podoff 1.1.0 → 1.1.1

data/CHANGELOG.txt

@@ -2,6 +2,12 @@
 = podoff CHANGELOG.txt
 
 
+== podoff 1.1.1  released 2015-10-26
+
+- reworked xref table output
+- FlateDecode stream if length > 98
+
+
 == podoff 1.1.0  released 2015-10-25
 
 - more tolerant at parsing (StringScanner)

data/lib/podoff.rb

@@ -23,13 +23,14 @@
 # Made in Japan.
 #++
 
+require 'zlib'
 require 'strscan'
 require 'stringio'
 
 
 module Podoff
 
-  VERSION = '1.1.0'
+  VERSION = '1.1.1'
 
   def self.load(path, encoding='iso-8859-1')
 
@@ -53,7 +54,7 @@ module Podoff
       Podoff::Document.new(s)
     end
 
-    attr_reader :source
+    attr_reader :scanner
     attr_reader :version
     attr_reader :xref
     attr_reader :objs
@@ -67,7 +68,7 @@ module Podoff
       fail ArgumentError.new('not a PDF file') \
         unless s.match(/\A%PDF-\d+\.\d+\s/)
 
-      @source = s
+      @scanner = ::StringScanner.new(s)
       @version = nil
       @xref = nil
       @objs = {}
@@ -76,15 +77,14 @@ module Podoff
 
       @additions = {}
 
-      sca = ::StringScanner.new(s)
-      @version = sca.scan(/%PDF-\d+\.\d+/)
+      @version = @scanner.scan(/%PDF-\d+\.\d+/)
 
       loop do
 
-        i = sca.skip_until(
+        i = @scanner.skip_until(
           /(startxref\s+\d+|\d+\s+\d+\s+obj|\/Root\s+\d+\s+\d+\s+R)/)
 
-        m = sca.matched
+        m = @scanner.matched
         break unless m
 
         if m[0] == 's'
@@ -92,22 +92,27 @@ module Podoff
         elsif m[0] == '/'
           @root = extract_ref(m)
         else
-          obj = Podoff::Obj.extract(self, sca)
+          obj = Podoff::Obj.extract(self)
           @objs[obj.ref] = obj
           @obj_counters[obj.ref] = (@obj_counters[obj.ref] || 0) + 1
         end
       end
 
       if @root == nil
-        sca.pos = 0
+        @scanner.pos = 0
         loop do
-          i = sca.skip_until(/\/Root\s+\d+\s+\d+\s+R/)
-          break unless sca.matched
-          @root = extract_ref(sca.matched)
+          i = @scanner.skip_until(/\/Root\s+\d+\s+\d+\s+R/)
+          break unless @scanner.matched
+          @root = extract_ref(@scanner.matched)
         end
       end
     end
 
+    def source
+
+      @scanner.string
+    end
+
     def extract_ref(s)
 
       s.gsub(/\s+/, ' ').gsub(/[^0-9 ]+/, '').strip
@@ -124,7 +129,7 @@ module Podoff
 
       self.class.allocate.instance_eval do
 
-        @source = o.source
+        @scanner = ::StringScanner.new(o.source)
         @xref = o.xref
 
         @objs = o.objs.inject({}) { |h, (k, v)| h[k] = v.dup(self); h }
@@ -182,35 +187,32 @@ module Podoff
 
       name = name[1..-1] if name[0] == '/'
 
-      ref = new_ref
+      r = new_ref
+      s = "#{r} obj <</Type /Font /Subtype /Type1 /BaseFont /#{name}>> endobj"
 
-      add(
-        Obj.create(
-          self,
-          ref,
-          [
-            "#{ref} obj",
-            "<< /Type /Font /Subtype /Type1 /BaseFont /#{name} >>",
-            "endobj"
-          ].join(' ')))
+      add(Obj.new(self, r, source: s))
     end
 
