr br img p thead tbody tfoot tr td th ul ol dl li dl dt dd}
EMPHASIS_TYPE_MAP = 
{'em' => :em, 'i' => :em, 'strong' => :strong, 'b' => :strong}

Constants included from Utils::Entities

Utils::Entities::ENTITY_MAP, Utils::Entities::ENTITY_TABLE

Constants included from Constants

Constants::HTML_ATTRIBUTE_RE, Constants::HTML_BLOCK_ELEMENTS, Constants::HTML_COMMENT_RE, Constants::HTML_CONTENT_MODEL, Constants::HTML_CONTENT_MODEL_BLOCK, Constants::HTML_CONTENT_MODEL_RAW, Constants::HTML_CONTENT_MODEL_SPAN, Constants::HTML_DOCTYPE_RE, Constants::HTML_ELEMENTS_WITHOUT_BODY, Constants::HTML_ENTITY_RE, Constants::HTML_INSTRUCTION_RE, Constants::HTML_SPAN_ELEMENTS, Constants::HTML_TAG_CLOSE_RE, Constants::HTML_TAG_RE

Class Method Summary collapse

Instance Method Summary collapse

Methods included from Utils::Entities

entity

Constructor Details

#initialize(root) ⇒ ElementConverter

Returns a new instance of ElementConverter.



187
188
189
 
# File 'lib/kramdown/parser/html.rb', line 187

def initialize(root)
  @root = root
end

Class Method Details

.convert(root, el = root) ⇒ Object 



191
192
193
 
# File 'lib/kramdown/parser/html.rb', line 191

def self.convert(root, el = root)
  new(root).process(el)
end

Instance Method Details

#convert_a(el) ⇒ Object 



344
345
346
347
348
349
350
351
 
# File 'lib/kramdown/parser/html.rb', line 344

def convert_a(el)
  if el.attr['href']
    set_basics(el, :a)
    process_children(el)
  else
    process_html_element(el, false)
  end
end

#convert_code(el) ⇒ Object Also known as: convert_pre 



377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
 
# File 'lib/kramdown/parser/html.rb', line 377

def convert_code(el)
  raw = ''
  extract_text(el, raw)
  result = process_text(raw, true)
  begin
    str = result.inject('') do |mem, c|
      if c.type == :text
        mem << c.value
      elsif c.type == :entity
        if RUBY_VERSION >= '1.9'
          mem << c.value.char.encode(@root.options[:encoding])
        elsif [60, 62, 34, 38].include?(c.value.code_point)
          mem << c.value.code_point.chr
        end
      elsif c.type == :smart_quote || c.type == :typographic_sym
        mem << entity(c.value.to_s).char.encode(@root.options[:encoding])
      else
        raise "Bug - please report"
      end
    end
    result.clear
    result << Element.new(:text, str)
  rescue
  end
  if result.length > 1 || result.first.type != :text
    process_html_element(el, false, true)
  else
    if el.value == 'code'
      set_basics(el, :codespan)
    else
      set_basics(el, :codeblock)
    end
    el.value = result.first.value
    el.children.clear
  end
end

#convert_em(el) ⇒ Object 



354
355
356
357
358
359
360
361
362
363
 
# File 'lib/kramdown/parser/html.rb', line 354

def convert_em(el)
  text = ''
  extract_text(el, text)
  if text =~ /\A\s/ || text =~ /\s\z/
    process_html_element(el, false)
  else
    set_basics(el, EMPHASIS_TYPE_MAP[el.value])
    process_children(el)
  end
end

#convert_h1(el) ⇒ Object 



368
369
370
371
372
 
# File 'lib/kramdown/parser/html.rb', line 368

def convert_h1(el)
  set_basics(el, :header, :level => el.value[1..1].to_i)
  extract_text(el, el.options[:raw_text] = '')
  process_children(el)
end

#convert_script(el) ⇒ Object 



518
519
520
521
522
523
524
 
# File 'lib/kramdown/parser/html.rb', line 518

def convert_script(el)
  if !is_math_tag?(el)
    process_html_element(el)
  else
    handle_math_tag(el)
  end
end

#convert_table(el) ⇒ Object 



415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
 
# File 'lib/kramdown/parser/html.rb', line 415

def convert_table(el)
  if !is_simple_table?(el)
    process_html_element(el, false)
    return
  end
  remove_text_children(el)
  process_children(el)
  set_basics(el, :table)

  calc_alignment = lambda do |c|
    if c.type == :tr
      el.options[:alignment] = c.children.map do |td|
        if td.attr['style']
          td.attr['style'].slice!(/(?:;\s*)?text-align:\s+(center|left|right)/)
          td.attr.delete('style') if td.attr['style'].strip.empty?
          $1.to_sym
        else
          :default
        end
      end
    else
      c.children.each {|cc| calc_alignment.call(cc)}
    end
  end
  calc_alignment.call(el)
  el.children.delete_if {|c| c.type == :html_element}

  change_th_type = lambda do |c|
    if c.type == :th
      c.type = :td
    else
      c.children.each {|cc| change_th_type.call(cc)}
    end
  end
  change_th_type.call(el)

  if el.children.first.type == :tr
    tbody = Element.new(:tbody)
    tbody.children = el.children
    el.children = [tbody]
  end
end

