gistfile1.txt

#This converts a csv file to a slightly different csv. 

#usage: ruby csv_to_csv.rb <csv input filename> <csv output filename> 

# leaving out the output filename will result in a csv file with the same name as the input, and will have 'converted' in front of it.

require 'rubygems'
require 'csv'
require 'json'

raise Exception, 'you must provide a csv' unless ARGV[0]
file_in = ARGV[0]
file_out = ARGV[1] || "converted_#{file_in}"

headers = []
csv_arr = []

CSV.open(file_in, "r") do |lines|
  lines.each do |line|
    print "face_count: ", line[6], "\tface_count:confidence: ", line[7], "\timage_url: ", line[13], "\n"
    if headers.empty?
      headers = ["face_count", "face_count_confidence", "face_coordinates", "face_confidence", "image_url"]
    else
      face_count = line[6].to_i
      if face_count > 0
        image_annotation = JSON.parse(line[8])
        coor_confs = image_annotation["features"].map{|feature| {coordinates: feature["geometry"]["coordinates"], confidence: feature["properties"]["confidence"]}}.sort{|a,b| a[:confidence] <=> b[:confidence]}.reverse.take(face_count)
        coor_confs.each do |coor_conf|
          csv_arr << [face_count, line[7], coor_conf[:coordinates], coor_conf[:confidence], line[13]]
        end
      end
    end
  end
end

# write out a csv file with a header that is the union of all the json keys

CSV.open(file_out, "w", :headers => headers, :write_headers => true) do |csv|
  csv_arr.each do |line|
    csv << line
  end
end