Mailman 3 September 2015 - pytest-dev

pytest-capturelog status/merging to core?
by holger krekel Nov. 9, 2015

Nov. 9, 2015

Hi all, pytest-capturelog has been abandoned [1] but is still getting used a lot as far as i see. As it basically helps dealing with capturing logging and logging is part of the python standard lib i think we could see to merge it into pytest and maintain it from there. Is anyone up for trying to do this along the lines of pytest-cache? FWIW I expect less breakage with pytest-capturelog because there is no change like there was with the execnet->json move while merging. It should be a relatively straightforward task. Alternatively we could see to get maintainership from pypi admins but i think in this case it's not worth it. best, hogler [1] I and others tried to contact the author several times to no avail.

7 16

how can one add fork to CI?
by holger krekel Sept. 29, 2015

Sept. 29, 2015

Hi, how can i add my hpk42/pytest and its branches to CI (travis/appveyor)testing? holger -- about me: http://holgerkrekel.net/about-me/ contracting: http://merlinux.eu

2 1

pytest-2.8.1: fixes and more fixes
by holger krekel Sept. 29, 2015

Sept. 29, 2015

Just 11 days after 2.8 hit PyPI we published pytest-2.8.1 which fixes a host of regressions and bugs, see the changelog excerpt below. Docs are also freshly generated at http://pytest.org including a streamlined PDF. Note that the pytest team now follows more SEMVER style versioning which currently means that you can expect bugfixes with 2.8.X and new features with 2.9.0 and planned incompatibilities with 3.0. Many thanks to the PR contributors and mergers of 2.8.1: Bruno Oliveira Ronny Pfannschmidt Florian Bruhin Floris Bruynooghe Eric Hunsberger Tim Staley Simon Gomizelj Russel Winder Oleg Alexandrov Michael Howitz Michael Birtwell Ben Webb Andy Freeland Abhijeet Kasurde cheers, holger krekel 2.8.1 ----- - fix #1034: Add missing nodeid on pytest_logwarning call in addhook. Thanks Simon Gomizelj for the PR. - 'deprecated_call' is now only satisfied with a DeprecationWarning or PendingDeprecationWarning. Before 2.8.0, it accepted any warning, and 2.8.0 made it accept only DeprecationWarning (but not PendingDeprecationWarning). Thanks Alex Gaynor for the issue and Eric Hunsberger for the PR. - fix issue #1073: avoid calling __getattr__ on potential plugin objects. This fixes an incompatibility with pytest-django. Thanks Andreas Pelme, Bruno Oliveira and Ronny Pfannschmidt for contributing and Holger Krekel for the fix. - Fix issue #704: handle versionconflict during plugin loading more gracefully. Thanks Bruno Oliveira for the PR. - Fix issue #1064: ""--junitxml" regression when used with the "pytest-xdist" plugin, with test reports being assigned to the wrong tests. Thanks Daniel Grunwald for the report and Bruno Oliveira for the PR. - (experimental) adapt more SEMVER style versioning and change meaning of master branch in git repo: "master" branch now keeps the bugfixes, changes aimed for micro releases. "features" branch will only be be released with minor or major pytest releases. - Fix issue #766 by removing documentation references to distutils. Thanks Russel Winder. - Fix issue #1030: now byte-strings are escaped to produce item node ids to make them always serializable. Thanks Andy Freeland for the report and Bruno Oliveira for the PR. - Python 2: if unicode parametrized values are convertible to ascii, their ascii representation is used for the node id. - Fix issue #411: Add __eq__ method to assertion comparison example. Thanks Ben Webb. - fix issue 877: properly handle assertion explanations with non-ascii repr Thanks Mathieu Agopian for the report and Ronny Pfannschmidt for the PR. - fix issue 1029: transform errors when writing cache values into pytest-warnings

