pdfrb 0.8.3 → 0.8.5

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 386ea091bd58d931bc196bd3df335f26043feb5f70e02e33cd076b30566e157d
-  data.tar.gz: d41d511beef5d357dd5e86f0253cd932f11475dc173e0e29665462bb40fcab61
+  metadata.gz: 04a819d2ed2b54eb4ca8c0ac17ceb14cc1c45a79e359bb51c6b5194e2fdb7025
+  data.tar.gz: bcb6a670c04b7e68b20b90f8ac344ac0672a76f1fd3e16306ac46b4c44f728ec
 SHA512:
-  metadata.gz: 249ca1bde48ee762616405b716b93389fd6eea3950f981a59d8cc18b81620891190c1addae3009b1624065ed7ef3411304761fb1e0cc014bd60c02ba8db01eef
-  data.tar.gz: 69258188bf5f7cc3a3ffddbb89348b7a3589c54bb5a943fcb46a059c72aded574cdbfb43d212abca82eea0e44375a1f8caf863f51666dc79f95537b3c4f8945d
+  metadata.gz: f8eb286788ae52128019b39b18006776c9c9227ab92e76b0a72add85eb92b16daa390704d243627a0f79fcd8b9c0195a606f7f6878953db634ff1c9ad257b0f4
+  data.tar.gz: 1940d86297e8e20ae06d33b23bd7b36815d5249fce28eb120262ab33e5350668fa015247444ed905ac1ba5b73295fa092b74d6e477c9fd74a507252a78cb110f

data/lib/pdfrb/annotation/base.rb

@@ -39,7 +39,7 @@ module Pdfrb
             Type: :Annot,
             Subtype: subtype,
             Rect: rect,
-            P: Pdfrb::Model::Reference.new(page.oid, page.gen),
+            P: page.ref,
             Contents: contents,
             F: flags,
           }.compact
@@ -64,7 +64,7 @@ module Pdfrb
 
         def attach_to_page(page, annot)
           annots = page.value[:Annots]
-          ref = Pdfrb::Model::Reference.new(annot.oid, annot.gen)
+          ref = annot.ref
           if annots.nil?
             page.value[:Annots] = [ref]
           elsif annots.is_a?(::Array)

data/lib/pdfrb/appearance/generator.rb

@@ -195,7 +195,7 @@ module Pdfrb
 
       def set_appearance(field, form, state_name = :Normal)
         form_stream = form.stream
-        ref = Pdfrb::Model::Reference.new(form_stream.oid, form_stream.gen)
+        ref = form_stream.ref
 
         ap = field.value[:AP]
         if ap.nil?

data/lib/pdfrb/cli/images.rb

@@ -18,7 +18,7 @@ module Pdfrb
 
           xo = xo.value if xo.is_a?(Pdfrb::Model::Cos::Dictionary)
           xo&.each do |name, ref|
-            obj = ref.is_a?(Pdfrb::Model::Reference) ? doc.object(ref) : ref
+            obj = doc.resolve(ref)
             next unless obj&.value&.[](:Subtype) == :Image
 
             count += 1

data/lib/pdfrb/cli.rb

@@ -158,7 +158,7 @@ module Pdfrb
       return unless fields
 
       fields.each do |ref|
-        field = ref.is_a?(Pdfrb::Model::Reference) ? doc.object(ref) : ref
+        field = doc.resolve(ref)
         next unless field
 
         name = field[:T] || "(unnamed)"
@@ -330,7 +330,7 @@ module Pdfrb
           O: o_entry, U: u_entry },
         type: Pdfrb::Model::Cos::Dictionary
       )
-      doc.trailer[:Encrypt] = Pdfrb::Model::Reference.new(encrypt_dict.oid, encrypt_dict.gen)
+      doc.trailer[:Encrypt] = encrypt_dict.ref
 
