python中elasticsearch_dsl模塊的使用方法
更新時(shí)間：2022年09月22日 09:41:40 作者：IT之一小佬
這篇文章主要介紹了python中elasticsearch_dsl模塊的使用方法，elasticsearch-dsl是基于elasticsearch-py封裝實(shí)現(xiàn)的，提供了更簡(jiǎn)便的操作elasticsearch的方法
前言
elasticsearch-dsl是基于elasticsearch-py封裝實(shí)現(xiàn)的，提供了更簡(jiǎn)便的操作elasticsearch的方法。
安裝：
install elasticsearch_dsl
連接elasticsearch
from elasticsearch_dsl import connections, Search es = connections.create_connection(hosts=["127.0.0.1:9200"], timeout=20) print(es)
還可以通過(guò)alias給連接設(shè)置別名，后續(xù)可以通過(guò)別名來(lái)引用該連接，默認(rèn)別名為default。
from elasticsearch_dsl import connections, Search # 方式一：連接es es = connections.create_connection(hosts=["127.0.0.1:9200"], timeout=20) print(es) # 方式二：連接es connections.create_connection(alias="my_new_connection", hosts=["127.0.0.1:9200"], timeout=20)
elasticsearch_dsl.Search
search對(duì)象代表整個(gè)搜索請(qǐng)求，包括：queries、filters、aggregations、sort、pagination、additional parameters、associated client。
API被設(shè)置為可鏈接的即和用.連續(xù)操作。search對(duì)象是不可變的，除了聚合，對(duì)對(duì)象的所有更改都將導(dǎo)致創(chuàng)建包含該更改的淺表副本。
當(dāng)初始化Search對(duì)象時(shí)，傳遞elasticsearch客戶端作為using的參數(shù)
示例代碼1：
from elasticsearch_dsl import connections, Search
# 方式一：連接es
es = connections.create_connection(hosts=["127.0.0.1:9200"], timeout=20)
# print(es)
# 方式二：連接es
connections.create_connection(alias="my_new_connection", hosts=["127.0.0.1:9200"], timeout=20)
# 不使用別名使用
res = Search(using=es).index("test_index").query()
# print(res)
for data in res:
print(data.to_dict())
print("*" * 100)
# 使用別名后這樣使用
res2 = Search(using="my_new_connection").index('test_index').query()
# print(e)
for data in res2:
print(data.to_dict())運(yùn)行結(jié)果：
示例代碼2：
from elasticsearch_dsl import connections, Search
# 方式一：連接es
es = connections.create_connection(hosts=["127.0.0.1:9200"], timeout=20)
# print(es)
# 不使用別名使用
res = Search(using=es).index("test_index").query()
# print(res)
for data in res:
print(data.to_dict())
print("*" * 100)
# 書(shū)寫(xiě)方式一：按條件查詢數(shù)據(jù)
res2 = Search(using=es).index("test_index").query("match", name="張三") # 查詢時(shí)注意分詞器的使用
for data in res2:
print(data.to_dict())
print("*" * 100)
# 書(shū)寫(xiě)方式二：按條件查詢數(shù)據(jù)
res3 = Search(using=es).index("test_index").query({"match": {"name": "張三"}})
for data in res3:
print(data.to_dict())運(yùn)行結(jié)果：
在上述執(zhí)行execute方法將請(qǐng)求發(fā)送給elasticsearch：
response = res.execute()不需要執(zhí)行execute()方法，迭代后可以通過(guò)to_dict()方法將Search對(duì)象序列化為一個(gè)dict對(duì)象，這樣可以方便調(diào)試。
query方法
查詢，參數(shù)可以是Q對(duì)象，也可以是query模塊中的一些類，還可以是自已寫(xiě)上如何查詢。
示例代碼1：
from elasticsearch_dsl import connections, Search, Q import time # 方式一：連接es es = connections.create_connection(hosts=["127.0.0.1:9200"], timeout=20) # print(es) res = Search(using=es, index="test_index").query().query() # 當(dāng)調(diào)用.query()方法多次時(shí)，內(nèi)部會(huì)使用&操作符 print(res.to_dict())