1 0

pytest 2.8.1 issue status and blockers
by Ronny Pfannschmidt Sept. 28, 2015

Sept. 28, 2015

Hi everyone, over the weekend i had some time to do grooming and attempted a few fixes and did issue grooming the 1.8.1 milestone is down to 4 open issues https://github.com/pytest-dev/pytest/issues/1073 * a fix in pluggy is on the way thanks to Brono! https://github.com/pytest-dev/pytest/pull/1060 * that pull request is currently not merge-able since it brings back regressions we fixed pre-2.8, but the items it fixes are critical we def want that in Buck did a great job of discovering the cause and doing the fixes, unfortunately his refracting went too far and reintroduced an internal issue https://github.com/pytest-dev/pytest/issues/1037 * is a regression we def want to fix https://github.com/pytest-dev/pytest/issues/1026 * is not critical but nice to have documentation, ideally fixed in union with #1073 i consider #1060 is a release blocker, but it has a very high backward break potential as is, however i cant work on it myself this week, we can either try to include it or should make a big note as a known issue and try to do another patch release as soon as its ready current time-frames look like we should mark it as known issue -- Ronny

2 1

maintaining a synced git fork easily?
by holger krekel Sept. 28, 2015

Sept. 28, 2015

Hey git-users :) could someone help me with how to best manage my git fork of pytest? I currently have two separate directories for my pytest work -- one which was cloned from pytest-dev/pytest and one from hpk42/pytest which was forked using the github UI. Is it a good idea to try to maintain both "remotes" in a single directory on my laptop? I'd basically like to be able to make a branch from pytest-dev/pytest but have it stored in the hpk42/pytest fork repo. And only have a single "git pull/fetch". Instead currently i do this: cd pytest ; git pull cd ../pytest-hpk git co master ; git pull ../pytest ; git push origin git co features ; git pull ../pytest ; git push origin which seems a bit much -- also the last two "git co" commands usually involve a merge commit which seems wrong (i probably messed up at some point -- do i need to do a fresh fork/clone?). How are you solving this usage issue? holger

5 5

pytest-2.8.1 monday?
by holger krekel Sept. 26, 2015

Sept. 26, 2015

Hey all, i suggest we do a 2.8.1 release so that the fixes we already gathered reach a wider audience. Anyone wanting to do it? Any last PRs you want to do/see merged? Now is the time :) holger

3 2

Re: [pytest-dev] pytest-capturelog status/merging to core?
by holger krekel Sept. 25, 2015

Sept. 25, 2015

On Fri, Sep 25, 2015 at 12:24 +0200, Ronny Pfannschmidt wrote: > Before we can merge it we need to restructure capturing, since logcapture currently incurs a huge perf cost (at least it did the last time we tried to merge it) can't remember but ok :) On a cursory look it seems that capturelog creates and configures a LogHandler for each test. We can isntead probably just create one log handler per session and snapshot relevant loglines per test. Also we probably want to add an option "--log-capture=yes" (default no or "auto"). Lastly, i think we should consider merging capturelog into the core capture.py plugin as it's very related. Maybe I can give a PR a try next week. In any case, pytest-2.9 is not going to be out soon so we'd have some time to play with and settle mechanics. holger > Am 25. September 2015 11:43:16 MESZ, schrieb holger krekel <holger(a)merlinux.eu>: > >Hi all, > > > >pytest-capturelog has been abandoned [1] but is still getting used a > >lot > >as far as i see. As it basically helps dealing with capturing logging > >and logging is part of the python standard lib i think we could see > >to merge it into pytest and maintain it from there. Is anyone up > >for trying to do this along the lines of pytest-cache? > > > >FWIW I expect less breakage with pytest-capturelog because there is > >no change like there was with the execnet->json move while merging. > >It should be a relatively straightforward task. > > > >Alternatively we could see to get maintainership from pypi admins > >but i think in this case it's not worth it. > > > >best, > >hogler > > > > > >[1] I and others tried to contact the author several times to no avail. > > > >_______________________________________________ > >pytest-dev mailing list > >pytest-dev(a)python.org > >https://mail.python.org/mailman/listinfo/pytest-dev > > MFG Ronny > -- about me: http://holgerkrekel.net/about-me/ contracting: http://merlinux.eu