#extract_text(el, raw) ⇒ Object 



339
340
341
342
 
# File 'lib/kramdown/parser/html.rb', line 339

def extract_text(el, raw)
  raw << el.value.to_s if el.type == :text
  el.children.each {|c| extract_text(c, raw)}
end

#handle_math_tag(el) ⇒ Object 



530
531
532
533
534
 
# File 'lib/kramdown/parser/html.rb', line 530

def handle_math_tag(el)
  set_basics(el, :math, :category => (el.attr['type'] =~ /mode=display/ ? :block : :span))
  el.value = el.children.shift.value
  el.attr.delete('type')
end

#is_math_tag?(el) ⇒ Boolean

Returns:

  • (Boolean) 


526
527
528
 
# File 'lib/kramdown/parser/html.rb', line 526

def is_math_tag?(el)
  el.attr['type'].to_s =~ /\bmath\/tex\b/
end

#is_simple_table?(el) ⇒ Boolean

Returns:

  • (Boolean) 


458
459
460
461
462
463
464
465
466
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
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
 
# File 'lib/kramdown/parser/html.rb', line 458

def is_simple_table?(el)
  only_phrasing_content = lambda do |c|
    c.children.all? do |cc|
      (cc.type == :text || !HTML_BLOCK_ELEMENTS.include?(cc.value)) && only_phrasing_content.call(cc)
    end
  end
  check_cells = Proc.new do |c|
    if c.value == 'th' || c.value == 'td'
      return false if !only_phrasing_content.call(c)
    else
      c.children.each {|cc| check_cells.call(cc)}
    end
  end
  check_cells.call(el)

  nr_cells = 0
  check_nr_cells = lambda do |t|
    if t.value == 'tr'
      count = t.children.select {|cc| cc.value == 'th' || cc.value == 'td'}.length
      if count != nr_cells
        if nr_cells == 0
          nr_cells = count
        else
          nr_cells = -1
          break
        end
      end
    else
      t.children.each {|cc| check_nr_cells.call(cc)}
    end
  end
  check_nr_cells.call(el)
  return false if nr_cells == -1

  alignment = nil
  check_alignment = Proc.new do |t|
    if t.value == 'tr'
      cur_alignment = t.children.select {|cc| cc.value == 'th' || cc.value == 'td'}.map do |cell|
        md = /text-align:\s+(center|left|right|justify|inherit)/.match(cell.attr['style'].to_s)
        return false if md && (md[1] == 'justify' || md[1] == 'inherit')
        md.nil? ? :default : md[1]
      end
      alignment = cur_alignment if alignment.nil?
      return false if alignment != cur_alignment
    else
      t.children.each {|cc| check_alignment.call(cc)}
    end
  end
  check_alignment.call(el)

  check_rows = lambda do |t, type|
    t.children.all? {|r| (r.value == 'tr' || r.type == :text) && r.children.all? {|c| c.value == type || c.type == :text}}
  end
  check_rows.call(el, 'td') ||
    (el.children.all? do |t|
       t.type == :text || (t.value == 'thead' && check_rows.call(t, 'th')) ||
         ((t.value == 'tfoot' || t.value == 'tbody') && check_rows.call(t, 'td'))
     end && el.children.any? {|t| t.value == 'tbody'})
end

#process(el, do_conversion = true, preserve_text = false, parent = nil) ⇒ Object

Convert the element el and its children.



196
197
198
199
200
201
202
203
204
205
206
207
208
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
 
# File 'lib/kramdown/parser/html.rb', line 196

def process(el, do_conversion = true, preserve_text = false, parent = nil)
  case el.type
  when :xml_comment, :xml_pi
    ptype = if parent.nil?
              'div'
            else
              case parent.type
              when :html_element then parent.value
              when :code_span then 'code'
              when :code_block then 'pre'
              when :header then 'h1'
              else parent.type.to_s
              end
            end
    el.options.replace({:category => (HTML_CONTENT_MODEL[ptype] == :span ? :span : :block)})
    return
  when :html_element
  when :root
    el.children.each {|c| process(c)}
    remove_whitespace_children(el)
    return
  else return
  end

  mname = "convert_#{el.value}"
  if do_conversion && self.class.method_defined?(mname)
    send(mname, el)
  else
    type = el.value
    remove_text_children(el) if do_conversion && REMOVE_TEXT_CHILDREN.include?(type)

    if do_conversion && SIMPLE_ELEMENTS.include?(type)
      set_basics(el, type.intern)
      process_children(el, do_conversion, preserve_text)
    else
      process_html_element(el, do_conversion, preserve_text)
    end

    if do_conversion
      strip_whitespace(el) if STRIP_WHITESPACE.include?(type)
      remove_whitespace_children(el) if REMOVE_WHITESPACE_CHILDREN.include?(type)
      wrap_text_children(el) if WRAP_TEXT_CHILDREN.include?(type)
    end
  end
end

