bug-gnu-emacs
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

bug#55849: 29.0.50; process-async-https-with-delay error detection is wr


From: Ken Brown
Subject: bug#55849: 29.0.50; process-async-https-with-delay error detection is wrong
Date: Wed, 8 Jun 2022 12:30:31 -0400
User-agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:91.0) Gecko/20100101 Thunderbird/91.9.1

process-async-https-with-delay in test/src/process-tests.el fails on
Cygwin as follows:

Test process-async-https-with-delay condition:
    (ert-test-failed
     ((should
       (>
        (buffer-size buf)
        0))
      :form
      (> 0 0)
      :value nil))

I haven't yet figured out why, but in attempting to debug this I
discovered that the error reporting in the test is faulty.  What
actually happens is that the connection fails, and this failure is
indicated by the 'status' variable.  The latter is a plist, but the test
uses '(should-not (assq :error status))' as though it were an alist.

The following patch fixes the problem:

diff --git a/test/src/process-tests.el b/test/src/process-tests.el
index f5908d3cda..824c6da119 100644
--- a/test/src/process-tests.el
+++ b/test/src/process-tests.el
@@ -931,7 +931,7 @@ process-async-https-with-delay
                         (< (float-time) (+ t0 limit)))
               (sit-for 0.1)))
           (should status)
-          (should-not (assq :error status))
+          (should-not (plist-get status ':error))
           (should buf)
           (should (> (buffer-size buf) 0))
           )

With this patch the correct reason for the failure is shown:

Test process-async-https-with-delay condition:
    (ert-test-failed
     ((should-not
       (plist-get status ':error))
      :form
      (plist-get
       (:error
         (error connection-failed "deleted\n" :host "elpa.gnu.org" :service 
443))
       :error)
      :value
      (error connection-failed "deleted\n" :host "elpa.gnu.org" :service 443)))

OK to apply?  If so, to which branch?  The problem exists on both master
and the release branch.

Ken





reply via email to

[Prev in Thread] Current Thread [Next in Thread]