crawlee 0.6.13b31__py3-none-any.whl → 1.1.1b1__py3-none-any.whl

crawlee/_utils/robots.py

@@ -1,5 +1,6 @@
 from __future__ import annotations
 
+from logging import getLogger
 from typing import TYPE_CHECKING
 
 from protego import Protego
@@ -15,6 +16,9 @@ if TYPE_CHECKING:
     from crawlee.proxy_configuration import ProxyInfo
 
 
+logger = getLogger(__name__)
+
+
 class RobotsTxtFile:
     def __init__(
         self, url: str, robots: Protego, http_client: HttpClient | None = None, proxy_info: ProxyInfo | None = None
@@ -56,12 +60,20 @@ class RobotsTxtFile:
             http_client: The `HttpClient` instance used to perform the network request for fetching the robots.txt file.
             proxy_info: Optional `ProxyInfo` to be used when fetching the robots.txt file. If None, no proxy is used.
         """
-        response = await http_client.send_request(url, proxy_info=proxy_info)
-        body = (
-            b'User-agent: *\nAllow: /' if is_status_code_client_error(response.status_code) else await response.read()
-        )
+        try:
+            response = await http_client.send_request(url, proxy_info=proxy_info)
+
+            body = (
+                b'User-agent: *\nAllow: /'
+                if is_status_code_client_error(response.status_code)
+                else await response.read()
+            )
+            robots = Protego.parse(body.decode('utf-8'))
+
+        except Exception as e:
+            logger.warning(f'Failed to fetch from robots.txt from "{url}" with error: "{e}"')
 
-        robots = Protego.parse(body.decode('utf-8'))
+            robots = Protego.parse('User-agent: *\nAllow: /')
 
         return cls(url, robots, http_client=http_client, proxy_info=proxy_info)
 

crawlee/_utils/sitemap.py

@@ -335,7 +335,7 @@ async def _fetch_and_process_sitemap(
                         # Check if the first chunk is a valid gzip header
                         if first_chunk and raw_chunk.startswith(b'\x1f\x8b'):
                             decompressor = zlib.decompressobj(zlib.MAX_WBITS | 16)
-                            first_chunk = False
+                        first_chunk = False
 
                         chunk = decompressor.decompress(raw_chunk) if decompressor else raw_chunk
                         text_chunk = decoder.decode(chunk)

crawlee/_utils/system.py

@@ -36,7 +36,7 @@ else:
 class CpuInfo(BaseModel):
     """Information about the CPU usage."""
 
-    model_config = ConfigDict(populate_by_name=True)
+    model_config = ConfigDict(validate_by_name=True, validate_by_alias=True)
 
     used_ratio: Annotated[float, Field(alias='usedRatio')]
     """The ratio of CPU currently in use, represented as a float between 0 and 1."""
@@ -51,7 +51,7 @@ class CpuInfo(BaseModel):
 class MemoryUsageInfo(BaseModel):
     """Information about the memory usage."""
 
-    model_config = ConfigDict(populate_by_name=True)
+    model_config = ConfigDict(validate_by_name=True, validate_by_alias=True)
 
     current_size: Annotated[
         ByteSize,
@@ -71,7 +71,7 @@ class MemoryUsageInfo(BaseModel):
 class MemoryInfo(MemoryUsageInfo):
     """Information about system memory."""
 
-    model_config = ConfigDict(populate_by_name=True)
+    model_config = ConfigDict(validate_by_name=True, validate_by_alias=True)
 
     total_size: Annotated[
         ByteSize, PlainValidator(ByteSize.validate), PlainSerializer(lambda size: size.bytes), Field(alias='totalSize')

crawlee/_utils/urls.py

@@ -7,6 +7,7 @@ from yarl import URL
 
 if TYPE_CHECKING:
     from collections.abc import Iterator
+    from logging import Logger
 
 
 def is_url_absolute(url: str) -> bool:
@@ -22,13 +23,19 @@ def convert_to_absolute_url(base_url: str, relative_url: str) -> str:
     return str(URL(base_url).join(URL(relative_url)))
 
