PlaywrightCapture 1.25.3__tar.gz → 1.25.5__tar.gz

{playwrightcapture-1.25.3 → playwrightcapture-1.25.5}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: PlaywrightCapture
-Version: 1.25.3
+Version: 1.25.5
 Summary: A simple library to capture websites using playwright
 Home-page: https://github.com/Lookyloo/PlaywrightCapture
 License: BSD-3-Clause

{playwrightcapture-1.25.3 → playwrightcapture-1.25.5}/playwrightcapture/capture.py

@@ -165,10 +165,15 @@ class Capture():
         if proxy:
             if isinstance(proxy, str):
                 self.proxy = {'server': proxy}
-            else:
+            elif isinstance(proxy, dict):
                 self.proxy = {'server': proxy['server'], 'bypass': proxy.get('bypass', ''),
                               'username': proxy.get('username', ''),
                               'password': proxy.get('password', '')}
+            elif isinstance(proxy, int):
+                # This is clearly a mistake, just ignoring it
+                self.logger.warning('Proxy is an integer, this is a mistake, ignoring it.')
+            else:
+                raise InvalidPlaywrightParameter(f'Invalid proxy parameter: "{proxy}" ({type(proxy)})')
 
         self.should_retry: bool = False
         self.__network_not_idle: int = 2  # makes sure we do not wait for network idle the max amount of time the capture is allowed to take
@@ -682,8 +687,7 @@ class Capture():
             capturing_sub = False
             try:
                 page = await self.context.new_page()
-                await page.clock.install()
-                page.on("dialog", lambda dialog: dialog.accept())
+                # await page.clock.install()
             except Error as e:
                 self.logger.warning(f'The context is in a broken state: {e}')
                 self.should_retry = True
@@ -707,6 +711,7 @@ class Capture():
             page.set_default_timeout((self._capture_timeout - 2) * 1000)
             # trigger a callback on each request to store it in a dict indexed by URL to get it back from the favicon fetcher
             page.on("requestfinished", store_request)
+            page.on("dialog", lambda dialog: dialog.accept())
 
         try:
             # Parse the URL. If there is a fragment, we need to scroll to it manually
@@ -762,31 +767,31 @@ class Capture():
                 await self._wait_for_random_timeout(page, 5)  # Wait 5 sec after document loaded
                 self.logger.debug('Start instrumentation.')
 
-                # ==== recaptcha
-                # Same technique as: https://github.com/NikolaiT/uncaptcha3
-                if CAN_SOLVE_CAPTCHA:
-                    try:
-                        if (await page.locator("//iframe[@title='reCAPTCHA']").first.is_visible(timeout=3000)
-                                and await page.locator("//iframe[@title='reCAPTCHA']").first.is_enabled(timeout=2000)):
-                            self.logger.info('Found a captcha')
-                            await self._recaptcha_solver(page)
-                    except PlaywrightTimeoutError as e:
-                        self.logger.info(f'Captcha on {url} is not ready: {e}')
-                    except TargetClosedError as e:
-                        self.logger.warning(f'Target closed while resolving captcha on {url}: {e}')
-                    except Error as e:
-                        self.logger.warning(f'Error while resolving captcha on {url}: {e}')
-                    except Exception as e:
-                        self.logger.exception(f'General error with captcha solving on {url}: {e}')
-                # ======
-                # NOTE: testing
-                # await self.__cloudflare_bypass_attempt(page)
-                self.logger.debug('Done with captcha.')
-
                 # check if we have anything on the page. If we don't, the page is not working properly.
                 if await self._failsafe_get_content(page):
                     self.logger.debug('Got rendered content')
 
+                    # ==== recaptcha
+                    # Same technique as: https://github.com/NikolaiT/uncaptcha3
+                    if CAN_SOLVE_CAPTCHA:
+                        try:
+                            if (await page.locator("//iframe[@title='reCAPTCHA']").first.is_visible(timeout=3000)
+                                    and await page.locator("//iframe[@title='reCAPTCHA']").first.is_enabled(timeout=2000)):
+                                self.logger.info('Found a captcha')
+                                await self._recaptcha_solver(page)
+                        except PlaywrightTimeoutError as e:
+                            self.logger.info(f'Captcha on {url} is not ready: {e}')
+                        except TargetClosedError as e:
+                            self.logger.warning(f'Target closed while resolving captcha on {url}: {e}')
+                        except Error as e:
+                            self.logger.warning(f'Error while resolving captcha on {url}: {e}')
+                        except Exception as e:
+                            self.logger.exception(f'General error with captcha solving on {url}: {e}')
+                    # ======
+                    # NOTE: testing
+                    # await self.__cloudflare_bypass_attempt(page)
+                    self.logger.debug('Done with captcha.')
+
                     # move mouse
                     await page.mouse.move(x=random.uniform(300, 800), y=random.uniform(200, 500))
                     self.logger.debug('Moved mouse.')
@@ -812,7 +817,7 @@ class Capture():
                         try:
                             await page.locator(f'id={fragment}').first.scroll_into_view_if_needed(timeout=3000)
                             await self._wait_for_random_timeout(page, 2)
-                            async with timeout(3):
+                            async with timeout(5):
                                 await page.mouse.wheel(delta_y=random.uniform(150, 300), delta_x=0)
                             self.logger.debug('Jumped to fragment.')
                         except PlaywrightTimeoutError as e:
@@ -821,20 +826,24 @@ class Capture():
                             self.logger.warning(f'Target closed, unable to go to fragment "{fragment}": {e}')
                         except Error as e:
                             self.logger.exception(f'Unable to go to fragment "{fragment}": {e}')
-                        except TimeoutError:
+                        except (asyncio.TimeoutError, TimeoutError):
                             self.logger.debug('Unable to scroll due to timeout')
+                        except (asyncio.CancelledError):
+                            self.logger.debug('Unable to scroll due to timeout, call canceled')
                     else:
                         # scroll more
                         try:
                             # NOTE using page.mouse.wheel causes the instrumentation to fail, sometimes.
                             #   2024-07-08: Also, it sometimes get stuck.
-                            async with timeout(3):
+                            async with timeout(5):
                                 await page.mouse.wheel(delta_y=random.uniform(1500, 3000), delta_x=0)
                             self.logger.debug('Scrolled down.')
                         except Error as e:
                             self.logger.debug(f'Unable to scroll: {e}')
-                        except TimeoutError:
+                        except (TimeoutError, asyncio.TimeoutError):
                             self.logger.debug('Unable to scroll due to timeout')
+                        except (asyncio.CancelledError):
+                            self.logger.debug('Unable to scroll due to timeout, call canceled')
 
                     await self._wait_for_random_timeout(page, 3)
                     self.logger.debug('Keep going after moving on page.')
@@ -866,8 +875,12 @@ class Capture():
                         to_return["downloaded_file"] = mem_zip.getvalue()
 
                 # fast forward 30s
-                await page.clock.run_for("30")
-                self.logger.debug('Moved time forward.')
+                # try:
+                #    async with timeout(3):
+                #        await page.clock.run_for("47")
+                #        self.logger.debug('Moved time forward.')
+                # except (TimeoutError, asyncio.TimeoutError):
+                #    self.logger.warning('Unable to move time forward.')
 
                 self.logger.debug('Done with instrumentation, waiting for network idle.')
                 await self._wait_for_random_timeout(page, 5)  # Wait 5 sec after instrumentation
@@ -915,7 +928,7 @@ class Capture():
                                         rendered_hostname_only=rendered_hostname_only,
                                         max_depth_capture_time=max_capture_time)
                                     to_return['children'].append(child_capture)  # type: ignore[union-attr]
-                            except (TimeoutError, asyncio.exceptions.TimeoutError, asyncio.TimeoutError):
+                            except (TimeoutError, asyncio.TimeoutError):
                                 self.logger.info(f'Timeout error, took more than {max_capture_time}s. Unable to capture {url}.')
                                 consecutive_errors += 1
                             except Exception as e:
@@ -943,6 +956,9 @@ class Capture():
         except PlaywrightTimeoutError as e:
             to_return['error'] = f"The capture took too long - {e.message}"
             self.should_retry = True
+        except (asyncio.TimeoutError, TimeoutError):
+            to_return['error'] = "Something in the capture took too long"
+            self.should_retry = True
         except TargetClosedError as e:
             to_return['error'] = f"The target was closed - {e}"
             self.should_retry = True
@@ -1078,8 +1094,9 @@ class Capture():
         tries = 3
         while tries:
             try:
-                return await page.content()
-            except Error:
+                async with timeout(30):
+                    return await page.content()
+            except (Error, TimeoutError):
                 self.logger.debug('Unable to get page content, trying again.')
                 tries -= 1
                 await self._wait_for_random_timeout(page, 1)
@@ -1225,6 +1242,11 @@ class Capture():
             if ': ' in name:
                 _, name = name.split(': ', maxsplit=1)
             exception._name = name.strip()
+        else:
+            # The format changed in Playwright 1.43.0, the name of the method that failed is set before the exception itself.
+            if ': ' in exception.message:
+                _, name = exception.message.split(': ', maxsplit=1)
+            exception._name = name.strip()
 
     def _exception_is_network_error(self, exception: Error) -> bool:
         if exception.name in [

{playwrightcapture-1.25.3 → playwrightcapture-1.25.5}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "PlaywrightCapture"
-version = "1.25.3"
+version = "1.25.5"
 description = "A simple library to capture websites using playwright"
 authors = ["Raphaël Vinot <raphael.vinot@circl.lu>"]
 license = "BSD-3-Clause"