Class: HexaPDF::Type::Page

Inherits:
Dictionary show all
Defined in:
lib/hexapdf/type/page.rb

Overview

Represents a page of a PDF document.

A page object contains the meta information for a page. Most of the fields are independent from the page’s content like the /Dur field. However, some of them (like /Resources or /UserUnit) influence how or if the page’s content can be rendered correctly.

A number of field values can also be inherited: /Resources, /MediaBox, /CropBox, /Rotate. Field inheritance means that if a field is not set on the page object itself, the value is taken from the nearest page tree ancestor that has this value set.

See: PDF2.0 s7.7.3.3, s7.7.3.4, Pages

Constant Summary collapse

PAPER_SIZE =

The predefined paper sizes in points (1/72 inch):

  • ISO sizes: A0x4, A0x2, A0-A10, B0-B10, C0-C10

  • Letter, Legal, Ledger, Tabloid, Executive

{
  A0x4: [0, 0, 4768, 6741].freeze,
  A0x2: [0, 0, 3370, 4768].freeze,
  A0: [0, 0, 2384, 3370].freeze,
  A1: [0, 0, 1684, 2384].freeze,
  A2: [0, 0, 1191, 1684].freeze,
  A3: [0, 0, 842, 1191].freeze,
  A4: [0, 0, 595, 842].freeze,
  A5: [0, 0, 420, 595].freeze,
  A6: [0, 0, 298, 420].freeze,
  A7: [0, 0, 210, 298].freeze,
  A8: [0, 0, 147, 210].freeze,
  A9: [0, 0, 105, 147].freeze,
  A10: [0, 0, 74, 105].freeze,
  B0: [0, 0, 2835, 4008].freeze,
  B1: [0, 0, 2004, 2835].freeze,
  B2: [0, 0, 1417, 2004].freeze,
  B3: [0, 0, 1001, 1417].freeze,
  B4: [0, 0, 709, 1001].freeze,
  B5: [0, 0, 499, 709].freeze,
  B6: [0, 0, 354, 499].freeze,
  B7: [0, 0, 249, 354].freeze,
  B8: [0, 0, 176, 249].freeze,
  B9: [0, 0, 125, 176].freeze,
  B10: [0, 0, 88, 125].freeze,
  C0: [0, 0, 2599, 3677].freeze,
  C1: [0, 0, 1837, 2599].freeze,
  C2: [0, 0, 1298, 1837].freeze,
  C3: [0, 0, 918, 1298].freeze,
  C4: [0, 0, 649, 918].freeze,
  C5: [0, 0, 459, 649].freeze,
  C6: [0, 0, 323, 459].freeze,
  C7: [0, 0, 230, 323].freeze,
  C8: [0, 0, 162, 230].freeze,
  C9: [0, 0, 113, 162].freeze,
  C10: [0, 0, 79, 113].freeze,
  Letter: [0, 0, 612, 792].freeze,
  Legal: [0, 0, 612, 1008].freeze,
  Ledger: [0, 0, 792, 1224].freeze,
  Tabloid: [0, 0, 1224, 792].freeze,
  Executive: [0, 0, 522, 756].freeze,
}.freeze
INHERITABLE_FIELDS =

The inheritable fields.

[:Resources, :MediaBox, :CropBox, :Rotate].freeze

Constants included from DictionaryFields

DictionaryFields::Boolean, DictionaryFields::PDFByteString, DictionaryFields::PDFDate

Instance Attribute Summary

Attributes inherited from Object

#data, #document, #must_be_indirect

Class Method Summary collapse

Instance Method Summary collapse

Methods inherited from Dictionary

#[]=, define_field, define_type, #delete, #each, each_field, #empty?, field, #key?, #to_h, type, #type

Methods inherited from Object

#<=>, #==, #cache, #cached?, #clear_cache, deep_copy, #deep_copy, #document?, #eql?, field, #gen, #gen=, #hash, #indirect?, #initialize, #inspect, make_direct, #null?, #oid, #oid=, #type, #validate, #value, #value=

Constructor Details

This class inherits a constructor from HexaPDF::Object

Class Method Details

.media_box(paper_size, orientation: :portrait) ⇒ Object

Returns the media box for the given paper size or array.

If an array is specified, it needs to contain exactly four numbers. The orientation argument is not used in this case.

See PAPER_SIZE for the defined paper sizes.