-    def add_stream(s=nil, &block)
+    def add_stream(src=nil, &block)
 
       ref = new_ref
 
-      s = s || make_stream(&block)
+      src =
+        src &&
+        [
+          "#{ref} obj",
+          "<< /Length #{src.size} >>\nstream\n#{src}\nendstream",
+          "endobj"
+        ].join("\n")
 
-      s = [
-        "#{ref} obj",
-        "<< /Length #{s.length} >>",
-        "stream\n#{s}\nendstream",
-        "endobj"
-      ].join("\n") if s.is_a?(String)
+      str =
+        src ?
+        nil :
+        make_stream(&block)
 
-      o = add(Obj.create(self, ref, s))
+      obj = add(Obj.new(self, ref, source: src, stream: str))
 
-      s.is_a?(Podoff::Stream) ? s : o
+      str || obj
     end
 
     def re_add(obj_or_ref)
@@ -231,7 +233,7 @@ module Podoff
           else path
         end
 
-      f.write(@source)
+      f.write(source)
 
       if @additions.any?
 
@@ -239,27 +241,14 @@ module Podoff
 
         @additions.values.each do |o|
           f.write("\n")
-          pointers[o.ref] = f.pos + 1
-          if o.source.is_a?(String)
-            f.write(o.source)
-          else # Stream
-            s = o.source.to_s
-            f.write("#{o.ref} obj\n<< /Length #{s.length} >>\n")
-            f.write("stream\n#{s}\nendstream\nendobj")
-          end
+          pointers[o.ref.split(' ').first.to_i] = f.pos + 1
+          f.write(o.to_s)
         end
         f.write("\n\n")
 
         xref = f.pos + 1
 
-        f.write("xref\n")
-        f.write("0 1\n")
-        f.write("0000000000 65535 f\n")
-
-        pointers.each do |k, v|
-          f.write("#{k.split(' ').first} 1\n")
-          f.write(sprintf("%010d 00000 n\n", v))
-        end
+        write_xref(f, pointers)
 
         f.write("trailer\n")
         f.write("<<\n")
@@ -289,42 +278,17 @@ module Podoff
       f.write(v)
       f.write("\n")
 
-      ptrs = {}
+      pointers = {}
 
       objs.keys.sort.each do |k|
-        ptrs[k] = f.pos + 1
+        pointers[k.split(' ').first.to_i] = f.pos + 1
         f.write(objs[k].source)
         f.write("\n")
       end
 
       xref = f.pos + 1
-      max = objs.keys.inject(-1) { |i, k| [ i, k.split(' ')[0].to_i ].max }
-
-      #f.write("xref\n0 #{max}\n0000000000 65535 f\n")
-      f.write("xref\n0 1\n0000000000 65535 f\n")
-
-      partitions = [ [] ]
-      #
-      (1..max).each do |i|
-        k = "#{i} 0"
-        last = partitions.last
-        if ptrs.has_key?(k)
-          last << i
-        else
-          partitions << [] unless last == []
-        end
-      end
-      #
-      partitions.each do |part|
 
-        f.write("#{part.first} #{part.size}\n")
-
-        part.each do |i|
-          k = "#{i} 0"
-          #f.write(sprintf("%010d 00000 n %% %s\n", ptrs[k], k))
-          f.write(sprintf("%010d 00000 n\n", ptrs[k]))
-        end
-      end
+      write_xref(f, pointers)
 
       f.write("trailer\n")
       f.write("<<\n")
@@ -339,7 +303,27 @@ module Podoff
       f.is_a?(StringIO) ? f.string : nil
     end
 
-    private
+    protected
+
+    def write_xref(f, pointers)
+
+      f.write("xref\n")
+      f.write("0 1\n")
+      f.write("0000000000 65535 f\n")
+
+      pointers
+        .keys
+        .sort
+        .inject([ [] ]) { |ps, k|
+          ps << [] if ps.last != [] && k > ps.last.last + 1
+          ps.last << k
+          ps
+        }
+        .each { |part|
+          f.write("#{part.first} #{part.size}\n")
+          part.each { |k| f.write(sprintf("%010d 00000 n\n", pointers[k])) }
+        }
+    end
 
     def make_stream(&block)
 
