left_anti_join_rewriter.py

import os
import re

# Get the current directory
dir_path = os.getcwd()

pattern = re.compile(r'LEFT ANTI JOIN (\w+) t ON (t.\w+) = (\w+.\w+)')


def replace_anti_join(match):
    table_name = match.group(1)
    join_col = match.group(2)
    target_col = match.group(3)
    return f'LEFT JOIN {table_name} t ON {join_col} = {target_col} WHERE {join_col} IS NULL'


# Traverse all files and subdirectories in the current directory
for root, dirs, files in os.walk(dir_path):
    for file_name in files:
        # Rewrite TIMESTAMP()/timestamp() in the file content
        with open(os.path.join(root, file_name), 'r') as f:
            content = f.read()

        updated_query = pattern.sub(replace_anti_join, content)

        with open(os.path.join(root, file_name), 'w') as f:
            f.write(updated_query)