[parenscript-devel] Another CASE fall-through bug

[Daniel Gackle]

The recent bug fixes seem to have worked (thanks). It seems
I can't remove my warning-silencing workaround code, though (hacking
*defined-operators*). I mention this because it's not clear to me if the
intention was to not need that workaround anymore. What should the
current behavior be - you get a warning if you redefine a built-in operator,
but not an error?

I did encounter a new problem while removing other workarounds.
At least I think it's a new problem rather than a failure to fix
the related earlier bug - I'm not really sure.

This incorrectly falls through to the 456 case when foo1() returns
true and foo2() does not. The example can probably be pared down
further, but I don't have time to make it shorter.


(defun blah (x)
  (case x
    (123
     (cond ((foo1)
            (when (foo2)
              (when (foo3)
                (return-from blah nil))
              t))))
    (456 (foo7))))

function blah(x) {
    switch (x) {
    case 123:
        if (foo1()) {
            if (foo2()) {
                if (foo3()) {
                    return null;
                };
                return true;
            };
        };
    case 456:
        return foo7();
    };
};
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.common-lisp.net/pipermail/parenscript-devel/attachments/20110202/628c875a/attachment.html>