Re: [discuss] Add the default mode to the path in the Json functions.

2020-04-20 Thread Forward Xu
Hi Julian, Thank you for bringing such a good shared link, there are some areas in the JSON function that need to be improved and implemented. I will continue to improve it. Forward Julian Hyde 于2020年4月21日周二 上午1:28写道: > Speaking of JSON functions, JOOQ creator Lukas Eder has been giving JSON >

Re: [discuss] Add the default mode to the path in the Json functions.

2020-04-20 Thread Julian Hyde
Speaking of JSON functions, JOOQ creator Lukas Eder has been giving JSON functions in MySQL/MariaDB a good workout over the last few days. It’s amusing to read what he has discovered: https://twitter.com/lukaseder Julian > On Apr 20, 2020, at 8:27 AM,

Re: [discuss] Add the default mode to the path in the Json functions.

2020-04-20 Thread Stamatis Zampetakis
Hi, I know that Oracle uses lax by default [1] but I don't remember what other DBMS do. In any case adopting a default mode sounds like a reasonable thing to do. Best, Stamatis [1] https://docs.oracle.com/database/121/ADXDB/json.htm#ADXDB6259 On Mon, Apr 20, 2020, 7:28 AM Chunwei Lei wrote:

Re: [discuss] Add the default mode to the path in the Json functions.

2020-04-19 Thread Chunwei Lei
Thank you for proposing this, Forward. I am wondering whether it is really useful for users who want to use these functions since they probably know what the syntax is. But personally I would like to have the default mode since it is more convenient. Best, Chunwei On Sun, Apr 19, 2020 at

[discuss] Add the default mode to the path in the Json functions.

2020-04-18 Thread Forward Xu
hi everyone, I recently found some discussable optimizations when I contributed the flip-90 [1] Json functions: The current json functions in calcite: JSON_EXISTS, JSON_VALUE, JSON_QUERY, JSON_OBJECT, JSON_OBJECTAGG, JSON_ARRAY, JSON_ARRAYAGG and IS JSON predication functions. These functions are