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
The next merge after PR #2234 was @jdufresne's PR #2199, a few minutes later the same day, which also contained a change to .travis.yml, adding pip install -e .:
There's a difference around line 3838 (looking at the line numbers on the left) when running $ lcov --capture --directory . -b . --output-file coverage.info:
Scanning . for .gcda files ...
Found 1 data files in .
Processing _webp.gcda
Finished .info-file creation
->
Scanning . for .gcda files ...
geninfo: ERROR: no .gcda files found in .!
Note that local Python coverage reports worked, but not the stuff to combine the C and Python reports and send to Coveralls.
Replacing pip install -e . with pip install olefile fixes the build, and sends reasonable-looking coverage numbers to Coveralls.
I've sent a PR for this (#2300), but is it the correct thing to do? Can/should we fix whatever pip install -e . did wrong?
The last coverage report on Coveralls is from a week ago, 13 December 2016, when PR #2234 was merged: https://coveralls.io/repos/19620/builds
The next merge after PR #2234 was @jdufresne's PR #2199, a few minutes later the same day, which also contained a change to .travis.yml, adding
pip install -e .
:https://github.com/python-pillow/Pillow/pull/2199/files#diff-354f30a63fb0907d4ad57269548329e3R45
Here's a diff of the Py2.7 builds from the #2234 and #2199 merges:
https://www.diffchecker.com/GPJlMb1Y
There's a difference around line 3838 (looking at the line numbers on the left) when running
$ lcov --capture --directory . -b . --output-file coverage.info
:->
Note that local Python coverage reports worked, but not the stuff to combine the C and Python reports and send to Coveralls.
Replacing
pip install -e .
withpip install olefile
fixes the build, and sends reasonable-looking coverage numbers to Coveralls.I've sent a PR for this (#2300), but is it the correct thing to do? Can/should we fix whatever
pip install -e .
did wrong?pip install -e .
did this:Perhaps
python setup.py develop
is the problem?The text was updated successfully, but these errors were encountered: