refactor: change Scroll action to element-based navigation

Author: mmmay0722

tests/mocks/action_mocks.json

@@ -336,56 +336,53 @@
       "actions": [
         {
           "type": "Scroll",
-          "locate": null,
-          "param": {
-            "direction": "down",
-            "scrollType": "untilBottom",
-            "distance": null
-          }
+          "locate": {"id": "1"},
+          "param": null
         },
         {
           "type": "Scroll",
-          "locate": null,
-          "param": {
-            "direction": "up",
-            "scrollType": "untilTop",
-            "distance": null
-          }
+          "locate": {"id": "2"},
+          "param": null
         }
       ],
-      "id_map": {}
+      "id_map": {
+        "1": {
+          "id": 100,
+          "tagName": "footer",
+          "isInViewport": false,
+          "selector": "footer#site-footer",
+          "xpath": "//*[@id=\"site-footer\"]",
+          "center_y": 2000
+        },
+        "2": {
+          "id": 101,
+          "tagName": "header",
+          "isInViewport": false,
+          "selector": "header#site-header",
+          "xpath": "//*[@id=\"site-header\"]",
+          "center_y": 50
+        }
+      }
     },
     {
       "url": "https://arxiv.org/list/astro-ph/new",
       "actions": [
         {
           "type": "Scroll",
-          "locate": null,
-          "param": {
-            "direction": "down",
-            "scrollType": "once"
-          }
-        },
-        {
-          "type": "Scroll",
-          "locate": null,
-          "param": {
-            "direction": "down",
-            "scrollType": "once",
-            "distance": 400
-          }
-        },
-        {
-          "type": "Scroll",
-          "locate": null,
-          "param": {
-            "direction": "up",
-            "scrollType": "once",
-            "distance": 200
-          }
+          "locate": {"id": "1"},
+          "param": null
         }
       ],
-      "id_map": {}
+      "id_map": {
+        "1": {
+          "id": 200,
+          "tagName": "section",
+          "isInViewport": false,
+          "selector": "section.results",
+          "xpath": "//section[@class='results']",
+          "center_y": 1500
+        }
+      }
     }
   ],
   "SelectDropdown": [

tests/mocks/actions_negative_mocks.json

@@ -63,24 +63,30 @@
   ],
   "Scroll_Negative": [
     {
+      "description": "Scroll without locate.id should fail",
       "url": "https://arxiv.org/list/astro-ph/new",
       "actions": [
         {
           "type": "Scroll",
           "locate": null,
-          "param": {
-            "direction": "right",
-            "scrollType": "once"
-          }
+          "param": null
         },
         {
           "type": "Scroll",
-          "locate": null,
-          "param": {
-            "direction": "down",
-            "scrollType": "untilRight",
-            "distance": null
-          }
+          "locate": {},
+          "param": null
+        }
+      ],
+      "id_map": {}
+    },
+    {
+      "description": "Scroll with non-existent element should fail",
+      "url": "https://arxiv.org/list/astro-ph/new",
+      "actions": [
+        {
+          "type": "Scroll",
+          "locate": {"id": "999"},
+          "param": null
         }
       ],
       "id_map": {}

webqa_agent/actions/action_executor.py

@@ -68,8 +68,6 @@ def _validate_params(self, action, required_params):
             for key in keys:
                 value = value.get(key)
                 if value is None:
-                    if action["type"] == "Scroll" and key == "distance":
-                        continue
                     logging.error(f"Missing required parameter: {param}")
                     return False  # Return False to indicate validation failure
         return True  # Return True if all parameters are present
@@ -267,18 +265,38 @@ async def _execute_input(self, action):
             return {"success": False, "message": f"Input action failed with an exception: {e}"}
 
     async def _execute_scroll(self, action):
-        """Execute scroll action."""
-        if not self._validate_params(action, ["param.direction", "param.scrollType", "param.distance"]):
-            return {"success": False, "message": "Missing parameters for scroll action"}
-        direction = action.get("param").get("direction", "down")
-        scroll_type = action.get("param").get("scrollType", "once")
-        distance = action.get("param").get("distance", None)
-
-        success = await self._actions.scroll(direction, scroll_type, distance)
+        """Execute scroll action - scroll to a specific element."""
+        if not self._validate_params(action, ["locate.id"]):
+            return {"success": False, "message": "Missing locate.id for scroll action. Scroll requires an element ID to scroll to."}
+        
+        element_id = action.get("locate", {}).get("id")
+        
+        success = await self._actions.scroll(element_id)
+        
+        # Read action context for detailed error information
+        ctx = action_context_var.get()
+        
         if success:
-            return {"success": True, "message": f"Scrolled {direction} successfully."}
+            return {"success": True, "message": f"Scrolled to element {element_id} successfully."}
         else:
-            return {"success": False, "message": "Scroll action failed."}
+            # Enrich error message with context
+            base_message = f"Scroll to element {element_id} failed."
+            error_details = {}
+            
+            if ctx and ctx.error_type:
+                error_details = {
+                    "error_type": ctx.error_type,
+                    "error_reason": ctx.error_reason,
+                    "attempted_strategies": ctx.attempted_strategies,
+                    "element_info": ctx.element_info,
+                    "playwright_error": ctx.playwright_error
+                }
+            
+            return {
+                "success": False,
+                "message": base_message,
+                "error_details": error_details
+            }
 
     async def _execute_keyboard_press(self, action):
         """Execute keyboard press action."""