> --- /usr/local/stow/emacs-cvs/share/emacs/22.0.50/lisp/.backup/subr.el.~1~
> 2005-02-09 07:50:41.000000000 -0800
> +++ /usr/local/stow/emacs-cvs/share/emacs/22.0.50/lisp/subr.el
> 2005-02-16 19:50:57.000000000 -0800
> @@ -1969,13 +1969,15 @@
> (re-search-forward "\\(\\sw\\|\\s_\\)+"
> (save-excursion (end-of-line) (point))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(line-end-position)
> t))
> - (progn (goto-char (match-end 0))
> - (buffer-substring-no-properties
> - (point)
> - (progn (forward-sexp -1)
> - (while (looking-at "\\s'")
> - (forward-char 1))
> - (point))))
> + (condition-case nil
> + (progn (goto-char (match-end 0))
> + (buffer-substring-no-properties
> + (point)
> + (progn (forward-sexp -1)
> + (while (looking-at "\\s'")
> + (forward-char 1))
> + (point))))
> + (error nil))
> nil)))
I recommend to only catch "scan-error" so as not to hide other potential
bugs. But maybe a better fix is to let-bind forward-sexp-function to nil
around the call to forward-sexp. After all, we're using here forward-sexp
to skip over a symbol, not just any a sexp. Or maybe even replace
forward-sexp with (skip-syntax-backward "w_"), although it seems so obvious
that maybe there's a reason why it's not done: better check the file's log
to seen if it wasn't (skip-syntax-backward "w_") already in the past.
Stefan
_______________________________________________
Emacs-devel mailing list
Emacs-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/emacs-devel
