Tryke

tryke.mp4

Getting started

For more information, see the documentation.

Write a test.

from typing import Annotated

import tryke as t


# Fixtures with `per="scope"` are cached for their scope:
# - Module-level for globally defined fixtures
# - Per `describe` block for fixtures defined within a `describe` block
@t.fixture(per="scope")
def database():
    db = {}
    yield db
    db.clear()


with t.describe("users"):
    # By default, fixtures run per-test
    # Fixtures can be composed by requesting other fixtures
    @t.fixture
    def users(database: Annotated[dict[str, dict[str, str]], t.Depends(database)]):
        database["users"] = {}
        return database["users"]

    with t.describe("get"):
        # Define display labels for tests
        # Async tests are supported
        @t.test("returns a stored user")
        async def test_get(users: Annotated[dict[str, str], t.Depends(users)]):
            users["alice"] = "alice@example.com"
            # Pass a label as the second argument to expect() so reports
            # show "returns stored email" instead of the raw expression
            t.expect(users["alice"], "returns stored email").to_equal(
                "alice@example.com"
            )

    with t.describe("set"):

        @t.test("stores a new user")
        async def test_set(users: Annotated[dict[str, str], t.Depends(users)]):
            users["bob"] = "bob@example.com"
            t.expect(users["bob"], "stores email under user key").to_equal(
                "bob@example.com"
            )

Run your tests — tryke test --watch for an always-on loop, tryke test --changed for just what your working tree touched, or plain:

uvx tryke test

tryke test v0.0.25

sample.py:
  users
    get
      ✓ returns a stored user [0.00ms]
        ✓ returns stored email
    set
      ✓ stores a new user [0.00ms]
        ✓ stores email under user key

 Test Files  1 passed (1)
      Tests  2 passed (2)
   Start at  10:02:24
   Duration  36.36ms (discover 0.76ms, tests 35.60ms)

  PASS

Coming from pytest?

The migration guide has a side-by-side cheat sheet and — faster — a copy-paste LLM prompt that walks an AI coding assistant through a phased, gated pytest → Tryke migration with discovery- and results-parity checks built in.

License

This repository is licensed under the MIT License.