運(yùn)行結(jié)果：
filter方法
在過(guò)濾上下文中添加查詢，可以使用filter()函數(shù)來(lái)使之變的簡(jiǎn)單。
示例代碼1：
from elasticsearch_dsl import connections, Search, Q
# 方式一：連接es
es = connections.create_connection(hosts=["127.0.0.1:9200"], timeout=20)
# print(es)
# res = Search(using=es).index("test_index").filter({"match": {"name": "北"}})
# res = Search(using=es).index("test_index").filter("terms", tags=["name", "id"])
res = Search(using=es).index("test_index").query("bool", filter=[
Q("terms", tags=["name", "id"])]) # 上面代碼在背后會(huì)產(chǎn)生一個(gè)bool查詢，并將指定的條件查詢放入到filter分支
print(res)
for data in res:
print(data.to_dict())示例代碼2：
from elasticsearch_dsl import connections, Search, Q
import time
# 方式一：連接es
es = connections.create_connection(hosts=["127.0.0.1:9200"], timeout=20)
# print(es)
# 范圍查詢
# res = Search(using=es, index="test_index").filter("range", timestamp={"gte": 0, "lt": time.time()}).query({"match": {"name": "北"}})
res = Search(using=es, index="test_index").filter("range", id={"gte": 1, "lte": 4}).query({"match": {"name": "北"}})
print(res)
for data in res:
print(data.to_dict())
# 普通過(guò)濾
res2 = Search(using=es, index="test_index").filter("terms", id=["2", "4"]).execute()
print(res2)
for data in res2:
print(data.to_dict())運(yùn)行結(jié)果：
示例代碼3：
from elasticsearch_dsl import connections, Search, Q
# 方式一：連接es
es = connections.create_connection(hosts=["127.0.0.1:9200"], timeout=20)
# print(es)
# 方式一
q = Q('range', age={"gte": 25, "lte": 27})
res = Search(using=es, index="account_info").query(q)
print(res.to_dict())
for data in res:
print(data.to_dict())
print("*" * 100)
# 方式二
q2 = Q('range', **{"age": {"gte": 25, "lte": 27}})
res2 = Search(using=es, index="account_info").query(q2)
print(res2.to_dict())
for data in res2:
print(data.to_dict())運(yùn)行結(jié)果：
index方法
指定索引
usring方法
指定哪個(gè)elasticsearch
elasticsearch_dsl.query
該庫(kù)為所有的Elasticsearch查詢類型都提供了類。以關(guān)鍵字參數(shù)傳遞所有的參數(shù)，最終會(huì)把參數(shù)序列化后傳遞給Elasticsearch，這意味著在原始查詢和它對(duì)應(yīng)的dsl之間有這一個(gè)清理的一對(duì)一的映射。
示例代碼：
from elasticsearch_dsl import connections, Search, Q
from elasticsearch_dsl.query import MultiMatch, Match
# 方式一：連接es
es = connections.create_connection(hosts=["127.0.0.1:9200"], timeout=20)
# print(es)
# 相對(duì)與{"multi_match": {"query": "ha", "fields": ["firstname", "lastname"]}}
m1 = MultiMatch(query="Ha", fields=["firstname", "lastname"])
res = Search(using=es, index="test_index").query(m1)
print(res)
for data in res:
print(data.to_dict())
# 相當(dāng)于{"match": {"firstname": {"query": "Hughes"}}}
m2 = Match(firstname={"query": "Hughes"})
res = Search(using=es, index="test_index").query(m2)
print(res)
for data in res:
print(data.to_dict())elasticsearch_dsl.Q
使用快捷方式Q通過(guò)命名參數(shù)或者原始dict類型數(shù)據(jù)來(lái)構(gòu)建一個(gè)查詢實(shí)例。Q的格式一般是Q("查詢類型", 字段="xxx")或Q("查詢類型", query="xxx", fields=["字段1", "字段2"])
