Command line option to disable printing World when there are failed tests or panics

[yds12]

Hello,

First of all, thanks for the work on this awesome crate.

If I understand correctly, currently there is no way to disable printing the World in error messages. Whenever there is a failed test or panic, the whole World struct is printed. In some use cases this object can be quite large, making the errors difficult to navigate.

Would it be possible to add a command line flag that controls this behaviour?

[tyranron]

cc @ilslv
Let's do this as writer::Basic's CLI option and pub method before 0.11 release.

[tyranron]

@ilslv also, I think it's worth to bind it to the exisiting verbosity CLI option, so by default world is not printed, but with --verbose yes. What do you say?

[ilslv]

@tyranron I would split them, because I think that default behaviour should be "print failed world, but not the docstring". One reason for that is I saw some of users of this crate using docstrings as some kind of spec, including graphql query there for example

[tyranron]

@ilslv what about verbosity levels? -v prints world and -vv prints doc strings?

[ilslv]

@tyranron good idea 👍. I don't think there will be a case when someone wants to print docstrings, but not the world.

[tyranron]

@ilslv so, let's do it this way.

[theredfish]

If there is no verbosity option passed, nothing will be printed? Or will you add a default option?

[ilslv]

@tyranron

what about verbosity levels? -v prints world and -vv prints doc strings?

I believe that by default we should print World and problem this verbosity flag is that we can set -v by default, but there is no way to downgrade from here. We can make -v do nothing, -vv print World (and be default one), -vvv print both World and docstrings, but I'me sure this is idiomatic approach for verbosity flags. So I would stick with additional bool flag.

[tyranron]

@ilslv why not just -v 0, -v 1, -v 2 with a -v implied -v 1 by default?

[ilslv]

@yds12 @theredfish what do you think should happen by default on failed Steps: printing World or not?

[yds12]

@ilslv I would say not by default.

[theredfish]

I see printing the World as extra debug info, that I would prefer to opt-in if I need it. I think that the assertion message on failed Steps is enough as a default behavior - then if I need to debug more I can re-run failed tests with the World output.