Waits for a file or directory to change or appear, then just returns. The watched file or directory does not have to exist yet.

npm install --g just-wait

# Use --help or -h to see usage details
$ just-wait --help
Usage: just-wait [options]

  Options:

    -h, --help                  output usage information
    -p, --pattern <pattern>     glob pattern. "," separates multiple patterns.
                                More info: https://github.com/isaacs/minimatch
    -d, --delay <milliseconds>  delay returning for a number of milliseconds
    -t, --timeout <seconds>     timeout waiting after a number of seconds (default=30)
    -s, --silent                supress logging.

  Examples:

    # watch "lib" dir, return when something changes
    $ just-wait -p "lib/**"

    # watch "lib" and "src" dirs, return 500ms after something changes
    $ just-wait -p "lib/**,src/**" -d 500

    # watch "lib" dir, timeout after 10 seconds
    $ just-wait -p "lib/**,src/**" -t 10

In the case of a timeout, just-wait will return with exit code 1. In other cases it will return with exit code 0

By default just-wait will log two messages to the terminal to provide feedback to the user:

$ just-wait -p README.md && echo Done!
Waiting for README.md (max 30 seconds)
Ready. README.md changed
Done!
$

or, in the case of timeout:

$ just-wait -p README.md --timeout 10 && echo Done!
Waiting for README.md (max 10 seconds)
Timed out waiting for README.md after 10 seconds.
$

It uses picolog for logging, which allows us to influence logging verbosity. The Waiting for.. line is logged at level WARN (which means it is visible at picolog's default log level of WARN) and both the success and error message are logged at level ERROR. You can change the picolog log level by setting the environment variable PICOLOG_LEVEL, or you can completely suppress logging by passing the --silent (or -s) flag.

Based off of Rick Wong's wait-run, this simplified version just waits for the file or directory to appear/change and then returns. It does not execute any commands. Use it by chaining commands after it using &&, which works under *nix as well as Windows.

Credits also go to Fabian Eichenberger, who created watch-run, which was the basis for Rick's version.

Special thanks to Mark Reynolds for making all our lives easier with his contribution.

• © 2016, Stijn de Witt. (this project)
• © 2016, Rick Wong. (wait-run)
• © 2015, Fabian Eichenberger. (watch-run)

• Creative Commons Attribution 4.0 (CC-BY-4.0) (this project)
• BSD 3-Clause license (wait-run)
• MIT License (watch-run)