auctex-devel
[Top][All Lists]
Advanced

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

[AUCTeX-devel] LaTeX-math-abbrev-prefix


From: Ralf Angeli
Subject: [AUCTeX-devel] LaTeX-math-abbrev-prefix
Date: Wed, 13 Apr 2005 08:56:38 +0200
User-agent: Gnus/5.110003 (No Gnus v0.3) Emacs/22.0.50 (gnu/linux)

In order to make customization of `LaTeX-math-abbrev-prefix' easier, I
made it accept a string in the format suitable for `kbd'.  It now has
a mixed defcustom where you can choose between a string and a sexp.
Formerly we only had a sexp (which meant you had to add non-ASCII keys
with something like [?§]).  Contrary to the type of defcustom the
manual described it as a string.

The change I made is fully backwards compatible.  That means in case
somebody actually customized the value to [?§] or similar there won't
be an error and it will be displayed correctly in the customization
buffer.  The drawback is that it adds both complexity at the user
interface for customization and at the code level.

So I made up my mind and think it would be a good idea to reduce this
complexity.  The rigorous way would be to accept only a string both at
user interface and code level.  Of course this will result in an error
for people not using a string.  The alternative would be to remove
compatibility only at user interface level.  That means the defcustom
would accept only a string but the code below could still deal with
any Elisp expression.  People not using a string won't see any
difference unless they try to customize the variable.  Then they might
get a wrong display in the customization buffer.

Currently I favor the rigorous way accompanied by a note in the news
because it is cleaner.  Any other opinions?

-- 
Ralf





reply via email to

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