@midscene/shared 1.0.1-beta-20251024063839.0 → 1.0.1-beta-20251024064637.0

package/src/env/types.ts

@@ -1,10 +1,10 @@
 // config keys
-export const MIDSCENE_MODEL_INIT_CONFIG_JSON =
-  'MIDSCENE_MODEL_INIT_CONFIG_JSON';
+export const MIDSCENE_OPENAI_INIT_CONFIG_JSON =
+  'MIDSCENE_OPENAI_INIT_CONFIG_JSON';
 export const MIDSCENE_MODEL_NAME = 'MIDSCENE_MODEL_NAME';
 export const MIDSCENE_LANGSMITH_DEBUG = 'MIDSCENE_LANGSMITH_DEBUG';
-export const MIDSCENE_DEBUG_MODEL_PROFILE = 'MIDSCENE_DEBUG_MODEL_PROFILE';
-export const MIDSCENE_DEBUG_MODEL_RESPONSE = 'MIDSCENE_DEBUG_MODEL_RESPONSE';
+export const MIDSCENE_DEBUG_AI_PROFILE = 'MIDSCENE_DEBUG_AI_PROFILE';
+export const MIDSCENE_DEBUG_AI_RESPONSE = 'MIDSCENE_DEBUG_AI_RESPONSE';
 export const MIDSCENE_DANGEROUSLY_PRINT_ALL_CONFIG =
   'MIDSCENE_DANGEROUSLY_PRINT_ALL_CONFIG';
 export const MIDSCENE_DEBUG_MODE = 'MIDSCENE_DEBUG_MODE';
@@ -15,20 +15,9 @@ export const MIDSCENE_MCP_ANDROID_MODE = 'MIDSCENE_MCP_ANDROID_MODE';
 export const DOCKER_CONTAINER = 'DOCKER_CONTAINER';
 export const MIDSCENE_FORCE_DEEP_THINK = 'MIDSCENE_FORCE_DEEP_THINK';
 
-export const MIDSCENE_MODEL_SOCKS_PROXY = 'MIDSCENE_MODEL_SOCKS_PROXY';
-export const MIDSCENE_MODEL_HTTP_PROXY = 'MIDSCENE_MODEL_HTTP_PROXY';
-
-// New primary names for public API
-export const MODEL_API_KEY = 'MODEL_API_KEY';
-export const MODEL_BASE_URL = 'MODEL_BASE_URL';
-
-/**
- * @deprecated Use MODEL_API_KEY instead. This is kept for backward compatibility.
- */
+export const MIDSCENE_OPENAI_SOCKS_PROXY = 'MIDSCENE_OPENAI_SOCKS_PROXY';
+export const MIDSCENE_OPENAI_HTTP_PROXY = 'MIDSCENE_OPENAI_HTTP_PROXY';
 export const OPENAI_API_KEY = 'OPENAI_API_KEY';
-/**
- * @deprecated Use MODEL_BASE_URL instead. This is kept for backward compatibility.
- */
 export const OPENAI_BASE_URL = 'OPENAI_BASE_URL';
 export const OPENAI_MAX_TOKENS = 'OPENAI_MAX_TOKENS';
 
@@ -61,48 +50,48 @@ export const MIDSCENE_CACHE_MAX_FILENAME_LENGTH =
 export const MIDSCENE_RUN_DIR = 'MIDSCENE_RUN_DIR';
 
 // default new
-export const MIDSCENE_MODEL_BASE_URL = 'MIDSCENE_MODEL_BASE_URL';
-export const MIDSCENE_MODEL_API_KEY = 'MIDSCENE_MODEL_API_KEY';
-export const MIDSCENE_LOCATOR_MODE = 'MIDSCENE_LOCATOR_MODE';
+export const MIDSCENE_OPENAI_BASE_URL = 'MIDSCENE_OPENAI_BASE_URL';
+export const MIDSCENE_OPENAI_API_KEY = 'MIDSCENE_OPENAI_API_KEY';
+export const MIDSCENE_VL_MODE = 'MIDSCENE_VL_MODE';
 
