https://github.com/eventbrite/nose-knows
Knows is a nose plugin for figuring out which tests you need to run based on the files you have modified
https://github.com/eventbrite/nose-knows
Last synced: 3 months ago
JSON representation
Knows is a nose plugin for figuring out which tests you need to run based on the files you have modified
- Host: GitHub
- URL: https://github.com/eventbrite/nose-knows
- Owner: eventbrite
- License: bsd-3-clause
- Created: 2013-03-25T17:16:29.000Z (about 12 years ago)
- Default Branch: master
- Last Pushed: 2023-06-13T16:47:32.000Z (almost 2 years ago)
- Last Synced: 2024-08-08T22:14:57.500Z (10 months ago)
- Language: Python
- Size: 18.6 KB
- Stars: 0
- Watchers: 21
- Forks: 2
- Open Issues: 1
-
Metadata Files:
- Readme: README.rst
- License: LICENSE
Awesome Lists containing this project
README
==========
nose-knows
==========
**nose-knows** is a nose plugin for figuring out which unit tests you should
run after modifying code. It works by tracing your code while you run unit
tests, and creating an output file that can be used later.
Installing
==========
You can install **nose-knows** through ``pip`` or ``easy-install``::
pip install nose-knows
Or you can download the `latest development version`_, which may
contain new features.
Using nose-knows
================
**nose-knows** can be invoked in either input or output mode. In output mode
(``--knows-out``) it will generate a ``.knows`` file that contains the mapping
from a file in your code tree to the set of unit tests that run it. In input
mode, it uses the ``.knows`` file to selectively run unit tests based on the
files you pass in. Note: here, ``$BASE_DIR`` is the name of the base project
directory, not the (direct / relative) path to it. It is used to figure out how
to rename the file names in the output file to make it more portable.
Creating a ``.knows`` file::
eyal-01575:src eyal$ nosetests --with-knows --knows-dir=$BASE_DIR --knows-out
.....................................................................
----------------------------------------------------------------------
Ran 62 tests in 0.444s
OK
The ``.knows`` file now contains the following::
warehouse/src/load_data/sql_utils/checkpoints.py:
src.tests.test_load_data_statements:TestSQLStatement.test_finalize
src.tests.test_load_data_checkpoints:TestCheckpoints.test_checkpoint_finalize_with_delete
src.tests.test_load_data_checkpoints:TestCheckpoints.test_checkpoint_finalize
src.tests.test_load_data_checkpoints:TestCheckpoints.test_checkpoint_with_previous_checkin
src.tests.test_load_data_checkpoints:TestCheckpoints.test_checkpoint
src.tests.test_load_data_statements:TestSQLStatement.test_to_sql_import
src.tests.test_load_data_statements:TestSQLStatement.test_to_sql_schema_update
You can now run **nose-knows** in input mode, passing in
``load_data/sql_utils/checkpoints.py``::
eyal-01575:src eyal$ nosetests --with-knows --knows-dir=$BASE_DIR load_data/sql_utils/checkpoints.py
.......
----------------------------------------------------------------------
Ran 7 tests in 0.003s
OK
There is also (experimental at this point) support for py.test. You can
generate your ``.knows`` file via::
eyal-01575:src eyal$ py.test --with-knows --knows-dir=$BASE_DIR --knows-out
===================== test session starts ======================
platform darwin -- Python 2.7.1 -- pytest-2.3.4
plugins: nose-knows
collected 62 items
tests/test_load_data_checkpoints.py ....
tests/test_load_data_from_mysql.py ..
tests/test_load_data_statements.py .....
tests/test_process_data_denormalize.py ........
tests/test_process_data_mapping.py .....
tests/test_transform_data_daemon.py .
tests/test_transform_data_table_follower.py .................
tests/test_transformers/test_avg_data.py ..
tests/test_transformers/test_count.py ...
tests/test_transformers/test_join.py ....
tests/test_transformers/test_json_data.py .....
tests/test_transformers/test_min_data.py ..
tests/test_transformers/test_std_data.py ..
tests/test_transformers/test_sum_data.py ..
================== 62 passed in 2.18 seconds ===================
And selectively run specific unit tests like so::
eyal-01575:src eyal$ py.test --with-knows --knows-dir=$BASE_DIR load_data/sql_utils/checkpoints.py
===================== test session starts ======================
platform darwin -- Python 2.7.1 -- pytest-2.3.4
plugins: nose-knows
collected 62 items
tests/test_load_data_checkpoints.py ....
tests/test_load_data_statements.py ...
=================== 7 passed in 0.30 seconds ===================
The best practice here is to have a system like Jenkins run the unit test suite
once in a while to create this map (we have it running daily), and then
creating a bash function/script to download the knows output file from Jenkins
and run it against the set of changed files from a commit. Ours looks like::
function grab_latest_knows_output() {
NOW=`date +%s`
if [ ! -f $KNOWS_FILE_TMP ] ; then
curl --compressed $KNOWS_FILE_URL > $KNOWS_FILE_TMP
else
KNOWS_FILE_AGE=`stat -c %Y $KNOWS_FILE_TMP`
if [ `expr $NOW - $KNOWS_FILE_AGE` -gt "86400" ] ; then
curl --compressed $KNOWS_FILE_URL > $KNOWS_FILE_TMP
else
echo "Using latest knows output file."
fi
fi
}
function test_changed() {
grab_latest_knows_output
nosetests $KNOWS_FLAGS `git diff --name-only --cached origin | xargs`
}
function run_tests_for() {
grab_latest_knows_output
nosetests $KNOWS_FLAGS $@
}
License
========
**nose-knows** is copyright 2013 Eventbrite and Contributors, and is made
available under BSD-style license; see LICENSE for details.
.. _`latest development version`: https://github.com/eventbrite/nose-knows/tarball/master#egg=nose-nose