all messages for Emacs-related lists mirrored at yhetil.org
 help / color / mirror / code / Atom feed
From: David Ponce via "Bug reports for GNU Emacs, the Swiss army knife of text editors" <bug-gnu-emacs@gnu.org>
To: Joseph Turner <joseph@breatheoutbreathe.in>
Cc: 70093@debbugs.gnu.org, Eli Zaretskii <eliz@gnu.org>
Subject: bug#70093: *** SPAM *** bug#70093: Remove unnecessary `image-tests--map-equal'
Date: Sun, 31 Mar 2024 12:08:27 +0200	[thread overview]
Message-ID: <d2e50c9a-c137-4c71-bc53-3fa2885ab37d@orange.fr> (raw)
In-Reply-To: <87frw6c0dv.fsf@breatheoutbreathe.in>

On 31/03/2024 09:37, Joseph Turner wrote:
> 
> David Ponce <da_vid@orange.fr> writes:
> 
>> On 30/03/2024 20:44, Joseph Turner wrote:
>>> With the resolution of bug#69992, it may be possible to remove the
>>> `image-tests--map-equal' equality predicate from lisp/image-tests.el.
>>> Eli and David, would you please apply the attached patch and then
>>> run
>>> the tests to see if we still need to allow for tolerances in
>>> `image--compute-map-and-original-map'?
>>> The tests pass for me, but then again, they never failed on my
>>> machine...
>>> Thank you!! I'm grateful for this fun collaboration. :)
>>> Joseph
>>>
>>
>> Hi Joseph,
>>
>> Did you forget to attach the patch? I cannot find it.
> 
> Apologies.  Here it is.  (This is the same patch as in my response to Eli)
> 
>> Anyway, I don't think the resolution of bug#69992 changed anything
>> regarding the test `image--compute-map-and-original-map'.  And to
>> be honest I don't understand why it gives a different result on our
>> configurations.
> 
> I thought perhaps calculating scale with
> 
> (/ (float (car size))
>                       (car (image-size
>                             (image--image-without-parameters image) t)))
> 
> instead of
> 
> (or (image-property image :scale) 1)
> 
> would fix the variations on our machines.
> 
>> To test, I also eval the following simple code in *scratch* buffer,
>> which gives me consistent results on my laptop:
>>
>> (let* ((svg (svg-create 200 100))
>>         (omap `(((rect . ((0 . 0) . (50 . 50)))
>>                  rect-test (help-echo "rect-test"))
>>                 ((circle . ((85 . 55) . 25))
>>                  circle-test (help-echo "circle-test"))
>>                 ((poly . [100 0 125 0 170 50 125 50])
>>                  poly-test (help-echo "poly-test")))))
>>    (svg-rectangle svg 0 0 50 50
>>                   :stroke-width 0 :fill "red")
>>    (svg-circle    svg 85 55 25
>>                   :stroke-width 0 :fill "green")
>>    (svg-polyline  svg '((100 . 0) (125 . 0) (170 . 50) (125 . 50))
>>                   :stroke-width 0 :fill "blue")
>>    (let ((image (svg-image svg
>>                            :original-map omap
>>                            :scale nil    ;; vary scale test
>>                            :rotation nil ;; vary rotation test
>>                            :flip nil     ;; vary flip test
>>                            )))
>>      ;; Must see the pointer and help-echo change when mouse hovers the
>>      ;; rect, circle and poly hot spots.
>>      (insert-image image)
>>      (insert ?\n)
>>      ;; Must see the same original map.
>>      (image--compute-original-map image)))
> 
> This manual test also works on my machine.  I had been doing something
> similar; I like your use of svg-create :)
> 
> Let me know if the unit tests pass with this patch applied.
> 
> Thank you!
> 
> Joseph
> 
Hello,

With the patch applied the test fails the same for me:

_______________________________________________________________
Selector: image--compute-map-and-original-map
Passed:  0
Failed:  1 (1 unexpected)
Skipped: 0
Total:   1/1

Started at:   2024-03-31 11:17:56+0200
Finished.
Finished at:  2024-03-31 11:17:56+0200

F

F image--compute-map-and-original-map
     Test ‘image--compute-map’ and ‘image--compute-original-map’.
     (ert-test-failed
      ((should (equal (image--compute-map image) rotated-map)) :form
       (equal
        (((circle ... . 24) "a" (help-echo "A"))
         ((rect ... 127 . 77) "b" (help-echo "B"))
         ((poly . [199 161 206 160 213 154 218 146 221 136 ...]) "c"
          (help-echo "C")))
        (((circle ... . 24) "a" (help-echo "A"))
         ((rect ... 54 . 77) "b" (help-echo "B"))
         ((poly . [126 161 133 160 140 154 145 146 148 136 ...]) "c"
          (help-echo "C"))))
       :value nil :explanation (list-elt 0 (list-elt 0 (cdr (car ...))))))
_______________________________________________________________