@@ -355,7 +339,9 @@ module Podoff
     ATTRIBUTES =
       { type: 'Type', contents: 'Contents', pagenum: 'pdftk_PageNum' }
 
-    def self.extract(doc, sca)
+    def self.extract(doc)
+
+      sca = doc.scanner
 
       re = sca.matched[0..-4].strip
       st = sca.pos - sca.matched.length
@@ -363,50 +349,48 @@ module Podoff
       i = sca.skip_until(/endobj/); return nil unless i
       en = sca.pos - 1
 
-      atts = {}
-      ATTRIBUTES.each do |k, v|
-        sca.pos = st
-        i = sca.skip_until(/\/#{v}\b/); next unless i
-        next if sca.pos > en
-        atts[k] = sca.scan(/ *\/?[^\n\r\/>]+/).strip
-      end
-
-      sca.pos = en
-
-      Podoff::Obj.new(doc, re, st, en, atts)
+      Podoff::Obj.new(doc, re, start_index: st, end_index: en)
     end
 
     attr_reader :document
     attr_reader :ref
     attr_reader :start_index, :end_index
+    attr_reader :stream
     attr_reader :attributes
 
-    def initialize(doc, ref, st, en, atts, source=nil)
+    def initialize(doc, ref, opts={})
 
       @document = doc
       @ref = ref
-      @start_index = st
-      @end_index = en
-      @attributes = atts
-      @source = source
 
-      recompute_attributes if @source.is_a?(String)
-      @source.obj = self if @source.is_a?(Podoff::Stream)
-    end
+      @start_index = opts[:start_index]
+      @end_index = opts[:end_index]
+      @attributes = nil
+      @source = opts[:source]
 
-    def dup(new_doc)
+      @stream = opts[:stream]
+      @stream.obj = self if @stream
 
-      self.class.new(new_doc, ref, start_index, end_index, attributes.dup)
+      recompute_attributes
+      #@source.obj = self if @source.is_a?(Podoff::Stream)
+
+      @document.scanner.pos = @end_index if @document.scanner && @end_index
     end
 
-    def self.create(doc, ref, source)
+    def dup(new_doc)
 
-      self.new(doc, ref, nil, nil, nil, source)
+      self.class.new(
+        new_doc, ref,
+        start_index: start_index, end_index: end_index)
     end
 
+    #def self.create(doc, ref, source)
+    #  self.new(doc, ref, nil, nil, nil, source)
+    #end
+
     def replicate
 
-      self.class.create(document, ref, source.dup)
+      self.class.new(document, ref, source: source.dup)
     end
 
     def to_a
@@ -416,7 +400,7 @@ module Podoff
 
     def source
 
-      @source || @document.source[@start_index..@end_index]
+      @source || (@start_index && @document.source[@start_index..@end_index])
     end
 
     def replica?
@@ -463,14 +447,29 @@ module Podoff
     end
     alias :insert_content :insert_contents
 
+    def to_s
+
+      source || stream.to_s
+    end
+
     protected
 
     def recompute_attributes
 
+      st, en, sca =
+        if @start_index
+          [ @start_index, @end_index, @document.scanner ]
+        elsif @source
+          [ 0, @source.length, ::StringScanner.new(@source) ]
+        end
+
+      return unless sca
+
       @attributes =
         ATTRIBUTES.inject({}) do |h, (k, v)|
-          m = @source.match(/\/#{v}\s+(\/?[^\/\n<>]+)/)
-          h[k] = m[1].strip if m
+          sca.pos = st
+          i = sca.skip_until(/\/#{v}\b/)
+          h[k] = sca.scan(/ *\/?[^\n\r\/>]+/).strip if i && sca.pos < en
           h
         end
     end
@@ -504,8 +503,9 @@ module Podoff
 
     attr_accessor :obj
 
-    def initialize
+    def initialize(obj=nil)
 
