es 创建索引 指定id_ES实践

最近使用ES做召回的一种工具，将实践分享给大家

1、ES安装和启动：

请参考博文：

MacOS中Elasticsearch的安装「借助Homebrew」 - niceyoo - 博客园​www.cnblogs.com

2、创建索引：

from elasticsearch import Elasticsearch

es = Elasticsearch(['127.0.0.1:9200'])
mappings = {
    "mappings": {
        "properties": {
            "sp_id": {
                "type": "long",
                "index": "false"
            },
            "product_name": {
                "type": "text",
                "index": True
            },
            "third_category_id": {
                "type": "long",
                "index": True
            }
        }
    }
}

res = es.indices.create(index='cate_index', body=mappings)

# 判断索引有没有写入成功
if es.indices.exists("cate_index"):
    print('索引存在')
else:
    print('索引不存在')

3、数据写入ES：

from elasticsearch import Elasticsearch
from elasticsearch.helpers import bulk
es = Elasticsearch(['127.0.0.1:9200'])
ACTIONS = []
file = open("数据.txt")

for line in file.readlines():
    line = line.strip("n")
    line_arr = line.split("t")
    print(line_arr)
    if line_arr[2] == 'NULL':
        continue
    action = {
          "_index": "cate_index",
          "_source": {
            "sp_id": line_arr[0],
            "product_name": line_arr[1],
            "third_category_id": line_arr[2]
          }
        }
    ACTIONS.append(action)

res, _ = bulk(es, ACTIONS, index="cate_index", raise_on_error=True)

注：使用bulk写入可以快速提高写入的速度，比使用index写入要快一个数量级

4、ES查询：

from elasticsearch import Elasticsearch

es = Elasticsearch(['127.0.0.1:9200'])

body = {
    "query": {
        "match": {
            "product_name": product_name
        }
    }
}

sort = {
    "_score": {
        "order": "desc"
    }
}
res = es.search(index="cate_index", body=body, size=50, sort=sort)
print(res)