2022 08 19

ElasticSearch REST API 使用示例

ElasticSearch Example 2022年08月19日

Index Api

创建 Index

http
PUT /index_name/

删除 Index

http
DELETE /index_name

Document Api

删除

http
DELETE /index_name/label/N310108198708302043

批量删除

http
POST /index_name/_delete_by_query
{
 "query": {
 "match": {
 "field_name": "field_value"
 }
 }
}

添加

http
POST /index_name/label/N310108198708302043
{
 "UNITED_CIB_MILES": 0,
 "CNX_1YEAR_RATIO": 0,
 "PVG6_8_CNT": 0,
 "MFM_1YEAR_RATIO": 0,
 "ACTUAL_YEAR_DIS": 0.15,
 "IDCARD_AGE": "65",
 "LATELY_DAY": 221,
 "PVG21_24_CNT": 0,
 "PRICE_SENSITIVE_FLAG": "是",
 "UNITED_ABC_MILES": 0,
 "BUSINESS_RATIO": 0
}

增量更新字段

http
POST /index_name/label/N310108198708302043/_update
{
 "doc": {
 "TOTAL_REFUND_CNT": 1
 }
}

Search Api

Range

http
POST /index_name/_search/
{
 "query": {
 "constant_score": {
 "filter": {
 "range": {
 "PK_MEMBER_ID": {
 "gt": 230,
 "lt": 240,
 "relation": "CONTAINS"
 }
 }
 }
 }
 },
 "size": 50
}

参数:

参数 key参数说明
gt大于
gte大于等于
lt小于
lte小于等于
format日期格式化
relation对于 range 对象的匹配规则
time_zone例如 +01:00 或 -08:00 或 America/Los_Angeles
boost用于减少或增加查询相关性分数的浮点数

relation:

  • INTERSECTS(默认)匹配具有与查询范围相交的范围字段值的文档。
  • CONTAINS 匹配具有完全包含查询范围的范围字段值的文档。
  • WITHIN 匹配具有完全在查询范围内的范围字段值的文档。

Regexp

http
POST /index_name/_search/
{
 "query": {
 "regexp": {
 "PK_MEMBER_ID": {
 "value": "23.",
 "flags": "ALL",
 "max_determinized_states": 10000,
 "rewrite": "constant_score"
 }
 }
 }
}

筛选字段

http
POST /index_name/_search/
{
 "_source": [
 "IS_MEMBER_FLAG",
 "REGIST_DATE",
 "IS_UNITED_CARD_FLAG",
 "MEMBER_LEVEL_CODE",
 "SOURCE_NAME",
 "PK_MEMBER_ID",
 "LEVEL_NAME1",
 "LEVEL_NAME2",
 "UPDATE_AT"
 ],
 "query": {
 "regexp": {
 "PK_MEMBER_ID": {
 "value": "23.",
 "flags": "ALL",
 "max_determinized_states": 10000,
 "rewrite": "constant_score"
 }
 }
 },
 "sort": [
 "PK_MEMBER_ID"
 ]
}

Exists

http
GET /index_name/_search
{
 "query": {
 "exists": {
 "field": "KC_EMPHASIS"
 }
 }
}

更新 mapping

http
PUT /index_name/_mapping
{
 "properties": {
 "name": {
 "type": "text"
 },
 "age": {
 "type": "integer"
 },
 "address": {
 "type": "keyword"
 }
 }
}