 
-def to_absolute_url_iterator(base_url: str, urls: Iterator[str]) -> Iterator[str]:
+def to_absolute_url_iterator(base_url: str, urls: Iterator[str], logger: Logger | None = None) -> Iterator[str]:
     """Convert an iterator of relative URLs to absolute URLs using a base URL."""
     for url in urls:
         if is_url_absolute(url):
             yield url
         else:
-            yield convert_to_absolute_url(base_url, url)
+            converted_url = convert_to_absolute_url(base_url, url)
+            # Skip the URL if conversion fails, probably due to an incorrect format, such as 'mailto:'.
+            if not is_url_absolute(converted_url):
+                if logger:
+                    logger.debug(f'Could not convert URL "{url}" to absolute using base URL "{base_url}". Skipping it.')
+                continue
+            yield converted_url
 
 
 _http_url_adapter = TypeAdapter(AnyHttpUrl)

crawlee/browsers/_browser_pool.py

@@ -118,7 +118,10 @@ class BrowserPool:
         """Initialize a new instance with a single `PlaywrightBrowserPlugin` configured with the provided options.
 
         Args:
-            browser_type: The type of browser to launch ('chromium', 'firefox', or 'webkit').
+            browser_type: The type of browser to launch:
+                - 'chromium', 'firefox', 'webkit': Use Playwright-managed browsers
+                - 'chrome': Use your locally installed Google Chrome browser. Requires Google Chrome to be installed on
+                    the system.
             user_data_dir: Path to a user data directory, which stores browser session data like cookies
                 and local storage.
             browser_launch_options: Keyword arguments to pass to the browser launch method. These options are provided

crawlee/browsers/_playwright_browser_controller.py

@@ -2,6 +2,7 @@
 
 from __future__ import annotations
 
+from asyncio import Lock
 from datetime import datetime, timedelta, timezone
 from typing import TYPE_CHECKING, Any, cast
 
@@ -77,6 +78,19 @@ class PlaywrightBrowserController(BrowserController):
 
         self._total_opened_pages = 0
 
+        self._context_creation_lock: Lock | None = None
+
+    async def _get_context_creation_lock(self) -> Lock:
+        """Get context checking and creation lock.
+
+        It should be done with lock to prevent multiple concurrent attempts to create context, which could lead to
+        memory leak as one of the two concurrently created contexts will become orphaned and not properly closed.
+        """
+        if self._context_creation_lock:
+            return self._context_creation_lock
+        self._context_creation_lock = Lock()
+        return self._context_creation_lock
+
     @property
     @override
     def pages(self) -> list[Page]:
@@ -137,12 +151,6 @@ class PlaywrightBrowserController(BrowserController):
         Raises:
             ValueError: If the browser has reached the maximum number of open pages.
         """
-        if not self._browser_context:
-            self._browser_context = await self._create_browser_context(
-                browser_new_context_options=browser_new_context_options,
-                proxy_info=proxy_info,
-            )
-
         if not self.has_free_capacity:
             raise ValueError('Cannot open more pages in this browser.')
 
@@ -154,11 +162,12 @@ class PlaywrightBrowserController(BrowserController):
             )
             page = await new_context.new_page()
         else:
-            if not self._browser_context:
-                self._browser_context = await self._create_browser_context(
-                    browser_new_context_options=browser_new_context_options,
-                    proxy_info=proxy_info,
-                )
+            async with await self._get_context_creation_lock():
+                if not self._browser_context:
+                    self._browser_context = await self._create_browser_context(
+                        browser_new_context_options=browser_new_context_options,
+                        proxy_info=proxy_info,
+                    )
             page = await self._browser_context.new_page()
 
         # Handle page close event
@@ -169,7 +178,6 @@ class PlaywrightBrowserController(BrowserController):
         self._last_page_opened_at = datetime.now(timezone.utc)
 
         self._total_opened_pages += 1
-
         return page
 
     @override
@@ -206,10 +214,9 @@ class PlaywrightBrowserController(BrowserController):
         `self._fingerprint_generator` is available.
         """
         browser_new_context_options = dict(browser_new_context_options) if browser_new_context_options else {}
-
         if proxy_info:
             if browser_new_context_options.get('proxy'):
-                logger.warning("browser_new_context_options['proxy'] overriden by explicit `proxy_info` argument.")
+                logger.warning("browser_new_context_options['proxy'] overridden by explicit `proxy_info` argument.")
 