113
114
115
116
117
118
119
120
121
122
123
124
 
# File 'lib/hexapdf/type/page.rb', line 113

def self.media_box(paper_size, orientation: :portrait)
  return paper_size if paper_size.kind_of?(Array) && paper_size.size == 4 &&
    paper_size.all?(Numeric)

  unless PAPER_SIZE.key?(paper_size)
    raise HexaPDF::Error, "Invalid paper size specified: #{paper_size}"
  end

  media_box = PAPER_SIZE[paper_size].dup
  media_box[2], media_box[3] = media_box[3], media_box[2] if orientation == :landscape
  media_box
end

Instance Method Details

#[](name) ⇒ Object

Returns the value for the entry name.

If name is an inheritable value and the value has not been set on the page object, its value is retrieved from the ancestor page tree nodes.

See: Dictionary#[]



173
174
175
176
177
178
179
180
181
 
# File 'lib/hexapdf/type/page.rb', line 173

def [](name)
  if value[name].nil? && INHERITABLE_FIELDS.include?(name)
    node = self
    node = node[:Parent] while node.value[name].nil? && node[:Parent]
    node == self || node.value[name].nil? ? super : node[name]
  else
    super
  end
end

#ancestor_nodes ⇒ Object

Returns all parent nodes of the page up to the root of the page tree.

The direct parent is the first node in the array and the root node the last.



406
407
408
409
410
411
 
# File 'lib/hexapdf/type/page.rb', line 406

def ancestor_nodes
  parent = self[:Parent]
  result = [parent]
  result << parent while (parent = parent[:Parent])
  result
end

#box(type = :crop, rectangle = nil) ⇒ Object

:call-seq:

page.box(type = :crop)              -> box
page.box(type = :crop, rectangle)   -> rectangle

If no rectangle is given, returns the rectangle defining a certain kind of box for the page. Otherwise sets the value for the given box type to rectangle (an array with four values or a HexaPDF::Rectangle).

This method should be used instead of directly accessing any of /MediaBox, /CropBox, /BleedBox, /ArtBox or /TrimBox because it also takes the fallback values into account!

The following types are allowed:

:media

The media box defines the boundaries of the medium the page is to be printed on.

:crop

The crop box defines the region to which the contents of the page should be clipped when it is displayed or printed. The default is the media box.

:bleed

The bleed box defines the region to which the contents of the page should be clipped when output in a production environment. The default is the crop box.

:trim

The trim box defines the intended dimensions of the page after trimming. The default value is the crop box.

:art

The art box defines the region of the page’s meaningful content as intended by the author. The default is the crop box.

See: PDF2.0 s14.11.2



227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
 
# File 'lib/hexapdf/type/page.rb', line 227

def box(type = :crop, rectangle = nil)
  if rectangle
    case type
    when :media, :crop, :bleed, :trim, :art
      self["#{type.capitalize}Box".to_sym] = rectangle
    else
      raise ArgumentError, "Unsupported page box type provided: #{type}"
    end
  else
    case type
    when :media then self[:MediaBox]
    when :crop then self[:CropBox] || self[:MediaBox]
    when :bleed then self[:BleedBox] || self[:CropBox] || self[:MediaBox]
    when :trim then self[:TrimBox] || self[:CropBox] || self[:MediaBox]
    when :art then self[:ArtBox] || self[:CropBox] || self[:MediaBox]
    else
      raise ArgumentError, "Unsupported page box type provided: #{type}"
    end
  end
end

#canvas(type: :page, translate_origin: true) ⇒ Object

Returns the requested type of canvas for the page.

There are potentially three different canvas objects, one for each of the types :underlay, :page, and :overlay. The canvas objects are cached once they are created so that their graphics states are correctly retained without the need for parsing the contents. This also means that on subsequent invocations the graphic states of the canvases might already be changed.

type

Can either be

  • :page for getting the canvas for the page itself (only valid for initially empty pages)

  • :overlay for getting the canvas for drawing over the page contents

  • :underlay for getting the canvas for drawing unter the page contents

translate_origin

Specifies whether the origin should automatically be translated into the lower-left corner of the crop box.

Note that this argument is only used for the first invocation for every canvas type. So if a canvas was initially requested with this argument set to false and then with true, it won’t have any effect as the cached canvas is returned.

To check whether the origin has been translated or not, use

canvas.pos(0, 0)

and check whether the result is [0, 0]. If it is, then the origin has not been translated.