 // VQA
 export const MIDSCENE_VQA_MODEL_NAME = 'MIDSCENE_VQA_MODEL_NAME';
-export const MIDSCENE_VQA_MODEL_SOCKS_PROXY = 'MIDSCENE_VQA_MODEL_SOCKS_PROXY';
-export const MIDSCENE_VQA_MODEL_HTTP_PROXY = 'MIDSCENE_VQA_MODEL_HTTP_PROXY';
-export const MIDSCENE_VQA_MODEL_BASE_URL = 'MIDSCENE_VQA_MODEL_BASE_URL';
-export const MIDSCENE_VQA_MODEL_API_KEY = 'MIDSCENE_VQA_MODEL_API_KEY';
-export const MIDSCENE_VQA_MODEL_INIT_CONFIG_JSON =
-  'MIDSCENE_VQA_MODEL_INIT_CONFIG_JSON';
-export const MIDSCENE_VQA_LOCATOR_MODE = 'MIDSCENE_VQA_LOCATOR_MODE';
+export const MIDSCENE_VQA_OPENAI_SOCKS_PROXY =
+  'MIDSCENE_VQA_OPENAI_SOCKS_PROXY';
+export const MIDSCENE_VQA_OPENAI_HTTP_PROXY = 'MIDSCENE_VQA_OPENAI_HTTP_PROXY';
+export const MIDSCENE_VQA_OPENAI_BASE_URL = 'MIDSCENE_VQA_OPENAI_BASE_URL';
+export const MIDSCENE_VQA_OPENAI_API_KEY = 'MIDSCENE_VQA_OPENAI_API_KEY';
+export const MIDSCENE_VQA_OPENAI_INIT_CONFIG_JSON =
+  'MIDSCENE_VQA_OPENAI_INIT_CONFIG_JSON';
+export const MIDSCENE_VQA_VL_MODE = 'MIDSCENE_VQA_VL_MODE';
 
 // PLANNING
 export const MIDSCENE_PLANNING_MODEL_NAME = 'MIDSCENE_PLANNING_MODEL_NAME';
-export const MIDSCENE_PLANNING_MODEL_SOCKS_PROXY =
-  'MIDSCENE_PLANNING_MODEL_SOCKS_PROXY';
-export const MIDSCENE_PLANNING_MODEL_HTTP_PROXY =
-  'MIDSCENE_PLANNING_MODEL_HTTP_PROXY';
-export const MIDSCENE_PLANNING_MODEL_BASE_URL =
-  'MIDSCENE_PLANNING_MODEL_BASE_URL';
-export const MIDSCENE_PLANNING_MODEL_API_KEY =
-  'MIDSCENE_PLANNING_MODEL_API_KEY';
-export const MIDSCENE_PLANNING_MODEL_INIT_CONFIG_JSON =
-  'MIDSCENE_PLANNING_MODEL_INIT_CONFIG_JSON';
-export const MIDSCENE_PLANNING_LOCATOR_MODE = 'MIDSCENE_PLANNING_LOCATOR_MODE';
+export const MIDSCENE_PLANNING_OPENAI_SOCKS_PROXY =
+  'MIDSCENE_PLANNING_OPENAI_SOCKS_PROXY';
+export const MIDSCENE_PLANNING_OPENAI_HTTP_PROXY =
+  'MIDSCENE_PLANNING_OPENAI_HTTP_PROXY';
+export const MIDSCENE_PLANNING_OPENAI_BASE_URL =
+  'MIDSCENE_PLANNING_OPENAI_BASE_URL';
+export const MIDSCENE_PLANNING_OPENAI_API_KEY =
+  'MIDSCENE_PLANNING_OPENAI_API_KEY';
+export const MIDSCENE_PLANNING_OPENAI_INIT_CONFIG_JSON =
+  'MIDSCENE_PLANNING_OPENAI_INIT_CONFIG_JSON';
+export const MIDSCENE_PLANNING_VL_MODE = 'MIDSCENE_PLANNING_VL_MODE';
 