示例代碼1：
from elasticsearch_dsl import connections, Search, Q
from elasticsearch_dsl.query import MultiMatch, Match
# 方式一：連接es
es = connections.create_connection(hosts=["127.0.0.1:9200"], timeout=20)
# print(es)
# q = Q("match", city="Summerfield")
q = Q("multi_match", query="Summerfield", fields=["city", "firstname"])
res = Search(using=es, index="test_index").query(q)
print(res)
for data in res:
print(data.to_dict())查詢對(duì)象可以通過(guò)邏輯運(yùn)算符組合起來(lái)：
Q("match", title="python") | Q("match", title="django")
# {"bool": {"should": [...]}}
Q("match", title="python") & Q("match", title="django")
# {"bool": {"must": [...]}}
~Q("match", title="python")
# {"bool": {"must_not": [...]}}示例代碼2：
from elasticsearch_dsl import connections, Search, Q
# 方式一：連接es
es = connections.create_connection(hosts=["127.0.0.1:9200"], timeout=20)
# print(es)
# q = Q("multi_match", query="123.244.101.255", fields=["clientip", "timestamp"])
q = Q('match', name='張') | Q("match", name="北")
res = Search(using=es, index="test_index").query(q)
# print(res)
for data in res:
print(data.to_dict(), data.name)
print("*" * 100)
q = Q('match', name='張') & Q("match", name="北")
res = Search(using=es, index="test_index").query(q)
# print(res)
for data in res:
print(data.to_dict(), data.name)
print("*" * 100)
q = ~Q('match', name='張')
res = Search(using=es, index="test_index").query(q)
# print(res)
for data in res:
print(data.to_dict(), data.name)運(yùn)行結(jié)果：
示例代碼3：
from elasticsearch_dsl import connections, Search, Q
# 連接es
es = connections.create_connection(hosts=["127.0.0.1:9200"], timeout=20)
# print(es)
s = Search(using=es, index="account_info")
# constant_score內(nèi)置屬性
q = Q({"constant_score": {"filter": {"term": {"age": 25}}}})
res = s.query(q).execute()
for hit in res:
print(hit.to_dict())
print("*" * 100)
q2 = Q("bool", must=[Q("match", address="山")], should=[Q("match", gender="男"), Q("match", emplyer="AAA")], minimum_should_match=1)
res2 = s.query(q2).execute()
for hit in res2:
print(hit.to_dict())運(yùn)行結(jié)果：
嵌套類型
有時(shí)候需要引用一個(gè)在其他字段中的字段，例如多字段（title.keyword）或者在一個(gè)json文檔中的address.city。為了方便，Q允許你使用雙下劃線‘__’代替關(guān)鍵詞參數(shù)中的‘.’
示例代碼：
from elasticsearch_dsl import connections, Search, Q
# 方式一：連接es
es = connections.create_connection(hosts=["127.0.0.1:9200"], timeout=20)
# print(es)
# res = Search(using=es, index="test_index").query("match", address__city="北京")
res = Search(using=es, index="test_index").filter("term", address__city="北京")
# print(res)
for data in res:
print(data.to_dict(), data.name)查詢
示例代碼：
from elasticsearch_dsl import Search
from elasticsearch import Elasticsearch
# 連接es
es = Elasticsearch(hosts=["127.0.0.1:9200"], sniffer_timeout=60, timeout=30)
# 獲取es中所有的索引
# 返回類型為字典，只返回索引名
index_name = es.cat.indices(format="json", h="index")
