ElasticSearch之通过search after和scroll解决深度分页问题

from elasticsearch import Elasticsearch
from elasticsearch import helpers
 
# 假设已经有了一个Elasticsearch客户端实例
es = Elasticsearch()
 
# 使用search_after进行分页
def search_after_pagination(scroll_id, size):
    # 获取上次查询的最后一条记录的sort值
    last_sort_value = scroll_id
    body = {
        "size": size,
        "query": {
            "function_score": {
                "query": {
                    "match_all": {}
                },
                "functions": [
                    {
                        "random_score": {
                            "seed": 123456,
                            "field": "_seq_no"
                        }
                    }
                ],
                "boost_mode": "replace"
            }
        },
        "sort": [
            {
                "_id": {
                    "order": "asc"
                }
            }
        ]
    }
    if last_sort_value:
        body['search_after'] = last_sort_value
    
    response = es.search(index='your_index', body=body)
    # 返回结果和新的sort值
    return response['hits']['hits'], response['hits']['sort'][0]
 
# 使用scroll进行深度分页
def scroll_pagination(size):
    # 初始查询
    scroll_id = 'initial'
    s = helpers.scan(
        client=es,
        query={
            "function_score": {
                "query": {
                    "match_all": {}
                },
                "functions": [
                    {
                        "random_score": {
                            "seed": 123456,
                            "field": "_seq_no"
                        }
                    }
                ],
                "boost_mode": "replace"
            }
        },
        index='your_index',
        size=size,
        scroll='5m',  # 设置scroll超时时间
    )
    
    for i, response in enumerate(s):
        # 处理返回的结果
        print(response)
        if i == 10:  # 假设我们只需要获取10个结果