Function escape_json_string

crawl4ai/utils.py:152–179

Escapes characters in a string to be JSON safe. Parameters: s (str): The input string to be escaped. Returns: str: The escaped string, safe for JSON encoding.

(s)

        return text.encode('ascii', errors='ignore').decode('ascii')

def escape_json_string(s):
    """
    Escapes characters in a string to be JSON safe.

    Parameters:
    s (str): The input string to be escaped.

    Returns:
    str: The escaped string, safe for JSON encoding.
    """
    # Replace problematic backslash first
    s = s.replace('\\', '\\\\')
    
    # Replace the double quote
    s = s.replace('"', '\\"')
    
    # Escape control characters
    s = s.replace('\b', '\\b')
    s = s.replace('\f', '\\f')
    s = s.replace('\n', '\\n')
    s = s.replace('\r', '\\r')
    s = s.replace('\t', '\\t')
    
    # Additional problematic characters
    # Unicode control characters
    s = re.sub(r'[\x00-\x1f\x7f-\x9f]', lambda x: '\\u{:04x}'.format(ord(x.group())), s)
    
    return s

class CustomHTML2Text(HTML2Text):
    def __init__(self, *args, **kwargs):