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
Doctest workflow is flaky and seems to fail randomly, because some tests either expect some long string and get nothing, or they expect nothing and get said string. The string in question starts with (2): Using non-preferred coordinate operation between EPSG:2964 and EPSG:4326. [...].
There is already another issue about doctests. But it seems to be concerned with a problem about externally managed environment during pip install, which I cannot reproduce and do not encounter. Therefore this is a new issue concerning this new behavior.
After reproducing the test workflow locally, it seems the string in questions is a warning message from QGIS that appears only the first time when the Alaska airport layer with EPSG:2964 is added to the qgis project, which by default has EPSG:4326, because there are multiple possible reprojection methods. It seems that the warning string was added as an expected test output mid-July.
A possible explanation for the observed flaky behavior could be a combination of Sphinx Doctest not guaranteeing a fixed execution order of test groups and the warning being raised only the first time the airports layer is being added.
A possible fix could be to use a dataset in EPSG:4326 to hopefully get rid of the warning and remove the warning as expected test output.
The text was updated successfully, but these errors were encountered:
Doctest workflow is flaky and seems to fail randomly, because some tests either expect some long string and get nothing, or they expect nothing and get said string. The string in question starts with
(2): Using non-preferred coordinate operation between EPSG:2964 and EPSG:4326. [...]
.There is already another issue about doctests. But it seems to be concerned with a problem about externally managed environment during pip install, which I cannot reproduce and do not encounter. Therefore this is a new issue concerning this new behavior.
After reproducing the test workflow locally, it seems the string in questions is a warning message from QGIS that appears only the first time when the Alaska airport layer with EPSG:2964 is added to the qgis project, which by default has EPSG:4326, because there are multiple possible reprojection methods. It seems that the warning string was added as an expected test output mid-July.
A possible explanation for the observed flaky behavior could be a combination of Sphinx Doctest not guaranteeing a fixed execution order of test groups and the warning being raised only the first time the airports layer is being added.
A possible fix could be to use a dataset in EPSG:4326 to hopefully get rid of the warning and remove the warning as expected test output.
The text was updated successfully, but these errors were encountered: