datasette/tests/conftest.py

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

421 lines
13 KiB
Python
Raw Normal View History

import importlib.metadata
import os
import pathlib
import re
import socket
import subprocess
import sys
import tempfile
import time
from dataclasses import dataclass
import httpx
import pytest
import pytest_asyncio
from datasette import Event, hookimpl
try:
import pysqlite3 as sqlite3
except ImportError:
import sqlite3
UNDOCUMENTED_PERMISSIONS = {
"this_is_allowed",
"this_is_denied",
"this_is_allowed_async",
"this_is_denied_async",
"no_match",
# Test actions from test_hook_register_actions_with_custom_resources
"manage_documents",
"view_document_collection",
"view_document",
}
2026-04-16 20:44:21 -07:00
def wait_until_responds(url, timeout=5.0, client=httpx, process=None, **kwargs):
start = time.time()
while time.time() - start < timeout:
# If the server died there is no point waiting out the timeout - fail
# now, with its output, instead of after `timeout` seconds of silence
if process is not None and process.poll() is not None:
raise AssertionError(
"Server exited early with returncode {}\n{}".format(
process.returncode, process.stdout.read().decode("utf-8")
)
)
try:
client.get(url, **kwargs)
return
except httpx.TransportError:
time.sleep(0.1)
raise AssertionError(f"Timed out waiting for {url} to respond")
def find_free_port():
with socket.socket() as sock:
sock.bind(("127.0.0.1", 0))
return sock.getsockname()[1]
Add a plugin telemetry kit: public registry API, linked_root_span_kwargs, test helpers, docs A survey of five plugin OTel plans (datasette-paper, -agent, -litestream, -accounts, -cron) found every one hand-copying the same core machinery: the registry classes, the conformance-test harness, the pytest fixtures, the bucket boundaries and the detached-root-with-Link recipe. This makes that machinery importable instead: - The registry classes are documented public API. Attribute gains values= (a closed enum the conformance helpers enforce - what makes an attribute safe as a metric dimension); SpanName gains prefix=True for span families like "chat {model}" whose names share a fixed prefix, matched by span_for() after exact names. span_for()/attribute helpers accept a spans= tuple so plugin registries can use them. - datasette.telemetry.linked_root_span_kwargs(): the root-span-with-Link shape for work a request caused without containing - background jobs, scheduled ticks, block=False writes. Core's own write thread now uses it instead of building the kwargs inline. - datasette.telemetry_testing: the session provider fixtures, otel_spans / otel_metrics, a two-way registry conformance checker (including enum and prefix handling, filtered by instrumentation scope) and an assert_package_never_imports_sdk() guard. Core's conftest now imports these instead of defining them, so the suite consumes the kit exactly as a plugin's would. - New "Telemetry for plugin authors" docs page: scope discipline, registry usage, privacy/cardinality rules, named-callable guidance, request_span(), the background root-with-link convention (one root per tick, always emitted), provider-ordering facts and known caveats. request_span() is now documented public API. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_012U7coQfVu8nK2R4q2mCULA
