Error at the end of the build

[andrei-mihaila]

The bug

Getting an error at the end of the build:

error: improper list

Usage:
 ps [options]

 Try 'ps --help <simple|list|output|threads|misc|all>'
  or 'ps --help <s|l|o|t|m|a>'
 for additional help text.

For more details see ps(1)

This probably just affects the message displayed at the end of the build ('You can activate this installation running the following command:') - not displayed.

Software versions

A list of software versions where the bug is apparent, as detailed as possible:

• Operating System: Gentoo & Fedora GNU/Linux (probably irrelevant)
• Erlang/OTP: 24, 25, 26 (irrelevant)
• kerl: latest, retrieved on 2023.10.05 from https://raw.githubusercontent.com/kerl/kerl/master/kerl
• other (where applicable): ps --version
  ps from procps-ng 4.0.4

How to replicate

An ordered list of steps to replicate the bug, e.g.:

1. run kerl build ...
2. search for error: improper list in the error output

Expected behaviour

The error should not be happening and the right message should be displayed at the end of the build.

Additional context

I think the problem is on this line: 1462.
The code is running ps to get the parent PID (used later to determine if the current terminal is bash or zsh or other). The result of that line can contain leading spaces, for example:

$ ps -p $$ -o ppid=
 2970

On the next line ps expects a list of comma separated PIDs (no spaces). For example:

$ ps -p " 2970" -o ucomm | tail -n 1
error: improper list

Usage:
 ps [options]

 Try 'ps --help <simple|list|output|threads|misc|all>'
  or 'ps --help <s|l|o|t|m|a>'
 for additional help text.

For more details see ps(1).
# it works with the spaces removed
$ ps -p " 2970" -o ucomm | tail -n 1
bash

Thanks!

[paulo-ferraz-oliveira]

Thanks for reporting. I can fix this based on your input, and using a Fedora image. I can't replicate it locally, but probably because my process id is already so high that no space is output.

Edit: the error was introduced by me, here.

[paulo-ferraz-oliveira]

@andrei-mihaila, I've pull requested to fix this. Would you care to check the proposed changes and let us know if they're effective for your use case: https://github.com/kerl/kerl/pull/474/files?

Thanks.

Ah, as a side note: I could just go back to the "old" code and add the shellcheck disable, but this way it's clearer as to what our intentions are. (the lack of comment there was what led me to believe we we're "just avoiding" ShellCheck).

[andrei-mihaila]

Tested and it works. Thank you!