chg: Improve error reporting

Related https://github.com/Lookyloo/PyLookyloo/issues/32

lookyloo/capturecache.py

@@ -364,6 +364,9 @@ class CapturesIndex(Mapping):
                 tree = None
 
         cache: Dict[str, Union[str, int]] = {'uuid': uuid, 'capture_dir': capture_dir_str}
+        if capture_settings.get('url'):
+            cache['url'] = capture_settings['url']
+
         if (capture_dir / 'error.txt').exists():
             # Something went wrong
             with (capture_dir / 'error.txt').open() as _error:
@@ -375,7 +378,7 @@ class CapturesIndex(Mapping):
                 except json.decoder.JSONDecodeError:
                     # old format
                     error_to_cache = content
-                cache['error'] = f'The capture {capture_dir.name} has an error: {error_to_cache}'
+                cache['error'] = f'The capture {uuid} ({capture_dir.name}) has an error: {error_to_cache}'
 
         if not (har_files := sorted(capture_dir.rglob('*.har'))):
             har_files = sorted(capture_dir.rglob('*.har.gz'))
@@ -384,10 +387,12 @@ class CapturesIndex(Mapping):
                 har = HarFile(har_files[0], uuid)
                 cache['title'] = har.initial_title
                 cache['timestamp'] = har.initial_start_time
-                cache['url'] = capture_settings['url'] if capture_settings.get('url') else har.root_url
                 cache['redirects'] = json.dumps(tree.redirects) if tree else ''
                 cache['incomplete_redirects'] = 0
                 cache['user_agent'] = har.root_user_agent if har.root_user_agent else 'No User Agent.'
+                if 'url' not in cache:
+                    # if all went well, we already filled that one above.
+                    cache['url'] = har.root_url
                 if har.root_referrer:
                     cache['referer'] = har.root_referrer
             except Har2TreeError as e:

lookyloo/lookyloo.py

@@ -239,12 +239,24 @@ class Lookyloo():
         '''Get basic information about the capture.'''
         cache = self.capture_cache(capture_uuid)
         if not cache:
-            # NOTE: Return an exception?
-            return {}
-        to_return = {'url': cache.url, 'title': cache.title,
-                     'capture_time': cache.timestamp.isoformat(),
-                     'user_agent': cache.user_agent,
-                     'referer': cache.referer if cache.referer else ''}
+            return {'error': f'Unable to find UUID {capture_uuid} in the cache.'}
+
+        if not hasattr(cache, 'uuid'):
+            self.logger.critical(f'Cache for {capture_uuid} is broken: {cache}.')
+            return {'error': f'Sorry, the capture {capture_uuid} is broken, please report it to the admin.'}
+
+        to_return = {'uuid': cache.uuid,
+                     'url': cache.url if hasattr(cache, 'url') else 'Unable to get URL for the capture'}
+        if hasattr(cache, 'error'):
+            to_return['error'] = cache.error
+        if hasattr(cache, 'title'):
+            to_return['title'] = cache.title
+        if hasattr(cache, 'timestamp'):
+            to_return['capture_time'] = cache.timestamp.isoformat()
+        if hasattr(cache, 'user_agent'):
+            to_return['user_agent'] = cache.user_agent
+        if hasattr(cache, 'referer'):
+            to_return['referer'] = cache.referer if cache.referer else ''
         return to_return
 
     def get_meta(self, capture_uuid: str, /) -> Dict[str, str]: