podoff 1.0.0 → 1.1.0

data/CHANGELOG.txt

@@ -2,6 +2,13 @@
 = podoff CHANGELOG.txt
 
 
+== podoff 1.1.0  released 2015-10-25
+
+- more tolerant at parsing (StringScanner)
+- bin/podoff
+- Document#rewrite(path)
+
+
 == podoff 1.0.0  released 2015-10-23
 
 - leverage incremental updates

data/lib/podoff.rb

@@ -23,10 +23,13 @@
 # Made in Japan.
 #++
 
+require 'strscan'
+require 'stringio'
+
 
 module Podoff
 
-  VERSION = '1.0.0'
+  VERSION = '1.1.0'
 
   def self.load(path, encoding='iso-8859-1')
 
@@ -38,13 +41,6 @@ module Podoff
     Podoff::Document.new(s)
   end
 
-  #OBJ_ATTRIBUTES =
-  #  { type: 'Type', subtype: 'Subtype',
-  #    parent: 'Parent', kids: 'Kids', contents: 'Contents', annots: 'Annots',
-  #    pagenum: 'pdftk_PageNum' }
-  OBJ_ATTRIBUTES =
-    { type: 'Type', contents: 'Contents', pagenum: 'pdftk_PageNum' }
-
   class Document
 
     def self.load(path, encoding='iso-8859-1')
@@ -58,6 +54,7 @@ module Podoff
     end
 
     attr_reader :source
+    attr_reader :version
     attr_reader :xref
     attr_reader :objs
     attr_reader :obj_counters
@@ -68,9 +65,10 @@ module Podoff
     def initialize(s)
 
       fail ArgumentError.new('not a PDF file') \
-        unless s.match(/\A%PDF-\d+\.\d+\n/)
+        unless s.match(/\A%PDF-\d+\.\d+\s/)
 
       @source = s
+      @version = nil
       @xref = nil
       @objs = {}
       @obj_counters = {}
@@ -78,47 +76,41 @@ module Podoff
 
       @additions = {}
 
-      index = 0
-      matches = {}
-      #
-      loop do
-
-        matches[:obj] ||= s.match(/^(\d+ \d+) obj\b/, index)
-        matches[:endobj] ||= s.match(/\bendobj\b/, index)
-        #
-        OBJ_ATTRIBUTES.each do |k, v|
-          matches[k] ||= s.match(/\/#{v} (\/?[^\/\n<>]+)/, index)
-        end
-        #
-        matches[:startxref] ||= s.match(/\bstartxref\s+(\d+)\s*%%EOF/, index)
-
-        objm = matches[:obj]
-        sxrm = matches[:startxref]
-
-        break unless sxrm || objm
+      sca = ::StringScanner.new(s)
+      @version = sca.scan(/%PDF-\d+\.\d+/)
 
-        fail ArgumentError.new('failed to find "startxref"') unless sxrm
+      loop do
 
-        @root = nil if @root && index > @root.offset(0).last
-        @root ||= s.match(/\/Root (\d+ \d+) R\b/, index)
+        i = sca.skip_until(
+          /(startxref\s+\d+|\d+\s+\d+\s+obj|\/Root\s+\d+\s+\d+\s+R)/)
 
-        sxri = sxrm.offset(0).first
-        obji = objm ? objm.offset(0).first : sxri + 1
+        m = sca.matched
+        break unless m
 
-        if obji < sxri
-          obj = Podoff::Obj.extract(self, matches)
+        if m[0] == 's'
+          @xref = m.split(' ').last.to_i
+        elsif m[0] == '/'
+          @root = extract_ref(m)
+        else
+          obj = Podoff::Obj.extract(self, sca)
           @objs[obj.ref] = obj
           @obj_counters[obj.ref] = (@obj_counters[obj.ref] || 0) + 1
-          index = obj.end_index + 1
-        else
-          @xref = sxrm[1].to_i
-          index = sxrm.offset(0).last + 1
-          matches.delete(:startxref)
         end
       end
 
-      fail ArgumentError.new('found no /Root') unless @root
-      @root = @root[1]
+      if @root == nil
+        sca.pos = 0
+        loop do
+          i = sca.skip_until(/\/Root\s+\d+\s+\d+\s+R/)
+          break unless sca.matched
+          @root = extract_ref(sca.matched)
+        end
+      end
+    end
+
+    def extract_ref(s)
+
+      s.gsub(/\s+/, ' ').gsub(/[^0-9 ]+/, '').strip
     end
 
     def updated?
@@ -232,7 +224,12 @@ module Podoff
 
     def write(path)
 