       # Set up the handler so Writer can encrypt per-object.
       handler = Pdfrb::Encryption::StandardSecurityHandler.new(
@@ -364,7 +364,7 @@ module Pdfrb
           O: o_entry, U: u_entry, OE: "".b, UE: "".b },
         type: Pdfrb::Model::Cos::Dictionary
       )
-      doc.trailer[:Encrypt] = Pdfrb::Model::Reference.new(encrypt_dict.oid, encrypt_dict.gen)
+      doc.trailer[:Encrypt] = encrypt_dict.ref
       doc.config["encryption.password"] = user_pw
     end
 

data/lib/pdfrb/color/default_profile.rb

@@ -47,7 +47,7 @@ module Pdfrb
           type: Pdfrb::Model::Cos::Stream
         )
         stream.stream = bytes
-        [:ICCBased, Pdfrb::Model::Reference.new(stream.oid, stream.gen)]
+        [:ICCBased, stream.ref]
       end
 
       # ---- Header construction ----

data/lib/pdfrb/conformance/ltv.rb

@@ -87,7 +87,7 @@ module Pdfrb
                         fields = Pdfrb::Conformance::Pades.signature_fields(doc)
                         has_archival = fields.any? do |field|
                           v = field[:V]
-                          v_obj = v.is_a?(Pdfrb::Model::Reference) ? doc.object(v) : v
+                          v_obj = doc.resolve(v)
                           v_obj && v_obj[:Type]&.to_sym == :DocTimeStamp
                         end
                         next nil if has_archival

data/lib/pdfrb/conformance/pades.rb

@@ -52,7 +52,7 @@ module Pdfrb
                           v = field[:V]
                           next unless v
 
-                          v_obj = v.is_a?(Pdfrb::Model::Reference) ? doc.object(v) : v
+                          v_obj = doc.resolve(v)
                           next if v_obj && ALLOWED_SIGNATURE_TYPES.include?(v_obj[:Type]&.to_sym)
 
                           vs << Violation.new(
@@ -81,7 +81,7 @@ module Pdfrb
                           v = field[:V]
                           next unless v
 
-                          v_obj = v.is_a?(Pdfrb::Model::Reference) ? doc.object(v) : v
+                          v_obj = doc.resolve(v)
                           has_timestamp = v_obj && (
                             v_obj[:Type]&.to_sym == :DocTimeStamp ||
                               (v_obj[:ByteRange] && timestamp_in_signature?(v_obj))
@@ -134,7 +134,7 @@ module Pdfrb
                         # protecting the DSS.
                         has_archival = signature_fields(doc).any? do |field|
                           v = field[:V]
-                          v_obj = v.is_a?(Pdfrb::Model::Reference) ? doc.object(v) : v
+                          v_obj = doc.resolve(v)
                           v_obj && v_obj[:Type]&.to_sym == :DocTimeStamp
                         end
                         next nil if has_archival
@@ -173,7 +173,7 @@ module Pdfrb
         return [] unless fields.is_a?(::Array)
 
         fields.filter_map do |ref|
-          obj = ref.is_a?(Pdfrb::Model::Reference) ? document.object(ref) : ref
+          obj = document.resolve(ref)
           next nil unless obj
 
           obj[:FT]&.to_sym == :Sig ? obj : nil

data/lib/pdfrb/conformance/pdf_2_af.rb

@@ -131,7 +131,7 @@ module Pdfrb
           af_value = af.is_a?(Pdfrb::Model::PdfArray) ? af.value : af
           entries = af_value.is_a?(::Array) ? af_value : [af_value]
           entries.each do |e|
-            resolved = e.is_a?(Pdfrb::Model::Reference) ? document.object(e) : e
+            resolved = document.resolve(e)
             yield container, resolved if resolved
           end
         end
@@ -153,7 +153,7 @@ module Pdfrb
           af_value = af.is_a?(Pdfrb::Model::PdfArray) ? af.value : af
           entries = af_value.is_a?(::Array) ? af_value : [af_value]
           entries.each do |e|
-            resolved = e.is_a?(Pdfrb::Model::Reference) ? document.object(e) : e
+            resolved = document.resolve(e)
             next unless resolved && resolved.value[:Type] == :FileSpec
 