print(index_name)
# 查詢多個(gè)索引
es_multi_index = Search(using=es, index=["personal_info_5000000", "grade", "test_index"])
print(es_multi_index.execute())
# 查詢一個(gè)索引
es_one_index = Search(using=es, index="test_index")
print(es_one_index.execute())
print("*" * 100)
# 條件查詢1
es_search1 = es_one_index.filter("range", id={"gte": 1, "lt": 5})
print(es_search1.execute())
# 條件查詢2
es_search2 = es_one_index.filter("term", name="張")
print(es_search2.execute())
print("*" * 100)
# 結(jié)果轉(zhuǎn)換為字典
es_search3 = es_search2.to_dict()
print(es_search3)
es_search4 = es_search2.execute().to_dict()
print(es_search4)運(yùn)行結(jié)果：
排序
示例代碼：
from elasticsearch_dsl import connections, Search, A
# 連接es
es = connections.create_connection(hosts=["127.0.0.1:9200"], timeout=20)
# print(es)
s = Search(using=es, index="account_info")
res = s.query().sort('-age').execute()
# print(res)
for data in res:
print(data.to_dict())運(yùn)行結(jié)果：
分頁(yè)
要指定from、size
示例代碼：
from elasticsearch_dsl import connections, Search, A
# 連接es
es = connections.create_connection(hosts=["127.0.0.1:9200"], timeout=20)
# print(es)
s = Search(using=es, index="account_info")
res = s.query()[2: 5].execute() # {"from": 2, "size": 5}
# print(res)
for data in res:
print(data.to_dict())運(yùn)行結(jié)果：
要訪問(wèn)匹配的所有文檔，可以使用scan()函數(shù)，scan()函數(shù)使用scan、scroll elasticsearch API，需要注意的是這種情況下結(jié)果是不會(huì)被排序的。
示例代碼：
from elasticsearch_dsl import connections, Search
# 連接es
es = connections.create_connection(hosts=["127.0.0.1:9200"], timeout=20)
# print(es)
s = Search(using=es, index="account_info")
res = s.query()
# print(res)
for hit in res.scan():
print(hit.age, hit.address)運(yùn)行結(jié)果：
聚合
使用A快捷方式來(lái)定義一個(gè)聚合。為了實(shí)現(xiàn)聚合嵌套，你可以使用.bucket()、.metirc()以及.pipeline()方法。
bucket 即為分組，其中第一個(gè)參數(shù)是分組的名字，自己指定即可，第二個(gè)參數(shù)是方法，第三個(gè)是指定的field。
metric 也是同樣，metric的方法有sum、avg、max、min等等，但是需要指出的是有兩個(gè)方法可以一次性返回這些值，stats和extended_stats，后者還可以返回方差等值。
示例代碼1：
from elasticsearch_dsl import connections, Search, A
# 連接es
es = connections.create_connection(hosts=["127.0.0.1:9200"], timeout=20)
# print(es)
s = Search(using=es, index="account_info")
a = A("terms", field="gender")
s.aggs.bucket("gender_terms", a)
res = s.execute()
# print(res)
for hit in res.aggregations.gender_terms:
print(hit.to_dict())運(yùn)行結(jié)果：
示例代碼2：
from elasticsearch_dsl import connections, Search, A
# 連接es
es = connections.create_connection(hosts=["127.0.0.1:9200"], timeout=20)
# print(es)
s = Search(using=es, index="account_info")
s.aggs.bucket("per_gender", "terms", field="gender")
s.aggs["per_gender"].metric("sum_age", "sum", field="age")
s.aggs["per_gender"].bucket("terms_balance", "terms", field="balance")
res = s.execute()
# print(res)
for hit in res.aggregations.per_gender:
print(hit.to_dict())運(yùn)行結(jié)果：
示例代碼3：
from elasticsearch_dsl import connections, Search, Q
# 連接es
es = connections.create_connection(hosts=["127.0.0.1:9200"], timeout=20)
# print(es)
s = Search(using=es, index="account_info")
res = s.aggs.bucket("aaa", "terms", field="gender").metric("avg_age", "avg", field="age")
print(res.to_dict())運(yùn)行結(jié)果：
示例代碼4： 【聚合，內(nèi)置排序】
from elasticsearch_dsl import connections, Search, Q
# 連接es
es = connections.create_connection(hosts=["127.0.0.1:9200"], timeout=20)
# print(es)
"""
{
'terms': {
'field': 'age',
'order': {
'_count': 'desc'
}
}
}
"""
s = Search(using=es, index="account_info")
res = s.aggs.bucket("agg_age", "terms", field="age", order={"_count": "desc"})
print(res.to_dict())
response = s.execute()
for hit in response.aggregations.agg_age:
print(hit.to_dict())
"""
{
'terms': {
'field': 'age',
'order': {
'_count': 'asc'
}
},
'aggs': {
'avg_age': {
'avg': {
'field': 'age'
}
}
}
}
"""
s2 = Search(using=es, index="account_info")
res2 = s2.aggs.bucket("agg_age", "terms", field="age", order={"_count": "asc"}).metric("avg_age", "avg", field="age")
print(res2.to_dict())
response = s2.execute()
for hit in response.aggregations.agg_age:
print(hit.to_dict())運(yùn)行結(jié)果：
示例代碼5：
from elasticsearch_dsl import connections, Search, Q
# 連接es
es = connections.create_connection(hosts=["127.0.0.1:9200"], timeout=20)
# print(es)
"""
{
'aggs': {
'avg_age': {
'avg': {
'field': 'age'
}
}
}
}
"""
s = Search(using=es, index="account_info").query("range", age={"gte": 28})
res = s.aggs.metric("avg_age", "avg", field="age")
print(res.to_dict())
response = s.execute()
print(response)
for hit in response:
print(hit.to_dict())運(yùn)行結(jié)果：
高亮顯示
示例代碼：【目前似乎沒(méi)有效果，待驗(yàn)證】
from elasticsearch_dsl import connections, Search, Q
# 連接es
es = connections.create_connection(hosts=["127.0.0.1:9200"], timeout=20)
# print(es)
s = Search(using=es, index="test_index")
res = s.highlight("id").execute().to_dict()
print(res)運(yùn)行結(jié)果：
source限制返回字段
示例代碼：
from elasticsearch_dsl import connections, Search, Q
# 連接es
es = connections.create_connection(hosts=["127.0.0.1:9200"], timeout=20)
# print(es)
s = Search(using=es, index="account_info")
res = s.query().execute()
for hit in res:
print(hit.to_dict())
# 限制返回字段
s2 = Search(using=es, index="account_info")
res2 = s2.query().source(['account_number', 'address']).execute()
for hit in res2:
print(hit.to_dict())運(yùn)行結(jié)果：
刪除
調(diào)用Search對(duì)象上的delete方法而不是execute來(lái)實(shí)現(xiàn)刪除匹配查詢的文檔
示例代碼：
from elasticsearch_dsl import connections, Search, Q
# 連接es
es = connections.create_connection(hosts=["127.0.0.1:9200"], timeout=20)
# print(es)
s = Search(using=es, index="test_index")
res = s.query("match", name="張").delete()
print(res)運(yùn)行結(jié)果：
案例分析
創(chuàng)建索引：
