Pyenchant spell checking with ElasticSearch

ElasticEnchant

class ElasticEnchant:

    def __init__(self, esDatabase):
        self.es_instance = esDatabase

    def produce_dictionary(self, output_file, **kwargs):
        """Produces a dictionary or updates it depending on kwargs

        If no kwargs are given then this method will write a full dictionary including all
        entries in all indices and types and output it in an enchant-friendly way to the output file.

        Accepted kwargs are index, and source_file. If you want to index multiple types
        or indices you should pass them in as comma delimited strings. Source file should be
        an absolute path to an existing enchant-friendly dictionary file.

        max_results will also set the maximum number of entries to be used in generating the dictionary.
        Set to 50k by default"""
        index = kwargs.get("index", "_all")
        max_results = kwargs.get("max_results", 50000)
        words = set()
        if kwargs.get("source_file", None):
            words = set(open(kwargs["source_file"]).readlines())
        url = "/".join([self.es_instance.url, index, "_search?size="+str(max_results)+"&q=*.*"])
        response = requests.get(url)
        for entry in json.loads(response._content)['hits']['hits']:
            if entry["_source"]["searchable_text"]:
                text = entry["_source"]["searchable_text"]
                words |= set(re.findall(r'[a-z]+', text.lower()))
            else:
                continue
        with open(output_file, 'wb') as dictionary:
            for word in words:
                dictionary.write(word+"\n")

Super