             browser_new_context_options['proxy'] = ProxySettings(
                 server=f'{proxy_info.scheme}://{proxy_info.hostname}:{proxy_info.port}',
@@ -244,5 +251,4 @@ class PlaywrightBrowserController(BrowserController):
         browser_new_context_options['extra_http_headers'] = browser_new_context_options.get(
             'extra_http_headers', extra_http_headers
         )
-
         return await self._browser.new_context(**browser_new_context_options)

crawlee/browsers/_playwright_browser_plugin.py

@@ -34,8 +34,8 @@ class PlaywrightBrowserPlugin(BrowserPlugin):
 
     It is a plugin designed to manage browser instances using the Playwright automation library. It acts as a factory
     for creating new browser instances and provides a unified interface for interacting with different browser types
-    (chromium, firefox, and webkit). This class integrates configuration options for browser launches (headless mode,
-    executable paths, sandboxing, ...). It also manages browser contexts and the number of pages open within each
+    (chromium, firefox, webkit and chrome). This class integrates configuration options for browser launches (headless
+    mode, executable paths, sandboxing, ...). It also manages browser contexts and the number of pages open within each
     browser instance, ensuring that resource limits are respected.
     """
 
@@ -55,7 +55,10 @@ class PlaywrightBrowserPlugin(BrowserPlugin):
         """Initialize a new instance.
 
         Args:
-            browser_type: The type of browser to launch ('chromium', 'firefox', or 'webkit').
+            browser_type: The type of browser to launch:
+                - 'chromium', 'firefox', 'webkit': Use Playwright-managed browsers
+                - 'chrome': Use your locally installed Google Chrome browser. Requires Google Chrome to be installed on
+                    the system.
             user_data_dir: Path to a User Data Directory, which stores browser session data like cookies and local
                 storage.
             browser_launch_options: Keyword arguments to pass to the browser launch method. These options are provided
@@ -80,6 +83,17 @@ class PlaywrightBrowserPlugin(BrowserPlugin):
             'chromium_sandbox': not config.disable_browser_sandbox,
         }
 
+        if browser_type == 'chrome' and default_launch_browser_options['executable_path']:
+            raise ValueError(
+                'Cannot use browser_type `chrome` with `Configuration.default_browser_path` or `executable_path` set.'
+            )
+
+        # Map 'chrome' to 'chromium' with the 'chrome' channel.
+        if browser_type == 'chrome':
+            browser_type = 'chromium'
+            # Chromium parameter 'channel' set to 'chrome' enables using installed Google Chrome.
+            default_launch_browser_options['channel'] = 'chrome'
+
         self._browser_type: BrowserType = browser_type
         self._browser_launch_options: dict[str, Any] = default_launch_browser_options | (browser_launch_options or {})
         self._browser_new_context_options = browser_new_context_options or {}

crawlee/browsers/_types.py

@@ -6,7 +6,7 @@ from typing import TYPE_CHECKING, Literal
 if TYPE_CHECKING:
     from playwright.async_api import Page
 
-BrowserType = Literal['chromium', 'firefox', 'webkit']
+BrowserType = Literal['chromium', 'firefox', 'webkit', 'chrome']
 
 
 @dataclass

crawlee/configuration.py

@@ -28,6 +28,8 @@ class Configuration(BaseSettings):
     Settings can also be configured via environment variables, prefixed with `CRAWLEE_`.
     """
 
+    # TODO: https://github.com/pydantic/pydantic-settings/issues/706
+    # Use `SettingsConfigDict(validate_by_name=True, validate_by_alias=True)` when issue is resolved.
     model_config = SettingsConfigDict(populate_by_name=True)
 
     internal_timeout: Annotated[timedelta | None, Field(alias='crawlee_internal_timeout')] = None

crawlee/crawlers/_abstract_http/_abstract_http_crawler.py

@@ -34,7 +34,9 @@ TStatisticsState = TypeVar('TStatisticsState', bound=StatisticsState, default=St
 
 @docs_group('Crawlers')
 class AbstractHttpCrawler(
-    Generic[TCrawlingContext, TParseResult, TSelectResult], BasicCrawler[TCrawlingContext, StatisticsState], ABC
+    BasicCrawler[TCrawlingContext, StatisticsState],
+    ABC,
+    Generic[TCrawlingContext, TParseResult, TSelectResult],
 ):
     """A web crawler for performing HTTP requests.
 
