Lays down the full test surface for the popout state machine ahead of
any transition logic. 62 collected tests across the four categories
from docs/POPOUT_REFACTOR_PLAN.md "Test plan":
1. Read-path queries (4 tests, all passing at commit 3 — these
exercise the parts of the skeleton that are already real:
compute_slider_display_ms, the terminal Closing guard, the
initial state defaults)
2. Per-state transition tests (~22 tests, all failing at commit 3
because the per-event handlers in state.py are stubs returning
[]. Each documents the expected new state and effects for one
specific (state, event) pair. These pass progressively as
commits 4-11 land.)
3. Race-fix invariant tests (6 tests — one for each of the six
structural fixes from the prior fix sweep: EOF race, double-
navigate, persistent viewport, F11 round-trip, seek pin,
pending mute replay. The EOF race test already passes because
dropping VideoEofReached in LoadingVideo is just "stub returns
[]", which is the right behavior for now. The others fail
until their transitions land.)
4. Illegal transition tests (17 parametrized cases — at commit 11
these become BOORU_VIEWER_STRICT_STATE-gated raises. At commits
3-10 they pass trivially because the stubs return [], which is
the release-mode behavior.)
All 62 tests are pure Python:
- Import only `booru_viewer.gui.popout.state` and `popout.viewport`
- Construct StateMachine() directly
- Use direct field mutation (`m.state = State.PLAYING_VIDEO`) for
setup, dispatch the event under test, assert the new state +
returned effects
- No QApplication, no mpv, no httpx, no filesystem outside tmp_path
- Sub-100ms total runtime (currently 0.31s including test discovery)
The forcing function: if state.py grows a PySide6/mpv/httpx import,
this test file fails to collect and the suite breaks. That's the
guardrail that keeps state.py pure as transitions land.
Test count breakdown (62 total):
- 4 trivially-passing (read-path queries + initial state)
- 22 transition tests (one per (state, event) pair)
- 6 invariant tests (mapped to the six race fixes)
- 17 illegal transition cases (parametrized over (state, event) pairs)
- 5 close-from-each-state cases (parametrized)
- 8 misc (state field persistence, window events)
Result at commit 3:
35 failed, 27 passed in 0.31s
The 27 passing are exactly the predicted set: trivial reads + the
illegal-transition pass-throughs (which work today because the stubs
return [] just like release-mode strict-state would). The 35 failing
are the transition handlers that need real implementations.
Phase A test suite (16 tests in tests/core/) still passes — this
commit only adds new tests, no existing test changed.
Test cases for state machine implementation (commits 4-11):
- Each failing test is its own commit acceptance criterion
- Commit N "passes" when the relevant subset of tests turns green
- Final state machine sweep (commit 11): all 62 tests pass
First regression-test layer for booru-viewer. Pure Python — no Qt, no
mpv, no network, no real filesystem outside tmp_path. Locks in the
security and concurrency invariants from the 54ccc40 + eb58d76 hardening
commits ahead of the upcoming popout state machine refactor (Prompt 3),
which needs a stable baseline to refactor against.
16 tests across five files mirroring the source layout under
booru_viewer/core/:
- tests/core/test_db.py (4):
- _validate_folder_name rejection rules (.., /foo, \\foo, .hidden,
~user, empty) and acceptance categories (unicode, spaces, parens)
- add_bookmark INSERT OR IGNORE collision returns the existing row
id (locks in the lastrowid=0 fix)
- get_bookmarks LIKE escaping (literal cat_ear does not match catear)
- tests/core/test_cache.py (7):
- _referer_for hostname suffix matching (gelbooru.com / donmai.us
apex rewrite, both exact-match and subdomain)
- _referer_for rejects substring attackers
(imgblahgelbooru.attacker.com does NOT pick up the booru referer)
- ugoira frame-count and uncompressed-size caps refuse zip bombs
before any decompression
- _do_download MAX_DOWNLOAD_BYTES enforced both at the
Content-Length pre-check AND in the chunk-loop running total
- _is_valid_media returns True on OSError (no delete + redownload
loop on transient EBUSY)
- tests/core/test_config.py (2):
- saved_folder_dir rejects literal .. and ../escape
- find_library_files walks root + 1 level, filters by
MEDIA_EXTENSIONS, exact post-id stem match
- tests/core/test_concurrency.py (2):
- get_app_loop raises RuntimeError before set_app_loop is called
- run_on_app_loop round-trips a coroutine result from a worker
thread loop back to the test thread
- tests/core/api/test_base.py (1):
- BooruClient._shared_client lazy singleton constructor-once under
10-thread first-call race
Plus tests/conftest.py with fixtures: tmp_db, tmp_library,
reset_app_loop, reset_shared_clients. All fixtures use tmp_path or
reset module-level globals around the test so the suite is parallel-
safe.
pyproject.toml:
- New [project.optional-dependencies] test extra: pytest>=8.0,
pytest-asyncio>=0.23
- New [tool.pytest.ini_options]: asyncio_mode = "auto",
testpaths = ["tests"]
README.md:
- Linux install section gains "Run tests" with the
pip install -e ".[test]" + pytest tests/ invocation
Phase B (post-sweep VideoPlayer regression tests for the seek slider
pin, _pending_mute lazy replay, and volume replay) is deferred to
Prompt 3's state machine work — VideoPlayer cannot be instantiated
without QApplication and a real mpv, which is out of scope for a
unit test suite. Once the state machine carves the pure-Python state
out of VideoPlayer, those tests become trivial against the helper
module.
Suite runs in 0.07s (16 tests). Independent of Qt/mpv/network/ffmpeg.
Test cases for Prompt 3:
- (already covered) — this IS the test suite Prompt 3 builds on top of
