Skip to main content

Web Platform Test

Deno uses a custom test runner for Web Platform Tests. It can be found at ./tools/wpt.ts.

Running tests

If you are on Windows, or your system does not support hashbangs, prefix all ./tools/wpt.ts commands with deno run --unstable --allow-write --allow-read --allow-net --allow-env --allow-run.

Before attempting to run WPT tests for the first time, please run the WPT setup. You must also run this command every time the ./test_util/wpt submodule is updated:

./tools/wpt.ts setup

To run all available web platform tests, run the following command:

./tools/wpt.ts run

# You can also filter which test files to run by specifying filters:
./tools/wpt.ts run -- streams/piping/general hr-time

The test runner will run each web platform test and record its status (failed or ok). It will then compare this output to the expected output of each test as specified in the ./tools/wpt/expectation.json file. This file is a nested JSON structure that mirrors the ./test_utils/wpt directory. It describes for each test file, if it should pass as a whole (all tests pass, true), if it should fail as a whole (test runner encounters an exception outside of a test or all tests fail, false), or which tests it expects to fail (a string array of test case names).

Updating enabled tests or expectations

You can update the ./tools/wpt/expectation.json file manually by changing the value of each of the test file entries in the JSON structure. The alternative and preferred option is to have the WPT runner run all, or a filtered subset of tests, and then automatically update the expectation.json file to match the current reality. You can do this with the ./wpt.ts update command. Example:

./tools/wpt.ts update -- hr-time

After running this command the expectation.json file will match the current output of all the tests that were run. This means that running wpt.ts run right after a wpt.ts update should always pass.

Subcommands

setup

Validate that your environment is configured correctly, or help you configure it.

This will check that the python3 (or python.exe on Windows) is actually Python 3.

You can specify the following flags to customize behaviour:

--rebuild
Rebuild the manifest instead of downloading. This can take up to 3 minutes.

--auto-config
Automatically configure /etc/hosts if it is not configured (no prompt will be shown).

run

Run all tests like specified in expectation.json.

You can specify the following flags to customize behaviour:

--release
Use the ./target/release/deno binary instead of ./target/debug/deno

--quiet
Disable printing of `ok` test cases.

--json=<file>
Output the test results as JSON to the file specified.

You can also specify exactly which tests to run by specifying one of more filters after a --:

./tools/wpt.ts run -- hr-time streams/piping/general

update

Update the expectation.json to match the current reality.

You can specify the following flags to customize behaviour:

--release
Use the ./target/release/deno binary instead of ./target/debug/deno

--quiet
Disable printing of `ok` test cases.

--json=<file>
Output the test results as JSON to the file specified.

You can also specify exactly which tests to run by specifying one of more filters after a --:

./tools/wpt.ts update -- hr-time streams/piping/general

FAQ

Upgrading the wpt submodule:

cd test_util/wpt/
git fetch origin
git checkout origin/epochs/daily
cd ../../
git add ./test_util/wpt

All contributors will need to rerun ./tools/wpt.ts setup after this.

Since upgrading WPT usually requires updating the expectations to cover all sorts of upstream changes, it's best to do that as a separate PR, rather than as part of a PR that implements a fix or feature.