441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
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
 
# File 'lib/hexapdf/type/page.rb', line 441

def canvas(type: :page, translate_origin: true)
  unless [:page, :overlay, :underlay].include?(type)
    raise ArgumentError, "Invalid value for 'type', expected: :page, :underlay or :overlay"
  end
  cache_key = "#{type}_canvas".intern
  return cache(cache_key) if cached?(cache_key)

  if type == :page && key?(:Contents)
    raise HexaPDF::Error, "Cannot get the canvas for a page with contents"
  end

  create_canvas = lambda do
    Content::Canvas.new(self).tap do |canvas|
      next unless translate_origin
      crop_box = box(:crop)
      if crop_box.left != 0 || crop_box.bottom != 0
        canvas.translate(crop_box.left, crop_box.bottom)
      end
    end
  end

  contents = self[:Contents]
  if contents.nil?
    page_canvas = cache(:page_canvas, create_canvas.call)
    self[:Contents] = document.add({Filter: :FlateDecode},
                                   stream: page_canvas.stream_data)
  end

  if type == :overlay || type == :underlay
    underlay_canvas = cache(:underlay_canvas, create_canvas.call)
    overlay_canvas = cache(:overlay_canvas, create_canvas.call)

    stream = HexaPDF::StreamData.new do
      Fiber.yield(" q ")
      fiber = underlay_canvas.stream_data.fiber
      while fiber.alive? && (data = fiber.resume)
        Fiber.yield(data)
      end
      " Q q "
    end
    underlay = document.add({Filter: :FlateDecode}, stream: stream)

    stream = HexaPDF::StreamData.new do
      Fiber.yield(" Q q ")
      fiber = overlay_canvas.stream_data.fiber
      while fiber.alive? && (data = fiber.resume)
        Fiber.yield(data)
      end
      " Q "
    end
    overlay = document.add({Filter: :FlateDecode}, stream: stream)

    self[:Contents] = [underlay, *self[:Contents], overlay]
  end

  cache(cache_key)
end

#contents ⇒ Object

Returns the concatenated stream data from the content streams as binary string.

Note: Any modifications done to the returned value *won’t* be reflected in any of the streams’ data!



343
344
345
346
347
348
 
# File 'lib/hexapdf/type/page.rb', line 343

def contents
  Array(self[:Contents]).each_with_object("".b) do |content_stream, content|
    content << " " unless content.empty?
    content << content_stream.stream
  end
end

#contents=(data) ⇒ Object

Replaces the contents of the page with the given string.

This is done by deleting all but the first content stream and reusing this content stream; or by creating a new one if no content stream exists.



354
355
356
357
358
359
360
361
362
363
 
# File 'lib/hexapdf/type/page.rb', line 354

def contents=(data)
  first, *rest = self[:Contents]
  rest.each {|stream| document.delete(stream) }
  if first
    self[:Contents] = first
    document.deref(first).stream = data
  else
    self[:Contents] = document.add({Filter: :FlateDecode}, stream: data)
  end
end

#copy_inherited_values ⇒ Object

Copies the page’s inherited values from the ancestor page tree nodes into a hash and returns the hash.

The hash can then be used to update the page itself (e.g. when moving a page from one position to another) or another page (e.g. when importing a page from another document).



188
189
190
191
192
 
# File 'lib/hexapdf/type/page.rb', line 188

def copy_inherited_values
  INHERITABLE_FIELDS.each_with_object({}) do |name, hash|
    hash[name] = HexaPDF::Object.deep_copy(self[name]) if value[name].nil?
  end
end

#each_annotation ⇒ Object

:call-seq:

page.each_annotation {|annotation| block}    -> page
page.each_annotation                         -> Enumerator

Yields each annotation of this page.



534
535
536
537
538
539
540
541
 
# File 'lib/hexapdf/type/page.rb', line 534

def each_annotation
  return to_enum(__method__) unless block_given?
  Array(self[:Annots]).each do |annotation|
    next unless annotation?(annotation)
    yield(document.wrap(annotation, type: :Annot))
  end
  self
end

#flatten_annotations(annotations = ) ⇒ Object

Flattens all or the given annotations of the page. Returns an array with all the annotations that couldn’t be flattened because they don’t have an appearance stream.

Flattening means making the appearances of the annotations part of the content stream of the page and deleting the annotations themselves. Invisible and hidden fields are deleted but not rendered into the content stream.

If an annotation is a form field widget, only the widget will be deleted but not the form field itself.



