ushahidi2pdf.rb

require 'rubygems'
$LOAD_PATH.unshift("/Users/chris/git/prawn/lib")
require 'prawn'
require 'json'
require "rubygems"
require "httparty"
require 'crack'
require 'ostruct'
require 'ap'

#1. call "scriptname cache" => cache.json is filled with instance data.
#2. call "scriptname print" => pdf is typeset and output

INSTANCE_URL= "http://haiti.ushahidi.com/api?task=incidents&by=sinceid&resp=json&id="
# INSTANCE_URL= "http://roguegenius.com/africa/api?task=incidents&by=sinceid&resp=json&id="

# You must adjust the limit to reflect the total number of instances.
READ_LIMIT=4040

#lower it for test printing
PRINT_LIMIT=100000

#anything larger than this will be trimmed
MAX_INCIDENT_SIZE= 1500

module PrintingPress

# ========
# = book =
# ========

  class Book
    def initialize
      @book = Prawn::Document.new
      Prawn.debug = true
      @x_pos = ((@book.bounds.width / 2) - 150) 
      @y_pos = ((@book.bounds.height / 2) + 200)
      
    end
  
    def typeset_header
      @book.font("Helvetica", :style => :bold)
      @book.font_size 10
    end
  
    def typeset_body
      @book.font_size 10
      @book.font("Courier", :style => :normal)
    end
  
    def typeset_timestamp
      @book.font_size 6
    end
  
    def print(incidents)
      incidents.each do |i|
        if i.empty? 
          p "empty incident, skipping ..."
        elsif i['incident']['incidentdescription'].length > MAX_INCIDENT_SIZE
          p "incident too big: #{i['incident']['incidentdescription']}"
        elsif i['incident']['incidentid'] == @previous_id
          p "duplicate incident, skipping ..."
        else
          @current_id= i['incident']['incidentid']
          if @current_id.to_i > PRINT_LIMIT
            @book.render_file("book.pdf")
            `open book.pdf`
            return true
          end
          p "printing! This incident: #{@current_id}"          
          @previous_id ||= "unset"
          p "previous incident: #{@previous_id}"
          @book.bounding_box([@x_pos, @y_pos], :width => 300, :height => 500) do  
            typeset_header
            @book.text((i['incident']['incidenttitle']))

            @book.text("\n")
            typeset_body
            @book.text((i['incident']['incidentdescription']).gsub(/IDUshahidi:\W+\d+/, '').gsub(/[\n]+/, "\n"))
            typeset_timestamp
            @book.text("\n")
            @book.text(i['incident']['incidentdate']) unless i['incident']['incidentdate'] == nil
            @book.text(i['incident']['locationname']) unless i['incident']['locationname'] == nil
            @previous_id= @current_id
          end
          @book.start_new_page
        end
      end
      p "done rendering, now printing ..."
      @book.render_file("book.pdf")
      `open book.pdf`
      
    end
  end

# =========
# = cache =
# =========

  class Cache
    def full?
      File.exist?("cache.json")
    end

    def read(filename="cache.json")
      p "... reading cache"
      jsonfile= File.open(filename, "r")
      results= jsonfile.read
      p "about to parse!"
      parsed_results= Crack::JSON.parse(results)
      jsonfile.close
      p "... incidents loaded"
      incidents= parsed_results['incidents']
      return incidents
    end
  
    def discover_file(filename)
      if File.exists?(filename)
        jsonfile=File.open(filename, "a")
      else
        jsonfile= File.new(filename, "w")
      end
      return jsonfile
    end
    
    
    #Worker.fill_cache relies on these two writers.
    def write_json(data, filename= "cache.json")
      jsonfile= discover_file(filename)
      jsonfile.write(JSON.pretty_generate(data))
      jsonfile.write(",")
      jsonfile.close
      p "... cache written"
    end
    def write_text(text, filename= "cache.json")
      jsonfile= discover_file(filename)      
      jsonfile.write(text)
      jsonfile.close
    end
    
  end

# ===========
# = crawler =
# ===========

  class Crawler
    def crawl(url)
      p "starting to crawl: #{url}"
      data= HTTParty.get(url).body
      parsed_data= Crack::JSON.parse(data)
      @incidents= parsed_data["payload"]["incidents"]
      return @incidents
    end
  end

# ==========
# = worker =
# ==========

  class Worker
    def fill_cache
      crawler= PrintingPress::Crawler.new  
      cache= PrintingPress::Cache.new
      sinceid=0
      cache.write_text('{"incidents":[')
      until sinceid > READ_LIMIT do
        # incrementing sinceid to work around API limits
        theurl= "#{INSTANCE_URL}#{sinceid}"
        incidents= crawler.crawl(theurl)
        incidents.each do |json|
          prev_json ||= "none"
          #write the json incident record unless it's the same as the last one
          cache.write_json(json)
          # increment the filter trap
          prev_json= json
          p prev_json
        end
        #be nice
        sleep 1
        sinceid = incidents.last["incident"]["incidentid"].to_i
      end
      p "writing the closing bit"
      cache.write_text('{}]}')
    end
  end
end

# ===========
# = routine =
# ===========

if ARGV[0] == "cache"
  worker=PrintingPress::Worker.new
  cache=PrintingPress::Cache.new
  if cache.full?
    p "looks like your cache has data in it. Delete it or print."
  else
    worker.fill_cache
  end  
elsif ARGV[0] == "print"
  book= PrintingPress::Book.new
  cache= PrintingPress::Cache.new
  book.print(incidents= cache.read)
else
  p "usage: ruby ushahidi2pdf.rb [cache|print]"
end