             yield container, resolved

data/lib/pdfrb/conformance/pdf_a.rb

@@ -65,7 +65,7 @@ module Pdfrb
                           next unless fonts
 
                           fonts.each_value do |ref|
-                            f = ref.is_a?(Pdfrb::Model::Reference) ? doc.object(ref) : ref
+                            f = doc.resolve(ref)
                             next unless f
                             next if f[:FontDescriptor]
 

data/lib/pdfrb/conformance/pdf_x.rb

@@ -94,7 +94,7 @@ module Pdfrb
 
                           fonts_hash = fonts.is_a?(Pdfrb::Model::Cos::Dictionary) ? fonts.value : fonts
                           fonts_hash&.each_value do |ref|
-                            font = ref.is_a?(Pdfrb::Model::Reference) ? doc.object(ref) : ref
+                            font = doc.resolve(ref)
                             next unless font
 
                             fd = font[:FontDescriptor]
@@ -174,7 +174,7 @@ module Pdfrb
 
         intents = [intents] unless intents.is_a?(::Array)
         intents.any? do |ref|
-          obj = ref.is_a?(Pdfrb::Model::Reference) ? doc.object(ref) : ref
+          obj = doc.resolve(ref)
           obj && obj.value[:S] == :GTS_PDFX
         end
       end
@@ -257,7 +257,7 @@ module Pdfrb
                           else [intents]
                           end
           has_pdfx = intents_array.any? do |i|
-            obj = i.is_a?(Pdfrb::Model::Reference) ? doc.object(i) : i
+            obj = doc.resolve(i)
             obj && obj[:S] == :GTS_PDFX
           end
           next nil if has_pdfx

data/lib/pdfrb/conformance/tagged_pdf.rb

@@ -145,7 +145,7 @@ module Pdfrb
         root_ref = document.catalog[:StructTreeRoot]
         return unless root_ref
 
-        root = root_ref.is_a?(Pdfrb::Model::Reference) ? document.object(root_ref) : root_ref
+        root = document.resolve(root_ref)
         return unless root
 
         first_kids = structure_kids(root, document)

data/lib/pdfrb/content/canvas.rb

@@ -531,7 +531,7 @@ module Pdfrb
         )
         r = @stream.value[:Resources] || {}
         eg = r[:ExtGState] || {}
-        eg[name] = Pdfrb::Model::Reference.new(gs.oid, gs.gen)
+        eg[name] = gs.ref
         r[:ExtGState] = eg
         @stream.value[:Resources] = r
         name

data/lib/pdfrb/content/hidden_text_detector.rb

@@ -69,7 +69,7 @@ module Pdfrb
                  end
 
           return obj if list.any? do |c|
-            c_obj = c.is_a?(Pdfrb::Model::Reference) ? @document.object(c) : c
+            c_obj = @document.resolve(c)
             c_obj == stream_obj
           end
         end

data/lib/pdfrb/content/shading/base.rb

@@ -31,7 +31,7 @@ module Pdfrb
         # @return [Symbol] resource name (e.g. :Sh1)
         def register_on(document, page)
           shading_obj = document.add(to_dict, type: Pdfrb::Model::Cos::Dictionary)
-          ref = Pdfrb::Model::Reference.new(shading_obj.oid, shading_obj.gen)
+          ref = shading_obj.ref
           resources = ensure_resources(document, page)
           shading_dict = resources.value[:Shading]
           if shading_dict.nil?

data/lib/pdfrb/content/tiling_pattern.rb

@@ -51,7 +51,7 @@ module Pdfrb
           yield canvas
         end
 
-        ref = Pdfrb::Model::Reference.new(pattern_stream.oid, pattern_stream.gen)
+        ref = pattern_stream.ref
         register_in_resources(document, page, ref)
       end
 