-      f = (path == :string) ? StringIO.new : File.open(path, 'wb')
+      f =
+        case path
+          when :string, '-' then StringIO.new
+          when String then File.open(path, 'wb')
+          else path
+        end
 
       f.write(@source)
 
@@ -274,9 +271,72 @@ module Podoff
         f.write("%%EOF\n")
       end
 
-      f.close
+      f.close if path.is_a?(String) || path.is_a?(Symbol)
 
-      path == :string ? f.string : nil
+      f.is_a?(StringIO) ? f.string : nil
+    end
+
+    def rewrite(path=:string)
+
+      f =
+        case path
+          when :string, '-' then StringIO.new
+          when String then File.open(path, 'wb')
+          else path
+        end
+
+      v = source.match(/%PDF-\d+\.\d+/)[0]
+      f.write(v)
+      f.write("\n")
+
+      ptrs = {}
+
+      objs.keys.sort.each do |k|
+        ptrs[k] = 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
+
+      f.write("trailer\n")
+      f.write("<<\n")
+      f.write("/Size #{objs.size}\n")
+      f.write("/Root #{root} R\n")
+      f.write(">>\n")
+      f.write("startxref #{xref}\n")
+      f.write("%%EOF\n")
+
+      f.close if path.is_a?(String) || path.is_a?(Symbol)
+
+      f.is_a?(StringIO) ? f.string : nil
     end
 
     private
@@ -292,24 +352,26 @@ module Podoff
 
   class Obj
 
-    def self.extract(doc, matches)
+    ATTRIBUTES =
+      { type: 'Type', contents: 'Contents', pagenum: 'pdftk_PageNum' }
 
-      re = matches[:obj][1]
-      st = matches[:obj].offset(0).first
-      en = matches[:endobj].offset(0).last - 1
+    def self.extract(doc, sca)
 
-      atts = {}
+      re = sca.matched[0..-4].strip
+      st = sca.pos - sca.matched.length
 
-      OBJ_ATTRIBUTES.keys.each do |k|
-        m = matches[k]
-        if m && m.offset(0).last < en
-          atts[k] = m[1].strip
-          matches.delete(k)
-        end
+      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
 
-      matches.delete(:obj)
-      matches.delete(:endobj)
+      sca.pos = en
 
       Podoff::Obj.new(doc, re, st, en, atts)
     end
@@ -373,62 +435,6 @@ module Podoff
       r ? r.to_i : nil
     end
 
-#    def parent
-#
-#      r = @attributes[:parent]
-#      r ? r[0..-2].strip : nil
-#    end
-#
-#    def kids
-#
-#      r = @attributes[:kids]
-#      (r || '').split(/[\[\]R]/).collect(&:strip).reject(&:empty?)
-#    end
-#
-#    def contents
-#
-#      r = @attributes[:contents]
-#      (r || '').split(/[\[\]R]/).collect(&:strip).reject(&:empty?)
-#    end
-
-#    def add_annotation(ref)
-#
-#      if annots = @attributes[:annots]
-#        fail "implement me!"
-#      else
-#        i = @source.index('/Type ')
-#        @source.insert(i, "/Annots [#{ref} R]\n")
-#      end
-#      recompute_attributes
-#    end
-
-#    def add_free_text(x, y, text, font, size)
-#
-#      fail ArgumentError.new('target is not a page') unless type == '/Page'
-#
-#      nref = document.new_ref
-#
-#      s = [
-#        "#{nref} obj <<",
-#        "/Type /Annot",
-#        "/Subtype /FreeText",
-#        "/Da (/F1 70 Tf 0 100 Td)",
-#        "/Rect [0 0 500 600]",
-#        "/Contents (#{text})",
-#        ">>",
-#        "endobj"
-#      ].join("\n")
-#      anno = Obj.create(document, nref, s)
-#
-#      page = self.replicate
-#      page.add_annotation(nref)
-#
-#      document.add(anno)
-#      document.add(page)
-#
-#      anno
-#    end
-
     def insert_font(nick, obj_or_ref)
 
       fail ArgumentError.new("target '#{ref}' not a replica") \
@@ -462,9 +468,9 @@ module Podoff
     def recompute_attributes
 
