kernel 0.107.0__tar.gz → 0.108.0__tar.gz

kernel-0.108.0/.release-please-manifest.json

@@ -0,0 +1,3 @@
+{
+  ".": "0.108.0"
+}

{kernel-0.107.0 → kernel-0.108.0}/CHANGELOG.md

@@ -1,5 +1,12 @@
 # Changelog
 
+## [0.108.0](https://github.com/kernel/kernel-python-sdk/compare/v0.107.0...v0.108.0) (2026-09-17)
+
+
+### Features
+
+* feat: add config registry analysis waiter ([e38c887](https://github.com/kernel/kernel-python-sdk/commit/e38c8878f9f80f5d28fce9f92df6f0a18a48e2bb))
+
 ## [0.107.0](https://github.com/kernel/kernel-python-sdk/compare/v0.106.0...v0.107.0) (2026-09-16)
 
 

{kernel-0.107.0 → kernel-0.108.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: kernel
-Version: 0.107.0
+Version: 0.108.0
 Summary: The official Python library for the kernel API
 Project-URL: Homepage, https://github.com/kernel/kernel-python-sdk
 Project-URL: Repository, https://github.com/kernel/kernel-python-sdk

{kernel-0.107.0 → kernel-0.108.0}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "kernel"
-version = "0.107.0"
+version = "0.108.0"
 description = "The official Python library for the kernel API"
 dynamic = ["readme"]
 license = "Apache-2.0"

{kernel-0.107.0 → kernel-0.108.0}/src/kernel/_version.py

@@ -1,4 +1,4 @@
 # File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
 
 __title__ = "kernel"
-__version__ = "0.107.0"  # x-release-please-version
+__version__ = "0.108.0"  # x-release-please-version

kernel-0.108.0/src/kernel/lib/config_registry_wait.py

@@ -0,0 +1,63 @@
+from __future__ import annotations
+
+import math
+import time
+import random
+from datetime import datetime
+
+from .._types import Omit, Headers
+from .._exceptions import KernelError
+from ..types.config_registry_response import ConfigRegistryResponse
+
+DEFAULT_CONFIG_REGISTRY_POLL_INTERVAL = 5.0
+_TERMINAL_STATUSES = frozenset({"completed", "failed", "canceled", "expired"})
+
+
+def validate_wait_options(poll_interval: float, max_wait_seconds: float | None) -> None:
+    if not math.isfinite(poll_interval) or poll_interval <= 0:
+        raise ValueError("Expected a finite, positive value for `poll_interval`")
+    if max_wait_seconds is not None and (not math.isfinite(max_wait_seconds) or max_wait_seconds < 0):
+        raise ValueError("Expected a finite, non-negative value for `max_wait_seconds`")
+
+
+def poll_headers(extra_headers: Headers | None) -> Headers:
+    headers: dict[str, str | Omit] = dict(extra_headers or {})
+    headers["X-Stainless-Poll-Helper"] = "true"
+    return headers
+
+
+def poll_delay(poll_interval: float) -> float:
+    return poll_interval * random.uniform(0.9, 1.1)
+
+
+def wait_timeout_error(id: str, polls: int, last_status: str | None, started_at: float) -> TimeoutError:
+    elapsed = time.monotonic() - started_at
+    return TimeoutError(
+        f"Timed out waiting for config registry analysis {id!r} after {elapsed:.1f}s "
+        f"and {polls} polls; last status was {last_status!r}"
+    )
+
+
+def analysis_finished(response: ConfigRegistryResponse, requested_id: str) -> tuple[bool, str]:
+    analysis = response.analysis
+    if analysis is None:
+        raise KernelError(f"Config registry response for {requested_id!r} is missing an analysis")
+
+    analysis_id = getattr(analysis, "id", None)
+    if not isinstance(analysis_id, str) or not analysis_id:
+        raise KernelError(f"Config registry response for {requested_id!r} has no valid analysis ID")
+    if analysis_id != requested_id:
+        raise KernelError(f"Config registry response for {requested_id!r} returned analysis {analysis_id!r}")
+
+    status = getattr(analysis, "status", None)
+    if not isinstance(status, str) or not status:
+        raise KernelError(f"Config registry analysis {requested_id!r} has no valid status")
+
+    if "finished_at" not in analysis.model_fields_set:
+        raise KernelError(f"Config registry analysis {requested_id!r} is missing `finished_at`")
+
+    finished_at = getattr(analysis, "finished_at", None)
+    if finished_at is not None and not isinstance(finished_at, datetime):
+        raise KernelError(f"Config registry analysis {requested_id!r} has an invalid `finished_at`")
+
+    return finished_at is not None or status in _TERMINAL_STATUSES, status

