load libreoffice writer eval -batch 2

desktop_env/controllers/setup.py

@@ -105,7 +105,6 @@ class SetupController:
             cache_path: str = os.path.join(self.cache_dir, "{:}_{:}".format(
                 uuid.uuid5(uuid.NAMESPACE_URL, url),
                 os.path.basename(path)))
-
             if not url or not path:
                 raise Exception(f"Setup Download - Invalid URL ({url}) or path ({path}).")
 

desktop_env/envs/desktop_env.py

@@ -47,8 +47,8 @@ class DesktopEnv(gym.Env):
             path_to_vm: str,
             action_space: str = "computer_13",
             task_config: Dict[str, Any] = None,
-            tmp_dir: str = "tmp",
-            cache_dir: str = "cache",
+            tmp_dir: str = "D:\\NJU\\HKUNLP\\Desktop-Env\\tmp",
+            cache_dir: str = "D:\\NJU\\HKUNLP\\Desktop-Env\\cache",
             screen_size: Tuple[int] = (1920, 1080)
     ):
         """

desktop_env/evaluators/getters/__init__.py

@@ -7,4 +7,4 @@ from .misc import get_rule, get_accessibility_tree
 from .replay import get_replay
 from .vlc import get_vlc_playing_info, get_vlc_config
 from .vscode import get_vscode_config
-from .impress import get_audio_in_slide
+# from .impress import get_audio_in_slide

desktop_env/evaluators/metrics/__init__.py

@@ -5,7 +5,7 @@ from .docs import find_default_font, contains_page_break, compare_docx_files, co
 from .docs import is_first_line_centered, check_file_exists, compare_contains_image
 from .docs import evaluate_colored_words_in_tables, check_highlighted_words, evaluate_strike_through_last_paragraph, \
     evaluate_conversion, evaluate_spacing, check_italic_font_size_14, evaluate_alignment, get_unique_train_ids, \
-    check_no_duplicates
+    check_no_duplicates, compare_init_lines
 from .general import exact_match, fuzzy_match
 from .general import check_csv, check_accessibility_tree, run_sqlite3, check_json
 from .gimp import increase_saturation, decrease_brightness, check_file_exists, compare_triangle_positions
@@ -16,7 +16,7 @@ from .pdf import check_pdf_pages
 #from .table import check_sheet_list, check_xlsx_freeze, check_xlsx_zoom, check_data_validations
 from .table import compare_table
 from .thunderbird import check_thunderbird_prefs, check_thunderbird_filter
-from .vlc import is_vlc_playing, is_vlc_recordings_folder, is_vlc_fullscreen, compare_images, compare_audios, \
-    compare_videos
+# from .vlc import is_vlc_playing, is_vlc_recordings_folder, is_vlc_fullscreen, compare_images, compare_audios, \
+    # compare_videos
 from .thunderbird import check_thunderbird_prefs, check_thunderbird_filter
 from .vscode import compare_text_file, compare_config, compare_answer, is_extension_installed, check_json_settings, check_json_keybindings

desktop_env/evaluators/metrics/docs.py

@@ -56,15 +56,34 @@ def compare_docx_files(file1, file2):
     doc2_paragraphs = [p.text for p in doc2.paragraphs]
 
     if len(doc1_paragraphs) != len(doc2_paragraphs):
+        # print(len(doc1_paragraphs))
+        # print(len(doc2_paragraphs))
         return 0
 
     # Compare each paragraph
     for p1, p2 in zip(doc1_paragraphs, doc2_paragraphs):
         if p1 != p2:
+            # print(p1)
+            # print(p2)
             return 0
 
     return 1
 
+def compare_init_lines(file1, file2):
+    doc1 = Document(file1)
+    doc2 = Document(file2)
+
+    doc1_paragraphs = [p.text for p in doc1.paragraphs]
+    doc2_paragraphs = [p.text for p in doc2.paragraphs]
+
+    # Compare each paragraph
+    for p1, p2 in zip(doc1_paragraphs, doc2_paragraphs):
+        if p1 != p2:
+            # print(p1)
+            # print(p2)
+            return 0
+
+    return 1
 
 def compare_docx_tables(docx_file1, docx_file2):
     doc1 = Document(docx_file1)
@@ -93,6 +112,8 @@ def compare_docx_tables(docx_file1, docx_file2):
 
 
 def compare_line_spacing(docx_file1, docx_file2):
+    if not compare_docx_files(docx_file1, docx_file2):
+        return 0
     doc1 = Document(docx_file1)
     doc2 = Document(docx_file2)
 
@@ -200,8 +221,10 @@ def compare_contains_image(docx_file1, docx_file2):
 # print(find_default_font("Ani", config_path))
 
 
-def evaluate_colored_words_in_tables(file_path):
-    document = Document(file_path)
+def evaluate_colored_words_in_tables(file_path1, file_path2):
+    if not compare_docx_files(file_path1, file_path2):
+        return 0
+    document = Document(file_path1)
 
     for table in document.tables:
         # Iterate through rows and cells in the table
@@ -221,8 +244,10 @@ def evaluate_colored_words_in_tables(file_path):
     return 1  # All words in tables are correctly colored
 
 
-def check_highlighted_words(file_path):
-    document = Document(file_path)
+def check_highlighted_words(file_path1, file_path2):
+    if not compare_docx_files(file_path1, file_path2):
+        return 0
+    document = Document(file_path1)
 
     for paragraph in document.paragraphs:
         for run in paragraph.runs:
@@ -232,8 +257,10 @@ def check_highlighted_words(file_path):
     return 1  # No highlighted words found
 
 
-def evaluate_strike_through_last_paragraph(file_path):
-    document = Document(file_path)
+def evaluate_strike_through_last_paragraph(file_path1, file_path2):
+    if not compare_docx_files(file_path1, file_path2):
+        return 0
+    document = Document(file_path1)
 
     # Get the last paragraph
     last_paragraph = document.paragraphs[-1]
@@ -278,8 +305,10 @@ def evaluate_spacing(file_path):
       return 0
 
 
-def check_italic_font_size_14(path):
-    document = Document(path)
+def check_italic_font_size_14(path1, path2):
+    if not compare_docx_files(path1, path2):
+        return 0
+    document = Document(path1)
     for paragraph in document.paragraphs:
         for run in paragraph.runs:
             if run.italic:
@@ -365,6 +394,8 @@ def compare_docx_lines(file1, file2):
 
     doc2 = Document(file2)
     doc2_lines = [p.text.strip() for p in doc2.paragraphs if p.text.strip()]
+    # print(doc1_lines)
+    # print(doc2_lines)
 
     # Convert the list of lines to sets and compare
     return set(doc1_lines) == set(doc2_lines)