Fix packaging and some code.

.gitignore

@@ -1 +1,4 @@
 bin/*.pdf
+
+*.sw*
+.DS_Store

Gemfile

@@ -0,0 +1,3 @@
+source 'https://rubygems.org'
+
+gemspec

Gemfile.lock

@@ -0,0 +1,53 @@
+PATH
+  remote: .
+  specs:
+    pdf-extract (0.1.1)
+      commander (>= 4.0.4)
+      json (>= 1.5.1)
+      nokogiri (>= 1.5.0)
+      pdf-reader (= 1.3.3)
+      prawn (>= 0.11.1)
+      rb-libsvm (>= 1.1.3)
+      sqlite3 (>= 1.3.4)
+
+GEM
+  remote: https://rubygems.org/
+  specs:
+    Ascii85 (1.0.2)
+    afm (0.2.0)
+    bson (1.9.2)
+    bson_ext (1.9.2)
+      bson (~> 1.9.2)
+    commander (4.1.5)
+      highline (~> 1.6.11)
+    hashery (2.1.1)
+    highline (1.6.20)
+    json (1.8.1)
+    mini_portile (0.5.2)
+    mongo (1.9.2)
+      bson (~> 1.9.2)
+    nokogiri (1.6.0)
+      mini_portile (~> 0.5.0)
+    pdf-reader (1.3.3)
+      Ascii85 (~> 1.0.0)
+      afm (~> 0.2.0)
+      hashery (~> 2.0)
+      ruby-rc4
+      ttfunk
+    prawn (0.12.0)
+      pdf-reader (>= 0.9.0)
+      ttfunk (~> 1.0.2)
+    rake (10.1.0)
+    rb-libsvm (1.1.3)
+    ruby-rc4 (0.1.5)
+    sqlite3 (1.3.8)
+    ttfunk (1.0.3)
+
+PLATFORMS
+  ruby
+
+DEPENDENCIES
+  bson_ext (>= 1.9.2)
+  mongo (>= 1.9.2)
+  pdf-extract!
+  rake (>= 10.1.0)

Rakefile

@@ -0,0 +1,2 @@
+require_relative 'tasks/assign.rb'
+require_relative 'tasks/train.rb'

bin/pdf-extract

@@ -2,8 +2,8 @@
 
 require 'commander/import'
 require 'json'
-require_relative '../lib/pdf-extract'
-require_relative '../lib/references/resolve'
+require_relative '../lib/pdf-extract.rb'
+require_relative '../lib/pdf/extract/references/resolve.rb'
 
 program :name, 'pdf-extract'
 program :version, '0.0.1'
@@ -146,7 +146,7 @@ command :settings do |c|
     else
       say "<%= color('AT DEFAULT:', BOLD) %>\n\n"
       s.unmodified.each_pair { |k, v| say "#{k}:\t#{v}" }
-    
+
       say "\n<%= color('OVERRIDDEN:', BOLD) %>\n\n"
       s.modified.each_pair { |k, v| say "#{k}:\t#{v}   (#{s.agent(k)})" }
     end

catalog.json

@@ -0,0 +1 @@
+{}

lib/pdf-extract.rb

@@ -1,75 +1 @@
-require_relative 'pdf'
-require_relative 'model/characters'
-require_relative 'model/chunks'
-require_relative 'model/regions'
-require_relative 'analysis/titles'
-require_relative 'analysis/margins'
-require_relative 'analysis/zones'
-require_relative 'analysis/columns'
-require_relative 'analysis/sections'
-require_relative 'references/references'
-require_relative 'references/resolved_references'
-require_relative 'view/pdf_view'
-require_relative 'view/xml_view'
-
-module PdfExtract
-
-  @views = {}
-
-  @parsers = []
-
-  def self.add_view name, view_class
-    @views[name] = view_class
-  end
-
-  def self.add_parser parser_class
-    @parsers << parser_class
-  end
-
-  def self.parse filename, &block
-    pdf = Pdf.new
-
-    @parsers.each do |p|
-      p.include_in pdf
-    end
-
-    yield pdf
-
-    pdf.spatial_calls.each do |spatial_call|
-      name = spatial_call[:name]
-      receiver = Receiver.new pdf
-      pdf.spatial_builders[name].call receiver
-      receiver.invoke_calls filename, pdf.spatial_options[name]
-    end
-
-    pdf
-  end
-
-  def self.view_class short_name
-    @views[short_name]
-  end
-
-  def self.view filename, options = {}, &block
-    pdf = parse filename, &block
-    view_class(options[:as]).new(pdf, filename).render options
-  end
-
-  def self.init
-    add_parser Characters
-    add_parser Chunks
-    add_parser Regions
-    add_parser Titles
-    add_parser Margins
-    add_parser Zones
-    add_parser Columns
-    add_parser Sections
-    add_parser References
-    add_parser ResolvedReferences
-
-    add_view :pdf, PdfView
-    add_view :xml, XmlView
-  end
-
-  init
-
-end
+require_relative 'pdf/extract.rb'

