Rudolf Adamkovič <salu...@me.com> writes: > Good points; I will fix everything.
Please see the attached (3rd) revision of the patch. However, the tests still fail when ran via `make test'. See below. > Ihor Radchenko <yanta...@posteo.net> writes: > >> Your tests are failing on my side... > > The tests *pass* when ran in-buffer but *fail* when ran via 'make > test'. Just the tests that call (current-message) have this problem. > Interesting. I cannot figure out why these tests pass when ran from inside of Emacs but fail when ran via 'make tests'. I found in `default.mk' that Make runs `ert-run-tests-batch', but even that *passes* from within Emacs: 1. load `testing/org-test.el', 2. evaluate `ox-html/mathjax-legacy-scale-message-in-buffer` test in `testing/lisp/test-ox-html.el', and 3. run `M-: ert-run-tests-batch'. It produces the following output in the `*Messages*' buffer: Running 1 tests (2022-11-05 23:17:56+0100, selector ‘t’) Converting legacy MathJax scale: 20 to 0.2 passed 1/1 ox-html/mathjax-legacy-scale-message-in-buffer (0.008186 sec) Ran 1 tests, 1 results as expected, 0 unexpected (2022-11-05 23:17:56+0100, 0.010790 sec) It seems that some part of the *batch* testing swallows the messages in a way that makes `(current-message)' work incorrectly in tests. Any ideas? Asking for an Emacs veteran. :) P.S. As a workaround I also tried to set `set-message-function', dependency-injection style, but it did not work either. Rudy -- "'Contrariwise,' continued Tweedledee, 'if it was so, it might be; and if it were so, it would be; but as it isn't, it ain't. That's logic.'" -- Lewis Carroll, Through the Looking Glass, 1871/1872 Rudolf Adamkovič <salu...@me.com> [he/him] Studenohorská 25 84103 Bratislava Slovakia