@@ -165,7 +167,9 @@ class AbstractHttpCrawler(
             kwargs.setdefault('strategy', 'same-hostname')
 
             links_iterator: Iterator[str] = iter(self._parser.find_links(parsed_content, selector=selector))
-            links_iterator = to_absolute_url_iterator(context.request.loaded_url or context.request.url, links_iterator)
+            links_iterator = to_absolute_url_iterator(
+                context.request.loaded_url or context.request.url, links_iterator, logger=context.log
+            )
 
             if robots_txt_file:
                 skipped, links_iterator = partition(lambda url: robots_txt_file.is_allowed(url), links_iterator)

crawlee/crawlers/_abstract_http/_abstract_http_parser.py

@@ -16,7 +16,7 @@ if TYPE_CHECKING:
 
 
 @docs_group('HTTP parsers')
-class AbstractHttpParser(Generic[TParseResult, TSelectResult], ABC):
+class AbstractHttpParser(ABC, Generic[TParseResult, TSelectResult]):
     """Parser used for parsing HTTP response and inspecting parsed result to find links or detect blocking."""
 
     @abstractmethod

crawlee/crawlers/_abstract_http/_http_crawling_context.py

@@ -31,7 +31,7 @@ class HttpCrawlingContext(BasicCrawlingContext, HttpCrawlingResult):
 
 @dataclass(frozen=True)
 @docs_group('Crawling contexts')
-class ParsedHttpCrawlingContext(Generic[TParseResult], HttpCrawlingContext):
+class ParsedHttpCrawlingContext(HttpCrawlingContext, Generic[TParseResult]):
     """The crawling context used by `AbstractHttpCrawler`.
 
     It provides access to key objects as well as utility functions for handling crawling tasks.

crawlee/crawlers/_adaptive_playwright/_adaptive_playwright_crawler.py

@@ -12,7 +12,7 @@ from bs4 import BeautifulSoup, Tag
 from parsel import Selector
 from typing_extensions import Self, TypeVar, override
 
-from crawlee._types import BasicCrawlingContext, JsonSerializable, RequestHandlerRunResult
+from crawlee._types import BasicCrawlingContext, ConcurrencySettings, JsonSerializable, RequestHandlerRunResult
 from crawlee._utils.docs import docs_group
 from crawlee._utils.wait import wait_for
 from crawlee.crawlers import (
@@ -71,7 +71,6 @@ class _NonPersistentStatistics(Statistics):
     async def __aenter__(self) -> Self:
         self._active = True
         await self._state.initialize()
-        self._after_initialize()
         return self
 
     async def __aexit__(
@@ -85,8 +84,8 @@ class _NonPersistentStatistics(Statistics):
 
 @docs_group('Crawlers')
 class AdaptivePlaywrightCrawler(
-    Generic[TStaticCrawlingContext, TStaticParseResult, TStaticSelectResult],
     BasicCrawler[AdaptivePlaywrightCrawlingContext, AdaptivePlaywrightCrawlerStatisticState],
+    Generic[TStaticCrawlingContext, TStaticParseResult, TStaticSelectResult],
 ):
     """An adaptive web crawler capable of using both static HTTP request based crawling and browser based crawling.
 
@@ -149,15 +148,15 @@ 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)
         self.result_comparator = result_comparator or create_default_comparator(result_checker)
 
+        # Set default concurrency settings for browser crawlers if not provided
+        if 'concurrency_settings' not in kwargs or kwargs['concurrency_settings'] is None:
+            kwargs['concurrency_settings'] = ConcurrencySettings(desired_concurrency=1)
+
         super().__init__(statistics=statistics, **kwargs)
 
         # Sub crawlers related.
@@ -166,11 +165,11 @@ 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)
@@ -315,7 +314,7 @@ class AdaptivePlaywrightCrawler(
                 ),
                 logger=self._logger,
             )
-            return SubCrawlerRun(result=result)
+            return SubCrawlerRun(result=result, run_context=context_linked_to_result)
         except Exception as e:
             return SubCrawlerRun(exception=e)
 
@@ -371,7 +370,8 @@ class AdaptivePlaywrightCrawler(
                 self.track_http_only_request_handler_runs()
 
                 static_run = await self._crawl_one(rendering_type='static', context=context)
-                if static_run.result and self.result_checker(static_run.result):
+                if static_run.result and static_run.run_context and self.result_checker(static_run.result):
+                    self._update_context_from_copy(context, static_run.run_context)
                     self._context_result_map[context] = static_run.result
                     return
                 if static_run.exception:
@@ -402,13 +402,10 @@ class AdaptivePlaywrightCrawler(
         if pw_run.exception is not None:
             raise pw_run.exception
 
