hexapdf 0.11.9 → 0.12.0

data/lib/hexapdf/font/true_type_wrapper.rb

@@ -4,7 +4,7 @@
 # This file is part of HexaPDF.
 #
 # HexaPDF - A Versatile PDF Creation and Manipulation Library For Ruby
-# Copyright (C) 2014-2019 Thomas Leitner
+# Copyright (C) 2014-2020 Thomas Leitner
 #
 # HexaPDF is free software: you can redistribute it and/or modify it
 # under the terms of the GNU Affero General Public License version 3 as
@@ -115,15 +115,19 @@ module HexaPDF
       # Returns the wrapped TrueType font object.
       attr_reader :wrapped_font
 
-      # Returns the PDF font dictionary representing the wrapped font.
-      attr_reader :dict
+      # Returns the PDF object associated with the wrapper.
+      attr_reader :pdf_object
 
       # Creates a new object wrapping the TrueType font for the PDF document.
       #
+      # The optional argument +pdf_object+ can be used to set the PDF font object that this wrapper
+      # should be associated with. If no object is set, a suitable one is automatically created.
+      #
       # If +subset+ is true, the font is subset.
-      def initialize(document, font, subset: true)
-        @document = document
+      def initialize(document, font, pdf_object: nil, subset: true)
         @wrapped_font = font
+        @missing_glyph_callable = document.config['font.on_missing_glyph']
+
         @subsetter = (subset ? HexaPDF::Font::TrueType::Subsetter.new(font) : nil)
 
         @cmap = font[:cmap].preferred_table
@@ -131,8 +135,7 @@ module HexaPDF
           raise HexaPDF::Error, "No mapping table for Unicode characters found for TTF " \
             "font #{font.full_name}"
         end
-        @dict = build_font_dict
-        @document.register_listener(:complete_objects, &method(:complete_font_dict))
+        @pdf_object = pdf_object || create_pdf_object(document)
 
         @id_to_glyph = {}
         @codepoint_to_glyph = {}
@@ -166,7 +169,7 @@ module HexaPDF
             if id >= 0 && id < @wrapped_font[:maxp].num_glyphs
               Glyph.new(@wrapped_font, id, str || (+'' << (@cmap.gid_to_code(id) || 0xFFFD)))
             else
-              @document.config['font.on_missing_glyph'].call("\u{FFFD}", font_type, @wrapped_font)
+              @missing_glyph_callable.call("\u{FFFD}", font_type, @wrapped_font)
             end
           end
       end
@@ -179,7 +182,7 @@ module HexaPDF
               if (gid = @cmap[c])
                 glyph(gid, +'' << c)
               else
-                @document.config['font.on_missing_glyph'].call(+'' << c, font_type, @wrapped_font)
+                @missing_glyph_callable.call(+'' << c, font_type, @wrapped_font)
               end
             end
         end
@@ -202,22 +205,21 @@ module HexaPDF
 
       private
 
-      # Builds a Type0 font object representing the TrueType font.
-      #
-      # The returned font object contains only information available at build time, so no
-      # information about glyph specific attributes like width.
+      # Creates a Type0 font object representing the TrueType font.
       #
-      # See: #complete_font_dict
-      def build_font_dict
-        fd = @document.add({Type: :FontDescriptor,
-                            FontName: @wrapped_font.font_name.intern,
-                            FontWeight: @wrapped_font.weight,
-                            Flags: 0,
-                            FontBBox: @wrapped_font.bounding_box.map {|m| m * scaling_factor },
-                            ItalicAngle: @wrapped_font.italic_angle || 0,
-                            Ascent: @wrapped_font.ascender * scaling_factor,
-                            Descent: @wrapped_font.descender * scaling_factor,
-                            StemV: @wrapped_font.dominant_vertical_stem_width})
+      # The returned font object contains only information available at creation time, so no
+      # information about glyph specific attributes like width. The missing information is added
+      # before the PDF document gets written.
+      def create_pdf_object(document)
+        fd = document.add({Type: :FontDescriptor,
+                           FontName: @wrapped_font.font_name.intern,
+                           FontWeight: @wrapped_font.weight,
+                           Flags: 0,
+                           FontBBox: @wrapped_font.bounding_box.map {|m| m * scaling_factor },
+                           ItalicAngle: @wrapped_font.italic_angle || 0,
+                           Ascent: @wrapped_font.ascender * scaling_factor,
+                           Descent: @wrapped_font.descender * scaling_factor,
+                           StemV: @wrapped_font.dominant_vertical_stem_width})
         if @wrapped_font[:'OS/2'].version >= 2
           fd[:CapHeight] = @wrapped_font.cap_height * scaling_factor
           fd[:XHeight] = @wrapped_font.x_height * scaling_factor
@@ -242,28 +244,29 @@ module HexaPDF
           @wrapped_font[:'OS/2'].selection_include?(:oblique)
         fd.flag(:symbolic)
 
