bash中set -t选项的作用是什么？为何测试与手册描述不符？

First, let's clear up the core behavior of set -t—the help text is a bit terse, so it's easy to misinterpret without digging into the details:

-t tells Bash to exit immediately after reading and executing one complete command once the option is enabled.

The key here is understanding what counts as a "complete command": it includes single commands, compound commands like { ... }, pipelines, or command lists separated by ;. Once the option is active, Bash will run that one command and then terminate the current shell session.

Why Your Test Scripts Didn't Behave As Expected #

Let's break down your two test cases to resolve the confusion:

Test Script 1 (No Output) #

Chances are your script had invalid syntax that prevented the echo from running. If you wrote it like this (all commands on the shebang line):

#!/usr/bin/env bash set -t echo "test"

This is wrong—the shebang line only specifies the interpreter path, not commands to run. Bash will try to execute a non-existent file named bash set -t echo "test", so nothing outputs.

If you split lines but grouped set -t with echo:

#!/usr/bin/env bash
set -t echo "test"

Here, echo and "test" are treated as arguments to the set command itself. Bash enables the -t option, then exits immediately after running this set command—your actual echo "test" never executes, hence no output.

The correct version would be:

#!/usr/bin/env bash
set -t
echo "test"

Running this will output test, then Bash exits right away (there's no further code to run anyway).

Test Script 2 (Variable Not Set) #

If your script was written as:

#!/usr/bin/env bash
set -t
export TEST_VAR="test"

The export command does run, and Bash exits afterward—but variables set in a script only exist in the script's subshell, not your interactive terminal. To make the variable available in your terminal, you'd need to source the script with . ./script.sh instead of running it directly.

If you wrote set -t export TEST_VAR="test" (grouping commands), the export is treated as an argument to set, so it never runs—hence the variable stays unset.

Practical Use Cases for set -t #

While set -t isn't as common as options like -e or -x, it has niche but useful applications:

• Run a single command in a temporary Bash session:
  Launch Bash with -t to execute one command and exit immediately, great for testing environment setups:

  bash -t -c 'source /path/to/setup.sh && echo $ENV_VAR'
  

• Auto-terminate a subshell after one command:
  Use it to isolate commands in a subshell and avoid writing an explicit exit:

  # Run commands in a subshell, then exit the subshell right away
  (set -t; echo "Running in isolated subshell"; ls -l ./temp)
  # Code here runs in your main shell
  

• Debug script execution up to a specific point:
  Stop execution after a critical command to verify state without running the rest of the script:

  #!/usr/bin/env bash
  echo "Step 1: Initialize workspace"
  mkdir -p ./build
  set -t
  echo "Step 2: Verify build dir exists"
  ls ./build
  # The lines below will NEVER execute
  echo "Step 3: This code won't run"
  rm -rf ./build
  

内容的提问来源于stack exchange，提问作者Taylor Clark