A small library used to write automated test for Discord bots.
Test framework originally written by DXsmiley, update to the rewrite and re-worked UI done by me, with tons of help from ALobsterDog
Want to see the bots in action? Join us here on discord for help and to see the example bots whenever they are being tested. See you there!
Just about ready for a 1.0, just need to add a few more features. Until I get everything to a 1.0 though, expect the command line arguments and everything else to be volatile. In addition to the example here, my main bot will be implementing this soon.
Best way to get this running is to git clone
it. The example bot's usage can be seen below.
In order to use this package, you'll have to create two bot accounts. The first, referred to as the target, is the bot that you wish to test. If you're already here you probably have a token for this bot already. The second one, called the tester is responsible for running the tests.
The most important thing to know about this bot is that it has two modes, interactive and CLI. Interactive mode is used if you want to manually run the tests from inside discord. I also use it when testing the new tests. CLI mode allows you to start the bot and run the tests without any interaction, which is the mode that you will probably want to use when using this package in most cases. More information on how these two modes are used is available further down the readme.
First, you need to run the bot that you wish to test. You can run the example bot supplied with this repo as follows:
python example_bot.py TARGET_TOKEN
Then, you have to run the tester bot (This will run it in interactive mode - CLI mode demo shown later):
python example_tester.py TARGET_USERNAME TESTER_TOKEN
Make sure to use the username of the bot without the discriminator (#1111) or the bot won't understand what the user is
Once both bots are running, go to any discord channel that both bots have access to and type ::run all
to run all the tests. Use ::help
to get more information on the commands that the tester can use in interactive mode.
Start by writing functions to do the tests, examples can be seen in the example tester. Decorate the functions you want to be run as tests with @distest.TestCollector()
and they will be available to be run. Instead of calling run_bot()
like you would normally do to start a bot with discord.py, use distest.run_dtest_bot()
and feed in the requested parameters. Based on the sysargs that are used to run the bot file, it will automatically run the bot in interactive or CLI mode. Basically, if you feed the bot the -c
parameter to specify the channel that the tests should be run in, the bot will run those automatically. Otherwise, it will wait for the commands in discord as described below.
CLI mode is designed to be used to run the tests in a normal way, such as on git hooks or a CI/CD pipeline. The following text is the usage snippet from the help command followed by some more general information from me. It isn't the same as the help message, as I tried to make it more in-depth.
usage: example_tester.py [-h] [-c channel]
[--run {all,test_reply_matches} | --stats]
target_bot_user tester_bot_token
Always Required
-
target_bot_user
: The username (no discriminator) of the target bot. Same as described above in the general section. -
tester_bot_token
: The token that will be used to run the tester bot. Also the same as described above in general.
CLI Mode
-
run
: Specifies if you will run all tests or a subset of them -
stats
: Runs the bot in stats mode. Mutually exclusive withrun
. (Not very useful, may be removed. If you use it in some way, open an issue and let me know!) -
channel
: The channel ID that the tests will be conducted in. Just need the int ID
Other
-h
: Just shows the help command. This is only the usage message, there is other information in the help.
Sample Command
The command I used to test this bot is available in run_tester.sh
Commands you can run in discord once the bot is running in interactive mode.
::stats
Gives details about which tests have been
run and what the results were
::run test_name
Run a particular test. Options are methods decorated
with `@distest.TestCollector()` in the tester bot.
::run all
Run all tests
::run unrun
Run all tests that have not yet been run
::run failed
Run all tests that failed on the most recent run
Please open an issue for your contribution and tag it with contribution to discuss it. I recommend waiting for a response before pouring hours and hours into the contribution, but it will likely be approved either way. The other thing is to make sure you check the github project to see if there is someone else already working on it who you can help. Other notes:
-
If you are adding new test types, please make sure you test them well to make sure they work as intended, and please add a demo of them in use to the
example_tests()
for others to see. When you are done, please open a PR and I'll add it in! -
I use Black for my code formatting, it would be appreciated if you could use it when contributing as well. I will remind you when you make a PR if you don't, it is essential to make sure that diffs aren't cluttered up with silly formatting changes. Additionally, CodeFactor should be tracking code quality and doing something with PRs. We will see soon exactly how that will work out.
-
Make sure that your issue goes onto the project, that's how we keep track of to-do and in progress things.
-
Also, if you just want to propose an idea, create an issue and tag it with enhancement. Thank you for your help!
I used to use this, now using the GitHib project instead. This will no longer be used.
- Update the Bot to the newly re-written discord.py 1.0
- Verify each test and add to
example_tests.py
- send_message
- edit_message
- wait_for_reaction
- wait_for_message
- wait_for_reply
- assert_message_equals
- assert_message_contains
- assert_message_matches
- assert_reply_equals
- assert_reply_contains
- assert_reply_matches
- assert_reaction_equals
- ensure_silence
- ask_human
- Allow running tests from the command line
- Add CLI mode section to the readme
- Have the bot return exit codes depending on test status
- More test types
- Prep for pip, get it packaged