Attempt to save last data retrieved- WIP

changedetectionio/__init__.py

@@ -503,7 +503,7 @@ def changedetection_app(config=None, datastore_o=None):
         from changedetectionio import fetch_site_status
 
         # Get the most recent one
-        newest_history_key = datastore.data['watching'][uuid].get('newest_history_key')
+        newest_history_key = datastore.get_val(uuid, 'newest_history_key')
 
         # 0 means that theres only one, so that there should be no 'unviewed' history available
         if newest_history_key == 0:

changedetectionio/fetch_site_status.py

@@ -15,6 +15,7 @@ urllib3.disable_warnings(urllib3.exceptions.InsecureRequestWarning)
 class perform_site_check():
     screenshot = None
     xpath_data = None
+    fetched_response = None
 
     def __init__(self, *args, datastore, **kwargs):
         super().__init__(*args, **kwargs)
@@ -63,11 +64,13 @@ class perform_site_check():
 
 
     def run(self, uuid):
+        timestamp = int(time.time())  # used for storage etc too
+
         changed_detected = False
         screenshot = False  # as bytes
         stripped_text_from_html = ""
 
-        watch = self.datastore.data['watching'].get(uuid)
+        watch = self.datastore.data['watching'][uuid]
 
         # Protect against file:// access
         if re.search(r'^file', watch['url'], re.IGNORECASE) and not os.getenv('ALLOW_FILE_URI', False):
@@ -78,7 +81,7 @@ class perform_site_check():
         # Unset any existing notification error
         update_obj = {'last_notification_error': False, 'last_error': False}
 
-        extra_headers =self.datastore.data['watching'][uuid].get('headers')
+        extra_headers = self.datastore.get_val(uuid, 'headers')
 
         # Tweak the base config with the per-watch ones
         request_headers = self.datastore.data['settings']['headers'].copy()
@@ -91,9 +94,9 @@ class perform_site_check():
             request_headers['Accept-Encoding'] = request_headers['Accept-Encoding'].replace(', br', '')
 
         timeout = self.datastore.data['settings']['requests']['timeout']
-        url = watch.get('url')
-        request_body = self.datastore.data['watching'][uuid].get('body')
-        request_method = self.datastore.data['watching'][uuid].get('method')
+        url = self.datastore.get_val(uuid, 'url')
+        request_body = self.datastore.get_val(uuid, 'body')
+        request_method = self.datastore.get_val(uuid, 'method')
         ignore_status_codes = self.datastore.data['watching'][uuid].get('ignore_status_codes', False)
 
         # source: support
@@ -129,6 +132,7 @@ class perform_site_check():
 
         self.screenshot = fetcher.screenshot
         self.xpath_data = fetcher.xpath_data
+        self.fetched_response = fetcher.content
 
         # Fetching complete, now filters
         # @todo move to class / maybe inside of fetcher abstract base?

changedetectionio/store.py

@@ -244,6 +244,10 @@ class ChangeDetectionStore:
 
         return False
 
+    def get_val(self, uuid, val):
+        # Probably their should be dict...
+        return self.data['watching'][uuid].get(val)
+
     # Remove a watchs data but keep the entry (URL etc)
     def clear_watch_history(self, uuid):
         import pathlib
@@ -371,6 +375,17 @@ class ChangeDetectionStore:
             f.write(json.dumps(data))
             f.close()
 
+    # Save whatever was returned from the fetcher
+    def save_last_response(self, watch_uuid, data):
+        if not self.data['watching'].get(watch_uuid):
+            return
+
+        target_path = os.path.join(self.datastore_path, watch_uuid, "last-response.bin")
+        # mimetype? binary? text? @todo
+        # gzip if its non-binary? auto get encoding?
+        with open(target_path, 'wb') as f:
+            f.write(data)
+            f.close()
 
     def sync_to_json(self):
         logging.info("Saving JSON..")

changedetectionio/update_worker.py

@@ -286,6 +286,9 @@ class update_worker(threading.Thread):
                         self.datastore.save_screenshot(watch_uuid=uuid, screenshot=update_handler.screenshot)
                     if update_handler.xpath_data:
                         self.datastore.save_xpath_data(watch_uuid=uuid, data=update_handler.xpath_data)
+                    if update_handler.fetched_response:
+                        # @todo mimetype?
+                        self.datastore.save_last_response(watch_uuid=uuid, data=update_handler.fetched_response)
 
 
                 self.current_uuid = None  # Done