from elasticsearch_dsl import Search
from elasticsearch import Elasticsearch
# 連接es
es = Elasticsearch(hosts=["127.0.0.1:9200"], sniffer_timeout=60, timeout=30)
body = {
"mappings": {
"properties": {
"account_number": {
"type": "integer"
},
"balance": {
"type": "integer"
},
"firstname": {
"type": "text"
},
"lastname": {
"type": "text"
},
"age": {
"type": "integer"
},
"gender": {
"type": "keyword"
},
"address": {
"type": "text"
},
"employer": {
"type": "text"
},
"email": {
"type": "text"
},
"province": {
"type": "text"
},
"state": {
"type": "text"
}
}
}
}
# 創(chuàng)建 index
es.indices.create(index="account_info", body=body)查看索引：
使用kibana批量生成數(shù)據(jù)：
POST account_info/_bulk
{"index": {"_index":"account_info"}}
{"account_number":1,"balance":20,"firstname":"三","lastname":"張","age":25,"gender":"男","address":"北京朝陽(yáng)","employer":"AAA","email":"123@qq.com","province":"北京","state":"正常"}
{"index": {"_index":"account_info"}}
{"account_number":2,"balance":70,"firstname":"二","lastname":"張","age":26,"gender":"男","address":"北京海淀","employer":"AAA","email":"123@qq.com","province":"北京","state":"正常"}
{"index": {"_index":"account_info"}}
{"account_number":3,"balance":80,"firstname":"四","lastname":"張","age":27,"gender":"女","address":"遼寧朝陽(yáng)","employer":"BBB","email":"123@qq.com","province":"遼寧","state":"正常"}
{"index": {"_index":"account_info"}}
{"account_number":4,"balance":60,"firstname":"五","lastname":"張","age":28,"gender":"男","address":"山東青島","employer":"AAA","email":"123@qq.com","province":"山東","state":"正常"}
{"index": {"_index":"account_info"}}
{"account_number":5,"balance":40,"firstname":"六","lastname":"張","age":29,"gender":"女","address":"山東濟(jì)南","employer":"AAA","email":"123@qq.com","province":"山東","state":"正常"}
{"index": {"_index":"account_info"}}
{"account_number":6,"balance":50,"firstname":"七","lastname":"張","age":30,"gender":"男","address":"河北唐山","employer":"BBB","email":"123@qq.com","province":"河北","state":"正常"}
{"index": {"_index":"account_info"}}
{"account_number":7,"balance":30,"firstname":"一","lastname":"張","age":31,"gender":"女","address":"河北石家莊","employer":"AAA","email":"123@qq.com","province":"河北","state":"正常"}查看生成的數(shù)據(jù)：
根據(jù)條件查詢：
1.查詢balance在40~70的信息
from elasticsearch_dsl import connections, Search, Q
# 連接es
es = connections.create_connection(hosts=["127.0.0.1:9200"], timeout=20)
# print(es)
s = Search(using=es, index="account_info")
# 查詢balance在40~70的信息
q = Q("range", balance={"gte": 40, "lte": 70})
res = s.query(q)
for data in res:
print(data.to_dict())
print("共查到%d條數(shù)據(jù)" % res.count())
2.查詢balance在40~70的男性信息
from elasticsearch_dsl import connections, Search, Q
# 連接es
es = connections.create_connection(hosts=["127.0.0.1:9200"], timeout=20)
# print(es)
s = Search(using=es, index="account_info")
