Add user specified id field from available fields

[rholder]

Allow the user to specify an id field for the document to uniquely identify it when sending it to Elasticsearch. By default Elasticsearch will generate its own unique document identifiers.

[TimCoder]

Just add 3 lines in the function 'documents_from_file'.
'id' is the specify 'id' in your xxx.csv file.

if 'id' in row.keys():   
    yield es.index_op(row, doc_type, True, index=index_name, id=row['id'])
else: 

The complete function worked is as follows:

def documents_from_file(es, filename, delimiter, quiet, index_name, doc_type):
    """
    Return a generator for pulling rows from a given delimited file.

    :param es: an ElasticSearch client
    :param filename: the name of the file to read from or '-' if stdin
    :param delimiter: the delimiter to use
    :param quiet: don't output anything to the console when this is True
    :return: generator returning document-indexing operations
    """
    def all_docs():
        with open(filename, 'rb') if filename != '-' else sys.stdin as doc_file:
            # delimited file should include the field names as the first row
            fieldnames = doc_file.next().strip().split(delimiter)
            echo('Using the following ' + str(len(fieldnames)) + ' fields:', quiet)
            for fieldname in fieldnames:
                echo(fieldname, quiet)

            reader = csv.DictReader(doc_file, delimiter=delimiter, fieldnames=fieldnames)
            count = 0
            for row in reader:
                count += 1
                if count % 10000 == 0:
                    echo('Sent documents: ' + str(count), quiet)
                if 'id' in row.keys():   
                    yield es.index_op(row, doc_type, True, index=index_name, id=row['id'])
                else: 
                    yield es.index_op(row)

    return all_docs