       @attributes =
-        OBJ_ATTRIBUTES.inject({}) do |h, (k, v)|
-          m = @source.match(/\/#{v} (\/?[^\/\n<>]+)/)
-          h[k] = m[1] if m
+        ATTRIBUTES.inject({}) do |h, (k, v)|
+          m = @source.match(/\/#{v}\s+(\/?[^\/\n<>]+)/)
+          h[k] = m[1].strip if m
           h
         end
     end
@@ -481,7 +487,7 @@ module Podoff
 
       fail ArgumentError.new("obj not replicated") unless @source
 
-      pkey = OBJ_ATTRIBUTES[key]
+      pkey = ATTRIBUTES[key]
 
       if v = @attributes[key]
         v = concat(v, ref)
@@ -504,10 +510,6 @@ module Podoff
       @content = StringIO.new
     end
 
-    #def document; obj.document; end
-    #def ref; obj.ref; end
-    #def source; self; end
-
     def tf(font_name, font_size)
 
       n = font_name[0] == '/' ? font_name[1..-1] : font_name

data/spec/core_spec.rb

@@ -17,8 +17,8 @@ describe Podoff do
       d = Podoff.load('pdfs/t0.pdf')
 
       expect(d.class).to eq(Podoff::Document)
-      expect(d.xref).to eq(414)
       expect(d.objs.keys).to eq([ '1 0', '2 0', '3 0', '4 0', '5 0', '6 0' ])
+      expect(d.xref).to eq(414)
 
       #pp d.objs.values.collect(&:to_a)
 
@@ -35,6 +35,8 @@ describe Podoff do
         [ 1, 1, 1, 1, 1, 1 ])
 
       expect(d.root).to eq('1 0')
+
+      expect(d.pages.size).to eq(1)
     end
 
     it 'loads a PDF document' do
@@ -48,6 +50,8 @@ describe Podoff do
       expect(d.objs.keys).to include('273 0')
 
       expect(d.root).to eq('65 0')
+
+      expect(d.pages.size).to eq(3)
     end
 
     it 'loads a PDF document with incremental updates' do
@@ -66,6 +70,25 @@ describe Podoff do
       expect(d.root).to eq('1 0')
     end
 
+    it 'loads a [re]compressed PDF documents' do
+
+      d = Podoff.load('pdfs/qdocument0.pdf')
+
+      expect(d.class).to eq(Podoff::Document)
+      expect(d.xref).to eq(1612815)
+      expect(d.objs.size).to eq(273)
+
+      expect(d.root).to eq('1 0')
+
+      #d.objs.each do |ref, o|
+      #  p [ o.ref, o.attributes ]
+      #end
+
+      expect(d.pages.size).to eq(3)
+      expect(d.pages.first.attributes[:pagenum]).to eq('1')
+      expect(d.objs['46 0'].attributes[:type]).to eq('/Annot')
+    end
+
     it 'rejects items that are not PDF documents' do
 
       expect {

data/spec/document_spec.rb

@@ -77,43 +77,6 @@ describe Podoff::Document do
     end
   end
 
-  describe '#write' do
-
-    it 'writes the document to a given path' do
-
-      @d.write('tmp/out.pdf')
-
-      s = File.open('tmp/out.pdf', 'r:iso8859-1') { |f| f.read }
-      lines = s.split("\n")
-
-      expect(lines.first).to match(/^%PDF-1.7$/)
-      expect(lines.last).to match(/^%%EOF$/)
-    end
-
-    it 'writes open streams as well' 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(
-        d.write(:string).index(%{
-7 0 obj
-<< /Length 37 >>
-stream
-BT 10 20 Td (hello open stream) Tj ET
-endstream
-endobj
-        }.strip)
-      ).to eq(722)
-    end
-  end
-
   describe '#dup' do
 
