computer-use-ootb-internal 0.0.108__py3-none-any.whl → 0.0.109__py3-none-any.whl

computer_use_ootb_internal/computer_use_demo/gui_agent/gui_parser/simple_parser/gui_parser.py

@@ -1,676 +0,0 @@
-import cv2
-import numpy as np
-import datetime
-import requests
-import importlib
-# import textdistance
-import glob
-from PIL import Image
-import ctypes
-
-from .text_detection import text_detection
-from .icon_detection.icon_detection import detect_icons, IconDetector
-from .utils import *
-
-import os
-
-from importlib import resources
-import pathlib
-import computer_use_ootb_internal
-
-class GUIParser():
-    name = "gui_parser"
-    description = """
-This tool can extract the information of screenshot.
-Invoke command: gui_parser(query, visual[i])
-:param query -> str, specific command. visual[i] -> image, the latest screenshot.
-"""
-
-    def __init__(self, cache_folder=".cache/", user_id=None, trace_id=None, scaleFactor="1.0x"):
-        # Use pathlib for platform-independent path handling
-        cache_path = pathlib.Path(cache_folder)
-        cache_path.mkdir(exist_ok=True)
-        ocr_cache_path = cache_path / "ocr"
-        ocr_cache_path.mkdir(exist_ok=True)
-        
-        super(GUIParser, self).__init__()
-        self.cache_folder = str(cache_path)
-        print(f"Cache folder: {self.cache_folder}")
-        
-        self.task_id = get_current_time()
-        self.parsers = {}
-        self.temperature = 0
-        self.gui_parser = None
-        self.icon_template_threshold = 0.7
-        try:
-            ootb_path = os.getenv("OOTB_PATH")
-            if ootb_path:
-                self.ootb_database_path = pathlib.Path(ootb_path) / "ootbdatabase"
-            else:
-                # Fallback to package resources if env var not set
-                with resources.path(__package__, "ootbdatabase") as db_path:
-                    self.ootb_database_path = db_path
-        except Exception as e:
-            print(f"Error: {e}. OOTB_PATH is not set correctly.")
-        self.user_id = user_id
-        self.trace_id = trace_id
-        self.scaleFactor = scaleFactor
-        self.width = None 
-        self.height = None
-
-    def __call__(self, uia_data=None, screenshot_path=None, query=None, mode="uia", ocr_mode="googleocr"):
-        screenshot_path = pathlib.Path(screenshot_path)
-        if not screenshot_path.exists():
-            raise FileNotFoundError(f"Screenshot not found: {screenshot_path}")
-            
-        image = Image.open(screenshot_path)
-        self.width, self.height = image.size
-
-        if mode == "teach":
-            if query and isinstance(query, list) and len(query) > 1:
-                specific_icon_names = [item for item in query if self.is_image_path(item)]
-                if specific_icon_names:
-                    icon_template_folder = self.ootb_database_path / self.user_id / self.trace_id / "icons"
-                    icon_template_folder.mkdir(parents=True, exist_ok=True)
-
-                    print(f"icon_template_folder: {icon_template_folder}")
-                    icon_detector = IconDetector(str(icon_template_folder))
-
-                    detected_items = icon_detector(
-                        screenshot_path=screenshot_path,
-                        mode="teach",
-                        threshold=self.icon_template_threshold, # TODO: setting on init
-                        scale_factor=self.scaleFactor,
-                        specific_icon_names=specific_icon_names,
-                    )
-                    
-                    if detected_items:
-                        icon_items = detected_items
-            print("detected_icons:", icon_items)
-            # 获取OCR文本结果
-            ocr_elements = []
-            if ocr and "texts" in ocr:
-                for text_item in ocr["texts"]:
-                    ocr_elements.append({
-                        "name": text_item["content"],
-                        "rectangle": text_item["bbox"],
-                        "is_text": True
-                    })
-            
-            # 合并所有元素
-            all_elements = []
-            all_elements.extend(icon_items)  # 添加图标元素
-            all_elements.extend(ocr_elements)  # 添加文本元素
-            
-            # 创建主面板
-            ocr_icon_parsed_gui = {
-                "Screen": [{
-                    "name": "Main Content",
-                    "rectangle": [0, 0, self.width, self.height],
-                    "elements": self.merge_elements({"combined_elements": all_elements})
-                }]
-            }
-            
-            # get uia parsed gui
-            uia_parsed_gui = self.get_panel_uia(uia_data)
-            self.postprocess_uia(uia_parsed_gui)
-            
-            # merge the code