{kernel-0.107.0 → kernel-0.108.0}/src/kernel/resources/config_registry/analyses.py

@@ -2,6 +2,8 @@
 
 from __future__ import annotations
 
+import time
+
 import httpx
 
 from ..._types import Body, Omit, Query, Headers, NotGiven, omit, not_given
@@ -18,6 +20,14 @@ from ...pagination import SyncOffsetPagination, AsyncOffsetPagination
 from ..._base_client import AsyncPaginator, make_request_options
 from ...types.config_registry import analysis_list_params
 from ...types.analysis_summary import AnalysisSummary
+from ...lib.config_registry_wait import (
+    DEFAULT_CONFIG_REGISTRY_POLL_INTERVAL,
+    poll_delay,
+    poll_headers,
+    analysis_finished,
+    wait_timeout_error,
+    validate_wait_options,
+)
 from ...types.config_registry_response import ConfigRegistryResponse
 
 __all__ = ["AnalysesResource", "AsyncAnalysesResource"]
@@ -79,6 +89,54 @@ class AnalysesResource(SyncAPIResource):
             cast_to=ConfigRegistryResponse,
         )
 
+    def wait_for_result(
+        self,
+        id: str,
+        *,
+        poll_interval: float = DEFAULT_CONFIG_REGISTRY_POLL_INTERVAL,
+        max_wait_seconds: float | None = None,
+        extra_headers: Headers | None = None,
+        extra_query: Query | None = None,
+        extra_body: Body | None = None,
+        timeout: float | httpx.Timeout | None | NotGiven = not_given,
+    ) -> ConfigRegistryResponse:
+        """Wait for an analysis to finish and return its complete result.
+
+        The first retrieval happens immediately. ``max_wait_seconds`` is a soft
+        polling deadline: an in-flight request and its normal retries may finish
+        after it. Timing out does not cancel the remote analysis.
+        """
+        validate_wait_options(poll_interval, max_wait_seconds)
+        started_at = time.monotonic()
+        deadline = started_at + max_wait_seconds if max_wait_seconds is not None else None
+        headers = poll_headers(extra_headers)
+        polls = 0
+        last_status: str | None = None
+
+        while True:
+            if polls > 0 and deadline is not None and time.monotonic() >= deadline:
+                raise wait_timeout_error(id, polls, last_status, started_at)
+
+            response = self.retrieve(
+                id,
+                extra_headers=headers,
+                extra_query=extra_query,
+                extra_body=extra_body,
+                timeout=timeout,
+            )
+            polls += 1
+            finished, last_status = analysis_finished(response, id)
+            if finished:
+                return response
+
+            delay = poll_delay(poll_interval)
+            if deadline is not None:
+                remaining = deadline - time.monotonic()
+                if remaining <= 0:
+                    raise wait_timeout_error(id, polls, last_status, started_at)
+                delay = min(delay, remaining)
+            self._sleep(delay)
+
     def list(
         self,
         *,
@@ -220,6 +278,55 @@ class AsyncAnalysesResource(AsyncAPIResource):
             cast_to=ConfigRegistryResponse,
         )
 