More in detail, these 3 tests pass:

1.
     (should (equal (image-property image :original-map)
                                     original-map))
2.
     (setf (image-property image :flip) nil)
     (setf (image-property image :rotation) 0)
     (setf (image-property image :scale) 2)
     (should (equal (image--compute-map image)
                                     scaled-map))
3.
     (setf (image-property image :scale) 2)
     (setf (image-property image :rotation) 90)
     (should (equal (image--compute-map image)
                                     scaled-rotated-flipped-map))

But these 2 tests fails:

1.
     (setf (image-property image :scale) 1)
     (setf (image-property image :rotation) 90)
     (should (equal (image--compute-map image)
                                     rotated-map))

     ;; Return
     ((((circle (199 . 41) . 24) "a" (help-echo "A"))
      ((rect (79 . 5) 127 . 77) "b" (help-echo "B"))
      ((poly . [199 161 206 160 213 154 218 146 221 136 223 125 221 114 218 104 213 96 206 91 199 89 191 91 185 96 179 104 176 114 175 125 176 136 179 146 185 154 191 160]) "c" (help-echo "C")))
     ;; Instead of
     (((circle (126 . 41) . 24) "a" (help-echo "A"))
      ((rect (6 . 5) 54 . 77) "b" (help-echo "B"))
      ((poly . [126 161 133 160 140 154 145 146 148 136 150 125 148 114 145 104 140 96 133 91 126 89 118 91 112 96 106 104 103 114 102 125 103 136 106 146 112 154 118 160]) "c" (help-echo "C"))))
     ;; All the result X coords are different, but Y coords are expected.

2.
     (setf (image-property image :rotation) 0)
     (setf (image-property image :flip) t)
     (should (equal (image--compute-map image)
                                     flipped-map))

     ;; Return
     ((((circle (205 . 29) . 24) "a" (help-echo "A"))
      ((rect (169 . 101) 241 . 149) "b" (help-echo "B"))
      ((poly . [85 29 86 22 92 15 100 10 110 7 121 5 132 7 142 10 150 15 155 22 157 29 155 37 150 43 142 49 132 52 121 53 110 52 100 49 92 43 86 37]) "c" (help-echo "C")))
     ;; Instead of
     (((circle (125 . 29) . 24) "a" (help-echo "A"))
      ((rect (89 . 101) 161 . 149) "b" (help-echo "B"))
      ((poly . [5 29 6 22 12 15 20 10 30 7 41 5 52 7 62 10 70 15 75 22 77 29 75 37 70 43 62 49 52 52 41 53 30 52 20 49 12 43 6 37]) "c" (help-echo "C"))))
     ;; All the result X coords are different, but Y coords are expected.

On my laptop:

   (image-size (image--image-without-parameters image) t) => (228 . 246)

Hope it will help to understand the reason.

Thanks!







  reply	other threads:[~2024-03-31 10:08 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-03-30 19:44 bug#70093: Remove unnecessary `image-tests--map-equal' Joseph Turner via Bug reports for GNU Emacs, the Swiss army knife of text editors
2024-03-30 22:04 ` bug#70093: *** SPAM *** " David Ponce via Bug reports for GNU Emacs, the Swiss army knife of text editors
2024-03-31  7:37   ` Joseph Turner via Bug reports for GNU Emacs, the Swiss army knife of text editors
2024-03-31 10:08     ` David Ponce via Bug reports for GNU Emacs, the Swiss army knife of text editors [this message]
2024-04-02 10:36       ` David Ponce via Bug reports for GNU Emacs, the Swiss army knife of text editors
2024-04-06  1:50         ` Joseph Turner via Bug reports for GNU Emacs, the Swiss army knife of text editors
2024-04-07 16:01           ` David Ponce via Bug reports for GNU Emacs, the Swiss army knife of text editors
2024-03-31  5:10 ` Eli Zaretskii
2024-03-31  5:32   ` Joseph Turner via Bug reports for GNU Emacs, the Swiss army knife of text editors
2024-03-31  7:33     ` Eli Zaretskii

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=d2e50c9a-c137-4c71-bc53-3fa2885ab37d@orange.fr \
    --to=bug-gnu-emacs@gnu.org \
    --cc=70093@debbugs.gnu.org \
    --cc=da_vid@orange.fr \
    --cc=eliz@gnu.org \
    --cc=joseph@breatheoutbreathe.in \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
Code repositories for project(s) associated with this external index

	https://git.savannah.gnu.org/cgit/emacs.git
	https://git.savannah.gnu.org/cgit/emacs/org-mode.git

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.