 // GROUNDING
 export const MIDSCENE_GROUNDING_MODEL_NAME = 'MIDSCENE_GROUNDING_MODEL_NAME';
-export const MIDSCENE_GROUNDING_MODEL_SOCKS_PROXY =
-  'MIDSCENE_GROUNDING_MODEL_SOCKS_PROXY';
-export const MIDSCENE_GROUNDING_MODEL_HTTP_PROXY =
-  'MIDSCENE_GROUNDING_MODEL_HTTP_PROXY';
-export const MIDSCENE_GROUNDING_MODEL_BASE_URL =
-  'MIDSCENE_GROUNDING_MODEL_BASE_URL';
-export const MIDSCENE_GROUNDING_MODEL_API_KEY =
-  'MIDSCENE_GROUNDING_MODEL_API_KEY';
-export const MIDSCENE_GROUNDING_MODEL_INIT_CONFIG_JSON =
-  'MIDSCENE_GROUNDING_MODEL_INIT_CONFIG_JSON';
-export const MIDSCENE_GROUNDING_LOCATOR_MODE =
-  'MIDSCENE_GROUNDING_LOCATOR_MODE';
+export const MIDSCENE_GROUNDING_OPENAI_SOCKS_PROXY =
+  'MIDSCENE_GROUNDING_OPENAI_SOCKS_PROXY';
+export const MIDSCENE_GROUNDING_OPENAI_HTTP_PROXY =
+  'MIDSCENE_GROUNDING_OPENAI_HTTP_PROXY';
+export const MIDSCENE_GROUNDING_OPENAI_BASE_URL =
+  'MIDSCENE_GROUNDING_OPENAI_BASE_URL';
+export const MIDSCENE_GROUNDING_OPENAI_API_KEY =
+  'MIDSCENE_GROUNDING_OPENAI_API_KEY';
+export const MIDSCENE_GROUNDING_OPENAI_INIT_CONFIG_JSON =
+  'MIDSCENE_GROUNDING_OPENAI_INIT_CONFIG_JSON';
+export const MIDSCENE_GROUNDING_VL_MODE = 'MIDSCENE_GROUNDING_VL_MODE';
 
 /**
  * env keys declared but unused
@@ -115,8 +104,8 @@ export const UNUSED_ENV_KEYS = [MIDSCENE_DANGEROUSLY_PRINT_ALL_CONFIG];
  */
 export const BASIC_ENV_KEYS = [
   MIDSCENE_DEBUG_MODE,
-  MIDSCENE_DEBUG_MODEL_PROFILE,
-  MIDSCENE_DEBUG_MODEL_RESPONSE,
+  MIDSCENE_DEBUG_AI_PROFILE,
+  MIDSCENE_DEBUG_AI_RESPONSE,
   MIDSCENE_RUN_DIR,
 ] as const;
 