# 查詢balance在40~70的信息
q1 = Q("range", balance={"gte": 40, "lte": 70})
# 男性
q2 = Q("term", gender="男")
# and
q = q1 & q2
res = s.query(q)
for data in res:
print(data.to_dict())
print("共查到%d條數(shù)據(jù)" % res.count())
3.省份為北京、25或30歲的男性信息
from elasticsearch_dsl import connections, Search, Q
# 連接es
es = connections.create_connection(hosts=["127.0.0.1:9200"], timeout=20)
# print(es)
s = Search(using=es, index="account_info")
# 方式一：
# 省份為北京
q1 = Q("match", province="北京")
# 25或30歲的男性信息
q2 = Q("bool", must=[Q("terms", age=[25, 30]), Q("term", gender="男")])
# and
q = q1 & q2
res = s.query(q)
for data in res:
print(data.to_dict())
print("共查到%d條數(shù)據(jù)" % res.count())
print("*" * 100)
# 方式二
# 省份為北京
q1 = Q("match", province="北京")
# 25或30歲的信息
# q2 = Q("bool", must=[Q("terms", age=[25, 30]), Q("term", gender="男")])
q2 = Q("term", age=25) | Q("term", age=30)
# 男性
q3 = Q("term", gender="男")
res = s.query(q1).query(q2).query(q3) # 多次query就是& ==> and 操作
for data in res:
print(data.to_dict())
print("共查到%d條數(shù)據(jù)" % res.count())
4.地址中有“山”字，年齡不在25~28歲的女性信息
from elasticsearch_dsl import connections, Search, Q
# 連接es
es = connections.create_connection(hosts=["127.0.0.1:9200"], timeout=20)
# print(es)
s = Search(using=es, index="account_info")
# 地址中有“山”字且為女性
q1 = Q("match", address="山") & Q("match", gender="女")
# 年齡在25~28歲
q2 = ~Q("range", age={"gte": 25, "lte": 28})
# 使用filter過(guò)濾
# query和filter的前后關(guān)系都行
res = s.filter(q2).query(q1)
for data in res:
print(data.to_dict())
print("共查到%d條數(shù)據(jù)" % res.count())
5.根據(jù)年齡進(jìn)行聚合，然后計(jì)算每個(gè)年齡的評(píng)價(jià)balance數(shù)值
示例代碼：
from elasticsearch_dsl import connections, Search, A
# 連接es
es = connections.create_connection(hosts=["127.0.0.1:9200"], timeout=20)
# print(es)
s = Search(using=es, index="account_info")
# 先用年齡聚合，然后拿到返平均數(shù)
# size指定最大返回多少條數(shù)據(jù)，默認(rèn)10條
# 實(shí)質(zhì)上account的數(shù)據(jù)中，age分組沒(méi)有100個(gè)這么多
q = A("terms", field="age", size=100).metric("age_per_balance", "avg", field="balance")
s.aggs.bucket("res", q)
# 執(zhí)行并拿到返回值
response = s.execute()
# res是bucket指定的名字
# response.aggregations.to_dict是一個(gè){'key': 25, 'doc_count': 1, 'age_per_balance': {'value': 20.0}}的數(shù)據(jù)，和用restful查詢的一樣
for data in response.aggregations.res:
print(data.to_dict())運(yùn)行結(jié)果：
6.根據(jù)年齡聚合，求25~28歲不同性別的balance值。
示例代碼：
from elasticsearch_dsl import connections, Search, A
# 連接es
es = connections.create_connection(hosts=["127.0.0.1:9200"], timeout=20)
# print(es)
s = Search(using=es, index="account_info")
# 這次就用這種方法
# range 要注意指定ranges參數(shù)和from to
a1 = A("range", field="age", ranges={"from": 25, "to": 28})
a2 = A("terms", field="gender")
a3 = A("avg", field="balance")
s.aggs.bucket("res", a1).bucket("gender_group", a2).metric("balance_avg", a3)
# 執(zhí)行并拿到返回值
response = s.execute()
