[Feature] How do you pipe?

[DevilDimon]

I'm trying the library out and essentially my question is how to execute the equivalent of the following:

grep 'text' . | wc -l

using this library without manually collecting the output of each step.

The run function signature only appears to accept one subprogram name as an argument to launch a single process (which is understandable).

The current outputs/inputs are types like String or Data, so it's not immediately clear how to 'pipe' the output of one subprocess to another, without resorting to doing it manually via an intermediate String or something.
I suppose all the shells on all supported platforms support piping so there could be a way to simplify this common task? (or maybe there already is?)

[cmcgee1024]

You can check out the implementation in #147 to see how you can set up a process pipeline manually. Basically, there is FileDescriptor.pipe() to set up read/write id's, and those can be provided to the input/output of run() using the .fileDescriptor() construction. Give each process its own Task, and you can block on the last one to get the final output in whichever form works best for you.