Re: font-locking and open parens in column 0

[martin rudalics <rudalics@gmx.at>]

 >> But I still think the
 >>"bug" is with the author who put the left paren in column zero of that
 >>comment.  That author should be warned just as in emacs-lisp-mode.
 >
 >
 > Should _HAVE_ been warned.  Martin Stjernholm has worked very hard to
 > remove that restriction from C Mode without sacrificing (much) speed.  It
 > would be a shame now to leave the restriction in Emacs 22.

Which means that Martin Stjernholm's change did not make it to
font-locking which, for c-mode, still uses the `beginning-of-defun' from
lisp-mode.  I see only two reasonable fixes to this:

(1) Set `syntax-begin-function' (or `beginning-of-defun-function') to
     `c-beginning-of-defun' if you trust that change.  This should,
     hopefully, populate your cache.

(2) Set `syntax-begin-function' to nil as Stefan suggested and let
     `syntax-ppss' do the work.  This should populate Stefan's cache.

Any other solution would be a shameful waste of resources since the
results of `parse-partial-sexp' from your original proposal don't get
stored anywhere (jumping to a position n in a c-mode buffer and
subsequently jumping to a position m < n would afford to rescan from
bob).  Also make sure that `font-lock-beginning-of-syntax-function'
nowhere interferes with these settings.

Ideally, there would be one and only one cache populated and consulted
by both - c-mode and `syntax-ppss'.