# res是bucket指定的名字
for data in response.aggregations.res:
print(data.to_dict())運(yùn)行結(jié)果： 【注意：不包含年齡28的值】
總結(jié)：
假如是數(shù)組，如：bool的must、terms，那么就要字段=[ ]假如是字典，如：range，那么就要字段={xxx: yyy, .... }假如是單值，如：term、match，那么就要字段=值假如查的是多個(gè)字段，如：multi_mathc，那么就要加上query="要查的值", fields=["字段1", "字段2", ...]然后各個(gè)條件的邏輯關(guān)系，可以通過(guò)多次query和filter或直接用Q("bool", must=[Q...], should=[Q...])再加上& | ~表示
到此這篇關(guān)于python中elasticsearch_dsl模塊的使用方法的文章就介紹到這了,更多相關(guān)python elasticsearch_dsl模塊內(nèi)容請(qǐng)搜索腳本之家以前的文章或繼續(xù)瀏覽下面的相關(guān)文章希望大家以后多多支持腳本之家！
您可能感興趣的文章:
- Python實(shí)戰(zhàn)之Elasticsearch的高級(jí)實(shí)現(xiàn)詳解
- python中的elasticsearch_dsl查詢語(yǔ)句轉(zhuǎn)換成es查詢語(yǔ)句詳解
- python 使用elasticsearch 實(shí)現(xiàn)翻頁(yè)的三種方式
- Python操作Elasticsearch處理timeout超時(shí)
- python3實(shí)現(xiàn)elasticsearch批量更新數(shù)據(jù)
- python更新數(shù)據(jù)庫(kù)中某個(gè)字段的數(shù)據(jù)(方法詳解)
- Python更新數(shù)據(jù)庫(kù)腳本兩種方法及對(duì)比介紹
- Python調(diào)用Elasticsearch更新數(shù)據(jù)庫(kù)的操作方法
相關(guān)文章
Python中類型關(guān)系和繼承關(guān)系實(shí)例詳解
這篇文章主要介紹了Python中類型關(guān)系和繼承關(guān)系,較為詳細(xì)的分析了Python中類型關(guān)系和繼承關(guān)系的原理與使用技巧,具有一定參考借鑒價(jià)值,需要的朋友可以參考下2015-05-05
Python實(shí)現(xiàn)UDP程序通信過(guò)程圖解
這篇文章主要介紹了Python實(shí)現(xiàn)UDP程序通信過(guò)程圖解,文中通過(guò)示例代碼介紹的非常詳細(xì)，對(duì)大家的學(xué)習(xí)或者工作具有一定的參考學(xué)習(xí)價(jià)值,需要的朋友可以參考下2020-05-05
pip安裝python庫(kù)時(shí)報(bào)Failed?building?wheel?for?xxx錯(cuò)誤的解決方法
最近在使用pip安裝python的時(shí)候遇到些問(wèn)題,所以下面這篇文章主要給大家介紹了關(guān)于pip安裝python庫(kù)時(shí)報(bào)Failed?building?wheel?for?xxx錯(cuò)誤的解決方法,需要的朋友可以參考下2023-01-01
詳談pandas中agg函數(shù)和apply函數(shù)的區(qū)別
下面小編就為大家分享一篇詳談pandas中agg函數(shù)和apply函數(shù)的區(qū)別，具有很好的參考價(jià)值，希望對(duì)大家有所幫助。一起跟隨小編過(guò)來(lái)看看吧2018-04-04
這篇文章主要介紹了Python程序流程控制實(shí)驗(yàn)，程序流程控制的使用在Python中具有非常重要的地位。在本文中我將布置Python程序流程控制的實(shí)驗(yàn)習(xí)題，希望您閱讀完本文后能夠有所收獲2022-01-01
Tensorflow2.1 MNIST圖像分類實(shí)現(xiàn)思路分析
這篇文章主要為大家介紹了Tensorflow2.1 MNIST圖像分類實(shí)現(xiàn)思路分析，有需要的朋友可以借鑒參考下，希望能夠有所幫助，祝大家多多進(jìn)步，早日升職加薪2022-11-11
Python中調(diào)用和運(yùn)行其他.py文件的多種實(shí)現(xiàn)方法
本文介紹了在Python中調(diào)用和運(yùn)行其他.py文件的四種方法：subprocess模塊、exec函數(shù)、import語(yǔ)句和os.system函數(shù),每種方法都有其適用場(chǎng)景和優(yōu)缺點(diǎn)2025-02-02
Python中很常用的函數(shù)map()用法實(shí)例
這篇文章主要給大家介紹了關(guān)于Python中很常用的函數(shù)map()用法的相關(guān)資料,map()函數(shù)是Python的內(nèi)置函數(shù),會(huì)根據(jù)提供的函數(shù)參數(shù),對(duì)傳入的序列數(shù)據(jù)進(jìn)行映射,需要的朋友可以參考下2023-10-10
記得我好幾年前剛學(xué)Python的時(shí)候，看裝飾器就覺(jué)得看九陰真經(jīng)里面的怪文字，完全看不懂，確實(shí)裝飾器是一個(gè)非常難以理解的概念，相信很多初學(xué)者一定也有這樣的困惑，所以篇文章主要介紹了關(guān)于Python中裝飾器的相關(guān)資料，需要的朋友可以參考下。2017-04-04
Python分支結(jié)構(gòu)和循環(huán)結(jié)構(gòu)示例代碼
在Python中,分支結(jié)構(gòu)通過(guò)if、elif和else關(guān)鍵字來(lái)實(shí)現(xiàn)條件判斷,在使用if語(yǔ)句時(shí),程序會(huì)根據(jù)條件表達(dá)式的真假執(zhí)行相應(yīng)的代碼塊,這篇文章主要介紹了Python分支結(jié)構(gòu)和循環(huán)結(jié)構(gòu),需要的朋友可以參考下2024-03-03