2026-09-02 12:24:42 -07:00
# The otel fixtures moved to datasette.telemetry_testing, which is public
# plugin API - core's suite consumes it exactly the way a plugin's would.
from datasette.telemetry_testing import ( # noqa: F401, E402
MetricsCollector,
otel_metrics,
otel_meter_provider,
otel_provider,
otel_spans,
)
Add OpenTelemetry metrics for SQL thread pool saturation and query latency Spans describe requests that have finished. They structurally cannot answer "am I saturating my 3 SQL threads right now", because that is a level rather than an event - and with num_sql_threads defaulting to 3, it is usually the first thing worth knowing about a busy Datasette. This adds the metrics that answer it. Five observable gauges, computed only when something is collecting, so an instance with no MeterProvider installed does no work for them at all: datasette.sql.threads.limit num_sql_threads datasette.sql.threads.queue_depth queries waiting for a free thread datasette.sql.queries.pending in-flight reads, by db.namespace datasette.write.queue_depth writes behind the single write thread datasette.connections.open tracked file connections Three instruments recorded inline, which matters because metrics survive trace sampling and spans do not - an operator sampling 1% of traces still gets 100% of the latency distribution: db.client.operation.duration semconv histogram, with error.type datasette.write.queue_wait the metric twin of the existing span datasette.sql.queries.interrupted sql_time_limit_ms kills The interrupted counter closes a gap the plan called out as unanswerable: "how often are we killing queries at the limit" is a rate, and a rate cannot be recovered from sampled spans. Core still creates no provider of any kind, so the architecture is unchanged; `grep -rn 'opentelemetry.sdk' datasette/` stays empty. One real difference from tracing is worth recording: _ProxyMeter and its instruments forward to a provider installed after they were created, whereas ProxyTracer permanently caches the first concrete tracer it resolves. Module-level instruments are therefore safe and the test fixture has no ordering constraint. Live instances are tracked in a lock-guarded WeakSet so instrumenting an instance never keeps it alive. The pool gauges carry no attribute saying which Datasette produced them: production runs one instance per process, and adding an id to disambiguate the test suite's hundreds of instances would buy unbounded attribute cardinality to fix a case that does not occur. The collision is documented instead, and the gauge callbacks are plain generator functions so tests can assert exact values by calling them directly rather than through the SDK's last-value aggregation. demos/otel/metrics_demo.py fires 12 concurrent 40ms queries at a 3-thread pool and samples the gauges mid-flight: queue_depth peaks at exactly 9, and the duration histogram reads max=0.1695s for a query whose work is 40ms. That gap is the queue, and it is the thing traces alone will not show you. Also corrects the demo README's privacy section, which still claimed parameter values are never recorded - that stopped being unconditionally true when trace_sql_parameters landed. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> (cherry picked from 6ef0dd8c and adapted to the rebuilt phase-1 stack: attribute names now come from telemetry_registry where entries exist, the meter carries the instrumentation-scope version and schema URL, and the interrupted-queries counter skips expected timeouts - callers that opted into a deliberately short budget, like facet suggestion - matching how those are excluded from span error status. The internals.rst reference lands with the registry commit that follows.) Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01F2h9ANGZ7paWSpqs5DUAcG
2026-07-30 09:43:30 -07:00
@pytest.fixture
def bare_ds():
"""
Minimal Datasette with no plugins, data, metadata, or config - for tests
that want to exercise core behavior (e.g. middleware) in isolation.
"""
from datasette.app import Datasette
return Datasette(memory=True)
@pytest_asyncio.fixture(scope="session")
async def ds_client():
import secrets
from datasette.app import Datasette
from datasette.database import Database
from .fixtures import CONFIG, METADATA, PLUGINS_DIR
ds = Datasette(
metadata=METADATA,
config=CONFIG,
plugins_dir=PLUGINS_DIR,
settings={
"default_page_size": 50,
"max_returned_rows": 100,
"sql_time_limit_ms": 200,
"facet_suggest_time_limit_ms": 200, # Up from 50 default
# Default is 3 but this results in "too many open files"
# errors when running the full test suite:
"num_sql_threads": 1,
},
)
from datasette.fixtures import populate_fixture_database
