fix(gcp_pubsub): remove conflicting `request_timeout` option

Since the buffer worker schema already contains that configuration,
having it two places can lead to quite confusing behavior.

lib-ee/emqx_ee_bridge/src/emqx_ee_bridge_gcp_pubsub.erl

@@ -79,15 +79,6 @@ fields(bridge_config) ->
                     desc => ?DESC("max_retries")
                 }
             )},
-        {request_timeout,
-            sc(
-                emqx_schema:duration_ms(),
-                #{
-                    required => false,
-                    default => <<"15s">>,
-                    desc => ?DESC("request_timeout")
-                }
-            )},
         {payload_template,
             sc(
                 binary(),

lib-ee/emqx_ee_bridge/test/emqx_ee_bridge_gcp_pubsub_SUITE.erl

@@ -282,7 +282,6 @@ gcp_pubsub_config(Config) ->
             "bridges.gcp_pubsub.~s {\n"
             "  enable = true\n"
             "  connect_timeout = 1s\n"
-            "  request_timeout = 1s\n"
             "  service_account_json = ~s\n"
             "  payload_template = ~p\n"
             "  pubsub_topic = ~s\n"

lib-ee/emqx_ee_connector/src/emqx_ee_connector_gcp_pubsub.erl

@@ -33,7 +33,7 @@
     connect_timeout := emqx_schema:duration_ms(),
     max_retries := non_neg_integer(),
     pubsub_topic := binary(),
-    request_timeout := emqx_schema:duration_ms(),
+    resource_opts := #{request_timeout := emqx_schema:duration_ms(), any() => term()},
     service_account_json := service_account_json(),
     any() => term()
 }.
@@ -71,7 +71,7 @@ on_start(
         payload_template := PayloadTemplate,
         pool_size := PoolSize,
         pubsub_topic := PubSubTopic,
-        request_timeout := RequestTimeout
+        resource_opts := #{request_timeout := RequestTimeout}
     } = Config
 ) ->
     ?SLOG(info, #{