data/lib/pdfrb/digital_signature/signing.rb

@@ -43,12 +43,12 @@ module Pdfrb
 
         field = document.add(
           { Type: :Annot, Subtype: :Widget, FT: :Sig, T: "Signature1",
-            V: Pdfrb::Model::Reference.new(sig_dict.oid, sig_dict.gen),
+            V: sig_dict.ref,
             Rect: [0, 0, 0, 0] },
           type: Pdfrb::Model::Cos::Dictionary
         )
         catalog.value[:AcroForm][:Fields] <<
-          Pdfrb::Model::Reference.new(field.oid, field.gen)
+          field.ref
       end
 
       def serialize(document)

data/lib/pdfrb/document/annotations.rb

@@ -16,14 +16,14 @@ module Pdfrb
             Type: :Annot,
             Subtype: subtype,
             Rect: rect,
-            P: Pdfrb::Model::Reference.new(page.oid, page.gen),
+            P: page.ref,
             Contents: opts[:contents],
             F: opts[:flags] || 0
           }.compact,
           type: Pdfrb::Model::Type::Annotation
         )
         annots = (page.value[:Annots] ||= [])
-        annots << Pdfrb::Model::Reference.new(annot.oid, annot.gen)
+        annots << annot.ref
         annot
       end
 
@@ -45,7 +45,7 @@ module Pdfrb
             { Type: :Action, S: :URI, URI: uri },
             type: Pdfrb::Model::Type::Action
           )
-          annot.value[:A] = Pdfrb::Model::Reference.new(action.oid, action.gen)
+          annot.value[:A] = action.ref
         end
         annot
       end
@@ -55,7 +55,7 @@ module Pdfrb
         annot = document.add(
           {
             Type: :Annot, Subtype: :FreeText, Rect: rect,
-            P: Pdfrb::Model::Reference.new(page.oid, page.gen),
+            P: page.ref,
             Contents: contents
           },
           type: Pdfrb::Model::Type::FreeTextAnnotation
@@ -65,7 +65,7 @@ module Pdfrb
         da_parts << "#{color.join(' ')} rg" if color
         annot.value[:DA] = da_parts.join(" ") unless da_parts.empty?
         annots = (page.value[:Annots] ||= [])
-        annots << Pdfrb::Model::Reference.new(annot.oid, annot.gen)
+        annots << annot.ref
         annot
       end
 
@@ -75,13 +75,13 @@ module Pdfrb
           {
             Type: :Annot, Subtype: :Highlight,
             QuadPoints: quad_points,
-            P: Pdfrb::Model::Reference.new(page.oid, page.gen),
+            P: page.ref,
             Contents: contents
           }.compact,
           type: Pdfrb::Model::Type::HighlightAnnotation
         )
         annots = (page.value[:Annots] ||= [])
-        annots << Pdfrb::Model::Reference.new(annot.oid, annot.gen)
+        annots << annot.ref
         annot
       end
 
@@ -90,13 +90,13 @@ module Pdfrb
         annot = document.add(
           {
             Type: :Annot, Subtype: :Stamp, Rect: rect, Name: name,
-            P: Pdfrb::Model::Reference.new(page.oid, page.gen),
+            P: page.ref,
             Contents: contents
           }.compact,
           type: Pdfrb::Model::Type::StampAnnotation
         )
         annots = (page.value[:Annots] ||= [])
-        annots << Pdfrb::Model::Reference.new(annot.oid, annot.gen)
+        annots << annot.ref
         annot
       end
 
@@ -114,7 +114,7 @@ module Pdfrb
         return self unless annots
 
         annots.each do |ref|
-          a = ref.is_a?(Pdfrb::Model::Reference) ? document.object(ref) : ref
+          a = document.resolve(ref)
           yield a if a
         end
         self

data/lib/pdfrb/document/associated_files.rb

@@ -54,7 +54,7 @@ module Pdfrb
         filespec = document.files.add(data, name: filename,
                                             description: description,
                                             mime_type: mime_type)
-        ref = Pdfrb::Model::Reference.new(filespec.oid, filespec.gen)
+        ref = filespec.ref
         filespec.value[:AFRelationship] = relationship
         add_to_catalog(ref, relationship: relationship)
         ref
@@ -95,7 +95,7 @@ module Pdfrb
 