-        cid_font = @document.add({Type: :Font, Subtype: :CIDFontType2,
-                                  BaseFont: fd[:FontName], FontDescriptor: fd,
-                                  CIDSystemInfo: {Registry: "Adobe", Ordering: "Identity",
-                                                  Supplement: 0},
-                                  CIDToGIDMap: :Identity})
-        @document.add({Type: :Font, Subtype: :Type0, BaseFont: cid_font[:BaseFont],
-                       Encoding: :"Identity-H", DescendantFonts: [cid_font]})
-      end
+        cid_font = document.add({Type: :Font, Subtype: :CIDFontType2,
+                                 BaseFont: fd[:FontName], FontDescriptor: fd,
+                                 CIDSystemInfo: {Registry: "Adobe", Ordering: "Identity",
+                                                 Supplement: 0},
+                                 CIDToGIDMap: :Identity})
+        dict = document.add({Type: :Font, Subtype: :Type0, BaseFont: cid_font[:BaseFont],
+                             Encoding: :"Identity-H", DescendantFonts: [cid_font]})
+        dict.font_wrapper = self
+
+        document.register_listener(:complete_objects) do
+          update_font_name(dict)
+          embed_font(dict, document)
+          complete_width_information(dict)
+          create_to_unicode_cmap(dict, document)
+        end
 
-      # Makes sure that the Type0 font object as well as the CIDFont object contain all the needed
-      # information.
-      def complete_font_dict
-        update_font_name
-        embed_font
-        complete_width_information
-        create_to_unicode_cmap
+        dict
       end
 
       UPPERCASE_LETTERS = ('A'..'Z').to_a.freeze #:nodoc:
 
       # Updates the font name with a unique tag if the font is subset.
-      def update_font_name
+      def update_font_name(dict)
         return unless @subsetter
 
         tag = +''
@@ -275,13 +278,13 @@ module HexaPDF
         end
 
         name = (tag << "+" << @wrapped_font.font_name).intern
-        @dict[:BaseFont] = name
-        @dict[:DescendantFonts].first[:BaseFont] = name
-        @dict[:DescendantFonts].first[:FontDescriptor][:FontName] = name
+        dict[:BaseFont] = name
+        dict[:DescendantFonts].first[:BaseFont] = name
+        dict[:DescendantFonts].first[:FontDescriptor][:FontName] = name
       end
 
       # Embeds the font.
-      def embed_font
+      def embed_font(dict, document)
         if @subsetter
           data = @subsetter.build_font
           length = data.size
@@ -290,22 +293,22 @@ module HexaPDF
           length = @wrapped_font.io.size
           stream = HexaPDF::StreamData.new(@wrapped_font.io, length: length)
         end
-        font = @document.add({Length1: length, Filter: :FlateDecode}, stream: stream)
-        @dict[:DescendantFonts].first[:FontDescriptor][:FontFile2] = font
+        font = document.add({Length1: length, Filter: :FlateDecode}, stream: stream)
+        dict[:DescendantFonts].first[:FontDescriptor][:FontFile2] = font
       end
 
       # Adds the /DW and /W fields to the CIDFont dictionary.
-      def complete_width_information
+      def complete_width_information(dict)
         default_width = glyph(3, " ").width.to_i
         widths = @encoded_glyphs.reject {|_, v| v[1].width == default_width }.map do |id, v|
           [(@subsetter ? @subsetter.subset_glyph_id(id) : id), v[1].width]
         end.sort!
-        @dict[:DescendantFonts].first.set_widths(widths, default_width: default_width)
+        dict[:DescendantFonts].first.set_widths(widths, default_width: default_width)
       end
 
