#!/usr/bin/env ruby require 'rubygems' require 'optparse' require 'socket' require 'biodiversity' # Get sockets from stdlib DEFAULT_PORT = 4334 RUBY_VERSION_INT = RUBY_VERSION.split('.')[0..1].join('').to_i OPTIONS = { output: 'json', canonical_with_rank: false, port: DEFAULT_PORT } options = {} ARGV.options do |opts| script_name = File.basename($0) opts.banner = "Usage: #{script_name} [options]" opts.separator '' opts.on('-r', '--canonical_with_rank', 'Adds infraspecies rank to canonical forms' ) { |rank| options[:canonical_with_rank] = rank } opts.separator '' opts.on('-o', '--output=output', String, 'Specifies the type of the output: json - parsed results in json canonical - canonical form only', 'Default: json') { |output| options[:output] = output } opts.separator '' opts.on('-p', '--port=port', String, 'Specifies the port number', "Default: #{DEFAULT_PORT}") { |port| options[:port] = port } opts.separator '' opts.on('-h', '--help', 'Show this help message.') { puts opts; exit } opts.parse! end OPTIONS[:output] = options[:output] if ['canonical'].include?(options[:output]) OPTIONS[:port] = options[:port].to_i if options[:port].to_i > 0 OPTIONS[:canonical_with_rank] = !!options[:canonical_with_rank] def get_output(name_string, parser) begin parsed = parser.parse(name_string) rescue parsed = ScientificNameParser::FAILED_RESULT.(name_string) end output = OPTIONS[:output] return parsed.to_json if output == 'json' parsed[:scientificName][:canonical].to_s end puts "Running parser service on port %s, output type is '%s'" % [OPTIONS[:port], OPTIONS[:output]] opts = {} opts = {canonical_with_rank: true} if OPTIONS[:canonical_with_rank] parser = ScientificNameParser.new(opts) server = TCPServer.open(OPTIONS[:port]) # Socket to listen on a port loop do # Servers run forever Thread.start(server.accept) do |client| puts 'opening client' count = 0 while a = client.readline rescue nil count += 1 puts "parsed %s'th name" % count if count % 1000 == 0 a.force_encoding('utf-8') if a && RUBY_VERSION_INT >= 19 if ['end','exit','q', '.'].include? a.strip client.close break end out = get_output(a, parser).strip client.write(out + "\n") rescue break end puts 'closing client' client.close end end