Pim (which stands for Process IMprover) is a task orchestrator meant for personal computers. Besides regular cron scheduling, it adds additional features like:

• Running on start (with optional delay).
• Running commands in specific directories.
• Running commands directly based on their set name (aliases).
• Doing catchup runs (if computer was turned off during scheduled run).
• Scheduling multiple commands at the same time.
• Access to logs.

Moreover, the goal of pim is to be an orchestrator that takes all scenarios into account, so ideas are always welcome.

Getting started

You can install pim with go install github.com/TimoKats/pim@latest (assuming you have go installed and GOPATH set correctly). Currently, pim is not available in any package repositories. Next, you can setup your schedule in ~/.pim/process.yaml.

Usage

process.yaml

All your tasks can be defined in your process file, which sits in ~/.pim/process.yaml. Here you can find a template of this file with some explanation of the possible attributes. In summary, there are some overall optional settings, and there is a list of "runs". Here, each run represents a command you want to schedule.

max_logs: 50 # optional: max logs pim will store. Trims based on FIFO logic. Defaults to none.
only_store_errors: true # optional: if true, only logs of erronious runs are stored. Defaults to false
process:
  - run:
    name: fetch-repo # name of your run, mandetory
    schedule: '@times;8:00;10:00;16:00;18:00;21:55' # schedules prefixed with @times will run every day at the selected time.
    directory: /home/user/code/ # optional: set a directory to run the code in.
    catchup: true # optional. if true, pim will do a catchup run on startup if the computer was off when last scheduled.
    command: git fetch # the actual command you want to run.

  - run:
    name: scraper
    schedule: '@start+20' # schedules prefixed with @start run on startup. +20 means wait 20 seconds after startup to run.
    duration: 60 # optional: max number of seconds the command can run. After which, program is gracefully exited
    command: python3 scraper.py

  - run:
    name: change-wallpaper
    schedule: '*/5 * * * *' # You can also have some good-ol cron strings :)
    command: ./change-wallpaper.sh

Commands

Pim is used from the command line. This is an overview of the commands/flags you can use. Note, the format is pim <<command>> and pim --<<flag>>. The gif below shows some pim commands being used in practice.

Commands:

• run <<command-name>>: Runs a command by the name defined in your process YAML.
• start: Starts the cron schedule defined in your process YAML. NOTE: Logs are sent to log files in ~/.pim/logs/ and not to standard output when using start.
• stop: Stops the cron schedule started by running: pim start.
• ls: Lists all the commands and their characteristics defined in your process YAML.
• log <<optional:run-id>>: Show all logs, or a log of a specific run.
• clean: Clean log files.

Flags:

• info/i: Outputs some information about this Pim installation.
• help/h: Well...if you see this message you probably typed this...
• version/v: Shows version of this Pim installation.
• license/l: Shows the license of this Pim installation.

Running in background

If you want to run pim start automatically you have a couple of options. First, you can add (pim start &) to your .bashrc file. This will start a pim session whenever you open your terminal (note, it prevents running multiple times if you have multiple windows). Next, you can also use something like crontab to start on reboot. A third option would be to run pim through a systemd file. Finally, it's also possible to just run pim start and pim stop whenever you want it to be on. E.g. in servers that never shut down.

For Windows users: Perhaps you can run pim as a service, but there's no way for me to test that given I don't have a Windows machine.