552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
 
# File 'lib/hexapdf/type/page.rb', line 552

def flatten_annotations(annotations = self[:Annots])
  not_flattened = Array(annotations) || []
  unless self[:Annots].kind_of?(PDFArray)
    return (not_flattened == [annotations] ? [] : not_flattened)
  end

  annotations = if annotations == self[:Annots]
                  not_flattened
                else
                  not_flattened & self[:Annots]
                end
  return not_flattened if annotations.empty?

  canvas = self.canvas(type: :overlay)
  if (pos = canvas.pos(0, 0)) != [0, 0]
    canvas.save_graphics_state
    canvas.translate(-pos[0], -pos[1])
  end

  to_delete = []
  not_flattened -= annotations
  annotations.each do |annotation|
    unless annotation?(annotation)
      self[:Annots].delete(annotation)
      next
    end

    annotation = document.wrap(annotation, type: :Annot)
    appearance = annotation.appearance
    if annotation.flagged?(:hidden) || annotation.flagged?(:invisible)
      to_delete << annotation
      next
    elsif !appearance
      not_flattened << annotation
      next
    end

    rect = annotation[:Rect]
    box = appearance.box

    # PDF2.0 12.5.5 algorithm
    # Step 1) Calculate smallest rectangle containing transformed bounding box
    matrix = HexaPDF::Content::TransformationMatrix.new(*appearance[:Matrix].value)
    llx, lly = matrix.evaluate(box.left, box.bottom)
    ulx, uly = matrix.evaluate(box.left, box.top)
    lrx, lry = matrix.evaluate(box.right, box.bottom)
    left, right = [llx, ulx, lrx, lrx + (ulx - llx)].minmax
    bottom, top = [lly, uly, lry, lry + (uly - lly)].minmax

    # Handle degenerate case of the transformed bounding box being a line or point
    if right - left == 0 || top - bottom == 0
      to_delete << annotation
      next
    end

    # Step 2) Fit calculated rectangle to annotation rectangle by translating/scaling

    # The final matrix is composed by translating the bottom-left corner of the transformed
    # bounding box to the bottom-left corner of the annotation rectangle and scaling from the
    # bottom-left corner of the transformed bounding box.
    sx = rect.width.fdiv(right - left)
    sy = rect.height.fdiv(top - bottom)
    tx = rect.left - left + left - left * sx
    ty = rect.bottom - bottom + bottom - bottom * sy

    # Step 3) Premultiply form matrix - done implicitly when drawing the XObject

    canvas.transform(sx, 0, 0, sy, tx, ty) do
      # Use [box.left, box.bottom] to counter default translation in #xobject since that
      # is already taken care of in matrix a
      canvas.xobject(appearance, at: [box.left, box.bottom])
    end
    to_delete << annotation
  end
  canvas.restore_graphics_state unless pos == [0, 0]

  to_delete.each do |annotation|
    if annotation[:Subtype] == :Widget
      annotation.form_field.delete_widget(annotation)
    else
      self[:Annots].delete(annotation)
      document.delete(annotation)
    end
  end

  not_flattened
end

#index ⇒ Object

Returns the index of the page in the page tree.



382
383
384
385
386
387
388
389
390
391
392
393
 
# File 'lib/hexapdf/type/page.rb', line 382

def index
  idx = 0
  node = self
  while (parent_node = node[:Parent])
    parent_node[:Kids].each do |kid|
      break if kid == node
      idx += (kid.type == :Page ? 1 : kid[:Count])
    end
    node = parent_node
  end
  idx
end

#label ⇒ Object

Returns the label of the page which is an optional, alternative description of the page index.

See HexaPDF::Document::Pages for details.



399
400
401
 
# File 'lib/hexapdf/type/page.rb', line 399

def label
  document.pages.page_label(index)
end

#must_be_indirect? ⇒ Boolean

Returns true since page objects must always be indirect.

Returns:



163
164
165
 
# File 'lib/hexapdf/type/page.rb', line 163

def must_be_indirect?
  true
end

#orientation(type = :crop) ⇒ Object

Returns the orientation of the specified box (default is the crop box), either :portrait or :landscape.



250
251
252
253
254
255
256
257
258
259
 
# File 'lib/hexapdf/type/page.rb', line 250

def orientation(type = :crop)
  box = self.box(type)
  rotation = self[:Rotate]
  if (box.height > box.width && (rotation == 0 || rotation == 180)) ||
      (box.height < box.width && (rotation == 90 || rotation == 270))
    :portrait
  else
    :landscape
  end
end

#process_contents(processor) ⇒ Object

Processes the content streams associated with the page with the given processor object.

See: HexaPDF::Content::Processor



375
376
377
378
379
 
# File 'lib/hexapdf/type/page.rb', line 375

def process_contents(processor)
  self[:Resources] = {} if self[:Resources].nil?
  processor.resources = self[:Resources]
  Content::Parser.parse(contents, processor)
end

#resources ⇒ Object

Returns the, possibly inherited, resource dictionary which is automatically created if it doesn’t exist.



