Consolidate tox run output into one report

The original implementation of erasing the coverage data file between
each test was the most accurate method, providing coverage stats in a
'cleanroom' setup, but resulted in extra reports being output. This
doesn't scale very well as more test files are added to the repository.

This patchset removes the extranenous erase command, and adds the `-a`
argument to the run commands, which allows coverage to accumulate data
across runs. Without the `-a` run, the coverage reports for each target
file are cross-contaminated, resulting in different coverage numbers.

Depends-On: I06729ac2bc1688a39255f2c8ea0d14131b5c2560

Change-Id: I4d52f7363e00bba62ed878fe8fa6e593eb21fdde
This commit is contained in:
Nolan Brubaker 2016-11-16 14:01:41 -05:00
parent 99e398125b
commit 010cadf114
1 changed files with 3 additions and 6 deletions

View File

@ -147,13 +147,10 @@ setenv =
PYTHONHASHSEED = 100
commands =
coverage erase
coverage run {toxinidir}/tests/test_inventory.py
coverage run -a {toxinidir}/tests/test_inventory.py
coverage run -a {toxinidir}/tests/test_manage.py
coverage run -a {toxinidir}/tests/test_ip.py
coverage report --show-missing --include={toxinidir}/playbooks/inventory/*,{toxinidir}/lib/*
coverage erase
coverage run {toxinidir}/tests/test_manage.py
coverage report --show-missing --include={toxinidir}/lib/*
coverage run {toxinidir}/tests/test_ip.py
coverage report --show-missing --include={toxinidir}/lib/*
[testenv:linters]