       # Creates the /ToUnicode CMap and updates the font dictionary so that text extraction works
       # correctly.
-      def create_to_unicode_cmap
+      def create_to_unicode_cmap(dict, document)
         stream = HexaPDF::StreamData.new do
           mapping = @encoded_glyphs.keys.map! do |id|
             # Using 0xFFFD as mentioned in Adobe #5411, last line before section 1.5
@@ -313,9 +316,9 @@ module HexaPDF
           end.sort_by!(&:first)
           HexaPDF::Font::CMap.create_to_unicode_cmap(mapping)
         end
-        stream_obj = @document.add({}, stream: stream)
+        stream_obj = document.add({}, stream: stream)
         stream_obj.set_filter(:FlateDecode)
-        @dict[:ToUnicode] = stream_obj
+        dict[:ToUnicode] = stream_obj
       end
 
     end

data/lib/hexapdf/font/type1.rb

@@ -4,7 +4,7 @@
 # This file is part of HexaPDF.
 #
 # HexaPDF - A Versatile PDF Creation and Manipulation Library For Ruby
-# Copyright (C) 2014-2019 Thomas Leitner
+# Copyright (C) 2014-2020 Thomas Leitner
 #
 # HexaPDF is free software: you can redistribute it and/or modify it
 # under the terms of the GNU Affero General Public License version 3 as

data/lib/hexapdf/font/type1/afm_parser.rb

@@ -4,7 +4,7 @@
 # This file is part of HexaPDF.
 #
 # HexaPDF - A Versatile PDF Creation and Manipulation Library For Ruby
-# Copyright (C) 2014-2019 Thomas Leitner
+# Copyright (C) 2014-2020 Thomas Leitner
 #
 # HexaPDF is free software: you can redistribute it and/or modify it
 # under the terms of the GNU Affero General Public License version 3 as

data/lib/hexapdf/font/type1/character_metrics.rb

@@ -4,7 +4,7 @@
 # This file is part of HexaPDF.
 #
 # HexaPDF - A Versatile PDF Creation and Manipulation Library For Ruby
-# Copyright (C) 2014-2019 Thomas Leitner
+# Copyright (C) 2014-2020 Thomas Leitner
 #
 # HexaPDF is free software: you can redistribute it and/or modify it
 # under the terms of the GNU Affero General Public License version 3 as

data/lib/hexapdf/font/type1/font.rb

@@ -4,7 +4,7 @@
 # This file is part of HexaPDF.
 #
 # HexaPDF - A Versatile PDF Creation and Manipulation Library For Ruby
-# Copyright (C) 2014-2019 Thomas Leitner
+# Copyright (C) 2014-2020 Thomas Leitner
 #
 # HexaPDF is free software: you can redistribute it and/or modify it
 # under the terms of the GNU Affero General Public License version 3 as

data/lib/hexapdf/font/type1/font_metrics.rb

@@ -4,7 +4,7 @@
 # This file is part of HexaPDF.
 #
 # HexaPDF - A Versatile PDF Creation and Manipulation Library For Ruby
-# Copyright (C) 2014-2019 Thomas Leitner
+# Copyright (C) 2014-2020 Thomas Leitner
 #
 # HexaPDF is free software: you can redistribute it and/or modify it
 # under the terms of the GNU Affero General Public License version 3 as

data/lib/hexapdf/font/type1/pfb_parser.rb

@@ -4,7 +4,7 @@
 # This file is part of HexaPDF.
 #
 # HexaPDF - A Versatile PDF Creation and Manipulation Library For Ruby
-# Copyright (C) 2014-2019 Thomas Leitner
+# Copyright (C) 2014-2020 Thomas Leitner
 #
 # HexaPDF is free software: you can redistribute it and/or modify it
 # under the terms of the GNU Affero General Public License version 3 as

data/lib/hexapdf/font/type1_wrapper.rb

@@ -4,7 +4,7 @@
 # This file is part of HexaPDF.
 #
 # HexaPDF - A Versatile PDF Creation and Manipulation Library For Ruby
-# Copyright (C) 2014-2019 Thomas Leitner
+# Copyright (C) 2014-2020 Thomas Leitner
 #
 # HexaPDF is free software: you can redistribute it and/or modify it
 # under the terms of the GNU Affero General Public License version 3 as
@@ -104,23 +104,33 @@ module HexaPDF
       # Returns the wrapped Type1 font object.
       attr_reader :wrapped_font
 
-      # The PDF font dictionary representing the wrapped font.
-      attr_reader :dict
+      # Returns the PDF object associated with the wrapper.
+      attr_reader :pdf_object
 