-            self.parsed_gui = self.merge_uia_and_teach_results(uia_parsed_gui, ocr_icon_parsed_gui)
-            return self.parsed_gui
-        
-        # only use uia mode
-        elif mode == "uia":
-            # Parse the UI elements of the current window
-            self.parsed_gui = {}
-            self.exclude_class_name_list = []
-            # if self.if_browser(window_name):
-            #     self.exclude_class_name_list = [
-            #         "Custom",
-            #         "Menu",
-            #         "Pane",
-            #         "TabControl",
-            #         "DataItem",
-            #     ]
-            # elif self.if_office(window_name):
-            #     self.exclude_class_name_list = [
-            #         "Custom",
-            #         "Menu",
-            #         "Pane",
-            #         "Toolbar",
-            #         "TabControl",
-            #         "TreeItem",
-            #         "DataItem",
-            #         "Hyperlink",
-            #     ]
-            # else:
-            #     self.exclude_class_name_list = []
-
-            self.parsed_gui = self.get_panel_uia(uia_data)
-            self.postprocess_uia(self.parsed_gui)
-
-            return self.parsed_gui
-        
-        # OCR + full icon detection mode
-        elif mode == "icon_ocr": 
-            ocr = text_detection(screenshot_path)
-            icon_template_folder = os.path.join(self.ootb_database_path, "ootbdatabase/teach_mode")
-            icon_detector = IconDetector(icon_template_folder)
-            icon_items = [
-                icon_detector(
-                    screenshot_path=screenshot_path,  mode="full_detect", threshold=0.6, scale_factor=self.scaleFactor
-                )
-            ]
-            self.parsed_gui = self.postprocess_icon(
-                "screen", icon_items, screenshot_path
-            )
-            for panel_item in self.parsed_gui["screen"]:
-                temp = {}
-                temp["editing_control"] = self.get_text(
-                    panel_item, ocr, screenshot_path
-                )
-
-                panel_item["elements"] += self.merge_elements(temp)
-                
-        # OCR only mode
-        else: 
-            ocr = text_detection(screenshot_path)
-            self.parsed_gui = self.postprocess_icon(
-                "screen", [], screenshot_path
-            )
-            for panel_item in self.parsed_gui["screen"]:
-                temp = {}
-                temp["editing_control"] = self.get_text(
-                    panel_item, ocr, screenshot_path
-                )
-
-                panel_item["elements"] += self.merge_elements(temp)
-        
-            
-            return self.parsed_gui
-
-
-
-    def if_browser(self, window_name):
-        # check if ["Edge", "Chrome", "Firefox", "Safari", "Opera"] in window_name
-        if any(browser in window_name for browser in ["Edge", "Chrome", "Firefox", "Safari", "Opera"]):
-            return True
-        return False
-    
-    def if_adobe(self, window_name):
-        # check if ["Adobe Premiere Pro", "Adobe After Effect", "Adobe Acrobat"] in window_name
-        if any(adobe in window_name for adobe in ["Adobe Premiere Pro", "Adobe After Effect", "Adobe Acrobat"]):
-            return True
-        return False
-
-    def if_office(self, window_name):
-        # check if ["Excel", "Word", "PowerPoint", "Outlook"] in window_name
-        if any(office in window_name for office in ["Excel", "Word", "PowerPoint", "Outlook"]):
-            return True
-        return False
-    
-    @staticmethod
-    def is_image_path(text):
-        # Checking if the input text ends with typical image file extensions
-        image_extensions = (".jpg", ".jpeg", ".png", ".gif", ".bmp", ".tiff", ".tif")
-        if text.endswith(image_extensions):
-            return True
-        else:
-            return False
-
-    @staticmethod
-    def get_text(panel_item, ocr, screenshot_path):
-        # Step1: Find all the texts in the panel
-        panel_rect = panel_item["rectangle"]
-        # 过滤出在panel区域内的文本
-        panel_texts = []
-        for item in ocr["texts"]:
-            bbox = item["bbox"]
-            # 检查文本是否在panel区域内
-            if (bbox[0] >= panel_rect[0] and 
-                bbox[1] >= panel_rect[1] and 
-                bbox[2] <= panel_rect[2] and 
-                bbox[3] <= panel_rect[3]):
-                panel_texts.append(item)
-
-        if not panel_texts:
-            return []
-            
-        # obtain information about panel texts
-        sorted_panel_texts = sorted(
-            panel_texts, key=lambda x: (x["bbox"][1], x["bbox"][0])
-        )
-        
-        # Step 2: Identify rows by grouping elements that are in approximately the same vertical position.
