chore: improve HttpClient isolating in tests (#1513)

### Description

### Issues

- Closes: #1441
- Closes: #1494
- Closes: #1460

Author: Mantisus

tests/unit/_utils/test_sitemap.py

@@ -1,9 +1,7 @@
 import base64
 import gzip
-import sys
 from datetime import datetime
 
-import pytest
 from yarl import URL
 
 from crawlee._utils.sitemap import Sitemap, SitemapUrl, parse_sitemap
@@ -95,8 +93,6 @@ async def test_gzipped_sitemap(server_url: URL, http_client: HttpClient) -> None
     assert set(sitemap.urls) == BASIC_RESULTS
 
 
-# TODO: Remove this skip when #1460 is resolved.
-@pytest.mark.skipif(sys.platform != 'linux', reason='Flaky with Curl on Windows, see #1460.')
 async def test_gzipped_sitemap_with_invalid_data(server_url: URL, http_client: HttpClient) -> None:
     """Test loading a invalid gzipped sitemap with correct type and .xml.gz url."""
     compress_data = compress_gzip(BASIC_SITEMAP)

tests/unit/conftest.py

@@ -6,7 +6,7 @@
 import logging
 import os
 import warnings
-from typing import TYPE_CHECKING, cast
+from typing import TYPE_CHECKING, Any, cast
 
 import pytest
 from curl_cffi import CurlHttpVersion
@@ -205,9 +205,16 @@ def redirect_server_url(redirect_http_server: TestServer) -> URL:
         pytest.param('curl', id='curl'),
     ]
 )
-async def http_client(request: pytest.FixtureRequest) -> HttpClient:
+async def http_client(request: pytest.FixtureRequest) -> AsyncGenerator[HttpClient, None]:
+    class_client: type[HttpClient]
     if request.param == 'curl':
-        return CurlImpersonateHttpClient(http_version=CurlHttpVersion.V1_1)
-    if request.param == 'impit':
-        return ImpitHttpClient(http3=False)
-    return HttpxHttpClient(http2=False)
+        class_client = CurlImpersonateHttpClient
+        kwargs: dict[str, Any] = {'http_version': CurlHttpVersion.V1_1}
+    elif request.param == 'impit':
+        class_client = ImpitHttpClient
+        kwargs = {'http3': False}
+    else:
+        class_client = HttpxHttpClient
+        kwargs = {'http2': True}
+    async with class_client(**kwargs) as client:
+        yield client

tests/unit/http_clients/test_http_clients.py

@@ -1,7 +1,6 @@
 from __future__ import annotations
 
 import os
-import sys
 from typing import TYPE_CHECKING
 
 import pytest
@@ -164,10 +163,6 @@ async def test_send_request_allow_redirects_false(custom_http_client: HttpClient
 
 
 async def test_stream(http_client: HttpClient, server_url: URL) -> None:
-    # TODO: Remove this skip when #1494 is resolved.
-    if isinstance(http_client, CurlImpersonateHttpClient) and sys.platform != 'linux':
-        pytest.skip('Flaky with Curl on Windows, see #1494.')
-
     content_body: bytes = b''
 
     async with http_client.stream(str(server_url)) as response:

tests/unit/storages/test_request_queue.py

@@ -449,8 +449,9 @@ async def test_add_requests_wait_for_all(
         # Immediately after adding, the total count may be less than 15 due to background processing
         assert await rq.get_total_count() <= 15
 
-        # Wait a 500 milliseconds for background tasks to complete
-        await asyncio.sleep(0.5)
+        # Wait for background tasks to complete
+        while await rq.get_total_count() < 15:  # noqa: ASYNC110
+            await asyncio.sleep(0.1)
 
     # Verify all requests were added
     assert await rq.get_total_count() == 15