+      @obj = obj
       @font = nil
       @content = StringIO.new
     end
@@ -535,7 +535,16 @@ module Podoff
 
     def to_s
 
-      @content.string
+      s = @content.string
+      f = ''
+      if s.length > 98
+        f = ' /Filter /FlateDecode'
+        s = Zlib::Deflate.deflate(s)
+      end
+
+      "#{obj.ref} obj\n" +
+      "<</Length #{s.size}#{f}>>\nstream\n#{s}\nendstream\n" +
+      "endobj"
     end
 
     protected

data/spec/document_spec.rb

@@ -97,6 +97,15 @@ describe Podoff::Document do
 
       expect(d.root).to eq('65 0')
     end
+
+    it 'sports objs with properly recomputed attributes' do
+
+      pa = @d.page(1)
+
+      d = @d.dup
+
+      expect(d.objs[pa.ref].attributes).to eq(pa.attributes)
+    end
   end
 
   context 'additions' do
@@ -120,13 +129,12 @@ describe Podoff::Document do
         expect(fo.ref).to eq('7 0')
 
         expect(fo.source).to eq(
-          '7 0 obj ' +
-          '<< /Type /Font /Subtype /Type1 /BaseFont /Helvetica >> endobj')
+          '7 0 obj <</Type /Font /Subtype /Type1 /BaseFont /Helvetica>> endobj')
 
         s = @d.write(:string)
         d = Podoff.parse(s)
 
-        expect(d.xref).to eq(682)
+        expect(d.xref).to eq(680)
       end
 
       it 'doesn\'t mind a slash in front of the font name' do
@@ -141,8 +149,7 @@ describe Podoff::Document do
         expect(fo.ref).to eq('7 0')
 
         expect(fo.source).to eq(
-          '7 0 obj ' +
-          '<< /Type /Font /Subtype /Type1 /BaseFont /Helvetica >> endobj')
+          '7 0 obj <</Type /Font /Subtype /Type1 /BaseFont /Helvetica>> endobj')
       end
     end
 
@@ -185,15 +192,20 @@ endobj
         expect(st.obj.document).to eq(@d)
         expect(st.obj.ref).to eq('7 0')
 
-        expect(st.obj.source.to_s).to eq(%{
+        expect(st.to_s).to eq(%{
+7 0 obj
+<</Length 97>>
+stream
 BT /Helvetica 35 Tf 10 20 Td (thirty here) Tj ET
 BT /Helvetica 35 Tf 40 50 Td (sixty there) Tj ET
+endstream
+endobj
         }.strip)
 
         d = Podoff.parse(@d.write(:string))
 
-        expect(d.source.index('<< /Length 97 >>')).to eq(618)
-        expect(d.xref).to eq(759)
+        expect(d.source.index('<</Length 97>>')).to eq(618)
+        expect(d.xref).to eq(757)
       end
 
       it 'returns the open stream when no arg given' do
@@ -273,9 +285,9 @@ BT /Helvetica 35 Tf 40 50 Td (sixty there) Tj ET
       s = d.write(:string)
 
       expect(
-        d.write(:string).index(%{
+        s.index(%{
 7 0 obj
-<< /Length 37 >>
+<</Length 37>>
 stream
 BT 10 20 Td (hello open stream) Tj ET
 endstream
@@ -283,6 +295,36 @@ endobj
         }.strip)
       ).to eq(722)
     end
+
+    it 'writes a proper xref table' do
+
+      d = Podoff.load('pdfs/t0.pdf')
+
+      pa = d.re_add(d.page(1))
+      st = d.add_stream
+      st.bt(10, 20, 'hello open stream')
+      pa.insert_contents(st)
+
+      s = d.write(:string)
+
+      expect(s[808..-1].strip).to eq(%{
+xref
+0 1
+0000000000 65535 f
+3 1
+0000000611 00000 n
+7 1
+0000000723 00000 n
+trailer
+<<
+/Prev 414
+/Size 7
+/Root 1 0 R
+>>
+startxref 809
+%%EOF
+      }.strip)
+    end
   end
 