         refs = af.is_a?(::Array) ? af : [af]
         refs.filter_map do |ref|
-          obj = ref.is_a?(Pdfrb::Model::Reference) ? document.object(ref) : ref
+          obj = document.resolve(ref)
           obj if obj.is_a?(Pdfrb::Model::Cos::Dictionary)
         end
       end

data/lib/pdfrb/document/colors.rb

@@ -30,7 +30,7 @@ module Pdfrb
         stream.stream = profile.raw_data
         Pdfrb::Model::PdfArray.new([
                                      :ICCBased,
-                                     Pdfrb::Model::Reference.new(stream.oid, stream.gen),
+                                     stream.ref,
                                    ])
       end
 

data/lib/pdfrb/document/destinations.rb

@@ -79,7 +79,7 @@ module Pdfrb
         d = catalog.value[:Dests]
         return nil unless d
 
-        d.is_a?(Pdfrb::Model::Reference) ? document.object(d) : d
+        document.resolve(d)
       end
 
       def resolved_name_tree
@@ -93,7 +93,7 @@ module Pdfrb
         d = names.is_a?(Pdfrb::Model::Cos::Dictionary) ? names.value[:Dests] : names[:Dests]
         return nil unless d
 
-        d.is_a?(Pdfrb::Model::Reference) ? document.object(d) : d
+        document.resolve(d)
       end
 
       def lookup_in_name_tree(node, name)
@@ -107,7 +107,7 @@ module Pdfrb
           arr = node.value[:Kids]
           arr = arr.to_a if arr.is_a?(Pdfrb::Model::PdfArray)
           arr.each do |kid_ref|
-            kid = kid_ref.is_a?(Pdfrb::Model::Reference) ? document.object(kid_ref) : kid_ref
+            kid = document.resolve(kid_ref)
             result = lookup_in_name_tree(kid, name)
             return result if result
           end
@@ -130,7 +130,7 @@ module Pdfrb
           arr = node.value[:Kids]
           arr = arr.to_a if arr.is_a?(Pdfrb::Model::PdfArray)
           arr.each do |kid_ref|
-            kid = kid_ref.is_a?(Pdfrb::Model::Reference) ? document.object(kid_ref) : kid_ref
+            kid = document.resolve(kid_ref)
             each_name_tree_key(kid, &block)
           end
         elsif node.is_a?(Hash) && node[:Names]

data/lib/pdfrb/document/encryption.rb

@@ -176,7 +176,7 @@ module Pdfrb
       def register(encrypt_hash, user_password:, prebuilt_handler: nil)
         dict = document.add(encrypt_hash,
                             type: Pdfrb::Model::Type::EncryptionStandard)
-        ref = Pdfrb::Model::Reference.new(dict.oid, dict.gen)
+        ref = dict.ref
         trailer = document.trailer
         if trailer.is_a?(Pdfrb::Model::Cos::Dictionary)
           trailer.value[:Encrypt] = ref

data/lib/pdfrb/document/files.rb

@@ -42,8 +42,8 @@ module Pdfrb
           {
             Type: :FileSpec,
             UF: name.to_s,
-            EF: { UF: Pdfrb::Model::Reference.new(ef_stream.oid, ef_stream.gen),
-                  F: Pdfrb::Model::Reference.new(ef_stream.oid, ef_stream.gen) },
+            EF: { UF: ef_stream.ref,
+                  F: ef_stream.ref },
           },
           type: Pdfrb::Model::Cos::Dictionary
         )
@@ -66,7 +66,7 @@ module Pdfrb
         return self unless names_array
 
         names_array.each_slice(2) do |name, ref|
-          resolved = ref.is_a?(Pdfrb::Model::Reference) ? @document.object(ref) : ref
+          resolved = @document.resolve(ref)
           yield(name.to_s, resolved) if resolved
         end
         self
@@ -100,7 +100,7 @@ module Pdfrb
         ef_tree = names[:EmbeddedFiles] ||= {}
         names_array = ef_tree[:Names] ||= []
         names_array << name
-        names_array << Pdfrb::Model::Reference.new(filespec.oid, filespec.gen)
+        names_array << filespec.ref
       end
 
