Re: [Emacs-diffs] master 51e7e46: Font-lock elisp macros/special forms dynamically

[Tassilo Horn]

Stefan Monnier <address@hidden> writes:

>> Ok.  I just tried it anyway, and with jit-lock disabled,
>> `font-lock-flush' does nothing, e.g., the buffer portion you have
>
> I think you've just found a bug.  How did you disable jit-lock?

Finding vhdl-mode.el (because that's large) and then doing M-:
(jit-lock-mode nil).

> Can you try "emacs -Q --eval '(setq font-lock-support-mode nil)'" and
> retry your test.

Yes.  So I did that and found vhdl-mode.el again.  Nothing is fontified.
(font-lock-flush) doesn't change that, but (font-lock-ensure) fontifies
the complete buffer.

>> If no one comes up with a (semi-)automated way of identifying
>> function-like macros, I'll go check the ~1300 macros in emacs
>> manually.
>
> We could also do it lazily: wait for people to complain and fix those
> that come up.

That's a good plan.  :-)

So Daniel and Drew, feel free to complain about concrete macros.  Or
just add a (declare (no-font-lock-keyword t)) to them yourself.

>>     + just there to save a funcall (ad-get-advice-info-macro,
>>       shadow-cluster-name)
>>       Shouldn't those be defsubsts instead of macros?
>
> Yes, but sometimes having them as macros is better (e.g., the code
> should be a bit faster than with a defsubst; also macros are expanded
> by gv.el so defining them as macros lets you use them as places in
> `setf' and `push' as well.  The new `define-inline' does address both
> of those problems, OTOH).

Oh, cool.  Is that going to be an alternative to `defsubst' or is the
plan to make it `defsubst' at some point in time?  (At least from a very
brief look, it seems to provide the same interface for the same
use-case.)

Bye,
Tassilo