-
Notifications
You must be signed in to change notification settings - Fork 79
BUG: Some tests using scripts:
non-functional on windows
#1152
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Comments
scripts:
non-functional (both false positive & false negatives)scripts:
non-functional on windows
I am not sure I understand. You want to test that Would you be able to make a small reproducer? Are you sure that your condition is correct in the windows test? It looks to me like you are testing the inverse of the unix tests here (on one you test for non-existence, and on the other for existence). We are just executing the script as given, with a few env vars. No big magic involved :) |
Can't reproduce right now because I'm running into #1054 during the zlib build, now also on x64. |
@zelosleone when you have time, can you make a very small reproducer and see if this is true? Some thing like should fail (as there is no package:
name: foobar
version: "0.1.0"
tests:
- script:
- if not exist %LIBRARY_BIN%\zlib.dll exit 1 |
Was not able to reproduce this at all with the same recipe, tried different dll names (with dummy ones too) with different if/else conditions too just in case |
Just in case, both checking zlib.dll is there or checking its not there works correctly. with |
Let's close this issue then. |
There are situations where the
package_contents
tests are not applicable (especially because file exclusion isn't implemented yet).In conda-forge/zlib-feedstock#83, I currently have
During the build (logs) on windows, we see that the file we're testing for is definitely there,
but the test fails
I understand that the positive tests can be formulated as
but what worries me far more is that anything under
tests: scripts:
produces false positives or negatives. If we can't trust that correctly specified, benign-looking tests are executed correctly (and doubly so if they're working underconda-build
), we're blocked from migrating at scale.The text was updated successfully, but these errors were encountered: