bug#21072: Brave new mark-defun (and a testing tool)

[Marcin Borkowski <mbork@mbork.pl>]

On 2017-04-27, at 23:48, Noam Postavsky <npostavs@users.sourceforge.net> wrote:

> On Thu, Apr 27, 2017 at 12:43 PM, Marcin Borkowski <mbork@mbork.pl> wrote:
>>
>> On 2017-04-24, at 14:52, npostavs@users.sourceforge.net wrote:
>>
>>> Actually, now that I've applied this thing properly, I notice 2 test
>>> failures in batch mode:
>>>
>>> 2 unexpected results:
>>>    FAILED  mark-defun-arg-region-active
>>>    FAILED  mark-defun-no-arg-region-active
>>>
>>> They pass when run in interactive mode (I haven't looked into why).
>>
>> Strangely enough, they also fail with my original code.  Does anyone
>> have any idea why that might be the case?
>
> Oh I see what it is, your test depends on `transient-mark-mode' being
> enabled. With --batch, that defaults to nil. Should we be testing in
> both modes or just explicitly set it to t?

Good catch, thanks!  I'll set it to t, since this functionality is
intended to be used in t-t-m anyway.  OTOH, I'm curious where it breaks
with that disabled - I'll look into it now, I guess.

> PS I noticed some problems with elisp-tests-with-temp-buffer while
> looking at this, I pushed a fix to fix/bug-21072 [1: 04741f0].
>
> 1: 2017-04-27 17:38:09 -0400 04741f02cca95147581e05cb49b54e6dbf8bed56
>   Fix elisp-tests-with-temp-buffer compilation

Thanks a lot, I'll take a look!

Best,

-- 
Marcin Borkowski