batching-and-chaining.md

Batching and chaining

Every snippet here is executed by ExamplesTest.

Both put several commands into one tmux invocation.

A batch: several commands, each with its own outcome

BatchResult result = server.batch()
        .add("new-window", "-d", "-n", "one")
        .add("new-window", "-d", "-n", "two")
        .run();

result.succeeded();                               // → true
result.operations().size();                       // → 2
result.operations().get(0).outcome();             // → COMPLETE

tmux discards a group after its first failure, so a single exit status cannot say which command failed or which never ran. Each operation is reported for itself: COMPLETE, FAILED, SKIPPED or UNKNOWN.

That distinction is the whole point. A batch of five where the third failed tells you exactly that — two done, one failed, two never attempted — rather than one unhelpful nonzero.

A chain: each step acts on what the last one made

server.chain()
        .newWindow("built")
        .splitLeftRight()
        .sendLine("echo chained")
        .run();

// One request built a window and split it, with no round trip to learn the id.
server.windows().stream().anyMatch(w -> w.name().equals("built"));   // → true

No step names a target. tmux moves its own current target as a group runs, so the split lands in the window just created and the keys land in the pane just split — with no round trip to learn either id.

That is the difference from a batch: a batch is several commands that happen to travel together, a chain is several commands that depend on each other.