Greetings,
the -k EXPRESSION matching behaviour of the pytest CLI, which deselects
tests based on the given expression, is case-sensitive. The relevant code
of the KeywordMapping is just a few lines:
https://github.com/pytest-dev/pytest/blob/master/src/_pytest/mark/legacy.py#L59-L63
That means u
Hi Christoph,
On Wed, Dec 4, 2019 at 9:38 AM Christoph Buelter <
buelter.christ...@gmail.com> wrote:
> Greetings,
>
> the -k EXPRESSION matching behaviour of the pytest CLI, which deselects
> tests based on the given expression, is case-sensitive.
>
More than once I've myself wished that it woul
Hey,
On Wed, Dec 04, 2019 at 10:58:09AM -0300, Bruno Oliveira wrote:
> I would actually vote to change the existing behavior to be
> case-insensitive. I doubt it would break many suites, if at all, and would
> be more in alignment with the intent of the option, more of it being a
> "partial match"
I think switching to case insensitive by default would be fine. Do we need a
flag for case sensitive? Not sure.
- Brian
> On Dec 4, 2019, at 6:26 AM, Florian Bruhin wrote:
>
> Hey,
>
>> On Wed, Dec 04, 2019 at 10:58:09AM -0300, Bruno Oliveira wrote:
>> I would actually vote to change the ex
Hi all,
it seems to me that a change to
https://github.com/pytest-dev/pytest/blob/master/src/_pytest/mark/legacy.py#L28
in addition to ensuring lowercase of the expression is all that's needed
on the technical side
on the release management side its a breaking change that needs a major
r
Thanks everyone for their input!
Christoph feel free to open a PR targeting the `features` branch with this
change.
For now I don't think we need an option to revert to the old behavior; if
people claim for it we can consider adding it in the future though.
Cheers,
Bruno
On Wed, Dec 4, 2019 at
As an aside, I would often like to use -k and have an exact match so e.g. I do
$ pytest -k test_xxx
and I want that to run test_xxx only and not test_xxx_1 etc. Is there
a way to do that?
--
Oscar
On Wed, 4 Dec 2019 at 20:56, Bruno Oliveira wrote:
>
> Thanks everyone for their input!
>
> C
Hi Oscar,
Not right now, the matching is always partial... not sure how we could
signal to `-k` that we want an exact match of the last part of the node id
(which seems what you want).
Cheers,
On Wed, Dec 4, 2019 at 7:40 PM Oscar Benjamin
wrote:
> As an aside, I would often like to use -k and
Using the full test id as shown by -v is what I always use for this case
instead of using -k
On Wed, 4 Dec 2019, 23:58 Bruno Oliveira, wrote:
> Hi Oscar,
>
> Not right now, the matching is always partial... not sure how we could
> signal to `-k` that we want an exact match of the last part of th
Ah thanks. I guess you mean
$ pytest somefile.py::test_xxx
which does indeed work. It would be nice if that string were included
in the output from test failures.
--
Oscar
On Wed, 4 Dec 2019 at 23:12, Floris Bruynooghe wrote:
>
> Using the full test id as shown by -v is what I always use f
Oscar,
> " It would be nice if that string were included in the output from test
failures. "
I'm pretty sure it is.
...
=== short test summary info
===
FAILED test_foo.py::test_one_more
= 1 failed, 1 passed in 0.16 seconds
Oops. My bad.
I'm used to running with -ra or -rf, which will produce the summary with
the test names.
- Brian
On Wed, Dec 4, 2019 at 3:37 PM Brian Okken wrote:
> Oscar,
> > " It would be nice if that string were included in the output from test
> failures. "
> I'm pretty sure it is.
> ...
> ===
Hi everyone,
in https://github.com/pytest-dev/pytest/issues/6080 it was noted that
https://github.com/apps/psf-chronographer ensures towncrier notes are in
prs,
i would like to propose enabling it for all of pytest-dev
Ronny
___
pytest-dev mail
13 matches
Mail list logo
