Serve different storefronts per test

mock.endpoints.storefront is a static setter you re-assign per test. The mock reads the field on every request, so re-assignment after intercept(...) takes effect immediately — no need to rebuild the mock or rebind the page.

One mock, multiple tests

from musickit_api_mock import (
    MusicKitApiMock,
    Storefront,
    StorefrontResponseSuccess,
)


def _storefront(code: str, name: str, language: str) -> StorefrontResponseSuccess:
    return StorefrontResponseSuccess(
        storefront=Storefront(
            id=code,
            name=name,
            default_language_tag=language,
            supported_language_tags=[language],
            explicit_content_policy="allowed",
        ),
    )


mock = MusicKitApiMock()


def test_us_storefront(page):
    mock.endpoints.storefront = _storefront("us", "United States", "en-US")
    # ... drive the page ...


def test_jp_storefront(page):
    mock.endpoints.storefront = _storefront("jp", "Japan", "ja")
    # ... drive the page ...

Decide at request time

When the storefront should depend on something the test computes lazily (e.g. a fixture parameter), pass a zero-argument callable instead of a value:

mock.endpoints.storefront = lambda: _storefront(
    current_country.id,
    current_country.name,
    current_country.language,
)

current_country is your test's active country — a fixture parameter, dataclass instance, or shared state holding id / name / language attributes. The callable runs on every /v1/me/storefront request, so updating current_country between requests changes what the page sees on the next fetch.

Keep account and storefront aligned

Account.subscription_storefront and Storefront.id are independent fields — the mock does not enforce a relationship. When your app reads both, set them to the same storefront identifier to avoid the page receiving a mismatched pair MusicKit's docs never describe.