Accessor for the location of a diagnostic

[bjchambers]

Accessor for the location of a diagnostic

[Johann150]

The issue with this is that you would need the Files because the Diagnostic itself only stores a byte offset into the file. This is why the locus calculation is currently done in codespan-reporting/src/term/views.rs ll. 101-134 which is more or less while rendering.

What I think could be a sensible approach to solving this problem is to add RichDiagnostic to the public API of codespan-reporting. This would also go towards the goal of easily allowing for alternative renderers. There isnt really an issue for this, but some discussion can be found in #293 and #176. I'm not really sure if it would be a good idea to add that to the public API though.

[bjchambers]

Depending on how the location is reported it shouldn't need the Files. For instance, if it just reports the minimum (FileId, usize) of the maximum primary label, it would only need the byte offset. It would only need the Files if it wanted to convert the location to a line number / column offset, right?

[Johann150]

I think so, but I don't see how that would be useful.

But it might make sense to implement such a function on the Diagnostic anyway because that would separate some of the code into more sensible pieces and probably also reduce some duplication between RichDiagnostic and ShortDiagnostic.