-      # Creates a new object wrapping the Type1 font for the PDF document.
+      # Creates a new Type1Wrapper object wrapping the Type1 font.
       #
-      # The optional argument +custom_encoding+ can be set to +true+ so that a custom encoding
-      # instead of the WinAnsiEncoding is used.
-      def initialize(document, font, custom_encoding: false)
-        @document = document
+      # The optional argument +pdf_object+ can be used to set the PDF font object that this wrapper
+      # should be associated with. If no object is set, a suitable one is automatically created.
+      #
+      # If +pdf_object+ is provided, the PDF object's encoding is used. Otherwise, the
+      # WinAnsiEncoding or, for 'Special' fonts, the font's internal encoding is used. The optional
+      # argument +custom_encoding+ can be set to +true+ so that a custom encoding is used (only
+      # respected if +pdf_object+ is not provided).
+      def initialize(document, font, pdf_object: nil, custom_encoding: false)
         @wrapped_font = font
+        @pdf_object = pdf_object || create_pdf_object(document)
+        @missing_glyph_callable = document.config['font.on_missing_glyph']
 
-        @dict = build_font_dict
-        @document.register_listener(:complete_objects, &method(:complete_font_dict))
-        if @wrapped_font.metrics.character_set == 'Special' || custom_encoding
+        if pdf_object
+          @encoding = pdf_object.encoding
+          @max_code = 255 # Encoding is not modified
+        elsif custom_encoding
           @encoding = Encoding::Base.new
           @encoding.code_to_name[32] = :space
           @max_code = 32 # 32 = space
+        elsif @wrapped_font.metrics.character_set == 'Special'
+          @encoding = @wrapped_font.encoding
+          @max_code = 255 # Encoding is not modified
         else
           @encoding = Encoding.for_name(:WinAnsiEncoding)
           @max_code = 255 # Encoding is not modified
@@ -150,17 +160,25 @@ module HexaPDF
             if @wrapped_font.metrics.character_metrics.key?(name)
               Glyph.new(@wrapped_font, name, str)
             else
-              @document.config['font.on_missing_glyph'].call(str, font_type, @wrapped_font)
+              @missing_glyph_callable.call(str, font_type, @wrapped_font)
             end
           end
       end
 
       # Returns an array of glyph objects representing the characters in the UTF-8 encoded string.
+      #
+      # If a Unicode codepoint is not available as glyph object, it is tried to map the codepoint
+      # using the font's internal encoding. This is useful, for example, for the ZapfDingbats font
+      # to use ASCII characters for accessing the glyphs.
       def decode_utf8(str)
         str.codepoints.map! do |c|
           @codepoint_to_glyph[c] ||=
             begin
               name = Encoding::GlyphList.unicode_to_name(+'' << c, **@zapf_dingbats_opt)
+              if @wrapped_font.metrics.character_set == 'Special' &&
+                  (name == :'.notdef' || !@wrapped_font.metrics.character_metrics.key?(name))
+                name = @encoding.name(c)
+              end
               name = +"u" << c.to_s(16).rjust(6, '0') if name == :'.notdef'
               glyph(name)
             end
@@ -189,49 +207,47 @@ module HexaPDF
 
       private
 
-      # Builds a generic Type1 font dictionary for the wrapped font.
-      #
-      # Generic in the sense that no information regarding the encoding or widths is included.
-      def build_font_dict
-        unless defined?(@fd)
-          @fd = @document.wrap({Type: :FontDescriptor,
-                                FontName: @wrapped_font.font_name.intern,
-                                FontWeight: @wrapped_font.weight_class,
-                                FontBBox: @wrapped_font.bounding_box,
-                                ItalicAngle: @wrapped_font.italic_angle || 0,
-                                Ascent: @wrapped_font.ascender || 0,
-                                Descent: @wrapped_font.descender || 0,
-                                CapHeight: @wrapped_font.cap_height,
-                                XHeight: @wrapped_font.x_height,
-                                StemH: @wrapped_font.dominant_horizontal_stem_width,
-                                StemV: @wrapped_font.dominant_vertical_stem_width || 0})
-          @fd.flag(:fixed_pitch) if @wrapped_font.metrics.is_fixed_pitch
-          @fd.flag(@wrapped_font.metrics.character_set == 'Special' ? :symbolic : :nonsymbolic)
-          @fd.must_be_indirect = true
-        end
-
-        @document.wrap({Type: :Font, Subtype: :Type1,
-                        BaseFont: @wrapped_font.font_name.intern, Encoding: :WinAnsiEncoding,
-                        FontDescriptor: @fd})
-      end
-
       # Array of valid encoding names in PDF
       VALID_ENCODING_NAMES = [:WinAnsiEncoding, :MacRomanEncoding, :MacExpertEncoding].freeze
 