-        editing_controls = []
-        current_row = []
-        if sorted_panel_texts:
-            previous_y = sorted_panel_texts[0]["bbox"][1]
-
-            for item in sorted_panel_texts:
-                y1 = item["bbox"][1]
-                # If the vertical position of the current item is significantly different from the previous one, start a new row.
-                if abs(y1 - previous_y) > 15:
-                    if current_row:  # 只有当current_row非空时才添加
-                        editing_controls.append(current_row)
-                    current_row = []
-
-                current_row.append(
-                    {
-                        "name": item["content"],
-                        "rectangle": item["bbox"],
-                    }
-                )
-                previous_y = y1
-                
-            # Add the last row if not empty.
-            if current_row:
-                editing_controls.append(current_row)
-                
-        # Step 3: Sort elements within each row based on their horizontal position.
-        for i, row in enumerate(editing_controls):
-            editing_controls[i] = sorted(row, key=lambda x: x["rectangle"][0])
-            
-        return editing_controls
-        
-    def postprocess_icon(self, window_name, icon_items, screenshot_path):
-        image = Image.open(screenshot_path)
-        width, height = image.size
-        result = {
-            window_name: [
-                {
-                    "name": "Main Content",
-                    "rectangle": [0, 0, width, height],
-                    "elements": None
-                }
-            ]
-        }
-        return result
-
-    @staticmethod
-    def merge_elements(panel_item):
-        # 检查是否为空
-        if not panel_item or not any(panel_item.values()):
-            return []
-        
-        # 收集所有元素
-        all_elements = []
-        for key, value in panel_item.items():
-            if isinstance(value, list):
-                all_elements.extend(value)
-        
-        if not all_elements:
-            return []
-        
-        # 按Y轴分组,同一行的阈值为15像素
-        y_sorted = sorted(all_elements, key=lambda x: (x["rectangle"][1] + x["rectangle"][3]) / 2)
-        merged_control = []
-        current_row = []
-        previous_y = None
-        
-        for element in y_sorted:
-            y_center = (element["rectangle"][1] + element["rectangle"][3]) / 2
-            
-            if previous_y is not None and abs(y_center - previous_y) > 15:
-                if current_row:
-                    # 对当前行按X轴排序
-                    current_row.sort(key=lambda x: x["rectangle"][0])
-                    merged_control.append(current_row)
-                current_row = []
-                
-            current_row.append(element)
-            previous_y = y_center
-        
-        # 添加最后一行
-        if current_row:
-            current_row.sort(key=lambda x: x["rectangle"][0])
-            merged_control.append(current_row)
-            
-        # 对所有行按Y轴中心点排序
-        merged_control.sort(
-            key=lambda row: (row[0]["rectangle"][1] + row[0]["rectangle"][3]) / 2
-        )
-        
-        return merged_control
-
-    def postprocess_uia(self, metadata):
-        # iterate each window
-        for window_data in metadata.values():
-
-            for element_data in window_data:
-                # get all elements of current level
-                elements = element_data.get("elements", [])
-
-                # if the class name is TitleBar, regain the rectangle of the whole panel
-                if element_data["class_name"] == "TitleBar" and len(elements) > 0:
-                    left = min(e.get("rectangle", [0, 0, 0, 0])[0] for e in elements)
-                    top = min(e.get("rectangle", [0, 0, 0, 0])[1] for e in elements)
-                    right = max(e.get("rectangle", [0, 0, 0, 0])[2] for e in elements)
-                    bottom = max(e.get("rectangle", [0, 0, 0, 0])[3] for e in elements)
-                    element_data["rectangle"] = [left, top, right, bottom]
-                # sort elements by x and y
-                element_data["elements"] = sort_elements_by_xy(elements)
-
-        return metadata
-    
-    def merge_uia_and_teach_results(self,uia_result, teach_result):
-        def calculate_overlap(rect1, rect2):
-            # 计算两个矩形的重叠面积
-            x1 = max(rect1[0], rect2[0])
-            y1 = max(rect1[1], rect2[1])
-            x2 = min(rect1[2], rect2[2])
-            y2 = min(rect1[3], rect2[3])
-            
-            if x1 >= x2 or y1 >= y2:
-                return 0
-                
-            overlap_width = x2 - x1
-            overlap_height = y2 - y1
-            
-            return overlap_width * overlap_height
-
