decoder: formatting improvements

[andresovela]

Besides refactoring the log formatting code, this PR introduces a few new features:

{c} metadata specifier

Now it's possible to add {c} to the log format in order to print the crate name. I'm currently parsing this out of the module path, there may be better ways to do this.

More detailed {f} metadata specifiers

I found the {f} specifier to be not very useful in most cases, mainly because a lot of files are called mod.rs. Now it's possible to increase how detailed to print the path to the file by adding {ff}, {fff}, etc... to the log format. With {ff} the Formatter would print format/mod.rs instead of just mod.rs, which I think is much more useful.

Zero-padding support

I added the zero-padding functionality from #780 to this PR. Now it's possible to zero-pad metadata specifiers, such as the timestamp or the line number.

[Urhengulas]

I am actually a bit unsure how to go about it. Yes, the default format logic is duplicated in knurling-rs/probe-run#425, but it is actually a bit different than in probe-run and also different than in this PR. Yes they check if there is a timestamp or not, but they also have a --no-location flag. So the code on that side would not necessarily get more simple by this PR. Also note that they are not using log and therefore not defmt_decoder::log::init_logger.

[andresovela]

Yes they check if there is a timestamp or not, but they also have a --no-location flag.

I'd then propose to add a no-location flag to DefmtLoggerConfig. I don't see why not, it seems like it would be a fairly common configuration option.

Also note that they are not using log and therefore not defmt_decoder::log::init_logger

I think all of this logic would end up in the Formatter introduced in #781, so I don't see this as an issue.

[Urhengulas]

Makes sense

[andresovela]

I did a relatively big refactor of all the formatting stuff.

[Urhengulas]

I did a relatively big refactor of all the formatting stuff.

Cool, thank you! I will hand over the project to a colleague this week. He will come back to you.

[andresovela]

Hi @Urhengulas, do you know if your colleague will have time for some reviews soon?

[Urhengulas]

Hi @Urhengulas, do you know if your colleague will have time for some reviews soon?

@BriocheBerlin @amanjeev

[jonathanpallant]

I'll take a look now.

[jonathanpallant]

Looks good to me. It'll need a rebase to fix the CHANGELOG, and perhaps we could squash the first three commits, just for neatness.

[andresovela]

Rebased now :)

[andresovela]

Ping @jonathanpallant