pdf-extract 0.0.1

data/bin/assign.rb

@@ -0,0 +1,72 @@
+#!/usr/bin/env ruby
+
+require "json"
+require "highline"
+require_relative "../lib/pdf-extract"
+
+class Assign
+  
+  def initialize features, categories
+    # The features of the data we wish to learn from.
+    @features = features
+
+    # The possible categorizations of data items.
+    @categories = categories
+
+    @hl = HighLine.new
+  end
+
+  def data_entry category, section
+    entry = {}
+    @features.each { |f| entry[f] = section[f] }
+    entry[:file] = File.split(ARGV[0]).last
+    entry[:category] = category
+    entry[:word_count] = section[:word_count]
+
+    puts entry
+  end
+  
+  def with_category
+    @hl.choose do |menu|
+      menu.prompt = "Category?"
+      @categories.each do |category|
+        menu.choice(category) do
+          yield category
+        end
+      end
+    end
+  end
+
+end
+
+# Display each section and ask if it is ref or non-ref.
+features = [:letter_ratio, :name_ratio, :year_ratio, :cap_ratio]
+categories = [:reference, :body, :mix, :none]
+data = []
+pdf = PdfExtract.parse(ARGV[0]) { |pdf| pdf.sections }
+assign = Assign.new(features, categories)
+  
+pdf[:sections].each do |section|
+  if section[:word_count] < 5
+    # Low word count sections are definitely not ref sections.
+    # Don't show them to the user.
+    data << assign.data_entry(:none, section)
+  else
+    puts ""
+    puts "-----"
+    puts ""
+    puts PdfExtract::Spatial.get_text_content(section)
+    puts ""
+    
+    assign.with_category { |category| data << assign.data_entry(category, section) }
+  end
+end
+
+puts data
+
+# File.open(ARGV[1]).write(data.to_json)
+
+
+
+
+

data/bin/config.json

@@ -0,0 +1,4 @@
+{
+  "body_ratio": 15, 
+  "line_slop": -5
+}