-        def is_significant_overlap(ocr_element, uia_element):
-            ocr_rect = ocr_element["rectangle"]
-            uia_rect = uia_element["rectangle"]
-            
-            overlap_area = calculate_overlap(ocr_rect, uia_rect)
-            ocr_area = (ocr_rect[2] - ocr_rect[0]) * (ocr_rect[3] - ocr_rect[1])
-            
-            return overlap_area > 0.5 * ocr_area if ocr_area > 0 else False
-
-        merged_result = uia_result.copy()
-
-        all_ocr_elements = []
-        # 遍历teach模式下的OCR文本元素
-        for screen in teach_result.get("Screen", []):
-            for elements_row in screen.get("elements", []):
-                for element in elements_row:
-                    if element["is_text"]:
-                        should_add = True
-                        # 检查是否与任何UIA元素有显著重叠
-                        for window_name, panels in merged_result.items():
-                            for panel in panels:
-                                for elements_list in panel.get("elements", []):
-                                    for uia_element in elements_list:
-                                        if is_significant_overlap(element, uia_element):
-                                            should_add = False
-                                            break
-                                    if not should_add:
-                                        break
-                                if not should_add:
-                                    break
-                            if not should_add:
-                                break
-                                
-                        # 如果没有显著重叠，将OCR元素添加到最合适的面板中
-                        if should_add:
-                            # 找到最合适的面板（这里简单地添加到第一个面板）
-                            all_ocr_elements.append(element)
-                    else:
-                        all_ocr_elements.append(element)
-                        
-        merged_result["OCR"] = [{
-                    "name": "Main Content",
-                    "rectangle": [0, 0, self.width, self.height],
-                    "elements": [all_ocr_elements]
-                }]
-        return merged_result
-    def get_panel_uia(self, control_info_list):
-        # 定义文本到面板名称的映射
-        text2panel_name = {
-            "新建会话属性": "New session properties",
-            "会话管理器": "Session manager",
-            "任务栏": "Taskbar",
-        }
-
-        # 递归处理控件信息
-        def recurse_controls(control_info, dialog_components, depth, software_name):
-            children = control_info.get("children", [])
-            if not children:
-                return
-            
-            for child_control in children:
-                child_properties = child_control.get("properties", {})
-                child_friendly_class_name = child_properties.get("friendly_class_name", "")
-                
-                # 如果控件类型在排除列表中，跳过处理
-                if child_friendly_class_name in self.exclude_class_name_list:
-                    continue
-                    
-                # 获取控件名称
-                child_texts = child_properties.get("texts", [])
-                if not child_texts:
-                    child_properties_name = ""
-                elif isinstance(child_texts[0], list):
-                    result = []
-                    for item in child_texts:
-                        if isinstance(item, list) and item and isinstance(item[0], str):
-                            result.append("".join(item))
-                        elif isinstance(item, str):
-                            result.append(item)
-                    child_properties_name = "".join(result)
-                else:
-                    child_properties_name = child_texts[0]
-                
-                # 特殊处理编辑框和组合框
-                if child_friendly_class_name in ["Edit", "ComboBox"] and not child_properties_name:
-                    child_properties_name = "Search Bar"
-                
-                # 判断是否需要处理该控件
-                rectangle = child_properties.get("rectangle", [0, 0, 0, 0])
-                automation_id = child_properties.get("automation_id", "")
-                
-                # 放宽控件过滤条件
-                browser_list = ["Edge", "Chrome", "Firefox", "Opera", "Safari"]
-                is_browser = False
-                for browser in browser_list:
-                    if browser in software_name:
-                        is_browser = True
-                        break
-                should_process = (
-                    # 任务栏特殊处理
-                    (software_name == "Taskbar" and child_friendly_class_name in ["Button", "Static", "Image"]) or
-                    # 浏览器特殊处理
-                    (is_browser and child_friendly_class_name in ["Pane", "Button", "Edit", "Tab", "TabItem", "Document", "Group"]) or
