From: Eric Abrahamsen <eric@ericabrahamsen.net>
Date: Fri, 6 May 2022 17:14:08 +0000 (-0700)
Subject: Fix handling of IMAP search strings
X-Git-Tag: emacs-29.0.90~1931^2~10
X-Git-Url: http://git.eshelyaron.com/gitweb/?a=commitdiff_plain;h=c57a6644ef97b3197c35c0c3ade60acd5607eef4;p=emacs.git

Fix handling of IMAP search strings

* lisp/gnus/gnus-search.el (gnus-search-imap-handle-string): This was
a misunderstanding of what `multibyte-string-p' means.  The check was
actually supposed to be whether the string was non-ascii or not.
---

diff --git a/lisp/gnus/gnus-search.el b/lisp/gnus/gnus-search.el
index 17724c3a514..369df81d9bd 100644
--- a/lisp/gnus/gnus-search.el
+++ b/lisp/gnus/gnus-search.el
@@ -1337,7 +1337,11 @@ elements are present."
 (cl-defmethod gnus-search-imap-handle-string ((engine gnus-search-imap)
 					      (str string))
   (with-slots (literal-plus) engine
-    (if (multibyte-string-p str)
+    ;; TODO: Figure out how Exchange IMAP servers actually work.  They
+    ;; do not accept any CHARSET but US-ASCII, but they do report
+    ;; Literal+ capability.  So what do we do?  Will quoted strings
+    ;; always work?
+    (if (string-match-p "[^[:ascii:]]" str)
 	;; If LITERAL+ is available, use it and encode string as
 	;; UTF-8.
 	(if literal-plus