cch123 / elasticsql Goto Github PK

current:

select * from a where id = missing

the expected:

select * from a where missing(id) = true

original SQL :

select title, price from products where title = '%A%'

expect like that :

{
    "_source": [
        "title",
        "price",
    ],
    "query": {
        "bool": {
            "must": [
                {
                    "match_phrase": {
                        "title": {
                            "query": "%A%"
                        }
                    }
                }
            ]
        }
    },
    "from": 0,
    "size": 1
} 

actually, elasticsql.Convert(sql) didn't generate "_source":["title", "price"]

输入sql转化成dsl语句？

eg.
select count(*) from worksheet;
select sum(account) from worksheet;

Does this library have the capability to perform a conversion from DSL to SQL?

implement the group by

range
date_range
date_histogram(already supported)
...

First of all thanks for the great work!

The functionality check-list mentions that like operator is implemented for phrase search.

However, in the example, although the strict equality = is used, the generated elasticsearch query is a phrase match. In that case, as far as I understood, shouldn't it generate a term search instead?

If the example is correct, what is the way to generate a term search query?

select count(DISTINCT deviceId) from uv where deviceId in (select deviceId from uv where userId = 'unknown') and uv.userId != 'unknown';

eg:
select * from `order`.paymen_info

select `order`, `timestamp` from ccc

没有版本列表，是否从 elasticsearch 2.x 支持到 6.x？

https://elasticsearch.cn/question/5524
你好，如果想要使用multi_match，sql应该使用什么符号进行查询？

elaticsearch sql have below dsl,
{
"aggs" : {
"articles_over_time" : {
"date_histogram" : {
"field" : "date",
"interval" : "1.5h"
}
}
}
}

is it possible?

select * from aaa where remote_addr="111.206.36.1" limit 10;
转换为:
{"query" : {"bool" : {"must" : [{"match" : {"remote_addr" : {"query" : "111.206.36.1", "type" : "phrase"}}}]}},"from" : 0,"size" : 10}

用该dsl语句查询的时候:
{ "error": { "root_cause": [ { "type": "parsing_exception", "reason": "[match] query does not support [type]", "line": 1, "col": 84 } ], "type": "parsing_exception", "reason": "[match] query does not support [type]", "line": 1, "col": 84 }, "status": 400 }

必须要删除 "type" : "phrase" 才行

bigint 自动转成字符串了， 拿进去查询，查出来的结果是错的

Hello, very nice work.

I was wondering if, is there a way to convert QDSL json object into the SQL?

Hello i need to know when we make it large query have alot of join inside SQL possible to make by your Go solution to Elastic Query match ?

Thanks,
George

请问一下 opendistro的_sql/_translate 您了解吗？我实验sql转换为dsl失败了

sql like : select count(id) from A where A.id > 5 group by A.a, A.b;
because of
innerMap["terms"] = msi{ "field": colName.Name.String(), "size": 0, }
will gen the dsl like: "terms":{"field":"field2","size":0}
and then make a error: "reason":"[size] must be greater than 0
using es7.X

原始sql

SELECT
    `familyCode`
FROM
    `myTable`
WHERE
    (`admin` LIKE '%v!_xxx%' ESCAPE '!')
ORDER BY
    `updateTime` DESC
LIMIT
    10

其中要对admin字段双侧LIKE字符串v_xxx。
输出的结果：

{
    "query": {
        "bool": {
            "must": [
                {
                    "match_phrase": {
                        "admin": {
                            "query": "v!_xxx"
                        }
                    }
                }
            ]
        }
    },
    "from": 0,
    "size": 10,
    "sort": [
        {
            "updateTime": "desc"
        }
    ]
}

其中query.bool.must[0].match_phrase.admin.query字段的值中，! 没有去掉。

使用例程转换下列语句为DSL的时候，来源
select * from table_name group by date_histogram(field='create_time', interval='1h')
报错
syntax error at position 81

interval要改成value么？