# Use a unique memory_name to avoid collisions between different
# Datasette instances in the same process, but use "fixtures" for routing
unique_memory_name = f"fixtures_{secrets.token_hex(8)}"
db = ds.add_database(Database(ds, memory_name=unique_memory_name), name="fixtures")
ds.remove_database("_memory")
def prepare(conn):
if not conn.execute("select count(*) from sqlite_master").fetchone()[0]:
populate_fixture_database(conn)
await db.execute_write_fn(prepare)
await ds.invoke_startup()
return ds.client
def pytest_report_header(config):
conn = sqlite3.connect(":memory:")
version = conn.execute("select sqlite_version()").fetchone()[0]
conn.close()
sqlite_utils_version = importlib.metadata.version("sqlite-utils")
headers = [
f"SQLite: {version}",
f"sqlite-utils: {sqlite_utils_version}",
]
if config.getoption("--playwright"):
try:
browsers = config.getoption("--browser")
except ValueError:
browsers = None
if isinstance(browsers, str):
browsers = [browsers]
if browsers:
headers.append("Playwright browsers: {}".format(", ".join(browsers)))
return headers
def pytest_addoption(parser):
parser.addoption(
"--playwright",
action="store_true",
default=False,
help="run Playwright browser automation tests",
)
2019-05-01 22:10:23 -07:00
def pytest_configure(config):
import sys
sys._called_from_test = True
def pytest_unconfigure(config):
import sys
del sys._called_from_test
def pytest_collection_modifyitems(config, items):
if not config.getoption("--playwright"):
skip_playwright = pytest.mark.skip(reason="need --playwright option to run")
for item in items:
if "playwright" in item.keywords:
item.add_marker(skip_playwright)
# Ensure test_cli.py and test_black.py and test_inspect.py run first before any asyncio code kicks in
move_to_front(items, "test_cli")
2019-05-11 14:45:59 -07:00
move_to_front(items, "test_black")
move_to_front(items, "test_inspect_cli")
move_to_front(items, "test_serve_with_get")
2020-09-11 15:04:23 -07:00
move_to_front(items, "test_serve_with_get_exit_code_for_error")
2019-05-11 15:03:52 -07:00
move_to_front(items, "test_inspect_cli_writes_to_file")
move_to_front(items, "test_spatialite_error_if_attempt_to_open_spatialite")
move_to_front(items, "test_package")
move_to_front(items, "test_package_with_port")
Add opentelemetry-api dependency and datasette/telemetry.py scaffolding Datasette core is gaining OpenTelemetry spans alongside the existing hand-rolled tracer. This commit only lays the groundwork - no span is emitted yet. Core takes a runtime dependency on opentelemetry-api and nothing more. It deliberately never creates a TracerProvider, configures an exporter, or touches sampling: that belongs to whoever runs Datasette, normally via an opentelemetry-instrument agent. Owning a provider in core was tried in an earlier design and produced a cross-request span leak, a process-global provider that tests could not tear down, and a sampling env var that silently blanked output. With no provider installed every span is a NonRecordingSpan and costs approximately nothing. datasette/telemetry.py exposes the module-level tracer plus sql_attribute(), which truncates SQL to 2048 characters. On a public instance the SQL is attacker-controlled and unbounded - someone can paste a 10MB query into ?sql= - so it must never reach a telemetry pipeline verbatim. opentelemetry-sdk goes in the dev dependency group only, because the test suite needs it to assert on spans while the package itself must not import it. tests/test_telemetry.py enforces that by importing datasette in a fresh interpreter and inspecting sys.modules, which catches a lazy import inside a function body that a grep would miss. conftest.py gains a session-scoped autouse fixture installing an SDK provider with an InMemorySpanExporter. It has to be session-scoped because set_tracer_provider() is effectively once-per-process - a second call logs a warning and is ignored. SimpleSpanProcessor rather than BatchSpanProcessor, so assertions made right after a request never race a background export thread. The otel_spans fixture that later tickets assert against is added here too. test_datasette_package_never_imports_the_sdk is moved to the front of the run. Late in a serial run the pytest process holds enough threads that the fork half of subprocess' fork+exec segfaults the interpreter on macOS/CPython 3.13. That reproduces with any subprocess call in that position on an unmodified tree, so it is a pre-existing hazard rather than something this commit introduces; the repo already moves its other subprocess-spawning tests to the front for related reasons. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-07-30 17:30:52 -07:00