367
368
369
370
 
# File 'lib/hexapdf/type/page.rb', line 367

def resources
  self[:Resources] ||= document.wrap({ProcSet: [:PDF, :Text, :ImageB, :ImageC, :ImageI]},
                                     type: :XXResources)
end

#rotate(angle, flatten: false) ⇒ Object

Rotates the page angle degrees counterclockwise where angle has to be a multiple of 90.

Positive values rotate the page to the left, negative values to the right. If flatten is true, the rotation is not done via the page’s meta (i.e. the /Rotate key) data but by rotating the canvas itself and all other necessary objects like the various page boxes and annotations.

Notes:

  • The given angle is applied in addition to a possibly already existing rotation (specified via the /Rotate key) and does not replace it.

  • Specifying 0 for angle is valid and means that no additional rotation should be applied. The only meaningful usage of 0 for angle is when flatten is set to true (so that the /Rotate key is removed and the existing rotation information incorporated into the canvas, page boxes and annotations).

  • The /Rotate key of a page object describes the angle in a clockwise orientation but this method uses counterclockwise rotation to be consistent with other rotation methods (e.g. HexaPDF::Content::Canvas#rotate).



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
326
327
328
329
330
331
332
333
334
335
336
337
 
# File 'lib/hexapdf/type/page.rb', line 281

def rotate(angle, flatten: false)
  if angle % 90 != 0
    raise ArgumentError, "Page rotation has to be multiple of 90 degrees"
  end

  # /Rotate and therefore cw_angle is angle in clockwise orientation
  cw_angle = (self[:Rotate] - angle) % 360

  if flatten
    delete(:Rotate)
    return if cw_angle == 0

    matrix = case cw_angle
             when 90  then Content::TransformationMatrix.new(0, -1, 1, 0, -box.bottom, box.right)
             when 180 then Content::TransformationMatrix.new(-1, 0, 0, -1, box.right, box.top)
             when 270 then Content::TransformationMatrix.new(0, 1, -1, 0, box.top, -box.left)
             end

    rotate_box = lambda do |box|
      llx, lly, urx, ury = \
        case cw_angle
        when 90  then [box.right, box.bottom, box.left, box.top]
        when 180 then [box.right, box.top, box.left, box.bottom]
        when 270 then [box.left, box.top, box.right, box.bottom]
        end
      box.value.replace(matrix.evaluate(llx, lly).concat(matrix.evaluate(urx, ury)))
    end

    [:MediaBox, :CropBox, :BleedBox, :TrimBox, :ArtBox].each do |box_name|
      next unless key?(box_name)
      rotate_box.call(self[box_name])
    end

    each_annotation do |annot|
      rotate_box.call(annot[:Rect])
      if (quad_points = annot[:QuadPoints])
        quad_points = quad_points.value if quad_points.respond_to?(:value)
        result = []
        quad_points.each_slice(2) {|x, y| result.concat(matrix.evaluate(x, y)) }
        quad_points.replace(result)
      end
      if (appearance = annot.appearance)
        appearance[:Matrix] = matrix.dup.premultiply(*appearance[:Matrix].value).to_a
      end
      if annot[:Subtype] == :Widget
        app_ch = annot[:MK] ||= document.wrap({}, type: :XXAppearanceCharacteristics)
        app_ch[:R] = (app_ch[:R] + 360 - cw_angle) % 360
      end
    end

    before_contents = document.add({}, stream: " q #{matrix.to_a.join(' ')} cm ")
    after_contents = document.add({}, stream: " Q ")
    self[:Contents] = [before_contents, *self[:Contents], after_contents]
  else
    self[:Rotate] = cw_angle
  end
end

#to_form_xobject(reference: true) ⇒ Object

Creates a Form XObject from the page’s dictionary and contents for the given PDF document.

If reference is true, the page’s contents is referenced when possible to avoid unnecessary decoding/encoding.

Note 1: The created Form XObject is not added to the document automatically!

Note 2: If reference is false and if a canvas is used on this page (see #canvas), this method should only be called once the contents of the page has been fully defined. The reason is that during the copying of the content stream data the contents may be modified to make it a fully valid content stream.



510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
 
# File 'lib/hexapdf/type/page.rb', line 510

def to_form_xobject(reference: true)
  first, *rest = self[:Contents]
  stream = if !first
             nil
           elsif !reference || !rest.empty? || first.raw_stream.kind_of?(String)
             contents
           else
             first.raw_stream
           end
  dict = {
    Type: :XObject,
    Subtype: :Form,
    BBox: HexaPDF::Object.deep_copy(box(:crop)),
    Resources: HexaPDF::Object.deep_copy(self[:Resources]),
    Filter: :FlateDecode,
  }
  document.wrap(dict, stream: stream)
end