   describe '#rewrite' do

data/spec/spec_helper.rb

@@ -6,6 +6,7 @@
 #
 
 require 'pp'
+require 'ostruct'
 
 require 'podoff'
 

data/spec/stream_spec.rb

@@ -14,7 +14,7 @@ describe Podoff::Stream do
 
     it 'sets the current font' do
 
-      st = Podoff::Stream.new
+      st = Podoff::Stream.new(OpenStruct.new(ref: '1 0'))
 
       st.tf('/Helvetica', 35)
       st.bt(10, 20, 'helvetic')
@@ -22,8 +22,13 @@ describe Podoff::Stream do
       st.bt(10, 50, 'zapfesque')
 
       expect(st.to_s).to eq(%{
+1 0 obj
+<</Length 95>>
+stream
 BT /Helvetica 35 Tf 10 20 Td (helvetic) Tj ET
 BT /ZapfDingbats 21 Tf 10 50 Td (zapfesque) Tj ET
+endstream
+endobj
       }.strip)
     end
   end
@@ -32,18 +37,32 @@ BT /ZapfDingbats 21 Tf 10 50 Td (zapfesque) Tj ET
 
     it 'works' do
 
-      st = Podoff::Stream.new
+      st = Podoff::Stream.new(OpenStruct.new(ref: '1 0'))
       st.bt(10, 20, 'hello world')
 
-      expect(st.to_s).to eq('BT 10 20 Td (hello world) Tj ET')
+      expect(st.to_s).to eq(%{
+1 0 obj
+<</Length 31>>
+stream
+BT 10 20 Td (hello world) Tj ET
+endstream
+endobj
+      }.strip)
     end
 
     it 'escapes the text' do
 
-      st = Podoff::Stream.new
+      st = Podoff::Stream.new(OpenStruct.new(ref: '1 0'))
       st.bt(10, 20, 'hello()world')
 
-      expect(st.to_s).to eq('BT 10 20 Td (hello\(\)world) Tj ET')
+      expect(st.to_s).to eq(%{
+1 0 obj
+<</Length 34>>
+stream
+BT 10 20 Td (hello\\(\\)world) Tj ET
+endstream
+endobj
+      }.strip)
     end
   end
 
@@ -51,17 +70,34 @@ BT /ZapfDingbats 21 Tf 10 50 Td (zapfesque) Tj ET
 
     it 'injects text into the stream' do
 
-      st = Podoff::Stream.new
+      st = Podoff::Stream.new(OpenStruct.new(ref: '1 0'))
       st.bt(10, 20, 'abc')
       st.write("\nBT 25 35 Td (ABC) Tj ET")
       st.bt(30, 40, 'def')
 
       expect(st.to_s).to eq(%{
+1 0 obj
+<</Length 71>>
+stream
 BT 10 20 Td (abc) Tj ET
 BT 25 35 Td (ABC) Tj ET
 BT 30 40 Td (def) Tj ET
+endstream
+endobj
       }.strip)
     end
   end
+
+  describe '#to_s' do
+
+    it 'applies /Filter /FlateDecode if stream.size > 98' do
+
+      st = Podoff::Stream.new(OpenStruct.new(ref: '1 0'))
+      st.write("BT /Helvetica 35 Tf 123 456 Td (Hello Nada) Tj ET\n" * 4)
+
+      expect(st.to_s).to match(
+        /^1 0 obj\n<<\/Length 60 \/Filter \/FlateDecode>>/)
+    end
+  end
 end
 

data/todo.txt

@@ -11,7 +11,7 @@
 
 
 [ ] stop using the st-mark thing
-[ ] recompress idea? uncompress with pdftk, recompress with podoff
+[o] recompress idea? uncompress with pdftk, recompress with podoff
     smaller docs anyway... /Filter /FlatDecode
 
 decompress:

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: podoff
 version: !ruby/object:Gem::Version
-  version: 1.1.0
+  version: 1.1.1
   prerelease: 
 platform: ruby
 authors:
@@ -9,7 +9,7 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2015-10-24 00:00:00.000000000 Z
+date: 2015-10-25 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rake