+    async def wait_for_result(
+        self,
+        id: str,
+        *,
+        poll_interval: float = DEFAULT_CONFIG_REGISTRY_POLL_INTERVAL,
+        max_wait_seconds: float | None = None,
+        extra_headers: Headers | None = None,
+        extra_query: Query | None = None,
+        extra_body: Body | None = None,
+        timeout: float | httpx.Timeout | None | NotGiven = not_given,
+    ) -> ConfigRegistryResponse:
+        """Wait for an analysis to finish and return its complete result.
+
+        The first retrieval happens immediately. ``max_wait_seconds`` is a soft
+        polling deadline: an in-flight request and its normal retries may finish
+        after it. Timing out does not cancel the remote analysis. Cancelling the
+        calling task stops the wait without cancelling the remote analysis.
+        """
+        validate_wait_options(poll_interval, max_wait_seconds)
+        started_at = time.monotonic()
+        deadline = started_at + max_wait_seconds if max_wait_seconds is not None else None
+        headers = poll_headers(extra_headers)
+        polls = 0
+        last_status: str | None = None
+
+        while True:
+            if polls > 0 and deadline is not None and time.monotonic() >= deadline:
+                raise wait_timeout_error(id, polls, last_status, started_at)
+
+            response = await self.retrieve(
+                id,
+                extra_headers=headers,
+                extra_query=extra_query,
+                extra_body=extra_body,
+                timeout=timeout,
+            )
+            polls += 1
+            finished, last_status = analysis_finished(response, id)
+            if finished:
+                return response
+
+            delay = poll_delay(poll_interval)
+            if deadline is not None:
+                remaining = deadline - time.monotonic()
+                if remaining <= 0:
+                    raise wait_timeout_error(id, polls, last_status, started_at)
+                delay = min(delay, remaining)
+            await self._sleep(delay)
+
     def list(
         self,
         *,

