跑自动化时经常会出现这样一个情况，一轮自动化跑完后零星出现了几个失败case，无法断定失败的原因，所以需要重新跑一下失败的case去debug，那我们要做的是就去修改脚本把那几个case筛选出来重新run。在pytest 中，你就无需这样做，因为cache功能，他能把上次跑的情况记录下来。不知道你们有没有注意到，用pytest run完case后会在当前目录下生成.pytest_cache,里面就保存了上一次run的信息。
跟cache有关的命令参数
- --last-failed, 如果run的时候跟这个参数只会运行上次失败的用例，这就解决了上面说的需求。
- --failed-first，如果run的时候跟这个参数会先运行上次失败的case，然后再run其余的case。
- --cache-show，跟上个参数，会显示上次run的信息。
- --cache-clear, 在run前先把之前的cache清除。
例如：
E:\jenkins\workspace\pytest_junit_report\learnpytest\ping_test>pytest -q --tb=no Test session starts (platform: win32, Python 3.7.0, pytest 5.1.3, pytest-sugar 0.9.2) rootdir: E:\jenkins\workspace\pytest_junit_report\learnpytest\ping_test, inifile: pytest.ini plugins: allure-pytest-2.8.11, assume-2.2.0, html-2.0.0, metadata-1.8.0, sugar-0.9.2 product_a/test_smoke.py ? 14% █▌ product_a/test_smoke.py ?? 29% ██▉ product_a/test_smoke.py ??? 43% ████▍ product_b/test_smoke.py ? 57% █████ product_b/test_smoke.py ?? 71% █████ product_b/test_smoke.py ??? 86% █████ product_b/test_smoke.py ???? 100% █████ █████ Results (0.12s): 6 passed 1 failed - product_b/test_smoke.py:1 test_smoket_1
一共七个case，一个failed。
--last-failed
E:\jenkins\workspace\pytest_junit_report\learnpytest\ping_test>pytest --tb=no --lf Test session starts (platform: win32, Python 3.7.0, pytest 5.1.3, pytest-sugar 0.9.2) rootdir: E:\jenkins\workspace\pytest_junit_report\learnpytest\ping_test, inifile: pytest.ini plugins: allure-pytest-2.8.11, assume-2.2.0, html-2.0.0, metadata-1.8.0, sugar-0.9.2 collecting ... run-last-failure: rerun previous 1 failure (skipped 9 files) product_b/test_smoke.py ? 100% █████ █████ Results (0.07s): 1 failed - product_b/test_smoke.py:1 test_smoket_1 3 deselected
这里只run了上一次失败的case
--failed-first
E:\jenkins\workspace\pytest_junit_report\learnpytest\ping_test>pytest --tb=no --ff Test session starts (platform: win32, Python 3.7.0, pytest 5.1.3, pytest-sugar 0.9.2) rootdir: E:\jenkins\workspace\pytest_junit_report\learnpytest\ping_test, inifile: pytest.ini plugins: allure-pytest-2.8.11, assume-2.2.0, html-2.0.0, metadata-1.8.0, sugar-0.9.2 collecting ... run-last-failure: rerun previous 1 failure first product_b/test_smoke.py ? 14% █▌ product_a/test_smoke.py ? 29% ██▉ product_a/test_smoke.py ?? 43% ████▍ product_b/test_smoke.py ???? 100% █████ █████ Results (0.11s): 6 passed 1 failed - product_b/test_smoke.py:1 test_smoket_1
这里先会run之前失败的case 然后在run其他的。
--cache-show
E:\jenkins\workspace\pytest_junit_report\learnpytest\ping_test>pytest --cache-show Test session starts (platform: win32, Python 3.7.0, pytest 5.1.3, pytest-sugar 0.9.2) rootdir: E:\jenkins\workspace\pytest_junit_report\learnpytest\ping_test, inifile: pytest.ini plugins: allure-pytest-2.8.11, assume-2.2.0, html-2.0.0, metadata-1.8.0, sugar-0.9.2 cachedir: E:\jenkins\workspace\pytest_junit_report\learnpytest\ping_test\.pytest_cache -------------------------------------- cache values for ‘*‘ --------------------------------------- cache\lastfailed contains: {‘product_b/test_smoke.py::test_smoket_1‘: True} cache\nodeids contains: [‘product_a/test_smoke.py::test_smoket_1‘, ‘product_a/test_smoke.py::test_smoket_2‘, ‘product_a/test_smoke.py::test_smoket_3‘, ‘product_b/test_smoke.py::test_smoket_1‘, ‘product_b/test_smoke.py::test_smoket_2‘, ‘product_b/test_smoke.py::test_smoket_3‘, ‘product_b/test_smoke.py::test_smoket_4‘] cache\stepwise contains: []