JSON-format activation #8727
JSON-format activation #8727
Conversation
|
@kalefranz Just a heads up to say this is the feature I discussed with you a while back. I believe this is one of the ideas you suggested and I definitely support this PR! |
|
This is very cool. We have really needed this for a long time. I'm going to spend some proper time reviewing this after I get the 4.7.1 release out. Thanks @necaris ! |
|
@jlstevens thanks for the ping about it! @msarahan y'all are very welcome! I'm looking forward to being able to use this functionality in my own editor integration ;-) By the way, I'm not wedded to any of the design decisions made in this PR so happy to change things up as you get time to review & comment. |
|
@necaris mind revisiting this? I think rebasing on master is a good start. 4.7.x is maintenance now, and this will be in the next feature release, 4.8. |
|
@msarahan yep, I'll have a look at it shortly! |
bcc2ee7 to
bd23d63
Compare
|
@msarahan could use a hand getting tests to pass -- I'm seeing a number of errors in the conda-build tests. The others all pass, and I'm not able to replicate the failures locally either. |
bd23d63 to
e4ddf47
Compare
|
@msarahan rebased this just now but still seeing test failures -- not sure what's causing them, could still use a hand figuring it out. |
|
@msarahan is this still of interest? |
|
Hey Rami, I'm moving back over from the enterprise side back to Conda for a while. Just starting as of yesterday, so need some time to ramp back up. Yes, this is still of interest. |
|
@kalefranz thanks! I'll rebase it against current master then, but will probably need a hand 8000 figuring out why some tests have continued to fail on CI (including code paths that I never touched 😉 ) |
|
Yeah I'm starting to dig into the tests now actually |
e4ddf47 to
5c4be6f
Compare
5c4be6f to
d1de3ed
Compare
|
@kalefranz don't know if this is still on your radar, but I've just rebased it against current master, FYI, so it should be up to date. |
|
Hey @necaris, I'm taking a fresh look at this |
|
@mcg1969 thanks! I could never quite figure out what was causing the test failures -- would love a hand getting this across the line. |
|
@kalefranz @chenghlee The test failures are all centered around failure to connect to the CodeCov service. Can we please work around those? The functional tests look good |
|
In the interest of keeping the blockers off the critical path for this I've reverted the restructuring in aef6861 -- I'll bring it back once I've had more of a chance to test things locally. |
|
Note that from what I can see, the AppVeyor tests are failing because |
Basically a copy of the XonshActivator tests, something to start from
This way activators for various shells can exist, but output things in JSON format!
aef6861 to
c0894ee
Compare
|
I've brought back the restructured dynamic activator class in c0894ee with a helper function, and correctly updated tests which were missing before. For good measure I've also rebased against upstream master. |
80c220c to
90a0772
Compare
This adds a separate formatters map to allow JSON-format output to be dynamically mixed in to whichever base activator is used. With this, passing `shell.posix+json` to `conda` will get the PosixActivator base class as well as the `JSONFormatMixin`.
...none of which I had anything to do with, but that's rebasing for you.
90a0772 to
7e5191d
Compare
|
Looks like the codecov issued is fixed -- thanks! The only thing that's not passing is AppVeyor, for the same reason as noted above: not IMHO related to this PR at all. |
|
This is great. We're getting very close. Will see if I can merge without AppVeyor passage, but I don't want to step on any toes on the team. |
|
OK, I know the problem we're seeing in CI |
|
boom! thanks folks |
|
Hi there, thank you for your contribution to Conda! This pull request has been automatically locked since it has not had recent activity after it was closed. Please open a new issue or pull request if needed. |
This is to serve the use-case of tools (e.g. editors) that want to know about conda environments, and potentially even activate them for the user. Rather than attempt to parse a shell script, why not simply provide the necessary values directly in a machine-readable format?