# Same reason: this one shells out to a fresh interpreter. Late in a serial
# run the pytest process holds enough threads that the fork half of
# subprocess' fork+exec crashes the interpreter on macOS/CPython 3.13
# (SIGSEGV/SIGBUS inside _execute_child). Reproduces with any subprocess
# call placed there, on an unmodified tree - running it first avoids it.
move_to_front(items, "test_datasette_package_never_imports_the_sdk")
Name the request span after the route it matched The request span was created at the ASGI edge, before anything knew which route would match, so it carried nothing but the method: every request in a trace UI showed up as "GET", and the only URL on it was url.path, which is unbounded on a public instance and useless as a grouping key. Routing resolves in DatasetteRouter, so that is where the span gets http.route and its semconv `{method} {route}` name. http.route is the compiled route pattern, not a prettified /{database}/{table} template. Datasette routes with compiled regexes and the route table is fixed when the app is built, so the pattern is exact, bounded and needs no parsing; the transform into something prettier accretes edge cases, and Django's instrumentation ships regex-flavoured routes for the same reason. A request that matches no route gets no http.route and keeps its bare method name, which is what semantic conventions ask for. Two things the obvious implementation gets wrong, both found by testing it: - The router must not read `get_current_span()`. A plugin asgi_wrapper() runs *inside* the request middleware, so an instrumented plugin makes its own span current for the whole request - and the route then lands on that plugin's INTERNAL span, renaming it, while the actual request span never gets the one attribute a trace UI groups by. It reproduces with a five-line plugin. The span is passed through the ASGI scope instead, falling back to the current span so an externally-created SERVER span is still enriched. - The method has to be clamped again here. The middleware clamps it for the attribute, but the name is rebuilt from request.method, which is the raw client string - so an unclamped rename put `FROB /(?P<database>...` back into the span name that the middleware had just kept it out of. Both guards are `is_recording()`, not `get_span_context().is_valid`: with no provider but an inbound traceparent the API returns a NonRecordingSpan carrying the remote context, which is valid and records nothing, so an is_valid guard would do the work on every request from a traced caller. Tests cover the route and name, the unrouted 404 fallback, the full attribute set, db.query spans reaching the request span by parent walk, a 500, an inbound traceparent becoming a remote parent, ?sql= never reaching a span attribute, and - in a subprocess, because the suite's provider fixture is session-scoped and unavoidable - the no-provider fast path handing the app the original `send`. The streaming test uses a table larger than one page so the export genuinely issues queries during the body send; without that it passes however early the span ends. Measured on this branch against fixtures.db: a faceted table page went from 112 spans in 56 traces to 113 spans in 1. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-07-30 19:52:12 -07:00
move_to_front(items, "test_no_provider_takes_the_fast_path")
2019-05-11 14:45:59 -07:00
def move_to_front(items, test_name):
test = [fn for fn in items if fn.name == test_name]
if test:
items.insert(0, items.pop(items.index(test[0])))
@pytest.fixture
def restore_working_directory(tmpdir, request):
try:
previous_cwd = os.getcwd()
except OSError:
# https://github.com/simonw/datasette/issues/1361
previous_cwd = None
tmpdir.chdir()
def return_to_previous():
os.chdir(previous_cwd)
if previous_cwd is not None:
request.addfinalizer(return_to_previous)
@pytest.fixture(scope="session", autouse=True)
def check_actions_are_documented():
2026-06-12 12:51:40 -07:00
from datasette.default_actions import register_actions as default_register_actions
from datasette.plugins import pm
content = (
pathlib.Path(__file__).parent.parent / "docs" / "authentication.rst"
).read_text()
permissions_re = re.compile(r"\.\. _actions_([^\s:]+):")
documented_actions = set(permissions_re.findall(content)).union(
UNDOCUMENTED_PERMISSIONS
)
2026-06-12 12:51:40 -07:00
# Only Datasette core actions need to be documented - actions registered
# by (test) plugins are checked for registration but not documentation
core_actions = {action.name for action in default_register_actions()}
def before(hook_name, hook_impls, kwargs):
if hook_name == "permission_resources_sql":
datasette = kwargs["datasette"]
assert kwargs["action"] in datasette.actions, (
"'{}' has not been registered with register_actions()".format(
kwargs["action"]
)
+ " (or maybe a test forgot to do await ds.invoke_startup())"
)
action = kwargs.get("action").replace("-", "_")
2026-06-12 12:51:40 -07:00
if kwargs["action"] in core_actions:
assert (
action in documented_actions
), f"Undocumented permission action: {action}"
pm.add_hookcall_monitoring(
before=before, after=lambda outcome, hook_name, hook_impls, kwargs: None
)
class TrackEventPlugin:
__name__ = "TrackEventPlugin"
@dataclass
class OneEvent(Event):
name = "one"
extra: str
@hookimpl
def register_events(self, datasette):
async def inner():
return [self.OneEvent]
return inner
@hookimpl
def track_event(self, datasette, event):
datasette._tracked_events = getattr(datasette, "_tracked_events", [])
datasette._tracked_events.append(event)
@pytest.fixture(scope="session", autouse=True)
def install_event_tracking_plugin():
from datasette.plugins import pm
pm.register(TrackEventPlugin(), name="TrackEventPlugin")
@pytest.fixture(scope="session")
def ds_localhost_http_server():
ds_proc = subprocess.Popen(
[sys.executable, "-m", "datasette", "--memory", "-p", "8041"],
stdout=subprocess.PIPE,
stderr=subprocess.STDOUT,
# Avoid FileNotFoundError: [Errno 2] No such file or directory:
cwd=tempfile.gettempdir(),
)
wait_until_responds("http://localhost:8041/")
# Check it started successfully
assert not ds_proc.poll(), ds_proc.stdout.read().decode("utf-8")
yield ds_proc
# Shut it down at the end of the pytest session
ds_proc.terminate()
@pytest.fixture(scope="session")
def ds_unix_domain_socket_server(tmp_path_factory):
