--- Begin Message ---
|
Subject: |
28.0.50; uudecode-decode-region-internal is broken |
|
Date: |
Tue, 03 Nov 2020 17:27:41 +0900 |
|
User-agent: |
Wanderlust/2.15.9 (Almost Unreal) SEMI-EPG/1.14.7 (Harue) FLIM-LB/1.14.9 (Gojō) APEL-LB/10.8 Emacs/28.0.50 (x86_64-w64-mingw32) MULE/6.0 (HANACHIRUSATO) (with unibyte mode) |
When I call uudecode-decode-region-internal in multibyte buffer, it
fails to decode eight-bit characters.
The function makes string from uuencoded text by passing unsigned char
vlue (0-255) to char-to-string function, which makes multibyte-string.
After that, string is decoded as binary. But eight-bit characters are
never made in that way.
(let ((ch #xc8))
(decode-coding-string (char-to-string ch) 'binary))
-> "8"
Additionally, concat and char-to-string functions are called so
frequently that deocder is very slow for large data.
Please see the below patch.
diff --git a/lisp/mail/uudecode.el b/lisp/mail/uudecode.el
index bcbd571b53..f9254aee75 100644
--- a/lisp/mail/uudecode.el
+++ b/lisp/mail/uudecode.el
@@ -149,12 +149,10 @@ uudecode-decode-region-internal
(setq counter (1+ counter)
inputpos (1+ inputpos))
(cond ((= counter 4)
- (setq result (cons
- (concat
- (char-to-string (ash bits -16))
- (char-to-string (logand (ash bits -8) 255))
- (char-to-string (logand bits 255)))
- result))
+ (setq result (cons (logand bits 255)
+ (cons (logand (ash bits -8) 255)
+ (cons (ash bits -16)
+ result))))
(setq bits 0 counter 0))
(t (setq bits (ash bits 6)))))))
(cond
@@ -166,26 +164,21 @@ uudecode-decode-region-internal
;;(error "uucode ends unexpectedly")
(setq done t))
((= counter 3)
- (setq result (cons
- (concat
- (char-to-string (logand (ash bits -16) 255))
- (char-to-string (logand (ash bits -8) 255)))
- result)))
+ (setq result (cons (logand (ash bits -8) 255)
+ (cons (logand (ash bits -16) 255)
+ result))))
((= counter 2)
- (setq result (cons
- (char-to-string (logand (ash bits -10) 255))
- result))))
+ (setq result (cons (logand (ash bits -10) 255)
+ result))))
(skip-chars-forward non-data-chars end))
+ (setq result (apply #'unibyte-string (nreverse result)))
(if file-name
(with-temp-file file-name
(set-buffer-multibyte nil)
- (insert (apply #'concat (nreverse result))))
+ (insert result))
(or (markerp end) (setq end (set-marker (make-marker) end)))
(goto-char start)
- (if enable-multibyte-characters
- (dolist (x (nreverse result))
- (insert (decode-coding-string x 'binary)))
- (insert (apply #'concat (nreverse result))))
+ (insert result)
(delete-region (point) end))))))
;;;###autoload
--
Kazuhiro Ito
--- End Message ---
--- Begin Message ---
|
Subject: |
Re: bug#44411: 28.0.50; uudecode-decode-region-internal is broken |
|
Date: |
Sat, 07 Nov 2020 11:42:58 +0200 |
> Date: Thu, 05 Nov 2020 19:48:08 +0900
> From: Kazuhiro Ito <kzhr@d1.dion.ne.jp>
> Cc: 44411@debbugs.gnu.org
>
> > Can you augment your patch along these lines, please?
>
> Here is a revised one.
Thanks, pushed to the emacs-27 branch.
Since with this patch you have exhausted the amount of changes we can
accept from you without copyright assignment, would you like to start
the assignment process at this time?
--- End Message ---