chg: Improve recaptures (WiP)

lookyloo/capturecache.py

@@ -28,7 +28,7 @@ from pyipasnhistory import IPASNHistory  # type: ignore[attr-defined]
 from redis import Redis
 
 from .context import Context
-from .helpers import get_captures_dir, is_locked
+from .helpers import get_captures_dir, is_locked, load_capture_settings
 from .indexing import Indexing
 from .default import LookylooException, try_make_file, get_config
 from .exceptions import MissingCaptureDirectory, NoValidHarFile, MissingUUID, TreeNeedsRebuild
@@ -413,12 +413,7 @@ class CapturesIndex(Mapping):  # type: ignore[type-arg]
             raise MissingCaptureDirectory(f'Unable to find the UUID file in {capture_dir}.')
 
         # Get capture settings as they were submitted
-        capture_settings_file = capture_dir / 'capture_settings.json'
-        if capture_settings_file.exists():
-            with capture_settings_file.open() as f:
-                capture_settings = json.load(f)
-        else:
-            capture_settings = {}
+        capture_settings = load_capture_settings(capture_dir)
 
         logger = LookylooCacheLogAdapter(self.logger, {'uuid': uuid})
         try:
@@ -435,7 +430,7 @@ class CapturesIndex(Mapping):  # type: ignore[type-arg]
                 tree = None
 
         cache: dict[str, str | int] = {'uuid': uuid, 'capture_dir': capture_dir_str}
-        if capture_settings.get('url'):
+        if capture_settings.get('url') and capture_settings['url'] is not None:
             cache['url'] = capture_settings['url']
 
         if (capture_dir / 'error.txt').exists():

lookyloo/helpers.py

@@ -82,6 +82,15 @@ def get_email_template() -> str:
         return f.read()
 
 
+@lru_cache(256)
+def load_capture_settings(capture_dir: Path) -> CaptureSettings:
+    capture_settings_file = capture_dir / 'capture_settings.json'
+    if capture_settings_file.exists():
+        with capture_settings_file.open() as f:
+            return json.load(f)
+    return {}
+
+
 @lru_cache
 def load_takedown_filters() -> tuple[re.Pattern[str], re.Pattern[str], dict[str, list[str]]]:
     filter_ini_file = get_homedir() / 'config' / 'takedown_filters.ini'

website/web/__init__.py

@@ -41,7 +41,9 @@ from werkzeug.wrappers.response import Response as WerkzeugResponse
 from lookyloo import Lookyloo, CaptureSettings
 from lookyloo.default import get_config
 from lookyloo.exceptions import MissingUUID, NoValidHarFile, LacusUnreachable
-from lookyloo.helpers import get_taxonomies, UserAgents, load_cookies, UserCaptureSettings, load_user_config
+from lookyloo.helpers import (get_taxonomies, UserAgents, load_cookies,
+                              UserCaptureSettings, load_user_config,
+                              load_capture_settings)
 
 if sys.version_info < (3, 9):
     from pytz import all_timezones_set
@@ -1424,14 +1426,14 @@ def search() -> str | Response | WerkzeugResponse:
     return render_template('search.html')
 
 
-def _prepare_capture_template(user_ua: str | None, predefined_url: str | None=None, *,
+def _prepare_capture_template(user_ua: str | None, predefined_settings: CaptureSettings | None=None, *,
                               user_config: UserCaptureSettings | None=None) -> str:
     return render_template('capture.html', user_agents=user_agents.user_agents,
                            default=user_agents.default,
                            personal_ua=user_ua,
                            default_public=get_config('generic', 'default_public'),
                            devices=lookyloo.get_playwright_devices(),
-                           predefined_url_to_capture=predefined_url if predefined_url else '',
+                           predefined_settings=predefined_settings if predefined_settings else {},
                            user_config=user_config,
                            show_project_page=get_config('generic', 'show_project_page'),
                            version=pkg_version,
@@ -1441,9 +1443,10 @@ def _prepare_capture_template(user_ua: str | None, predefined_url: str | None=No
 @app.route('/recapture/<string:tree_uuid>', methods=['GET'])
 def recapture(tree_uuid: str) -> str | Response | WerkzeugResponse:
     cache = lookyloo.capture_cache(tree_uuid)
-    if cache and hasattr(cache, 'url'):
+    if cache and hasattr(cache, 'capture_dir'):
+        capture_settings = load_capture_settings(cache.capture_dir)
         return _prepare_capture_template(user_ua=request.headers.get('User-Agent'),
-                                         predefined_url=cache.url)
+                                         predefined_settings=capture_settings)
     flash(f'Unable to find the capture {tree_uuid} in the cache.', 'error')
     return _prepare_capture_template(user_ua=request.headers.get('User-Agent'))
 

website/web/templates/capture.html

@@ -82,7 +82,8 @@
     <div class="row mb-3">
       <div class="col-sm-10">
         <div class="form-check">
-          <input class="form-check-input" type="checkbox" id="listing" name="listing" {% if default_public %}checked="checked"{% endif %}>
+          <input class="form-check-input" type="checkbox" id="listing" name="listing"
+            {% if default_public or predefined_settings.get('listing') is true %}checked="checked"{% endif %}>
           <label for="listing" class="form-check-label">Display results on public page</label>
         </div>
       </div>
@@ -102,7 +103,7 @@
         <div class="row input-group mb-3">
           <label for="url" class="col-sm-1 col-form-label">URL:</label>
           <input type="text" class="form-control col-auto" name="url" id=singleCaptureField
-                 placeholder="URL to capture" value="{{predefined_url_to_capture}}" required>
+                 placeholder="URL to capture" value="{{predefined_settings.get('url', '')}}" required>
 
           <textarea class="form-control col-auto d-none" placeholder="URLs to capture, one per line"
                     name="urls" id=multipleCapturesField></textarea>
@@ -280,7 +281,8 @@
           <label for="allow_tracking" class="col-sm-2 col-form-check-label">Allow tracking:</label>
           <div class="col-sm-10">
             <div class="form-check">
-              <input class="form-check-input" type="checkbox" id="allow_tracking" name="allow_tracking" aria-describedby="allow_tracking_help">
+              <input class="form-check-input" type="checkbox" id="allow_tracking" name="allow_tracking" aria-describedby="allow_tracking_help"
+                {% if predefined_settings.get('allow_tracking') is true %}checked="checked"{% endif %}>
               <div id="allow_tracking_help" class="form-text">We'll attempt to click on the button allowing the website captured to violate your privacy.</div>
             </div>
           </div>