Add --verbose flag to lint, showing copyright/license info sources

[robinkrahl]

If a project uses multiple licenses, it is hard to get an overview of which files are under which license without manually inspecting the files. reuse lint only lists all used licenses, and while reuse spdx lists the license information for all files, it is rather hard to read for a human.

I think it would be good to have a reuse listcommand that prints a table with the copyright information for all files (or maybe report to increase the edit distance from lint) :

$ reuse list
File	License		Copyright Text
1.c	MIT		2020 John Doe
2.c	MIT		2020 John Doe
3.txt	CC0-1.0		2019 Jane Smith

It would be even better if the command could group files with the same copyright information:

$ reuse list
File	License		Copyright Text
*.c	MIT		2020 John Doe
3.txt	CC0-1.0		2019 Jane Smith

Use cases:

1. Getting an overview of a third-party project.
2. Making sure that you added the correct annotations for your own project.

I can try to implement this if you would merge it.

[mxmehl]

Indeed, that'd be a nice view, also for debugging.

Perhaps related to #183 since reuse spdx should already include all information. Outputting it as JSON (and/or YAML) would allow for a simpler manipulation that could also be done manually, e.g. with tools like jq.

[robinkrahl]

Yes, using a separate parser for the reuse spdx output would also be a way to implement this. Does this mean you’d rather not have this feature in reuse-tool directly?

[mxmehl]

Does this mean you’d rather not have this feature in reuse-tool directly?

I am fine with having this function, either as a full list (as per your first post), or also with the option to have a detailed view for a single file (e.g. explaining the chain of precedence in case there are multiple explicit/glob coverages for this file).

Regarding spdx, I just wanted to make sure that we do not forget that users also may have this function to build their own custom query if necessary.

[mxmehl]

In today' hackathon we concluded that we would like to have this verbose listing as part of the lint subcommand, e.g. reuse lint --verbose, perhaps also combinable with --json.

Aside from the detected license and copyright it should also provide info on where this information has been derived from, e.g. the file itself, the .license file, or dep5 (or future REUSE.yaml).

[mxmehl]

So us maintainers would like to see this feature in the next release.

What's unclear yet:

• Should this be a separate subcommand (e.g. list) or a flag to lint like --verbose? We had a preference for the flag in the January call but in the light of recent developments this may have changed.
• Can we make it work together with Allow specifying a set of files/directories to lint #512 and Add --json flag to lint #183?

What we agreed on:

• This should not be the default output for lint if it would become a flag for this subcommand. Also not when just linting a single file, like proposed in Allow specifying a set of files/directories to lint #512

[mxmehl]

In today's call we clarified a few more questions:

• The desired functionality shall happen via the --verbose flag to the lint subcommand, not to a separate subcommand.
• It shall happen in the same step as Add --json flag to lint #183 as the --json flag shall imply --verbose already.

So this issue can be used to track the effort of adding the verbose output to the plain, human-readable output of reuse lint.