-                    # 常规控件处理
-                    (not child_control.get("children", []) or  # 没有子控件的元素保留
-                     automation_id or  # 有automation_id的控件保留
-                     child_friendly_class_name not in self.exclude_class_name_list)  # 不在排除列表中的控件保留
-                ) and (
-                    # 基本条件检查
-                    len(rectangle) == 4 and
-                    not all(element == 0 for element in rectangle) and
-                    (child_properties_name not in ["", '"'] or automation_id)  # 允许有automation_id的空名称控件
-                )
-                
-                if should_process:
-                    # 获取控件的矩形边界
-                    left, top, right, bottom = rectangle
-                    dialog_rect = dialog_components.get("rectangle", [0, 0, 0, 0])
-                    if len(dialog_rect) != 4:
-                        dialog_rect = [0, 0, 0, 0]
-                    left_bound, top_bound, right_bound, bottom_bound = dialog_rect
-
-                    # 根据软件名称进行特殊逻辑处理
-                    if software_name not in ["Taskbar"] and not is_browser:  # 任务栏和浏览器不裁剪坐标
-                        child_properties["rectangle"] = [
-                            max(left, left_bound),
-                            max(top, top_bound),
-                            min(right, right_bound),
-                            min(bottom, bottom_bound)
-                        ]
-                    else:
-                        child_properties["rectangle"] = [left, top, right, bottom]
-                    
-                    # 检查矩形有效性
-                    if (child_properties["rectangle"][0] < child_properties["rectangle"][2] 
-                        and child_properties["rectangle"][1] < child_properties["rectangle"][3]):
-                        
-                        # 添加控件信息到对话框组件
-                        element_info = {
-                            "name": child_properties_name.replace("\u200b", ""),
-                            "rectangle": child_properties["rectangle"],
-                            "class_name": child_friendly_class_name,
-                            "type": ["Click", "rightClick"],
-                            "depth": f"{depth}-{dialog_components.get('count', 1)}"
-                        }
-                        
-                        # 添加automation_id（如果存在）
-                        if automation_id:
-                            element_info["automation_id"] = automation_id
-                            
-                        dialog_components.setdefault("elements", []).append(element_info)
-                        dialog_components["count"] = dialog_components.get("count", 1) + 1
-
-                # 递归处理子控件
-                recurse_controls(child_control, dialog_components, depth, software_name)
-
-        dialog_components = {}
-
-        # 遍历每个软件的控件信息
-        for software_name, controls in control_info_list.items():
-            if not controls:  # 跳过空的控件列表
-                continue
-                
-            dialog_components[software_name] = []
-            
-            for control_info in controls:
-                # 判断是否为需要处理的对话框类型
-                control_properties = control_info.get("properties", {})
-                friendly_class_name = control_properties.get("friendly_class_name", "")
-                
-                # 放宽面板类型的限制
-                panel_types = ["Dialog", "Pane", "GroupBox", "TitleBar", "Menu", "Document", "ListBox", "AppBar", "Tab", "Group"]
-                if (friendly_class_name in panel_types and control_info.get("children")):
-                    
-                    control_texts = control_properties.get("texts", [""])
-                    control_name = control_texts[0] if control_texts else ""
-                    automation_id = control_properties.get("automation_id", "")
-                    
-                    # 设置控件名称
-                    if not control_name and automation_id:
-                        control_name = automation_id  # 使用automation_id作为备选名称
-                    elif not control_name:
-                        if friendly_class_name == "TitleBar":
-                            control_name = "Title Bar"
-                        elif friendly_class_name == "Document":
-                            control_name = "Main Content"
-                        elif friendly_class_name == "Pane":
-                            if software_name == "Taskbar":
-                                control_name = "Taskbar"
-                            elif "Edge" in software_name or "Chrome" in software_name or "Firefox" in software_name:
-                                control_name = "Browser Content"
-                            else:
-                                control_name = "Navigation Bar" if software_name in ["web", "web video"] else "Main Content"
