luulinh90s/Tabular-LLM-Study-Preference-Ranking
0
1import os2import shutil3from pathlib import Path4import re5 6method = 'POS'7 8def process_html_file(file_path, output_path):9 with open(file_path, 'r', encoding='utf-8') as file:10 content = file.read()11 12 # Remove lines containing 'Prediction'13 content = re.sub(r'<h3>Prediction:.*?</h3>\n?', '', content, flags=re.DOTALL)14 15 with open(output_path, 'w', encoding='utf-8') as file:16 file.write(content)17 18 19def process_directory(input_dir, output_dir):20 for root, dirs, files in os.walk(input_dir):21 for file in files:22 if file.endswith('.html'):23 input_path = Path(root) / file24 relative_path = input_path.relative_to(input_dir)25 output_path = Path(output_dir) / relative_path26 27 output_path.parent.mkdir(parents=True, exist_ok=True)28 process_html_file(input_path, output_path)29 30 31# Define input and output directories32input_directory = f"htmls_{method}"33output_directory = f"htmls_{method}_mod"34 35# Process the files36process_directory(input_directory, output_directory)37 38print(f"Processing complete for {method}. Modified files are in the output directory.")