Distinct exit code for invalid master password

[msladek]

Hey there, firstly thanks a lot for the tool, it is very helpful when working mostly with CLI.

I recently noticed while incorporating enpasscli in some bash scripts that all non-successful exits return 1, e.g. when unable to copy because the card wasn't found:

$ enp copy asdf || echo "Exit code: $?"
Enter master password: <VALID>
FATA[0004] card not found                               
Exit code: 1

However it would be useful to have distinct exit code for providing an invalid master password. Currently it also returns 1:

$ enp copy asdf || echo "Exit code: $?"
Enter master password: <INVALID>
FATA[0001] could not retrieve cards                      error="could not retrieve cards from database: file is not a database"
Exit code: 1

Now I'm not at all familiar with golang, but if I'm not mistaken we'd simply need to add e.g. os.Exit(2) in the error checks after getting the entries from the vault. I guess that'd be sufficient for my needs at least.

WDYT?

[hazcod]

Hi @msladek!
Thanks for reporting this feature. I've added it in 0c57602.
It should pop up in a new release in a couple of minutes.

[msladek]

Wow that was quick, thanks @hazcod!

Quick question: if I provide an invalid master password (with keyfile) I don't get the message could not open vault, instead I get the one above could not retrieve cards. Will it then still exit with code 1 in my case?

[hazcod]

Ohh, my mistake in being too fast.
I can look into this further for you or would appreciate a PR.

[msladek]

I guess the problem is a different one now: why doesn't it in my case (pw+keyfile) fail already within vault.Initialize? I guess an invalid password should already fail at vault.openEncryptedDatabase.

Can you reproduce this issue or do you have a clue what could be the cause? I'd be happy to provide a PR if I knew how to solve it.

[hazcod]

I was thinking the same thing, when looking at https://github.com/hazcod/enpass-cli/blob/master/pkg/enpass/vault.go#L131

[msladek]

@hazcod I've found the issue, opened a PR with a solution for this.

[hazcod]

@msladek ahh that makes sense, thank you for the work! Closing.