On 26.12.2019 23:56, Radey Shouman wrote:
> Dmitry Gutov writes:
>   > Hi Radey,
>   >
>   > sorry for the late reply.
> 
> Not at all.

And once again. Upon some consideration, I have settled on the most 
backward-compatible option, I think.

> No.  I believe that completion-ignore-case should always be bound to
> the value of tags-case-fold-search when doing a tags completion.  If
> this is not so then completion is not consistent with the tags search,
> which is the bug in question.

Yes, OK. See the attached patch.

The alternative would be to create a new user option local to the xref 
package that would affect both identifier completion and search, but a) 
that will likely be an incompatible change unless tags-case-fold-search 
becomes an obsolete-alias, b) it seems unrealistic to have all backends 
adhere to it anyway.

So let's ask the backends instead.

> On the other hand, it might well be useful to make
> tags-case-fold-search or any xref replacement a safe local variable.

Yes. The patch also does that.

> BTW, how might one find the list of safe local variables?

Well, um. You can evaluate this expression:

(seq-filter
  (lambda (s) (get (intern-soft s) 'safe-local-variable))
  (all-completions "" obarray))

Eli, is this patch okay for emacs-27?

On the one hand, the issue seems minor-ish.

On the other, it's a backward compatibility problem. And the patch must 
be pretty safe because it's basically additive: the only real change is 
to xref--read-identifier, and it calls 
xref-backend-identifier-completion-ignore-case which will default to the 
current behavior for all backends except etags.