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
We have a multi-step build which has an "integration" test phase as well as a unit test phase.
We use scoverage to instrument these runs.
Currently (see a9aaf621e7, where no justification is given), a side-effect of scoverage:compile is to wipe the coverage data so far.
This is a surprising side-effect of a "compile" command.
Why does scoverage do this?
Should it stop doing so?
Thanks
(This doesn't currently cause problems for us in general, as we don't recompile the app between the various test runs described above. However we do have one dev machine which is playing up and recompiling the app when it shouldn't. That machine is reporting incorrect coverage data, and it took me ages to track down why.)
The text was updated successfully, but these errors were encountered:
I think it was because if you run scoverage:compile over the top and a statement was removed / added, the id numbers used for statements in the subsequent run would clash.
We have a multi-step build which has an "integration" test phase as well as a unit test phase.
We use
scoverage
to instrument these runs.Currently (see a9aaf621e7, where no justification is given), a side-effect of
scoverage:compile
is to wipe the coverage data so far.This is a surprising side-effect of a "compile" command.
scoverage
do this?Thanks
(This doesn't currently cause problems for us in general, as we don't recompile the app between the various test runs described above. However we do have one dev machine which is playing up and recompiling the app when it shouldn't. That machine is reporting incorrect coverage data, and it took me ages to track down why.)
The text was updated successfully, but these errors were encountered: