lacuscore 1.13.5__tar.gz → 1.13.7__tar.gz

{lacuscore-1.13.5 → lacuscore-1.13.7}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: lacuscore
-Version: 1.13.5
+Version: 1.13.7
 Summary: Core of Lacus, usable as a module
 License: BSD-3-Clause
 Author: Raphaël Vinot
@@ -26,7 +26,7 @@ Requires-Dist: async-timeout (>=5.0.1) ; python_version < "3.11"
 Requires-Dist: defang (>=0.5.3)
 Requires-Dist: dnspython (>=2.7.0)
 Requires-Dist: eval-type-backport (>=0.2.2) ; python_version < "3.10"
-Requires-Dist: playwrightcapture[recaptcha] (>=1.28.2)
+Requires-Dist: playwrightcapture[recaptcha] (>=1.28.3)
 Requires-Dist: pydantic (>=2.10.6)
 Requires-Dist: redis[hiredis] (>=5.2.1)
 Requires-Dist: requests (>=2.32.3)

{lacuscore-1.13.5 → lacuscore-1.13.7}/lacuscore/helpers.py

@@ -5,6 +5,7 @@ from __future__ import annotations
 import json
 import sys
 
+from datetime import datetime, timedelta
 from enum import IntEnum, unique
 from logging import LoggerAdapter
 from typing import Any, Literal
@@ -127,6 +128,7 @@ class CaptureSettings(BaseModel):
     java_script_enabled: bool = True
     viewport: dict[str, int] | None = None
     referer: str | None = None
+    with_screenshot: bool = True
     with_favicon: bool = True
     allow_tracking: bool = False
     headless: bool = True
@@ -218,6 +220,28 @@ class CaptureSettings(BaseModel):
     @field_validator('cookies', mode='before')
     @classmethod
     def load_cookies_json(cls, cookies: Any) -> list[dict[str, Any]] | None:
+
+        def __prepare_cookie(cookie: dict[str, Any]) -> dict[str, str | float | bool]:
+            if len(cookie) == 1:
+                # {'name': 'value'} => {'name': 'name', 'value': 'value'}
+                name, value = cookie.popitem()
+                if name and value:
+                    cookie = {'name': name, 'value': value}
+            if not cookie.get('name') or not cookie.get('value'):
+                # invalid cookie, ignoring
+                return {}
+
+            if 'expires' in cookie and isinstance(cookie['expires'], str):
+                # Make it a float, as expected by Playwright
+                try:
+                    cookie['expires'] = datetime.fromisoformat(cookie['expires']).timestamp()
+                except ValueError:
+                    # if it ends with a Z, it fails in python < 3.12
+                    # And we don't really care.
+                    # make it expire 10 days from now
+                    cookie['expires'] = (datetime.now() + timedelta(days=10)).timestamp()
+            return cookie
+
         if not cookies:
             return None
         if isinstance(cookies, str):
@@ -236,16 +260,7 @@ class CaptureSettings(BaseModel):
             to_return = []
             for cookie in cookies:
                 if isinstance(cookie, dict):
-                    if 'name' in cookie and 'value' in cookie:
-                        to_return.append(cookie)
-                    elif len(cookie) == 1:
-                        # {'name': 'value'} => {'name': 'name', 'value': 'value'}
-                        name, value = cookie.popitem()
-                        if name and value:
-                            to_return.append({'name': name, 'value': value})
-                    else:
-                        # invalid cookie, ignoring
-                        pass
+                    to_return.append(__prepare_cookie(cookie))
             return to_return
         return None
 

{lacuscore-1.13.5 → lacuscore-1.13.7}/lacuscore/lacuscore.py

@@ -144,6 +144,7 @@ class LacusCore():
                 viewport: dict[str, int] | None=None,
                 referer: str | None=None,
                 rendered_hostname_only: bool=True,
+                with_screenshot: bool=True,
                 with_favicon: bool=False,
                 allow_tracking: bool=False,
                 headless: bool=True,
@@ -176,6 +177,7 @@ class LacusCore():
                 viewport: dict[str, int] | None=None,
                 referer: str | None=None,
                 rendered_hostname_only: bool=True,
+                with_screenshot: bool=True,
                 with_favicon: bool=False,
                 allow_tracking: bool=False,
                 headless: bool=True,
@@ -210,6 +212,7 @@ class LacusCore():
         :param viewport: The viewport of the browser used for capturing
         :param referer: The referer URL for the capture
         :param rendered_hostname_only: If depth > 0: only capture URLs with the same hostname as the rendered page
+        :param with_screenshot: If False, PlaywrightCapture won't take a screenshot of the rendered URL
         :param with_favicon: If True, PlaywrightCapture will attempt to get the potential favicons for the rendered URL. It is a dirty trick, see this issue for details: https://github.com/Lookyloo/PlaywrightCapture/issues/45
         :param allow_tracking: If True, PlaywrightCapture will attempt to click through the cookie banners. It is totally dependent on the framework used on the website.
         :param headless: Whether to run the browser in headless mode. WARNING: requires to run in a graphical environment.
@@ -232,7 +235,8 @@ class LacusCore():
                         'http_credentials': http_credentials, 'geolocation': geolocation,
                         'timezone_id': timezone_id, 'locale': locale,
                         'color_scheme': color_scheme, 'java_script_enabled': java_script_enabled,
-                        'viewport': viewport, 'referer': referer, 'with_favicon': with_favicon,
+                        'viewport': viewport, 'referer': referer,
+                        'with_screenshot': with_screenshot, 'with_favicon': with_favicon,
                         'allow_tracking': allow_tracking,
                         # Quietly force it to true if headed is not allowed.
                         'headless': headless if self.headed_allowed else True,
@@ -530,6 +534,7 @@ class LacusCore():
                                 url, referer=to_capture.referer,
                                 depth=to_capture.depth,
                                 rendered_hostname_only=to_capture.rendered_hostname_only,
+                                with_screenshot=to_capture.with_screenshot,
                                 with_favicon=to_capture.with_favicon,
                                 allow_tracking=to_capture.allow_tracking,
                                 max_depth_capture_time=self.max_capture_time)

{lacuscore-1.13.5 → lacuscore-1.13.7}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "lacuscore"
-version = "1.13.5"
+version = "1.13.7"
 description = "Core of Lacus, usable as a module"
 authors = [
     {name="Raphaël Vinot", email="raphael.vinot@circl.lu"}
@@ -15,7 +15,7 @@ dynamic = [ "classifiers" ]
 
 dependencies = [
     "requests (>=2.32.3)",
-    "playwrightcapture[recaptcha] (>=1.28.2)",
+    "playwrightcapture[recaptcha] (>=1.28.3)",
     "defang (>=0.5.3)",
     "ua-parser[regex] (>=1.0.1)",
     "redis [hiredis] (>=5.2.1)",