# This used to use tmp_path_factory.mktemp("uds") but that turned out to
# produce paths that were too long to use as UDS on macOS, see
# https://github.com/simonw/datasette/issues/1407 - so I switched to
# using tempfile.gettempdir() with a per-process filename.
uds = str(pathlib.Path(tempfile.gettempdir()) / f"datasette-{os.getpid()}.sock")
try:
os.unlink(uds)
except FileNotFoundError:
pass
ds_proc = subprocess.Popen(
[sys.executable, "-m", "datasette", "--memory", "--uds", uds],
stdout=subprocess.PIPE,
stderr=subprocess.STDOUT,
cwd=tempfile.gettempdir(),
)
2022-10-25 12:42:21 -07:00
# Poll until available
transport = httpx.HTTPTransport(uds=uds)
client = httpx.Client(transport=transport)
try:
wait_until_responds(
"http://localhost/_memory.json", timeout=30.0, client=client
)
# Check it started successfully
assert not ds_proc.poll(), ds_proc.stdout.read().decode("utf-8")
yield ds_proc, uds
finally:
client.close()
# Shut it down at the end of the pytest session
ds_proc.terminate()
try:
ds_proc.wait(timeout=5)
except subprocess.TimeoutExpired:
ds_proc.kill()
ds_proc.wait()
try:
os.unlink(uds)
except FileNotFoundError:
pass
@pytest.fixture
def serve_with_plugins(tmp_path):
"""Factory fixture for starting ``datasette serve`` in a subprocess with
plugins written to a temporary ``--plugins-dir``.
For tests that need the real serve path: event-loop wiring, exit codes,
signals. The usual in-process ``pm.register`` plugin pattern can't reach
a subprocess, so plugin source is written out as importable files instead.
Unlike ``ds_localhost_http_server`` this is function-scoped and takes a
fresh port each time, because each test needs its own plugins. Call it as::
proc, port = serve_with_plugins({"my_plugin": PLUGIN_SOURCE})
``plugins`` maps module name to Python source. Pass
``wait_for_startup=False`` when the server is expected to fail during
startup rather than begin serving. Extra CLI arguments are passed through.
Every process started is terminated when the test ends.
"""
processes = []
def start(plugins, *extra_args, wait_for_startup=True):
plugins_dir = tmp_path / "plugins"
plugins_dir.mkdir(exist_ok=True)
for module_name, source in plugins.items():
(plugins_dir / f"{module_name}.py").write_text(source, "utf-8")
port = find_free_port()
proc = subprocess.Popen(
[
sys.executable,
"-m",
"datasette",
"--memory",
"--plugins-dir",
str(plugins_dir),
"-h",
"127.0.0.1",
"-p",
str(port),
*extra_args,
],
stdout=subprocess.PIPE,
stderr=subprocess.STDOUT,
# Avoid FileNotFoundError: [Errno 2] No such file or directory:
cwd=tempfile.gettempdir(),
)
processes.append(proc)
if wait_for_startup:
wait_until_responds(
f"http://127.0.0.1:{port}/-/versions.json", process=proc
)
return proc, port
yield start
for proc in processes:
if proc.poll() is None:
proc.terminate()
try:
proc.wait(timeout=5)
except subprocess.TimeoutExpired:
proc.kill()
proc.wait()
# Import fixtures from fixtures.py to make them available
from .fixtures import ( # noqa: F401
TEMP_PLUGIN_SECRET_FILE,
app_client,
app_client_base_url_prefix,
app_client_conflicting_database_names,
app_client_csv_max_mb_one,
app_client_immutable_and_inspect_file,
app_client_larger_cache_size,
app_client_no_files,
app_client_returned_rows_matches_page_size,
app_client_shorter_time_limit,
app_client_two_attached_databases,
app_client_two_attached_databases_crossdb_enabled,
app_client_two_attached_databases_one_immutable,
app_client_with_cors,
app_client_with_dot,
app_client_with_trace,
make_app_client,
)