GitHub

Copying Processes

This page documents the process-copy helpers.

A process copy is rebuilt from saved construction data, not by copying the live context directly. This matters when a context was initialized from external storage, buffers, or data views.

If you want to run many independent jobs through reusable processes on Julia threads, see Threaded Process Managers.

Why Copy Instead of deepcopy

Process contexts are often built from Init(...) values that point at external storage, buffers, or data views. A raw deepcopy of the live context can therefore copy the wrong thing or preserve sharing that should be rebuilt per process.

The copying helpers work from the initialized loop algorithm and the normal lifecycle init pipeline instead:

  • copy the stored loop algorithm recipe,
  • replace selected inputs and overrides,
  • initialize a fresh context for each copy.

The saved recipe is the initialized loop algorithm: it stores the resolved algorithm, persistent context, and replayable Init/Override specs.

Copy APIs

Processes.copyinputs — Function
copyinputs(process_or_la)

Return the stored Init specs for an initialized process or loop algorithm.

source
Processes.copyoverrides — Function
copyoverrides(process_or_la)

Return the stored Override specs for an initialized process or loop algorithm.

source
Processes.copyprocess — Function
copyprocess(p::Process, Init(...), Override(...); kwargs...)

Build a new process from the stored initialized loop algorithm recipe. Extra init/override specs are merged by init.

source

Typical Copy Pattern

template = Process(
    MyAlgo,
    Init(MyAlgo; start = 0, buffer = Int[]),
    Override(MyAlgo; delta = 2);
    repeats = 10,
)

p = copyprocess(
    template,
    Init(MyAlgo; start = 100, buffer = Int[]),
)

run(p)
wait(p)
close(p)

If the context needs custom rebuilding logic, provide a fully prepared context = ... directly.