WIP: Initial cargo-grcov command

[gilescope]

PR for cargo-grcov. Will fix #326

I'm a big fan of early review to make sure I'm headed in the right direction. Wasn't qute sure how grcov created errors - they seem to be logged with log.error!(), but didn't spot any custom errors - should I just create a specific error type where needed - or is there one that converts from a string?

[marco-c]

Sorry for the delay, I hope I'll be able to get to this by the end of this week.

[dnaka91]

Is there any progress on this?

I'm recently using grcov a lot and it would be great to have a dedicated cargo-grcov command that does all of the heavy lifting within a single command. Currently have to copy over a bunch of setup scripts for each project and this would simplify things A LOT.

[marco-c]

Unfortunately I didn't have time to review this :(

If somebody can help me with reviewing the code in detail, I can review the overall architecture.

(Assuming @gilescope is still interested in finishing this).

[dnaka91]

I looked through the code and it is pretty much a wrapper around the grcov binary with the arguments and env vars described in the README.md directly applied by this new binary. It seems to miss support for source-based coverage yet.

I personally prefer to have cargo-grcov being a second binary that is independent of the grcov binary and simply uses the same components directly. That avoids version mismatches and failures trying to resolve the location of grcov.

Pretty much what was done for the flamegraph tool is what I think is more effective and safer towards errors. As we have a good abstraction already, it's easy to implement as well.

If you, @marco-c, don't mind, I would start a new PR that uses the same technique as in the flamegraph crate and combine it with the effort already done here, then adding in more features. For that PR I would probably strip many options from the CLI and only expose what make sense in the context of cargo.

What do you think?

[marco-c]

@dnaka91 sounds good to me!

[gilescope]

Sounds good to me too - I just want to use it :-) so happy to test out, feedback etc.

…

On Tue, 29 Jun 2021 at 10:06, Marco Castelluccio ***@***.***> wrote: @dnaka91 <https://github.com/dnaka91> sounds good to me! — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#533 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAGEJCEGEHLHHZEYKCMVXOTTVGEIRANCNFSM4UKIMFHA> .

[marco-c]

@dnaka91 still interested?

[dnaka91]

Definitely! Sorry, I didn't post any update on this for a while. Had a few other things going on and simply didn't get to it yet. Already checked the current implementation of this PR and I think I'll go with a very similar approach and then add a few cli args from the original grcov binary.

Also, I was hoping that #648 would maybe be merged before I start on this (or decided not to, just so I don't have to implement the CLI logic twice). Is there any chance you could give it a look soonish? 🙇‍♂️

[gilescope]

@dnaka91 structopt is merged in now if that helps...

[gilescope]

I've updated this PR to structopt. @dnaka91 as I said if you want to go on from here that's cool with me.

[dnaka91]

@gilescope thanks for letting me know. I didn't work on this for quite a long time but will likely be able to pick up on this again soon-ish.

[gilescope]

Great news. Happy to test out and give feedback.

…

On Sun, 17 Oct 2021 at 16:10, Dominik Nakamura ***@***.***> wrote: @gilescope <https://github.com/gilescope> thanks for letting me know. I didn't work on this for quite a long time but will likely be able to pick up on this again soon-ish. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#533 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAGEJCASSVKTDIPKCHJJVD3UHLRPJANCNFSM4UKIMFHA> .

[marco-c]

@dnaka91 @gilescope does any of you want to pick this up again?

[dnaka91]

Sorry, I didn't do any further work on this. Reason is, that I moved to cargo-llvm-cov which just works better for me and seems more active compared to grcov. Therefore, I don't really have the motivation to push this further.