bug#7006: Reader oddity after ?\^ --- should signal error if invalid instead of being garbage

bug#7006: Reader oddity after ?\^ --- should signal error if invalid instead of being garbage

[Daniel Colascione]

?\^ is supposed to be the prefix for a control character --- for
example, ?\^g. However, the character following the ^ can be a
non-existent control character like ' ' or ')'. Currently, the reader
outputs a nonsense value in this case. Instead, the reader ought to
signal an error.

bug#7006: Reader oddity after ?\^ --- should signal error if invalid instead of being garbage

[MON KEY]

,----
| Currently, the reader outputs a nonsense value in this case.
| Instead, the reader ought to signal an error.
`----

Not entirely nonsensical.

 (let ((^5 (get-buffer-create "#^")))
   (with-current-buffer (get-buffer ^5)
     (save-excursion
       (erase-buffer)
       (princ "#@4    ?\\^5" (current-buffer))))
   (unwind-protect
       (char-to-string
        (- (read (get-buffer ^5)) (expt 2 26)))
     (with-current-buffer (get-buffer ^5)
       (kill-buffer (current-buffer)))))
 ;=> "5"

:NOTE :bit 27 :byte 4 :2^ 26 :bit-weight 67108864
      :bit-oct #o400000000   :bit-hex #x4000000

--
/s_P\

bug#7006: Reader oddity after ?\^ --- should signal error if invalid instead of being garbage

[Lars Magne Ingebrigtsen]

Daniel Colascione <dan.colascione@gmail.com> writes:

> ?\^ is supposed to be the prefix for a control character --- for
> example, ?\^g. However, the character following the ^ can be a
> non-existent control character like ' ' or ')'. Currently, the reader
> outputs a nonsense value in this case. Instead, the reader ought to
> signal an error.

Hm, yes:

?\^9
=> 67108921

That seems awfully confusing.  

(format "%x" 67108921)
=> "4000039"

Does anybody know whether this a feature or a bug?

-- 
(domestic pets only, the antidote for overdose, milk.)
  bloggy blog http://lars.ingebrigtsen.no/

bug#7006: Reader oddity after ?\^ --- should signal error if invalid instead of being garbage

[Andreas Schwab]

Lars Magne Ingebrigtsen <larsi@gnus.org> writes:

> ?\^9
> => 67108921
>
> That seems awfully confusing.  
>
> (format "%x" 67108921)
> => "4000039"
>
> Does anybody know whether this a feature or a bug?

That's the same as ?\C-9, so not a bug.

Andreas.

-- 
Andreas Schwab, schwab@linux-m68k.org
GPG Key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

bug#7006: Reader oddity after ?\^ --- should signal error if invalid instead of being garbage

[Stefan Monnier]

>> ?\^ is supposed to be the prefix for a control character --- for
>> example, ?\^g. However, the character following the ^ can be a
>> non-existent control character like ' ' or ')'. Currently, the reader
>> outputs a nonsense value in this case. Instead, the reader ought to
>> signal an error.

It's not "nonsense".

> ?\^9
> => 67108921
> That seems awfully confusing.
> (format "%x" 67108921)
> => "4000039"
> Does anybody know whether this a feature or a bug?

It's a feature.  The ?<something> syntax can be used for more than
chars: it can be used for any char and any combination of modifiers
(such as the control modifier above), where the modifiers are stored in
the upper bits of the number.

An exception is made for those few chars whose combination which control
exists in ASCII, where the ?\C-<char> is then represented by the
corresponding ASCII char rather than use the "control" modifier bit.


        Stefan