-                    elif friendly_class_name == "Document" and software_name in ["web", "web video"]:
-                        control_name = "Main Content"
-                    
-                    # 转换文本名称
-                    control_name = text2panel_name.get(control_name, control_name)
-
-                    # 初始化对话框组件并递归处理控件
-                    panel_info = {
-                        "name": control_name,
-                        "rectangle": control_properties.get("rectangle", [0, 0, 0, 0]),
-                        "class_name": friendly_class_name,
-                        "depth": "1",
-                        "elements": [],
-                        "count": 1
-                    }
-                    
-                    # 添加automation_id（如果存在）
-                    if automation_id:
-                        panel_info["automation_id"] = automation_id
-                        
-                    dialog_components[software_name].append(panel_info)
-                    recurse_controls(control_info, dialog_components[software_name][-1], "1", software_name)
-                    
-        return dialog_components
-    
-import matplotlib.pyplot as plt
-import cv2
-import matplotlib.font_manager as fm
-
-def show_img_with_box_points(img, box, ax, color="red"):
-    x0, y0 = box[0], box[1]
-    w, h = box[2] - box[0], box[3] - box[1]
-    ax.add_patch(plt.Rectangle((x0, y0), w, h, edgecolor=color, facecolor=(0, 0, 0, 0), lw=0.4))
-
-def extract_second_underscore_content(filename):
-    parts = filename.split('_')
-    if len(parts) > 1 and "@" in filename:
-        return parts[1]
-    return filename
-
-def show_guiParserd(gui, img_path):
-    img = cv2.imread(img_path)
-    img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
-    plt.figure(dpi=500)
-    plt.axis('off')
-    plt.imshow(img)
-
-    # 设置微软雅黑字体
-    # font_path = "C:/Windows/Fonts/msyh.ttc"  # Windows系统上的微软雅黑字体路径
-    # prop = fm.FontProperties(fname=font_path)
-    item_cache = []
-    for windows_name, panel_list in gui.items():
-        print("windows_name: ", windows_name)
-        print("panel_list: ", panel_list)
-        for sequence in panel_list:
-            # print("sequence: ", sequence)
-            show_img_with_box_points(img, sequence["rectangle"], plt.gca(), color="green")
-            if len(sequence["name"]) < 100:
-                plt.text(sequence["rectangle"][0], sequence["rectangle"][1], sequence["name"], 
-                        fontsize=4, color='green')
-            for _, items in enumerate(sequence['elements']):
-                item_cache.extend(items)
-    for item in item_cache:
-        try:
-            if "rectangle" in item:
-                if len(item["name"]) > 200:
-                    continue
-                box = item["rectangle"]
-                if "current_step_icon" in item.keys() and item["current_step_icon"]:
-                    show_img_with_box_points(img, box, plt.gca(),color="green")
-                    plt.text(box[0], box[3]+25, "Current Step Icon", fontsize=3, color='green')
-                else:
-                    show_img_with_box_points(img, box, plt.gca(),color="blue")
-                plt.text(box[0], box[1], extract_second_underscore_content(item["name"]), 
-                        fontsize=2, color='red')
-        except Exception as e:
-            print(e)
-    # plt.savefig('gui_parser5.png', bbox_inches='tight', pad_inches=0)
-    plt.show()
-
-
-
-def get_screen_info():
-    try:
-        scaleFactor = ctypes.windll.shcore.GetScaleFactorForDevice(0) / 100
-        scaleFactor = str(scaleFactor) + "x"
-    except Exception as e:
-        print(e)
-        scaleFactor = "1.0x"
-    return scaleFactor
-        
-def parse_gui(user_id, trace_id, screenshot_path, user_scaleFactor="auto", uia_data=None, query=None, mode="uia", ocr_mode="googleocr"):
-    # TODO: check mult-screens
-    if user_scaleFactor == "auto":
-        user_scaleFactor = get_screen_info()
-    print("user_scaleFactor: ", user_scaleFactor)
-    gui_parser = GUIParser(user_id=user_id, trace_id=trace_id, scaleFactor=user_scaleFactor)
-    return gui_parser(uia_data=uia_data, screenshot_path=screenshot_path, query=query, mode=mode, ocr_mode=ocr_mode)
-
-import json
-import sys
-if __name__ == "__main__":
-
-    screenshot_path = r'D:\develop\computer_use_ootb_internal-main\.cache\20241222_041910\screenshot-0.png'
-    meta_data = json.load(open("test_uia_data.json", "r"))
-    gui_parser = GUIParser(user_id="test", trace_id="test", scaleFactor="1.0x")
-    uia_gui = gui_parser.get_panel_uia(meta_data)
-    uia_gui = gui_parser.postprocess_uia(uia_gui)
-    with open("test_uia_gui.json", "w") as f:
-        json.dump(uia_gui, f, indent=4)
-
-