trying to fix the https requests

ontologytimemachine/custom_proxy.py

@@ -1,6 +1,7 @@
-from proxy.http.proxy import HttpProxyBasePlugin
+from proxy.http.proxy import HttpProxyBasePlugin, HttpProxyPlugin
 from proxy.http.parser import HttpParser
 from proxy.common.utils import build_http_response
+from proxy.http.codes import httpStatusCodes
 import requests
 import sys
 import proxy
@@ -24,6 +25,16 @@ def __init__(
         super().__init__(*args, **kwargs)
 
     def before_upstream_connection(self, request: HttpParser):
+        if request.method == b'CONNECT':
+            # Handle HTTPS connection establishment
+            host, port = request.host, request.port
+            self.client.queue(
+                build_http_response(
+                    httpStatusCodes.SWITCHING_PROTOCOLS, reason=b'Connection established', headers={
+                        b'Proxy-Agent': b'Python Proxy'
+                    }
+                )
+            )
         return None
 
 
@@ -38,12 +49,10 @@ def handle_client_request(self, request: HttpParser):
                 )
             )
             return None
-
         # Process other requests using proxy logic
         self.proxy_logic(request)
         return None
 
-
     def proxy_logic(self, request: HttpParser):
         self.failover_mode(request)
         self.time_based_mode(request)
@@ -52,17 +61,38 @@ def proxy_logic(self, request: HttpParser):
 
     def failover_mode(self, request):
         logging.info('Failover mode')
-        ontology = str(request._url)
+        logging.info(f'Method: {request.method}')
+        logging.info(f'URL: {request._url}')
+        scheme = 'https' if request.method == b'CONNECT' else 'http'
+        logging.info(f'Scheme: {scheme}')
+        if scheme == 'https':
+            host = request.host.decode()
+            logging.info(f'Host: {host}')
+            port = request.port if request.port else (443 if scheme == 'https' else 80)
+            logging.info(f'Port: {port}')
+            path = request.path.decode() if request.path else '/'
+            #logging.info(f'{request.url.path}')
+            logging.info(f'Path: {path}')
+            full_url = f'{scheme}://{host}:{port}{path}'
+
+            #logging.info(response.text)
+            logging.info(full_url)
+            ontology = full_url
+        else:
+            ontology = str(request._url)
+
         logging.info(f'Ontology: {ontology}')
 
         try:
             response = requests.get(ontology)
+            #logging.info(f'{response.text}')
             logging.info('Response received')
             logging.info(f'Response status code: {response.status_code}')
             content_type = response.headers.get('Content-Type', '')
-            logging.info(f'Content type: {content_type}')
-
-            if response.status_code == 200 and content_type in ['text/turtle']:
+            logging.info(f'Content type: {content_type}')   
+            logging.info(f'Ontology: {ontology}')
+            if response.status_code == 200 and (content_type in ['text/turtle', 'text/html; charset=utf-8'] or 'text/html' in content_type):
+                logging.info(f'Send answer')
                 self.client.queue(
                     build_http_response(
                         200, reason=b'OK', headers={
@@ -130,9 +160,14 @@ def on_client_connection_close(self):
 
 
 if __name__ == '__main__':
+    sys.argv += [
+        '--ca-key-file', 'ca-key.pem',
+        '--ca-cert-file', 'ca-cert.pem',
+        '--ca-signing-key-file', 'ca-signing-key.pem'
+    ]
     sys.argv += [
         '--hostname', IP,
         '--port', PORT,
-        '--plugins', __name__ + '.OntologyTimeMachinePlugin',
+        '--plugins', __name__ + '.OntologyTimeMachinePlugin'
     ]
     proxy.main()