Fix one bug in Chrome getter; fix one erro for corner case in doc

desktop_env/evaluators/getters/chrome.py

@@ -558,39 +558,54 @@ def get_active_url_from_accessTree(env, config):
         Return
             url: str
     """
-    accessibility_tree: str = env.controller.get_accessibility_tree()
+    # Ensure the controller and its method are accessible and return a valid result
-    # download accessibility tree to "/home/user/Desktop"
+    if hasattr(env, 'controller') and callable(getattr(env.controller, 'get_accessibility_tree', None)):
-    logger.debug("AT@eval: %s", accessibility_tree)
+        accessibility_tree = env.controller.get_accessibility_tree()
-    # first, use accessibility API to get the active tab URL
+        if accessibility_tree is None:
-    at: _Element = lxml.etree.fromstring(accessibility_tree)
+            print("Failed to get the accessibility tree.")
-    arch = platform.machine()
+            return None
-    print("Your architecture is: {}".format(arch))
+    else:
-    try:
+        print("Controller or method 'get_accessibility_tree' not found.")
-        if "arm" in arch:
-            selector = CSSSelector("application[name=Chromium] entry[name=Address\\ and\\ search\\ bar]",
-                                namespaces=_accessibility_ns_map)
-        else:
-            selector = CSSSelector("application[name=Google\\ Chrome] entry[name=Address\\ and\\ search\\ bar]",
-                                namespaces=_accessibility_ns_map)
-    except:
-        logger.error("Failed to parse the selector for active tab URL")
-        return None
-    elements: List[_Element] = selector(at)
-    # if "xpath" in config:
-    #     elements: List[_Element] = at.xpath(config["xpath"], namespaces=_accessibility_ns_map)
-    # elif "selectors" in config:
-    #     selector = CSSSelector(", ".join(config["selectors"]), namespaces=_accessibility_ns_map)
-    #     elements: List[_Element] = selector(at)
-    if len(elements) == 0:
-        print("no elements found")
-        return None
-    elif elements[-1].text is None:
-        print("no text found")
         return None
 
-    active_tab_url = config["goto_prefix"] + elements[0].text if "goto_prefix" in config.keys() else "https://" + \
+    logger.debug("AT@eval: %s", accessibility_tree)
-                                                                                                     elements[0].text
+
-    print("active tab url now: {}".format(active_tab_url))
+    at = None
+    try:
+        at = lxml.etree.fromstring(accessibility_tree)
+    except ValueError as e:
+        logger.error(f"Error parsing accessibility tree: {e}")
+        return None
+
+    # Determine the correct selector based on system architecture
+    selector = None
+    arch = platform.machine()
+    print(f"Your architecture is: {arch}")
+
+    if "arm" in arch:
+        selector_string = "application[name=Chromium] entry[name=Address\\ and\\ search\\ bar]"
+    else:
+        selector_string = "application[name=Google\\ Chrome] entry[name=Address\\ and\\ search\\ bar]"
+
+    try:
+        selector = CSSSelector(selector_string, namespaces=_accessibility_ns_map)
+    except Exception as e:
+        logger.error(f"Failed to parse the selector for active tab URL: {e}")
+        return None
+
+    elements = selector(at) if selector else []
+    if not elements:
+        print("No elements found.")
+        return None
+    elif not elements[-1].text:
+        print("No text found in the latest element.")
+        return None
+
+    # Use a default prefix if 'goto_prefix' is not specified in the config
+    goto_prefix = config.get("goto_prefix", "https://")
+
+    active_tab_url = f"{goto_prefix}{elements[0].text}"
+    print(f"Active tab url now: {active_tab_url}")
     return active_tab_url
 
 

desktop_env/evaluators/metrics/docs.py

@@ -6,6 +6,7 @@ import zipfile
 from io import BytesIO
 from typing import List, Dict, Any
 
+import easyocr
 from PIL import Image
 from docx import Document
 from docx.enum.text import WD_PARAGRAPH_ALIGNMENT, WD_TAB_ALIGNMENT
@@ -247,8 +248,9 @@ def compare_docx_images(docx_file1, docx_file2):
     return 1
 
 
-import easyocr
 def compare_image_text(image_path, rule):
+    if not image_path:
+        return 0
     reader = easyocr.Reader(['en'])
     result = reader.readtext(image_path)
     extracted_text = ' '.join([entry[1] for entry in result])