@@ -167,47 +156,45 @@ export const GLOBAL_ENV_KEYS = [
 export const MODEL_ENV_KEYS = [
   // model default
   MIDSCENE_MODEL_NAME,
-  MIDSCENE_MODEL_INIT_CONFIG_JSON,
-  MIDSCENE_MODEL_API_KEY,
-  MIDSCENE_MODEL_BASE_URL,
-  MIDSCENE_MODEL_SOCKS_PROXY,
-  MIDSCENE_MODEL_HTTP_PROXY,
+  MIDSCENE_OPENAI_INIT_CONFIG_JSON,
+  MIDSCENE_OPENAI_API_KEY,
+  MIDSCENE_OPENAI_BASE_URL,
+  MIDSCENE_OPENAI_SOCKS_PROXY,
+  MIDSCENE_OPENAI_HTTP_PROXY,
   MIDSCENE_USE_VLM_UI_TARS,
   MIDSCENE_USE_QWEN_VL,
   MIDSCENE_USE_QWEN3_VL,
   MIDSCENE_USE_DOUBAO_VISION,
   MIDSCENE_USE_GEMINI,
   MIDSCENE_USE_VL_MODEL,
-  MIDSCENE_LOCATOR_MODE,
+  MIDSCENE_VL_MODE,
   // model default legacy
   OPENAI_API_KEY,
   OPENAI_BASE_URL,
-  MODEL_API_KEY,
-  MODEL_BASE_URL,
   // VQA
   MIDSCENE_VQA_MODEL_NAME,
-  MIDSCENE_VQA_MODEL_SOCKS_PROXY,
-  MIDSCENE_VQA_MODEL_HTTP_PROXY,
-  MIDSCENE_VQA_MODEL_BASE_URL,
-  MIDSCENE_VQA_MODEL_API_KEY,
-  MIDSCENE_VQA_MODEL_INIT_CONFIG_JSON,
-  MIDSCENE_VQA_LOCATOR_MODE,
+  MIDSCENE_VQA_OPENAI_SOCKS_PROXY,
+  MIDSCENE_VQA_OPENAI_HTTP_PROXY,
+  MIDSCENE_VQA_OPENAI_BASE_URL,
+  MIDSCENE_VQA_OPENAI_API_KEY,
+  MIDSCENE_VQA_OPENAI_INIT_CONFIG_JSON,
+  MIDSCENE_VQA_VL_MODE,
   // PLANNING
   MIDSCENE_PLANNING_MODEL_NAME,
-  MIDSCENE_PLANNING_MODEL_SOCKS_PROXY,
-  MIDSCENE_PLANNING_MODEL_HTTP_PROXY,
-  MIDSCENE_PLANNING_MODEL_BASE_URL,
-  MIDSCENE_PLANNING_MODEL_API_KEY,
-  MIDSCENE_PLANNING_MODEL_INIT_CONFIG_JSON,
-  MIDSCENE_PLANNING_LOCATOR_MODE,
+  MIDSCENE_PLANNING_OPENAI_SOCKS_PROXY,
+  MIDSCENE_PLANNING_OPENAI_HTTP_PROXY,
+  MIDSCENE_PLANNING_OPENAI_BASE_URL,
+  MIDSCENE_PLANNING_OPENAI_API_KEY,
+  MIDSCENE_PLANNING_OPENAI_INIT_CONFIG_JSON,
+  MIDSCENE_PLANNING_VL_MODE,
   // GROUNDING
   MIDSCENE_GROUNDING_MODEL_NAME,
-  MIDSCENE_GROUNDING_MODEL_SOCKS_PROXY,
-  MIDSCENE_GROUNDING_MODEL_HTTP_PROXY,
-  MIDSCENE_GROUNDING_MODEL_BASE_URL,
-  MIDSCENE_GROUNDING_MODEL_API_KEY,
-  MIDSCENE_GROUNDING_MODEL_INIT_CONFIG_JSON,
-  MIDSCENE_GROUNDING_LOCATOR_MODE,
+  MIDSCENE_GROUNDING_OPENAI_SOCKS_PROXY,
+  MIDSCENE_GROUNDING_OPENAI_HTTP_PROXY,
+  MIDSCENE_GROUNDING_OPENAI_BASE_URL,
+  MIDSCENE_GROUNDING_OPENAI_API_KEY,
+  MIDSCENE_GROUNDING_OPENAI_INIT_CONFIG_JSON,
+  MIDSCENE_GROUNDING_VL_MODE,
 ] as const;
 
 export const ALL_ENV_KEYS = [
@@ -240,14 +227,14 @@ export interface IModelConfigForVQA {
   // model name
   [MIDSCENE_VQA_MODEL_NAME]: string;
   // proxy
-  [MIDSCENE_VQA_MODEL_SOCKS_PROXY]?: string;
-  [MIDSCENE_VQA_MODEL_HTTP_PROXY]?: string;
+  [MIDSCENE_VQA_OPENAI_SOCKS_PROXY]?: string;
+  [MIDSCENE_VQA_OPENAI_HTTP_PROXY]?: string;
   // OpenAI
-  [MIDSCENE_VQA_MODEL_BASE_URL]?: string;
-  [MIDSCENE_VQA_MODEL_API_KEY]?: string;
-  [MIDSCENE_VQA_MODEL_INIT_CONFIG_JSON]?: string;
+  [MIDSCENE_VQA_OPENAI_BASE_URL]?: string;
+  [MIDSCENE_VQA_OPENAI_API_KEY]?: string;
+  [MIDSCENE_VQA_OPENAI_INIT_CONFIG_JSON]?: string;
   // extra
-  [MIDSCENE_VQA_LOCATOR_MODE]?: TVlModeValues;
+  [MIDSCENE_VQA_VL_MODE]?: TVlModeValues;
 }
 