     it 'produces a shallow copy of the document' do
@@ -272,6 +235,107 @@ BT /Helvetica 35 Tf 40 50 Td (sixty there) Tj ET
         expect(re.source).to eq(pa.source)
         expect(re.source).not_to equal(pa.source)
       end
+
+      it 'recomputes the attributes correctly' do
+
+        d = Podoff.load('pdfs/qdocument0.pdf')
+
+        pa = d.re_add(d.page(1))
+
+        expect(pa.attributes).to eq(
+          { type: '/Page', contents: '151 0 R', pagenum: '1' })
+      end
+    end
+  end
+
+  describe '#write' do
+
+    it 'writes the document to a given path' do
+
+      @d.write('tmp/out.pdf')
+
+      s = File.open('tmp/out.pdf', 'r:iso8859-1') { |f| f.read }
+      lines = s.split("\n")
+
+      expect(lines.first).to match(/^%PDF-1.7$/)
+      expect(lines.last).to match(/^%%EOF$/)
+    end
+
+    it 'writes open streams as well' 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(
+        d.write(:string).index(%{
+7 0 obj
+<< /Length 37 >>
+stream
+BT 10 20 Td (hello open stream) Tj ET
+endstream
+endobj
+        }.strip)
+      ).to eq(722)
+    end
+  end
+
+  describe '#rewrite' do
+
+    it 'rewrites a document in one go' do
+
+      d = Podoff.load('pdfs/t2.pdf')
+
+      s = d.rewrite(:string)
+
+      expect(s.strip).to eq(%{
+%PDF-1.4
+1 0 obj <</Type /Catalog /Pages 2 0 R>>
+endobj
+2 0 obj <</Type /Pages /Kids [3 0 R] /Count 1>>
+endobj
+3 0 obj <</Type /Page /Parent 2 0 R /Resources 4 0 R /MediaBox [0 0 500 800] /Contents [6 0 R 7 0 R]>>
+endobj
+4 0 obj <</Font <</F1 5 0 R>>>>
+endobj
+5 0 obj <</Type /Font /Subtype /Type1 /BaseFont /Helvetica>>
+endobj
+6 0 obj
+<</Length 44>>
+stream
+BT /F1 24 Tf 175 720 Td (Hello Nadaa!)Tj ET
+endstream
+endobj
+7 0 obj
+<</Length 44>>
+stream
+BT /F1 24 Tf 175 520 Td (Smurf Megane)Tj ET
+endstream
+endobj
+xref
+0 1
+0000000000 65535 f
+1 7
+0000000010 00000 n
+0000000057 00000 n
+0000000112 00000 n
+0000000222 00000 n
+0000000261 00000 n
+0000000329 00000 n
+0000000420 00000 n
+trailer
+<<
+/Size 7
+/Root 1 0 R
+>>
+startxref 511
+%%EOF
+      }.strip)
     end
   end
 end

data/spec/obj_spec.rb

@@ -232,6 +232,21 @@ endobj
         expect(pa.source).to match(/\/Font\s+<<\s+\/MyHelv #{fo.ref} R\s+/)
       end
     end
+
+    describe '#add_to_attribute' do
+
+      it 'adds to a list of references' do
+
+        d = Podoff.load('pdfs/qdocument0.pdf')
+
+        o = d.re_add('56 0')
+
+        o.send(:add_to_attribute, :contents, '9999 0')
+
+        expect(o.attributes).to eq(
+          { type: '/Page', contents: '[151 0 R 9999 0 R]', pagenum: '1' })
+      end
+    end
   end
 end
 

data/todo.txt

@@ -1,13 +1,21 @@
 
-[ ] stop using the st-mark thing
-[ ] doc.add_stream { |st| st.bt ... }
-
+[o] implement Document#rewrite
+[o] doc.add_stream { |st| st.bt ... }
 [o] st = d.addstream
     st.tf '/helv', 12
     st.bt 5, 6, "hello"
     d.write('out.pdf') # closes the stream...
 [ ] fail if insert_content or insert_font on an unclosed stream obj
 
+[o] Document.write('-') (stdout inspiration)
+
+
+[ ] stop using the st-mark thing
 [ ] recompress idea? uncompress with pdftk, recompress with podoff
-    smaller docs anyway...
+    smaller docs anyway... /Filter /FlatDecode
+
+decompress:
+  qpdf --qdf --object-streams=disable in.pdf out.pdf
+recompress:
+  qpdf in.pdf out.pdf
 

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: podoff
 version: !ruby/object:Gem::Version
-  version: 1.0.0
+  version: 1.1.0
   prerelease: 
 platform: ruby
 authors:
@@ -9,7 +9,7 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2015-10-23 00:00:00.000000000 Z
+date: 2015-10-24 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rake