bug#27112: Setting comment-continue to " " causes uncomment-region-default to infloop

[npostavs@users.sourceforge.net]

[-- Attachment #1: Type: text/plain, Size: 409 bytes --]

tags 27112 patch
quit

Clément Pit--Claudel <clement.pitclaudel@live.com> writes:

>> Can you give a full recipe?
>
> Sure thing:
>
> emacs -Q --eval '(setq comment-continue "  ")' --eval '(save-excursion (insert ";; AAA\n"))' --eval '(comment-line 1)'

Oh it triggers in lisp mode, I somehow got the idea that it needed a
mode with C like comments from initial report.  How about this patch:


[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #2: patch --]
[-- Type: text/x-diff, Size: 1149 bytes --]

From d7030832e1aecbb04affe52453fa32ea19ed7ded Mon Sep 17 00:00:00 2001
From: Noam Postavsky <npostavs@gmail.com>
Date: Thu, 1 Jun 2017 23:09:36 -0400
Subject: [PATCH v1] Fix infloop in uncomment-region-default (Bug#27112)

When `comment-continue' has only blank lines, `comment-padright'
produces a regexp that matches the empty string, so
`uncomment-region-default' will loop infinitely.
* lisp/newcomment.el (comment-padright): Check for non-padding
characters in STR.
---
 lisp/newcomment.el | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/lisp/newcomment.el b/lisp/newcomment.el
index 4b261c34c6..118549f421 100644
--- a/lisp/newcomment.el
+++ b/lisp/newcomment.el
@@ -815,7 +815,7 @@ (defun comment-padright (str &optional n)
 If N is `re', a regexp is returned instead, that would match
 the string for any N."
   (setq n (or n 0))
-  (when (and (stringp str) (not (string= "" str)))
+  (when (and (stringp str) (string-match "\\S-" str))
     ;; Separate the actual string from any leading/trailing padding
     (string-match "\\`\\s-*\\(.*?\\)\\s-*\\'" str)
     (let ((s (match-string 1 str))	;actual string
-- 
2.11.1