#process_children(el, do_conversion = true, preserve_text = false) ⇒ Object 



242
243
244
245
246
247
248
249
250
251
 
# File 'lib/kramdown/parser/html.rb', line 242

def process_children(el, do_conversion = true, preserve_text = false)
  el.children.map! do |c|
    if c.type == :text
      process_text(c.value, preserve_text || !do_conversion)
    else
      process(c, do_conversion, preserve_text, el)
      c
    end
  end.flatten!
end

#process_html_element(el, do_conversion = true, preserve_text = false) ⇒ Object 



284
285
286
287
288
 
# File 'lib/kramdown/parser/html.rb', line 284

def process_html_element(el, do_conversion = true, preserve_text = false)
  el.options.replace(:category => HTML_SPAN_ELEMENTS.include?(el.value) ? :span : :block,
                     :content_model => (do_conversion ? HTML_CONTENT_MODEL[el.value] : :raw))
  process_children(el, do_conversion, preserve_text)
end

#process_text(raw, preserve = false) ⇒ Object

Process the HTML text raw: compress whitespace (if preserve is false) and convert entities in entity elements.



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
 
# File 'lib/kramdown/parser/html.rb', line 255

def process_text(raw, preserve = false)
  raw.gsub!(/\s+/, ' ') unless preserve
  src = StringScanner.new(raw)
  result = []
  while !src.eos?
    if tmp = src.scan_until(/(?=#{HTML_ENTITY_RE})/)
      result << Element.new(:text, tmp)
      src.scan(HTML_ENTITY_RE)
      val = src[1] || (src[2] && src[2].to_i) || src[3].hex
      result << if %w{lsquo rsquo ldquo rdquo}.include?(val)
                  Element.new(:smart_quote, val.intern)
                elsif %w{mdash ndash hellip laquo raquo}.include?(val)
                  Element.new(:typographic_sym, val.intern)
                else
                  begin
                    Element.new(:entity, entity(val), nil, :original => src.matched)
                  rescue ::Kramdown::Error
                    src.pos -= src.matched_size - 1
                    Element.new(:entity, ::Kramdown::Utils::Entities.entity('amp'))
                  end
                end
    else
      result << Element.new(:text, src.rest)
      src.terminate
    end
  end
  result
end

#remove_text_children(el) ⇒ Object 



290
291
292
 
# File 'lib/kramdown/parser/html.rb', line 290

def remove_text_children(el)
  el.children.delete_if {|c| c.type == :text}
end

#remove_whitespace_children(el) ⇒ Object 



323
324
325
326
327
328
329
330
331
 
# File 'lib/kramdown/parser/html.rb', line 323

def remove_whitespace_children(el)
  i = -1
  el.children = el.children.reject do |c|
    i += 1
    c.type == :text && c.value.strip.empty? &&
      (i == 0 || i == el.children.length - 1 || (Element.category(el.children[i-1]) == :block &&
                                                 Element.category(el.children[i+1]) == :block))
  end
end

#set_basics(el, type, opts = {}) ⇒ Object 



333
334
335
336
337
 
# File 'lib/kramdown/parser/html.rb', line 333

def set_basics(el, type, opts = {})
  el.type = type
  el.options.replace(opts)
  el.value = nil
end

#strip_whitespace(el) ⇒ Object 



313
314
315
316
317
318
319
320
321
 
# File 'lib/kramdown/parser/html.rb', line 313

def strip_whitespace(el)
  return if el.children.empty?
  if el.children.first.type == :text
    el.children.first.value.lstrip!
  end
  if el.children.last.type == :text
    el.children.last.value.rstrip!
  end
end

#wrap_text_children(el) ⇒ Object 



294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
 
# File 'lib/kramdown/parser/html.rb', line 294

def wrap_text_children(el)
  tmp = []
  last_is_p = false
  el.children.each do |c|
    if Element.category(c) != :block || c.type == :text
      if !last_is_p
        tmp << Element.new(:p, nil, nil, :transparent => true)
        last_is_p = true
      end
      tmp.last.children << c
      tmp
    else
      tmp << c
      last_is_p = false
    end
  end
  el.children = tmp
end