-        if pw_run.result:
-            self._context_result_map[context] = pw_run.result
-
+        if pw_run.result and pw_run.run_context:
             if should_detect_rendering_type:
                 detection_result: RenderingType
                 static_run = await self._crawl_one('static', context=context, state=old_state_copy)
-
                 if static_run.result and self.result_comparator(static_run.result, pw_run.result):
                     detection_result = 'static'
                 else:
@@ -417,6 +414,9 @@ class AdaptivePlaywrightCrawler(
                 context.log.debug(f'Detected rendering type {detection_result} for {context.request.url}')
                 self.rendering_type_predictor.store_result(context.request, detection_result)
 
+            self._update_context_from_copy(context, pw_run.run_context)
+            self._context_result_map[context] = pw_run.result
+
     def pre_navigation_hook(
         self,
         hook: Callable[[AdaptivePlaywrightPreNavCrawlingContext], Awaitable[None]] | None = None,
@@ -451,8 +451,32 @@ class AdaptivePlaywrightCrawler(
     def track_rendering_type_mispredictions(self) -> None:
         self.statistics.state.rendering_type_mispredictions += 1
 
+    def _update_context_from_copy(self, context: BasicCrawlingContext, context_copy: BasicCrawlingContext) -> None:
+        """Update mutable fields of `context` from `context_copy`.
+
+        Uses object.__setattr__ to bypass frozen dataclass restrictions,
+        allowing state synchronization after isolated crawler execution.
+        """
+        updating_attributes = {
+            'request': ('headers', 'user_data'),
+            'session': ('_user_data', '_usage_count', '_error_score', '_cookies'),
+        }
+
+        for attr, sub_attrs in updating_attributes.items():
+            original_sub_obj = getattr(context, attr)
+            copy_sub_obj = getattr(context_copy, attr)
+
+            # Check that both sub objects are not None
+            if original_sub_obj is None or copy_sub_obj is None:
+                continue
+
+            for sub_attr in sub_attrs:
+                new_value = getattr(copy_sub_obj, sub_attr)
+                object.__setattr__(original_sub_obj, sub_attr, new_value)
+
 
 @dataclass(frozen=True)
 class SubCrawlerRun:
     result: RequestHandlerRunResult | None = None
     exception: Exception | None = None
+    run_context: BasicCrawlingContext | None = None

crawlee/crawlers/_adaptive_playwright/_adaptive_playwright_crawler_statistics.py

@@ -12,7 +12,7 @@ from crawlee.statistics import StatisticsState
 class AdaptivePlaywrightCrawlerStatisticState(StatisticsState):
     """Statistic data about a crawler run with additional information related to adaptive crawling."""
 
-    model_config = ConfigDict(populate_by_name=True, ser_json_inf_nan='constants')
+    model_config = ConfigDict(validate_by_name=True, validate_by_alias=True, ser_json_inf_nan='constants')
 
     http_only_request_handler_runs: Annotated[int, Field(alias='http_only_request_handler_runs')] = 0
     """Number representing how many times static http based crawling was used."""

crawlee/crawlers/_adaptive_playwright/_adaptive_playwright_crawling_context.py

@@ -31,7 +31,8 @@ class AdaptiveContextError(RuntimeError):
 @dataclass(frozen=True)
 @docs_group('Crawling contexts')
 class AdaptivePlaywrightCrawlingContext(
-    Generic[TStaticParseResult, TStaticSelectResult], ParsedHttpCrawlingContext[TStaticParseResult]
+    ParsedHttpCrawlingContext[TStaticParseResult],
+    Generic[TStaticParseResult, TStaticSelectResult],
 ):
     _static_parser: AbstractHttpParser[TStaticParseResult, TStaticSelectResult]
     """The crawling context used by `AdaptivePlaywrightCrawler`.

crawlee/crawlers/_adaptive_playwright/_rendering_type_predictor.py

@@ -32,7 +32,7 @@ FeatureVector = tuple[float, float]
 
 
 class RenderingTypePredictorState(BaseModel):
-    model_config = ConfigDict(populate_by_name=True)
+    model_config = ConfigDict(validate_by_name=True, validate_by_alias=True)
 
     model: Annotated[
         LogisticRegression,