crawlee 1.0.4b5__tar.gz → 1.0.4b6__tar.gz

{crawlee-1.0.4b5 → crawlee-1.0.4b6}/CHANGELOG.md

@@ -9,6 +9,7 @@ All notable changes to this project will be documented in this file.
 
 - Respect `enqueue_strategy` in `enqueue_links` ([#1505](https://github.com/apify/crawlee-python/pull/1505)) ([6ee04bc](https://github.com/apify/crawlee-python/commit/6ee04bc08c50a70f2e956a79d4ce5072a726c3a8)) by [@Mantisus](https://github.com/Mantisus), closes [#1504](https://github.com/apify/crawlee-python/issues/1504)
 - Exclude incorrect links before checking `robots.txt` ([#1502](https://github.com/apify/crawlee-python/pull/1502)) ([3273da5](https://github.com/apify/crawlee-python/commit/3273da5fee62ec9254666b376f382474c3532a56)) by [@Mantisus](https://github.com/Mantisus), closes [#1499](https://github.com/apify/crawlee-python/issues/1499)
+- Resolve compatibility issue between `SqlStorageClient` and `AdaptivePlaywrightCrawler` ([#1496](https://github.com/apify/crawlee-python/pull/1496)) ([ce172c4](https://github.com/apify/crawlee-python/commit/ce172c425a8643a1d4c919db4f5e5a6e47e91deb)) by [@Mantisus](https://github.com/Mantisus), closes [#1495](https://github.com/apify/crawlee-python/issues/1495)
 
 
 <!-- git-cliff-unreleased-end -->

{crawlee-1.0.4b5 → crawlee-1.0.4b6}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: crawlee
-Version: 1.0.4b5
+Version: 1.0.4b6
 Summary: Crawlee for Python
 Project-URL: Apify Homepage, https://apify.com
 Project-URL: Changelog, https://crawlee.dev/python/docs/changelog

{crawlee-1.0.4b5 → crawlee-1.0.4b6}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "hatchling.build"
 
 [project]
 name = "crawlee"
-version = "1.0.4b5"
+version = "1.0.4b6"
 description = "Crawlee for Python"
 authors = [{ name = "Apify Technologies s.r.o.", email = "support@apify.com" }]
 license = { file = "LICENSE" }

{crawlee-1.0.4b5 → crawlee-1.0.4b6}/src/crawlee/crawlers/_adaptive_playwright/_adaptive_playwright_crawler.py

@@ -149,10 +149,6 @@ class AdaptivePlaywrightCrawler(
                 non-default configuration.
             kwargs: Additional keyword arguments to pass to the underlying `BasicCrawler`.
         """
-        # Some sub crawler kwargs are internally modified. Prepare copies.
-        basic_crawler_kwargs_for_static_crawler = deepcopy(kwargs)
-        basic_crawler_kwargs_for_pw_crawler = deepcopy(kwargs)
-
         # Adaptive crawling related.
         self.rendering_type_predictor = rendering_type_predictor or DefaultRenderingTypePredictor()
         self.result_checker = result_checker or (lambda _: True)
@@ -170,11 +166,11 @@ class AdaptivePlaywrightCrawler(
         # Each sub crawler will use custom logger .
         static_logger = getLogger('Subcrawler_static')
         static_logger.setLevel(logging.ERROR)
-        basic_crawler_kwargs_for_static_crawler['_logger'] = static_logger
+        basic_crawler_kwargs_for_static_crawler: _BasicCrawlerOptions = {'_logger': static_logger, **kwargs}
 
         pw_logger = getLogger('Subcrawler_playwright')
         pw_logger.setLevel(logging.ERROR)
-        basic_crawler_kwargs_for_pw_crawler['_logger'] = pw_logger
+        basic_crawler_kwargs_for_pw_crawler: _BasicCrawlerOptions = {'_logger': pw_logger, **kwargs}
 
         # Initialize sub crawlers to create their pipelines.
         static_crawler_class = AbstractHttpCrawler.create_parsed_http_crawler_class(static_parser=static_parser)

{crawlee-1.0.4b5 → crawlee-1.0.4b6}/tests/unit/crawlers/_adaptive_playwright/test_adaptive_playwright_crawler.py

@@ -30,10 +30,12 @@ from crawlee.crawlers._adaptive_playwright._adaptive_playwright_crawling_context
     AdaptiveContextError,
 )
 from crawlee.statistics import Statistics
+from crawlee.storage_clients import SqlStorageClient
 from crawlee.storages import KeyValueStore
 
 if TYPE_CHECKING:
     from collections.abc import AsyncGenerator, Iterator
+    from pathlib import Path
 
     from yarl import URL
 
@@ -726,3 +728,23 @@ async def test_adaptive_context_query_non_existing_element(test_urls: list[str])
     await crawler.run(test_urls[:1])
 
     mocked_h3_handler.assert_called_once_with(None)
+
+
+async def test_adaptive_playwright_crawler_with_sql_storage(test_urls: list[str], tmp_path: Path) -> None:
+    """Tests that AdaptivePlaywrightCrawler can be initialized with SqlStorageClient."""
+    storage_dir = tmp_path / 'test_table.db'
+
+    async with SqlStorageClient(connection_string=f'sqlite+aiosqlite:///{storage_dir}') as storage_client:
+        crawler = AdaptivePlaywrightCrawler.with_beautifulsoup_static_parser(
+            storage_client=storage_client,
+        )
+
+        mocked_handler = Mock()
+
+        @crawler.router.default_handler
+        async def request_handler(_context: AdaptivePlaywrightCrawlingContext) -> None:
+            mocked_handler()
+
+        await crawler.run(test_urls[:1])
+
+        mocked_handler.assert_called()

{crawlee-1.0.4b5 → crawlee-1.0.4b6}/uv.lock

@@ -705,7 +705,7 @@ toml = [
 
 [[package]]
 name = "crawlee"
-version = "1.0.4b5"
+version = "1.0.4b6"
 source = { editable = "." }
 dependencies = [
     { name = "cachetools" },