       def embedded_files_names_array
@@ -117,7 +117,7 @@ module Pdfrb
       end
 
       def add_to_af(target, filespec, relationship)
-        ref = Pdfrb::Model::Reference.new(filespec.oid, filespec.gen)
+        ref = filespec.ref
         af_entry = if relationship
                      { Type: :AssociatedFile, AFRelationship: relationship,
                        File: ref }

data/lib/pdfrb/document/fonts/subsetting.rb

@@ -0,0 +1,87 @@
+# frozen_string_literal: true
+
+module Pdfrb
+  class Document
+    class Fonts
+      # Font-program subsetting (s9.6.4). Mixed into the Fonts
+      # facade; the seam it owns is "replace every embedded font
+      # program with a subset covering only the used codepoints",
+      # dispatching on the outline format (TrueType vs CFF-in-OTF).
+      module Subsetting
+        # Entry point, invoked by Document#write. Fonts that were
+        # never drawn with (no recorded codepoints) keep their full
+        # program; failures fall back to the full font.
+        def subset_fonts!
+          @font_streams.each do |resource, data|
+            next unless valid_font_data?(data)
+
+            codepoints = @used_codepoints[resource]
+            next if codepoints.empty?
+
+            begin
+              subset_font(resource, data, codepoints)
+            rescue StandardError
+              next
+            end
+          end
+        end
+
+        protected
+
+        # Build the subsetted font bytes for +resource+ and replace the
+        # descriptor's font-file stream (FontFile2 for TrueType
+        # outlines, FontFile3/OpenType for CFF).
+        def subset_font(resource, data, codepoints)
+          subset, font_file_key =
+            if data.byteslice(0, 4) == "OTTO".b
+              # CFF outlines: map codepoints to glyph IDs via the OTF
+              # cmap (it addresses CFF glyphs too), subset the 'CFF '
+              # table, and rebuild the OTF container.
+              cmap = Pdfrb::Font::TrueType::File.new(data).cmap
+              gids = codepoints.filter_map { |cp| cmap.glyph_id_for(cp) }.uniq
+              [Pdfrb::Font::CFF::Subsetter.subset_otf(data, gids), :FontFile3]
+            else
+              ttf = Pdfrb::Font::TrueType::File.new(data)
+              subsetter = Pdfrb::Font::TrueType::Subsetter.new(ttf, codepoints.to_a)
+              [subsetter.subset, :FontFile2]
+            end
+          dict = @font_dicts[resource]
+          return unless dict
+
+          desc_ref = dict.value[:FontDescriptor]
+          return unless desc_ref
+
+          desc = document.resolve(desc_ref)
+          return unless desc
+
+          # Reuse the add-time stream object in place rather than
+          # allocating a replacement — a second stream would leave the
+          # full original orphaned in the file (doubling output size).
+          fd_stream = desc.value[font_file_key]
+          fd_stream = document.resolve(fd_stream)
+          if fd_stream.is_a?(Pdfrb::Model::Cos::Stream)
+            fd_stream.stream = subset
+            fd_stream.value[:Length] = subset.bytesize
+            fd_stream.value.delete(:Filter)
+            if font_file_key == :FontFile3
+              fd_stream.value[:Subtype] = :OpenType
+              fd_stream.value[:Length1] = subset.bytesize
+            end
+          end
+        end
+
+        # Generate a deterministic 6-letter subset tag from the font bytes
+        # so re-embedding the same font produces the same prefix. Format
+        # is six uppercase ASCII letters derived from the SHA-1 of the
+        # first 1 KB of the font file (per PDF spec s9.6.4).
+        def subset_tag_for(font_bytes)
+          require "digest"
+          digest = Digest::SHA1.digest(font_bytes.byteslice(0, 1024) || "")
+          digest.byteslice(0, 6).bytes.each_with_object(+"") do |b, s|
+            s << ((b % 26) + 65).chr
+          end
+        end
+      end
+    end
+  end
+end