check_highlighted_words modified

desktop_env/evaluators/metrics/docs.py

@@ -1,6 +1,8 @@
 import logging
 import os
 import xml.etree.ElementTree as ET
+import zipfile
+import re
 from typing import List, Dict, Any
 
 from docx import Document
@@ -247,14 +249,24 @@ def evaluate_colored_words_in_tables(file_path1, file_path2):
 def check_highlighted_words(file_path1, file_path2):
     if not compare_docx_files(file_path1, file_path2):
         return 0
-    document = Document(file_path1)
+        
+    # Extract content.xml from the .odt file
+    extract_dir = file_path1 + "_extracted"
+    with zipfile.ZipFile(file_path1, 'r') as zip_ref:
+        zip_ref.extractall(extract_dir)
+    content_xml_path = os.path.join(extract_dir, 'content.xml')
+    with open(content_xml_path, 'r') as file:
+        content_xml = file.read()
 
-    for paragraph in document.paragraphs:
-        for run in paragraph.runs:
-            if run.font.highlight_color is not None:
-                return 0  # Highlighted words found
+    # Check for yellow highlights in the content.xml
+    yellow_highlight_pattern = re.compile(r'(.{0,50}background-color="#ffff00"[^>]*>.{0,50})')
+    yellow_highlight_matches = yellow_highlight_pattern.findall(content_xml)
 
-    return 1  # No highlighted words found
+    # Return True if yellow highlights are NOT found, otherwise True
+    if yellow_highlight_matches:
+        return 0
+    else:
+        return 1
 
 
 def evaluate_strike_through_last_paragraph(file_path1, file_path2):
@@ -415,4 +427,4 @@ def compare_highlighted_text(file1, file2):
     doc2_highlighted = extract_highlighted_text(Document(file2))
 
     # Compare the sets of highlighted text to check if they are the same
-    return set(doc1_highlighted) == set(doc2_highlighted)
+    return set(doc1_highlighted) == set(doc2_highlighted)