import re


class ElasticsearchHelpers(object):
    """
    Sanitize a query string for Elasticsearch / Lucene.
    
    Based on:
    http://stackoverflow.com/questions/16205341/symbols-in-query-string-for-elasticsearch
    https://gist.github.com/bcoe/6505434
    """

    @staticmethod
    def sanitize_string(text):
        # Escape special characters
        # http://lucene.apache.org/core/old_versioned_docs/versions/2_9_1/queryparsersyntax.html#Escaping Special Characters
        text = re.sub('([{}])'.format(re.escape('\\+\-&|!(){}\[\]^~*?:\/')), r"\\\1", text)

        # AND, OR and NOT are used by lucene as logical operators. We need
        # to escape them
        for word in ['AND', 'OR', 'NOT']:
            escaped_word = "".join(["\\" + letter for letter in word])
            text = re.sub(r'\s*\b({})\b\s*'.format(word), r" {} ".format(escaped_word), text)

        # Escape odd quotes
        quote_count = text.count('"')
        return re.sub(r'(.*)"(.*)', r'\1\"\2', text) if quote_count % 2 == 1 else text