Adding ability to have --stdin overwrite --text

[ericloyd]

PR adds --stdin option to allow text to be piped in, overriding the --text option. This is per #62

[ChaoticWeg]

LGTM

[ChaoticWeg]

Looks like tests failed, I'm not familiar with the framework/setup used for testing here, so will defer to @fieu

[ericloyd]

I'm new to bats so I may have built the test wrong. You can test manually from the command line:

echo "this text will be used" | ./discord.sh --username "username test" --text="this text is ignored" --stdin

[ericloyd]

This is the bats test file that worked for me, without all the other tests in it.

#!/usr/bin/env bats

load pre

# Test that text can be sent via pipe
@test "text will be ignored, gathered instead from stdin (pip)" {
    run bats_pipe echo "this text will be used" \| discord.sh --username "username test" --text="this text is ignored" --stdin
    [ "$status" -eq 0 ]
}

[ericloyd]

Nevermind. I see I forgot a closing } in the 01 tests file. :-(

[ChaoticWeg]

Nevermind. I see I forgot a closing } in the 01 tests file. :-(

Yup, only just now noticed it myself, lol. That's what I get for reviewing code at 8 AM.

[ChaoticWeg]

No obvious errors to me, test still fails. Deferring to fieu

[ericloyd]

Weird. When I pull that single test out separately, it works. I didn't want to run them all on my testing, but I'll run through them to see if I can figure it out.

[ericloyd]

84 tests, 0 failures

[T0biii]

discord.sh cant be found for the pipe

✗ text will be ignored, gathered instead from stdin (pip)
   (in test file tests/01.text.bats, line 69)
     `[ "$status" -eq 0 ]' failed
   Last output:
   /usr/local/lib/bats-core/test_functions.bash: line 276: discord.sh: command not 

84 tests, 1 failure

[ericloyd]

Thanks for that catch. I had a symlink in my path for discord.sh that was finding the file in my local repo. Good catch!

[T0biii]

mh ci still failing maybe the secret for the weebhook is broken?
when i setup in my fork the secret and run the pipline it works fine:
https://github.com/T0biii/discord.sh/actions/runs/15165388096/job/42641721325

[ericloyd]

I'm not sure what to do here so I'll leave it hanging in the ether until someone smarter than me makes decisions. :-)

[ChaoticWeg]

That'll be @fieu with the knowledge on how the actions and secrets are set up here, the code looks fine to me, so up to him whether to merge with failing tests, whether to fix, whatever we wanna do there.

Thanks for the PR @ericloyd !

[ericloyd]

@fieu Any luck with you having a chance to check out why the automated tests are failing? Works from command line and the PR is pretty simple. Would love to see this go through.

[ericloyd]

@fieu Sheldon, it's been two months for this relatively simple PR. Any chance you can take a look and push it through? I'm confident that the tests are failing because of the build environment, not the code.

[fieu]

Hey @ericloyd,
I'm really sorry this has taken so long to get merged. I've been dealing with some real life stuff and haven't been able to allocate much time to the project, but I now have more time to work on it.

When I first saw the GitHub Action workflow fail, I was too preoccupied with other things to dig into why it was failing. It was puzzling considering your change was very minor, but I just didn't have the bandwidth to investigate at the time.

After some digging (albeit 5 months later), I found out that even when approving a pull request to run the action workflow, the action won't have access to the repository secrets (which we use for the $DISCORD_WEBHOOK environment variable). This wasn't apparent at the time and I didn't have the time to look into it. I've actually never dealt with pull requests and actions before. We used to be on Travis CI, so I've learned a lot since then.

I noticed @T0biii mentioned the secrets might not be working properly in the action, and he was spot on. GitHub does this (as they should) from a security perspective to avoid leaks of secrets.

Overall, after doing some testing locally and verifying the pipeline passed, it looks good to me. Again, I apologize for the huge delay, and thank you for your patience and contribution!

[ericloyd]

We're all good. I know how RL can be. Thanks?

[fieu]

Tagged v2.0.1, thanks again!