ZVJS SQLITE to ELK 3

#!/usr/bin/env python
# -*- coding: utf-8 -*-

import sqlite3
import json
import unicodedata
from elasticsearch import Elasticsearch

conn = sqlite3.connect('SK_Prison_and_Court_Guard.sqlite')
conn.row_factory = sqlite3.Row
curs = conn.cursor()
conn.commit()

es = Elasticsearch(host='elk.myserver.sk', port=9200, http_auth=('auth_user', 'auth_pass'),)

response = es.search(
    index="testpy",
    body={
        "query": {
            "bool": {
                "must": [{"match": {"_type": "testpy"}}]
            }
        },
        "aggs": {
            "max_price": {"max": {"field": "invoice_id"}}},
        "sort": {"invoice_id": {"order": "desc"}},
        "size": 1
    }
)

for hit in response['hits']['hits']:
    last_elastic_id = (hit['_source']['invoice_id'])
    print "Last Elasticsearch ID is: ",(last_elastic_id)

# Temp WorkAround if isn't nothing in index
# If is empty use counter with number 1
last_sqlite_id = 999999
counter = last_elastic_id + 1
try:
    while counter < last_sqlite_id :

        conn = sqlite3.connect('SK_Prison_and_Court_Guard.sqlite')
        conn.row_factory = sqlite3.Row
        d = str(counter)
        curs.execute("SELECT * FROM data WHERE invoice_id=" + d + "")
        recs = curs.fetchall()
        rows = [dict(rec) for rec in recs]
        rows_json = json.dumps(rows)
        chunk = unicodedata.normalize('NFKD', unicode(rows_json, 'utf-8', 'ignore')).encode('ASCII', 'ignore').replace('[','').replace(']', '').replace('"null"', 'null').replace('999999.99', '0.0')
        print "Next try ID: ",(counter)

        es = Elasticsearch(host='elk.myserver.sk', port=9200, http_auth=('auth_user', 'auth_pass'),)
        res = es.index(index="testpy", doc_type='testpy', body=chunk)
        print(res['created']), "OK\n"

        counter = counter + 1
except:
    print "Finish"