Files
sci-gui-agent-benchmark/desktop_env/evaluators/metrics/docs.py
2023-12-30 16:32:15 +08:00

44 lines
1.7 KiB
Python

import xml.etree.ElementTree as ET
from docx import Document
from lxml import etree
def find_default_font(expected, config_file_path):
"""Find the default font in LibreOffice Writer."""
default_font = None
try:
tree = ET.parse(config_file_path)
root = tree.getroot()
# Define the XML namespace used in the file
namespace = {'oor': 'http://openoffice.org/2001/registry'}
# Search for the node containing the default font setting for LibreOffice Writer
for elem in root.findall('.//item[@oor:path="/org.openoffice.Office.Writer/DefaultFont"]', namespace):
for prop in elem.findall('.//prop[@oor:name="Standard"]', namespace):
for value in prop.findall('value', namespace):
default_font = value.text
except Exception as e:
print(f"Error: {e}")
return 1 if default_font == expected else 0
def contains_page_break(docx_file):
doc = Document(docx_file)
namespaces = {'w': 'http://schemas.openxmlformats.org/wordprocessingml/2006/main'}
for paragraph in doc.paragraphs:
for run in paragraph.runs:
br_elems = run.element.findall('.//w:br', namespaces)
for br in br_elems:
if br is not None and '{http://schemas.openxmlformats.org/wordprocessingml/2006/main}type' in br.attrib and br.attrib['{http://schemas.openxmlformats.org/wordprocessingml/2006/main}type'] == 'page':
return 1
return 0
# Replace 'your_document.docx' with the path to your document
# result = contains_page_break('your_document.docx')
# print(result)
#config_path = "/home/[username]/.config/libreoffice/4/user/registrymodifications.xcu"
#print(find_default_font("Ani", config_path))