 /**
@@ -256,7 +243,7 @@ export interface IModelConfigForVQA {
  * IMPORTANT: Planning MUST use a vision language model (VL mode).
  * DOM-based planning is not supported.
  *
- * Required: MIDSCENE_PLANNING_LOCATOR_MODE must be set to one of:
+ * Required: MIDSCENE_PLANNING_VL_MODE must be set to one of:
  *   - 'qwen-vl'
  *   - 'qwen3-vl'
  *   - 'gemini'
@@ -269,56 +256,56 @@ export interface IModelConfigForPlanning {
   // model name
   [MIDSCENE_PLANNING_MODEL_NAME]: string;
   // proxy
-  [MIDSCENE_PLANNING_MODEL_SOCKS_PROXY]?: string;
-  [MIDSCENE_PLANNING_MODEL_HTTP_PROXY]?: string;
+  [MIDSCENE_PLANNING_OPENAI_SOCKS_PROXY]?: string;
+  [MIDSCENE_PLANNING_OPENAI_HTTP_PROXY]?: string;
   // OpenAI
-  [MIDSCENE_PLANNING_MODEL_BASE_URL]?: string;
-  [MIDSCENE_PLANNING_MODEL_API_KEY]?: string;
-  [MIDSCENE_PLANNING_MODEL_INIT_CONFIG_JSON]?: string;
+  [MIDSCENE_PLANNING_OPENAI_BASE_URL]?: string;
+  [MIDSCENE_PLANNING_OPENAI_API_KEY]?: string;
+  [MIDSCENE_PLANNING_OPENAI_INIT_CONFIG_JSON]?: string;
   // extra
-  [MIDSCENE_PLANNING_LOCATOR_MODE]?: TVlModeValues;
+  [MIDSCENE_PLANNING_VL_MODE]?: TVlModeValues;
 }
 
 export interface IModeConfigForGrounding {
   // model name
   [MIDSCENE_GROUNDING_MODEL_NAME]: string;
   // proxy
-  [MIDSCENE_GROUNDING_MODEL_SOCKS_PROXY]?: string;
-  [MIDSCENE_GROUNDING_MODEL_HTTP_PROXY]?: string;
+  [MIDSCENE_GROUNDING_OPENAI_SOCKS_PROXY]?: string;
+  [MIDSCENE_GROUNDING_OPENAI_HTTP_PROXY]?: string;
   // OpenAI
-  [MIDSCENE_GROUNDING_MODEL_BASE_URL]?: string;
-  [MIDSCENE_GROUNDING_MODEL_API_KEY]?: string;
-  [MIDSCENE_GROUNDING_MODEL_INIT_CONFIG_JSON]?: string;
+  [MIDSCENE_GROUNDING_OPENAI_BASE_URL]?: string;
+  [MIDSCENE_GROUNDING_OPENAI_API_KEY]?: string;
+  [MIDSCENE_GROUNDING_OPENAI_INIT_CONFIG_JSON]?: string;
   // extra
-  [MIDSCENE_GROUNDING_LOCATOR_MODE]?: TVlModeValues;
+  [MIDSCENE_GROUNDING_VL_MODE]?: TVlModeValues;
 }
 
 export interface IModelConfigForDefault {
   // model name
   [MIDSCENE_MODEL_NAME]: string;
   // proxy
-  [MIDSCENE_MODEL_SOCKS_PROXY]?: string;
-  [MIDSCENE_MODEL_HTTP_PROXY]?: string;
+  [MIDSCENE_OPENAI_SOCKS_PROXY]?: string;
+  [MIDSCENE_OPENAI_HTTP_PROXY]?: string;
   // OpenAI
-  [MIDSCENE_MODEL_BASE_URL]?: string;
-  [MIDSCENE_MODEL_API_KEY]?: string;
-  [MIDSCENE_MODEL_INIT_CONFIG_JSON]?: string;
+  [MIDSCENE_OPENAI_BASE_URL]?: string;
+  [MIDSCENE_OPENAI_API_KEY]?: string;
+  [MIDSCENE_OPENAI_INIT_CONFIG_JSON]?: string;
   // extra
-  [MIDSCENE_LOCATOR_MODE]?: TVlModeValues;
+  [MIDSCENE_VL_MODE]?: TVlModeValues;
 }
 
