You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I'm pushing some code to github that I think will fix a bug I'm tracking with github issues. I'd like the ability to embed the information that this push will maybe fix the bug. Something like:
git commit -m " #maybe_fixes #124"
We can imagine that e.g. using testthat, we can have the issue in the title of the test, e.g.
test_that( "this works #213", {
...
} )
Then, if the test passes and the issue is open, we close it.
And, if the test fails, but the issue is closed. we reopen it.
I'm pretty sure we can use github api for that (ping @cscheid). I guess this would need a user token.
The text was updated successfully, but these errors were encountered:
This would be fantastic! I'd love to help make it work.
I don't know how credential management works with travis, but I've found it nontrivial to test rgithub itself in the past because I don't know how to distribute a test suite that uses the nonpublic parts of the github API, but not distribute the user tokens or client id/client secrets.
If you have some private stuff (e.g. a github token), then you go in the repo you're dealing with and you use travis encrypt VARIABLE=VALUE
It generates something you can put in your .travis.yml so that the rest of the world does not see your actual key, but travis can use it on your behalf. I've used it with Dropbox.
Scenario:
I'm pushing some code to github that I think will fix a bug I'm tracking with github issues. I'd like the ability to embed the information that this push will maybe fix the bug. Something like:
We can imagine that e.g. using
testthat
, we can have the issue in the title of the test, e.g.Then, if the test passes and the issue is open, we close it.
And, if the test fails, but the issue is closed. we reopen it.
I'm pretty sure we can use github api for that (ping @cscheid). I guess this would need a user token.
The text was updated successfully, but these errors were encountered: