Skip to main content

Allows running Python tests (unittest, nose, pytest) in the context of VPL Moodle plugin (http://vpl.dis.ulpgc.es/)

Project description

Background

VPL (http://vpl.dis.ulpgc.es/) allows customizing evaluation process with custom execution files. This can be used for testing students’ submissions with professional testing frameworks. vpltest makes this easier for Python assignments by translating output from nose or pytest to the format required by VPL.

Basic usage

  1. Write normal unittest, nose or pytest tests without worrying about VPL.

  2. Test your tests without worrying about VPL.

  3. Create a VPL activity and upload your test files under Advanced settings => Execution files.

  4. Upload also vpltest.py

  5. To vpl_execution.sh write following lines:

#!/bin/bash
python3 vpltest.py

Replace python3 with python if you’re using Python 2.

Command line arguments

You can tweak vpltest with following command line arguments:

  • --nose and --pytest select the framework for collecting and running the tests. If neither flag is present, vpltest uses the one which happens to be installed. It prefers nose if both are installed;

  • --show-grade makes vpltest propose a grade based on how big part of the tests passed;

  • --show-stacktrace includes stacktraces in the reports;

  • --allow-deletion (see “Automatically keeping all execution files available”);

  • all other arguments are passed on to the testing framework.

Example. If you want vpltest to compute the grade, run your tests with pytest and make pytest run also doctests, then your vpl_evaluation.sh should be

#!/bin/bash
python3 vpltest.py --show-grade --pytest --doctest-modules

Test discovery

By default all execution files matching *test*.py (except vpltest.py) are considered to be test files and are passed on to testing framework. You can override this by specifying test files in vpl_evaluation.sh, eg:

#!/bin/bash
python3 vpltest.py --show-grade --pytest basic_tests.py style_checks.py

Test discovery inside the modules are up to chosen framework.

Installing to server

If you use vpltest a lot, then it makes sense to install it into your testing server so that you don’t need to upload it every time. It can be installed with pip, eg. sudo pip3 install pytest.

Note that now python3 vpltest.py in your vpl_execution.sh won’t do as there won’t be vpltest.py in testing directory anymore. Use python3 -m vpltest instead.

Automatically creating vpl_execution

VPL is designed so that vpl_execution.sh is not meant for running the tests but for executing the compilation phase, which includes preparing vpl_execution, which will be used to run the tests. This design makes sense with compiled languages, but with Python the compilation phase is just a nuisance. vpltest tries to make its usage experience smoother, by taking care of creating vpl_execution if necessary. That’s why you need to write only 2 lines into vpl_execution.sh.

Automatically keeping all execution files available

After compilation phase, VPL by default deletes all original execution files, unless you tick them under Advanced settings => Files to keep when running.

vpltest needs original test files to be present when running, but you don’t need remember it, because it uses a trick which makes originals always available. If you don’t like this, then run it with --allow-deletion.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

vpltest-0.6.zip (7.1 kB view details)

Uploaded Source

File details

Details for the file vpltest-0.6.zip.

File metadata

  • Download URL: vpltest-0.6.zip
  • Upload date:
  • Size: 7.1 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No

File hashes

Hashes for vpltest-0.6.zip
Algorithm Hash digest
SHA256 2c7bd09f57b19ce47a4f986bdd64af10888c59d62ebd297833f4bf298ac3ecbd
MD5 f77ea65129fb419b1d08e6c612e12033
BLAKE2b-256 17c0c8a501169037e278cdd7b10784ffeb82bab9c931e6ec9950ea719c300c2f

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page