class Aggkit::ChildProcess::JRuby::Process
Public Class Methods
new(args)
click to toggle source
Calls superclass method
Aggkit::ChildProcess::AbstractProcess::new
# File lib/aggkit/childprocess/jruby/process.rb, line 7 def initialize(args) super(args) @pumps = [] end
Public Instance Methods
exited?()
click to toggle source
# File lib/aggkit/childprocess/jruby/process.rb, line 17 def exited? return true if @exit_code assert_started @exit_code = @process.exitValue stop_pumps true rescue java.lang.IllegalThreadStateException => ex log(ex.class => ex.message) false ensure log(:exit_code => @exit_code) end
io()
click to toggle source
# File lib/aggkit/childprocess/jruby/process.rb, line 13 def io @io ||= JRuby::IO.new end
pid()
click to toggle source
Only supported in JRuby
on a Unix
operating system, thanks to limitations in Java's classes
@return [Integer] the pid of the process after it has started @raise [NotImplementedError] when trying to access pid on non-Unix platform
# File lib/aggkit/childprocess/jruby/process.rb, line 57 def pid if @process.getClass.getName != "java.lang.UNIXProcess" raise NotImplementedError, "pid is only supported by JRuby child processes on Unix" end # About the best way we can do this is with a nasty reflection-based impl # Thanks to Martijn Courteaux # http://stackoverflow.com/questions/2950338/how-can-i-kill-a-linux-process-in-java-with-sigkill-process-destroy-does-sigter/2951193#2951193 field = @process.getClass.getDeclaredField("pid") field.accessible = true field.get(@process) end
stop(timeout = nil)
click to toggle source
# File lib/aggkit/childprocess/jruby/process.rb, line 32 def stop(timeout = nil) assert_started @process.destroy wait # no way to actually use the timeout here.. end
wait()
click to toggle source
# File lib/aggkit/childprocess/jruby/process.rb, line 39 def wait if exited? exit_code else @process.waitFor stop_pumps @exit_code = @process.exitValue end end
Private Instance Methods
create_stdin()
click to toggle source
# File lib/aggkit/childprocess/jruby/process.rb, line 133 def create_stdin output_stream = @process.getOutputStream stdin = output_stream.to_io stdin.sync = true stdin.instance_variable_set(:@childprocess_java_stream, output_stream) class << stdin # The stream provided is a BufferedeOutputStream, so we # have to flush it to make the bytes flow to the process def __childprocess_flush__ @childprocess_java_stream.flush end [:flush, :print, :printf, :putc, :puts, :write, :write_nonblock].each do |m| define_method(m) do |*args| super(*args) self.__childprocess_flush__ end end end stdin end
launch_process(&blk)
click to toggle source
# File lib/aggkit/childprocess/jruby/process.rb, line 72 def launch_process(&blk) pb = java.lang.ProcessBuilder.new(@args) pb.directory java.io.File.new(@cwd || Dir.pwd) set_env pb.environment begin @process = pb.start rescue java.io.IOException => ex raise LaunchError, ex.message end setup_io end
redirect(input, output)
click to toggle source
# File lib/aggkit/childprocess/jruby/process.rb, line 103 def redirect(input, output) if output.nil? input.close return end @pumps << Pump.new(input, output.to_outputstream).run end
set_env(env)
click to toggle source
# File lib/aggkit/childprocess/jruby/process.rb, line 116 def set_env(env) merged = ENV.to_hash @environment.each { |k, v| merged[k.to_s] = v } merged.each do |k, v| if v env.put(k, v.to_s) elsif env.has_key? k env.remove(k) end end removed_keys = env.key_set.to_a - merged.keys removed_keys.each { |k| env.remove(k) } end
setup_io()
click to toggle source
# File lib/aggkit/childprocess/jruby/process.rb, line 87 def setup_io if @io redirect(@process.getErrorStream, @io.stderr) redirect(@process.getInputStream, @io.stdout) else @process.getErrorStream.close @process.getInputStream.close end if duplex? io._stdin = create_stdin else @process.getOutputStream.close end end
stop_pumps()
click to toggle source
# File lib/aggkit/childprocess/jruby/process.rb, line 112 def stop_pumps @pumps.each { |pump| pump.stop } end