-      # Completes the font dictionary by filling in the values that depend on the used encoding.
-      def complete_font_dict
-        min, max = @encoding.code_to_name.keys.minmax
-        @dict[:FirstChar] = min
-        @dict[:LastChar] = max
-        @dict[:Widths] = (min..max).map {|code| glyph(@encoding.name(code)).width }
-
-        if VALID_ENCODING_NAMES.include?(@encoding.encoding_name)
-          @dict[:Encoding] = @encoding.encoding_name
-        else
-          differences = [min]
-          (min..max).each {|code| differences << @encoding.name(code) }
-          @dict[:Encoding] = {Differences: differences}
+      # Creates a PDF object representing the wrapped font for the given PDF document.
+      def create_pdf_object(document)
+        fd = document.wrap({Type: :FontDescriptor,
+                            FontName: @wrapped_font.font_name.intern,
+                            FontWeight: @wrapped_font.weight_class,
+                            FontBBox: @wrapped_font.bounding_box,
+                            ItalicAngle: @wrapped_font.italic_angle || 0,
+                            Ascent: @wrapped_font.ascender || 0,
+                            Descent: @wrapped_font.descender || 0,
+                            CapHeight: @wrapped_font.cap_height,
+                            XHeight: @wrapped_font.x_height,
+                            StemH: @wrapped_font.dominant_horizontal_stem_width,
+                            StemV: @wrapped_font.dominant_vertical_stem_width || 0})
+        fd.flag(:fixed_pitch) if @wrapped_font.metrics.is_fixed_pitch
+        fd.flag(@wrapped_font.metrics.character_set == 'Special' ? :symbolic : :nonsymbolic)
+        fd.must_be_indirect = true
+
+        dict = document.wrap({Type: :Font, Subtype: :Type1,
+                              BaseFont: @wrapped_font.font_name.intern,
+                              FontDescriptor: fd})
+        dict.font_wrapper = self
+
+        document.register_listener(:complete_objects) do
+          min, max = @encoding.code_to_name.keys.minmax
+          dict[:FirstChar] = min
+          dict[:LastChar] = max
+          dict[:Widths] = (min..max).map {|code| glyph(@encoding.name(code)).width }
+
+          if VALID_ENCODING_NAMES.include?(@encoding.encoding_name)
+            dict[:Encoding] = @encoding.encoding_name
+          elsif @encoding != @wrapped_font.encoding
+            differences = [min]
+            (min..max).each {|code| differences << @encoding.name(code) }
+            dict[:Encoding] = {Differences: differences}
+          end
         end
+
+        dict
       end
 
     end

data/lib/hexapdf/font_loader.rb

@@ -4,7 +4,7 @@
 # This file is part of HexaPDF.
 #
 # HexaPDF - A Versatile PDF Creation and Manipulation Library For Ruby
-# Copyright (C) 2014-2019 Thomas Leitner
+# Copyright (C) 2014-2020 Thomas Leitner
 #
 # HexaPDF is free software: you can redistribute it and/or modify it
 # under the terms of the GNU Affero General Public License version 3 as

data/lib/hexapdf/font_loader/from_configuration.rb

@@ -4,7 +4,7 @@
 # This file is part of HexaPDF.
 #
 # HexaPDF - A Versatile PDF Creation and Manipulation Library For Ruby
-# Copyright (C) 2014-2019 Thomas Leitner
+# Copyright (C) 2014-2020 Thomas Leitner
 #
 # HexaPDF is free software: you can redistribute it and/or modify it
 # under the terms of the GNU Affero General Public License version 3 as

data/lib/hexapdf/font_loader/from_file.rb

@@ -4,7 +4,7 @@
 # This file is part of HexaPDF.
 #
 # HexaPDF - A Versatile PDF Creation and Manipulation Library For Ruby
-# Copyright (C) 2014-2019 Thomas Leitner
+# Copyright (C) 2014-2020 Thomas Leitner
 #
 # HexaPDF is free software: you can redistribute it and/or modify it
 # under the terms of the GNU Affero General Public License version 3 as