2 bugs in paren.el and 1 patch

Byrel and I found 2 bugs in paren.el in 21.5.32: 1st bug: starting with xemacs --vanilla, go to the menu, Options->Display->Paren Highlighting, and see 4 choices, and you can select any of them and observe that visually, they are selected correctly, that is, the radio buttons change correctly. However, notice in a buffer of lisp code that the highlighting of parens does not change with your choices from the menu. That is because the file paren.el is not loaded yet. Manually loading it with M-x load-file paren.el makes it so the menu choices do indeed work and change how they work with different menu choices. I do not know how to fix that. Any ideas? 2nd bug: step 1. starting with xemacs --vanilla, go to the menu, Options->Display->Paren Highlighting, and see 4 choices, and you can select any of them and observe that visually, they are selected correctly, that is, the radio buttons change correctly. (never mind for the moment that they do not work as described above) Try it several times and the radio buttons work correctly every time. step 2. open a customize buffer for paren-matching with: M-x customize <ret> paren-matching go down to the first variable, Mode and show it. it has 5 choices: * None (default) * Blinking Paren * Highlighted Paren * Highlighted Expression * Highlighted Surrounding Expression Choose one, say, the last one. Go to the top of the customize buffer and click the buttons: set, save, then Done. Now go to the menu, Options->Display->Paren Highlighting, and see 4 choices, select "none" Go back to the menu several more times each time selecting "none". Instead of the radio button "none" staying selected, each selection of "none" selects the radio button to the next lower menu item, and even incrementing to the 5th selection that is not shown except in the customize buffer, and not showing any of the 4 radio buttons selected, which is correct, they are not selected. Looking at paren.el, it looks like a case is missing in the function paren-set-mode for the case where arg is passed in with the value set to nil. When arg was passed as nil, it fell through to the default case which was not correct for nil. Here's a patch adding a cond test for that case. --- /usr/local/share/xemacs/xemacs-packages/lisp/edit-utils/paren.el 2012-12-31 01:55:30.000000000 -0500 +++ /home/steve/paren.el.patched 2012-12-31 01:59:40.000000000 -0500 ＠＠ -365,6 +365,7 ＠＠ ((and (numberp arg) (> arg 0)) (nth (1- arg) paren-modes)) ((numberp arg) nil) + ( (not arg) nil) (t (car (cdr (memq paren-mode paren-next-modes))))) ))) Steve Mitchell Byrel Mitchell _______________________________________________ XEmacs-Beta mailing list XEmacs-Beta(a)xemacs.org http://lists.xemacs.org/mailman/listinfo/xemacs-beta