On Fri, Mar 17, 2023 at 01:33:56AM +0200, Eduard Zingerman wrote: > On Thu, 2023-03-16 at 16:23 -0700, Andrii Nakryiko wrote: > > > [...] > > > > > > > In term of logical structure and maybe extensibility, this is more appropriate, > > > > in term of pragmatism maybe less. > > > > > > > > I don't have strong opinions and can see benefit for both. > > > > > > idk, I don't have a strong opinion either. > > > > me neither, flatter struct would be simple to work with either with jq > > or hacky grepping, so I guess the question would be how much do we > > lose by using flatter structure? > > Okay, okay, noone wants to read jq manual, I get it :) > > I assume that current plan is to consume this output by some CI script > (and e.g. provide a foldable list of failed tests/sub-tests in the > final output). Correct. > So, we should probably use whatever makes more sense > for those scripts. If you and Manu think that flat structure works > best -- so be it. Nested is the most logical representation of the data. Given the one-liner you provided, let me play a bit more with it and see what it would take to get minimal info out of it.