 export interface IModelConfigForDefaultLegacy {
   // model name
   [MIDSCENE_MODEL_NAME]: string;
   // proxy
-  [MIDSCENE_MODEL_SOCKS_PROXY]?: string;
-  [MIDSCENE_MODEL_HTTP_PROXY]?: string;
+  [MIDSCENE_OPENAI_SOCKS_PROXY]?: string;
+  [MIDSCENE_OPENAI_HTTP_PROXY]?: string;
   // OpenAI
   [OPENAI_BASE_URL]?: string;
   [OPENAI_API_KEY]?: string;
-  [MIDSCENE_MODEL_INIT_CONFIG_JSON]?: string;
+  [MIDSCENE_OPENAI_INIT_CONFIG_JSON]?: string;
   // extra
-  [MIDSCENE_LOCATOR_MODE]?: TVlModeValues;
+  [MIDSCENE_VL_MODE]?: TVlModeValues;
 }
 
 /**
@@ -329,11 +316,7 @@ export interface IModelConfigForDefaultLegacy {
  */
 export type TIntent = 'VQA' | 'planning' | 'grounding' | 'default';
 
-/**
- * Internal type with intent parameter for ModelConfigManager
- * @internal
- */
-export type TModelConfigFnInternal = (options: {
+export type TModelConfigFn = (options: {
   intent: TIntent;
 }) =>
   | IModelConfigForVQA
@@ -341,16 +324,6 @@ export type TModelConfigFnInternal = (options: {
   | IModeConfigForGrounding
   | IModelConfigForDefault;
 
-/**
- * User-facing model config function type
- * Users return config objects without needing to know about intent parameter
- */
-export type TModelConfigFn = () =>
-  | IModelConfigForVQA
-  | IModelConfigForPlanning
-  | IModeConfigForGrounding
-  | IModelConfigForDefault;
-
 export enum UITarsModelVersion {
   V1_0 = '1.0',
   V1_5 = '1.5',

package/src/extractor/dom-util.ts

@@ -1,4 +1,4 @@
-import { NodeType } from '../constants';
+import type { LocateResultElement, Rect } from '../types';
 import { generateHashId } from '../utils';
 
 export function isFormElement(node: globalThis.Node) {
@@ -132,20 +132,24 @@ function includeBaseElement(node: globalThis.Node) {
   return false;
 }
 
-export function generateElementByPosition(position: { x: number; y: number }) {
+export function generateElementByPosition(position: {
+  x: number;
+  y: number;
+}): LocateResultElement {
+  const edgeSize = 8;
   const rect = {
-    left: Math.max(position.x - 4, 0),
-    top: Math.max(position.y - 4, 0),
-    width: 8,
-    height: 8,
+    left: Math.round(Math.max(position.x - edgeSize / 2, 0)),
+    top: Math.round(Math.max(position.y - edgeSize / 2, 0)),
+    width: edgeSize,
+    height: edgeSize,
   };
   const id = generateHashId(rect);
   const element = {
     id,
-    attributes: { nodeType: NodeType.POSITION },
     rect,
     content: '',
-    center: [position.x, position.y],
+    center: [position.x, position.y] as [number, number],
+    isOrderSensitive: false, // actually it's 'unknown'
   };
 
   return element;

package/src/extractor/index.ts

@@ -35,8 +35,6 @@ export { extractTreeNode as webExtractNodeTree } from './web-extractor';
 
 export { extractTreeNodeAsString as webExtractNodeTreeAsString } from './web-extractor';
 
