emacs-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Fwd: Update earlier posted hideif.el enhancements


From: Stefan Monnier
Subject: Re: Fwd: Update earlier posted hideif.el enhancements
Date: Tue, 03 Dec 2013 08:45:09 -0500
User-agent: Gnus/5.13 (Gnus v5.13) Emacs/24.3.50 (gnu/linux)

> Umm, here I only comment out the "make-local-variable" for hide-ifdef-env.
> Otherwise the defined tokens works only for that buffer.  I would like it to
> work for all currently openned buffers in a "project" based manner.

Hmm... I see.

Affecting the value globally would also affect buffers that are
unrelated.  Sadly we don't really have an easy way currently to talk
about project-wide settings.

I guess we should use the CEDET infrastructure for that (which would
also make it possible to scan files which aren't currently visited).

But I guess for now we can try your "quick&dirty" solution of making the
settings global.

> The "set" form derived from the original code.

Ah, right.  Then just replace "(set 'hide-ifdef-env" with "(setq
hide-ifdef-env".  It's just a stylistic issue.

>> Indeed, there's been no real development on it in recently.  See my
>> comments below.  Overall, I think it looks good and could be included in
>> hideif.el, but I'd appreciate if you could clean up a few things first
>> and split the change into a few chunks (this is a bit large to review as
>> one single chunk).  IIUC one chunk would focus on extend the current
>> parser & evaluator to handle the full CPP syntax.
>> The remaining cleanups
> The remaining cleanups are done. The original code that I derived from
> contains a lot of ';;;' comments, I fixed most of them but left some since
> they looks like to be placed on purpose to become a separator there.

Thanks.  Fixing the pre-existing issues is gravy.  The main part is
to not make it worse ;-)

> Therefore it's extremely time consuming to cherry-pick among all those
> changes and group them into several valid topic patches, sorry about that...

I'm not asking you to cherry pick commits.  Instead you should take your
current code (call it "target") and remove "all but the parsing-related
modifications", for example (I usually do that with "C-x v d" then C-u
C-c C-a to undo the hunks that are not related, plus manual changes for
those hunks which contain both changes I want to keep and changes I want
to remove).  That gives you a first patch to send.  Then you diff the
result with your "target code" and remove "all but something else".

This does not use the commit history.  Instead it kind of "creates
a brand new and clean history".


        Stefan



reply via email to

[Prev in Thread] Current Thread [Next in Thread]