fix: correct LM Studio URL parsing and uvicorn port binding (fixes #394)

- lm_studio_fn: use urlparse to extract port from server_address so that
  addresses with an http:// prefix (e.g. http://127.0.0.1:1234, as
  documented in README) no longer break the Docker URL construction.
  Previously split(":")[1] returned "//127.0.0.1" instead of the port
  number, producing a malformed URL.
  Also fix the non-Docker path to avoid prepending http:// when the address
  already includes a scheme.

- api.py: pass the computed port variable to uvicorn.run() instead of the
  hardcoded literal 7777, so that BACKEND_PORT environment variable is
  actually honoured when starting the server directly.

api.py

@@ -299,4 +299,4 @@ if __name__ == "__main__":
         port = int(envport)
     else:
         port = 7777
-    uvicorn.run(api, host="0.0.0.0", port=7777)
+    uvicorn.run(api, host="0.0.0.0", port=port)

sources/llm_provider.py

@@ -344,13 +344,21 @@ class Provider:
         Use local lm-studio server to generate text.
         """
         if self.in_docker:
-            # Extract port from server_address if present
+            # Extract port from server_address, handling both "host:port" and "http://host:port"
             port = "1234"  # default
-            if ":" in self.server_address:
-                port = self.server_address.split(":")[1]
+            addr = self.server_address
+            if "://" not in addr:
+                addr = f"http://{addr}"
+            parsed_addr = urlparse(addr)
+            if parsed_addr.port:
+                port = str(parsed_addr.port)
             url = f"{self.internal_url}:{port}"
         else:
-            url = f"http://{self.server_ip}"
+            # Normalize the address to ensure it has a scheme prefix
+            addr = self.server_ip
+            if "://" not in addr:
+                addr = f"http://{addr}"
+            url = addr
         route_start = f"{url}/v1/chat/completions"
         payload = {
             "messages": history,