-export { setNodeHashCacheListOnWindow, getNodeFromCacheList } from './util';
-
 export {
   getXpathsByPoint,
   getNodeInfoByXpath,

package/src/extractor/locator.ts

@@ -1,7 +1,6 @@
 import type { ElementInfo } from '.';
 import type { Point } from '../types';
 import { isSvgElement } from './dom-util';
-import { getNodeFromCacheList } from './util';
 import { getRect, isElementPartiallyInViewport } from './util';
 import { collectElementInfo } from './web-extractor';
 

package/src/extractor/tree.ts

@@ -45,7 +45,7 @@ export function trimAttributes(
       res[currentKey] = truncateText(attributeVal, truncateTextLength);
       return res;
     },
-    {} as BaseElement['attributes'],
+    {} as Record<string, string>,
   );
   return tailorAttributes;
 }
@@ -106,8 +106,8 @@ export function descriptionOfTree<
           .replace(/\sNode$/, '')
           .toLowerCase();
       }
-      const markerId = node.node.indexId;
-      const markerIdString = markerId ? `markerId="${markerId}"` : '';
+      // const markerId = node.node.indexId;
+      // const markerIdString = markerId ? `markerId="${markerId}"` : '';
       const rectAttribute = node.node.rect
         ? {
             left: node.node.rect.left,
@@ -116,7 +116,7 @@ export function descriptionOfTree<
             height: node.node.rect.height,
           }
         : {};
-      before = `<${nodeTypeString} id="${node.node.id}" ${markerIdString} ${attributesString(trimAttributes(node.node.attributes || {}, truncateTextLength))} ${attributesString(rectAttribute)}>`;
+      before = `<${nodeTypeString} id="${node.node.id}" ${attributesString(trimAttributes(node.node.attributes || {}, truncateTextLength))} ${attributesString(rectAttribute)}>`;
       const content = truncateText(node.node.content, truncateTextLength);
       contentWithIndent = content ? `\n${indentStr}  ${content}` : '';
       after = `</${nodeTypeString}>`;

package/src/extractor/util.ts

@@ -399,42 +399,10 @@ export function midsceneGenerateHash(
 ): string {
   const slicedHash = generateHashId(rect, content);
 
-  if (node) {
-    if (!(window as any).midsceneNodeHashCacheList) {
-      setNodeHashCacheListOnWindow();
-    }
-
-    setNodeToCacheList(node, slicedHash);
-  }
-
   // Returns the first 10 characters as a short hash
   return slicedHash;
 }
 
-export function setNodeHashCacheListOnWindow() {
-  if (typeof window !== 'undefined') {
-    (window as any).midsceneNodeHashCacheList = [];
-  }
-}
-
-export function setNodeToCacheList(node: globalThis.Node, id: string) {
-  if (typeof window !== 'undefined') {
-    if (getNodeFromCacheList(id)) {
-      return;
-    }
-    (window as any).midsceneNodeHashCacheList?.push({ node, id });
-  }
-}
-
-export function getNodeFromCacheList(id: string) {
-  if (typeof window !== 'undefined') {
-    return (window as any).midsceneNodeHashCacheList?.find(
-      (item: { node: Node; id: string }) => item.id === id,
-    )?.node;
-  }
-  return null;
-}
-
 export function generateId(numberId: number) {
   //   const letters = 'ABCDEFGHIJKLMNPRSTUVXYZ';
   //   const numbers = '0123456789';

package/src/node/fs.ts

@@ -75,7 +75,7 @@ export async function getExtraReturnLogic(tree = false) {
     return null;
   }
 
-  const elementInfosScriptContent = `${getElementInfosScriptContent()}midscene_element_inspector.setNodeHashCacheListOnWindow();`;
+  const elementInfosScriptContent = `${getElementInfosScriptContent()};`;
 
   if (tree) {
     return `${elementInfosScriptContent}midscene_element_inspector.webExtractNodeTree()`;

package/src/types/index.ts

@@ -17,7 +17,7 @@ export type Rect = Point & Size & { zoom?: number };
 export abstract class BaseElement {
   abstract id: string;
 
-  abstract indexId?: number; // markerId for web
+  // abstract indexId?: number; // markerId for web
 
   abstract attributes: {
     nodeType: NodeType;
@@ -30,7 +30,7 @@ export abstract class BaseElement {
 
   abstract center: [number, number];
 
-  abstract xpaths?: string[];
+  // abstract xpaths?: string[];
 
   abstract isVisible: boolean;
 }
@@ -45,3 +45,10 @@ export interface ElementTreeNode<
 export interface WebElementInfo extends ElementInfo {
   zoom: number;
 }
+
+export type LocateResultElement = {
+  center: [number, number];
+  rect: Rect;
+  id: string;
+  isOrderSensitive?: boolean;
+};