kernel-0.108.0/tests/test_config_registry_wait.py

@@ -0,0 +1,181 @@
+from __future__ import annotations
+
+import asyncio
+from typing import Any
+from collections.abc import Iterator
+
+import httpx
+import pytest
+
+from kernel import Kernel, AsyncKernel
+from kernel._exceptions import KernelError
+
+
+def analysis_response(
+    status: str,
+    *,
+    analysis_id: str = "analysis-1",
+    finished_at: str | None = None,
+) -> dict[str, Any]:
+    return {
+        "analysis": {
+            "id": analysis_id,
+            "created_at": "2026-09-16T00:00:00Z",
+            "expires_at": "2026-09-16T00:45:00Z",
+            "failure": None,
+            "finished_at": finished_at,
+            "status": status,
+            "intent": None,
+        },
+        "recommendation": None,
+        "target": {
+            "domain": "example.com",
+            "host": "example.com",
+            "normalized": "https://example.com/",
+        },
+        "working_configurations": [],
+        "guidance": None,
+        "workload_outcome": None,
+    }
+
+
+def response_sequence(payloads: list[dict[str, Any]]) -> tuple[httpx.MockTransport, list[httpx.Request]]:
+    remaining: Iterator[dict[str, Any]] = iter(payloads)
+    requests: list[httpx.Request] = []
+
+    def handler(request: httpx.Request) -> httpx.Response:
+        requests.append(request)
+        return httpx.Response(200, json=next(remaining))
+
+    return httpx.MockTransport(handler), requests
+
+
+def test_wait_for_result_polls_unknown_unfinished_status_until_finished() -> None:
+    transport, requests = response_sequence(
+        [
+            analysis_response("running"),
+            analysis_response("queued"),
+            analysis_response("archived", finished_at="2026-09-16T00:01:00Z"),
+        ]
+    )
+
+    with httpx.Client(transport=transport) as http_client:
+        client = Kernel(api_key="test", base_url="https://api.example", http_client=http_client)
+        result = client.config_registry.analyses.wait_for_result(
+            "analysis-1",
+            poll_interval=0.001,
+            extra_headers={"X-Test": "preserved", "X-Stainless-Poll-Helper": "caller"},
+        )
+
+    assert result.analysis is not None
+    assert result.analysis.status == "archived"
+    assert len(requests) == 3
+    assert all(request.headers["X-Test"] == "preserved" for request in requests)
+    assert all(request.headers["X-Stainless-Poll-Helper"] == "true" for request in requests)
+
+
+@pytest.mark.parametrize("status", ["completed", "failed", "canceled", "expired"])
+def test_wait_for_result_returns_known_terminal_status_without_finished_at(status: str) -> None:
+    transport, requests = response_sequence([analysis_response(status)])
+
+    with httpx.Client(transport=transport) as http_client:
+        client = Kernel(api_key="test", base_url="https://api.example", http_client=http_client)
+        result = client.config_registry.analyses.wait_for_result("analysis-1")
+
+    assert result.analysis is not None
+    assert result.analysis.status == status
+    assert len(requests) == 1
+
+
+def test_wait_for_result_zero_max_wait_reads_once_then_times_out() -> None:
+    transport, requests = response_sequence([analysis_response("running")])
+
+    with httpx.Client(transport=transport) as http_client:
+        client = Kernel(api_key="test", base_url="https://api.example", http_client=http_client)
+        with pytest.raises(TimeoutError, match="analysis-1"):
+            client.config_registry.analyses.wait_for_result("analysis-1", max_wait_seconds=0)
+
+    assert len(requests) == 1
+
+
+@pytest.mark.parametrize(
+    "payload, message",
+    [
+        ({"analysis": None}, "missing an analysis"),
+        (analysis_response("running", analysis_id="analysis-2"), "analysis-2"),
+    ],
+)
+def test_wait_for_result_rejects_invalid_analysis_response(payload: dict[str, Any], message: str) -> None:
+    transport, _ = response_sequence([payload])
+
+    with httpx.Client(transport=transport) as http_client:
+        client = Kernel(api_key="test", base_url="https://api.example", http_client=http_client)
+        with pytest.raises(KernelError, match=message):
+            client.config_registry.analyses.wait_for_result("analysis-1")
+
+
+def test_wait_for_result_rejects_missing_finished_at() -> None:
+    payload = analysis_response("running")
+    analysis = payload["analysis"]
+    assert isinstance(analysis, dict)
+    del analysis["finished_at"]
+    transport, _ = response_sequence([payload])
+
+    with httpx.Client(transport=transport) as http_client:
+        client = Kernel(api_key="test", base_url="https://api.example", http_client=http_client)
+        with pytest.raises(KernelError, match="finished_at"):
+            client.config_registry.analyses.wait_for_result("analysis-1")
+
+
+@pytest.mark.parametrize(
+    "kwargs",
+    [
+        {"poll_interval": 0},
+        {"poll_interval": float("nan")},
+        {"max_wait_seconds": -1},
+        {"max_wait_seconds": float("inf")},
+    ],
+)
+def test_wait_for_result_rejects_invalid_timing_options(kwargs: dict[str, float]) -> None:
+    client = Kernel(api_key="test", base_url="https://api.example")
+
+    with pytest.raises(ValueError):
+        client.config_registry.analyses.wait_for_result("analysis-1", **kwargs)  # type: ignore[arg-type]
+
+
+async def test_async_wait_for_result_matches_sync_behavior() -> None:
+    transport, requests = response_sequence(
+        [
+            analysis_response("running"),
+            analysis_response("completed", finished_at="2026-09-16T00:01:00Z"),
+        ]
+    )
+
+    async with httpx.AsyncClient(transport=transport) as http_client:
+        client = AsyncKernel(api_key="test", base_url="https://api.example", http_client=http_client)
+        result = await client.config_registry.analyses.wait_for_result("analysis-1", poll_interval=0.001)
+
+    assert result.analysis is not None
+    assert result.analysis.status == "completed"
+    assert len(requests) == 2
+
+
+async def test_async_wait_for_result_is_cancellable_during_poll_sleep() -> None:
+    first_request = asyncio.Event()
+    requests: list[httpx.Request] = []
+
+    def handler(request: httpx.Request) -> httpx.Response:
+        requests.append(request)
+        first_request.set()
+        return httpx.Response(200, json=analysis_response("running"))
+
+    async with httpx.AsyncClient(transport=httpx.MockTransport(handler)) as http_client:
+        client = AsyncKernel(api_key="test", base_url="https://api.example", http_client=http_client)
+        task = asyncio.create_task(client.config_registry.analyses.wait_for_result("analysis-1", poll_interval=60))
+        await first_request.wait()
+        await asyncio.sleep(0)
+        task.cancel()
+        with pytest.raises(asyncio.CancelledError):
+            await task
+
+    assert len(requests) == 1

kernel-0.107.0/.release-please-manifest.json

@@ -1,3 +0,0 @@
-{
-  ".": "0.107.0"
-}