lib/pdf/extract.rb

@@ -0,0 +1,75 @@
+require_relative 'extract/pdf.rb'
+require_relative 'extract/model/characters.rb'
+require_relative 'extract/model/chunks.rb'
+require_relative 'extract/model/regions.rb'
+require_relative 'extract/analysis/titles.rb'
+require_relative 'extract/analysis/margins.rb'
+require_relative 'extract/analysis/zones.rb'
+require_relative 'extract/analysis/columns.rb'
+require_relative 'extract/analysis/sections.rb'
+require_relative 'extract/references/references.rb'
+require_relative 'extract/references/resolved_references.rb'
+require_relative 'extract/view/pdf_view.rb'
+require_relative 'extract/view/xml_view.rb'
+
+module PdfExtract
+
+  @views = {}
+
+  @parsers = []
+
+  def self.add_view name, view_class
+    @views[name] = view_class
+  end
+
+  def self.add_parser parser_class
+    @parsers << parser_class
+  end
+
+  def self.parse filename, &block
+    pdf = Pdf.new
+
+    @parsers.each do |p|
+      p.include_in pdf
+    end
+
+    yield pdf
+
+    pdf.spatial_calls.each do |spatial_call|
+      name = spatial_call[:name]
+      receiver = Receiver.new pdf
+      pdf.spatial_builders[name].call receiver
+      receiver.invoke_calls filename, pdf.spatial_options[name]
+    end
+
+    pdf
+  end
+
+  def self.view_class short_name
+    @views[short_name]
+  end
+
+  def self.view filename, options = {}, &block
+    pdf = parse filename, &block
+    view_class(options[:as]).new(pdf, filename).render options
+  end
+
+  def self.init
+    add_parser Characters
+    add_parser Chunks
+    add_parser Regions
+    add_parser Titles
+    add_parser Margins
+    add_parser Zones
+    add_parser Columns
+    add_parser Sections
+    add_parser References
+    add_parser ResolvedReferences
+
+    add_view :pdf, PdfView
+    add_view :xml, XmlView
+  end
+
+  init
+
+end

lib/font_metrics.rb → lib/pdf/extract/font_metrics.rb

@@ -1,5 +1,5 @@
 # Taken from pdfminer, in turn extracted from AFM files at:
-# 
+#
 #   http://www.ctan.org/tex-archive/fonts/adobe/afm/
 #
 
@@ -39,12 +39,13 @@ def initialize font
           @@base_fonts[base_font][:Widths].fetch(c.codepoints.first, 0)
         }
       else
-        @ascent = font.ascent
-        @descent = font.descent
-        @bbox = font.bbox
+        font_descriptor = font.font_descriptor
+        @ascent = font_descriptor.ascent
+        @descent = font_descriptor.descent
+        @bbox = font_descriptor.font_bounding_box
         @glyph_width_lookup = proc do |c|
           begin
-            font.glyph_width c.codepoints.first
+            font.glyph_width(c.codepoints.first) || 0
           rescue TypeError => e
             # It seems some fonts don't have a first char attribute in their
             # descriptor and this causes problems for pdf-reader.

lib/language.rb → lib/pdf/extract/language.rb

@@ -1,4 +1,4 @@
-require_relative "names"
+require_relative 'names.rb'
 
 module PdfExtract::Language