1 0

Re: [pytest-dev] (Help) How to replace __multicall__ with hookwrapper?
by Ernesto D. Luzon Jr. Sept. 23, 2015

Sept. 23, 2015

Hi Holger, Thanks a lot! That worked. Now I am no longer getting 'deprecated __multicall__' warnings from pytest 2.8. Many thanks, Ernesto D. Luzon Jr. On Wed, Sep 23, 2015 at 11:00 AM, <pytest-dev-request(a)python.org> wrote: > Send pytest-dev mailing list submissions to > pytest-dev(a)python.org > > To subscribe or unsubscribe via the World Wide Web, visit > https://mail.python.org/mailman/listinfo/pytest-dev > or, via email, send a message with subject or body 'help' to > pytest-dev-request(a)python.org > > You can reach the person managing the list at > pytest-dev-owner(a)python.org > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of pytest-dev digest..." > > > Today's Topics: > > 1. (Help) How to replace __multicall__ with hookwrapper? > (Ernesto D. Luzon Jr.) > 2. Re: (Help) How to replace __multicall__ with hookwrapper? > (holger krekel) > 3. Re: (Help) How to replace __multicall__ with hookwrapper? > (holger krekel) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Tue, 22 Sep 2015 20:27:06 +0100 > From: "Ernesto D. Luzon Jr." <edluzonjr(a)gmail.com> > To: pytest-dev(a)python.org > Subject: [pytest-dev] (Help) How to replace __multicall__ with > hookwrapper? > Message-ID: > < > CAEpTN6DN-vrKHhAq7oYkp9Wjyvhc7g4pV-WsV5hqa2yzwcYv3Q(a)mail.gmail.com> > Content-Type: text/plain; charset="utf-8" > > Hi All, > > I have code below that is based from the snippet provided by Holger: > > http://stackoverflow.com/questions/10754970/in-which-py-test-callout-can-i-… > > def pytest_runtest_makereport(item, call, __multicall__): > rep = __multicall__.execute() > setattr(item, "rep_" + rep.when, rep) > return rep > > Now, I'm struggling how to use hookwrapper instead of __multicall__. > I'd really appreciate if someone can help, thanks! > > Many thanks, > Ernesto D. Luzon Jr. >

1 0

(Help) How to replace __multicall__ with hookwrapper?
by Ernesto D. Luzon Jr. Sept. 22, 2015

Sept. 22, 2015

Hi All, I have code below that is based from the snippet provided by Holger: http://stackoverflow.com/questions/10754970/in-which-py-test-callout-can-i-… def pytest_runtest_makereport(item, call, __multicall__): rep = __multicall__.execute() setattr(item, "rep_" + rep.when, rep) return rep Now, I'm struggling how to use hookwrapper instead of __multicall__. I'd really appreciate if someone can help, thanks! Many thanks, Ernesto D. Luzon Jr.

2 2

pytest-2.8 / new branches
by holger krekel Sept. 22, 2015

Sept. 22, 2015

Hey all, just changed CONTRIBUTING to reflect the pytest-2.8 reality: https://github.com/pytest-dev/pytest/pull/1038 In it i suggest we use a "2.8.X" branch for bugfixes of the 2.8 series and master for features. We could also think about using "bug" and "feature" as branch names to relieve us from changing version numbers and to be more obvious to people at the command line. After that PR above is merged (or updated to the suggested alternative) we should actually create the branches and update version numbers accordingly. Also i noted that the 2.8.0 tag is missing unless i am missing something -- Bruno, did you actually follow HOWTORELEASE.rst as a checklist? :) cheers, holger -- about me: http://holgerkrekel.net/about-me/ contracting: http://merlinux.eu

2 4