Re: [DISCUSS] FLIP-517: Better Handling of Dynamic Table Primitives with PTFs

2025-04-23 Thread Shengkai Fang
Hi, Timo I believe I may have misapplied the query hints and table hints. The current proposal only applies to query hints. For table hints, I think we would need to create a view, apply the table hints to the view, and use it as the SEARCH_KEY function's input. Best, Shengkai Shengkai Fang 于2

Re: [DISCUSS] FLIP-517: Better Handling of Dynamic Table Primitives with PTFs

2025-04-17 Thread Shengkai Fang
Thanks Timo's explain. > Can you give an example for a complex temporal condition? For example, consider the following SQL statement: SELECT * FROM src LEFT JOIN dim FOR SYSTEM_TIME AS OF PROCTIME() ON src.key = dim.key1 AND dim.key2 = 'literal_value'; In this scenario, I am uncertain how to re

Re: [DISCUSS] FLIP-517: Better Handling of Dynamic Table Primitives with PTFs

2025-04-16 Thread Timo Walther
Hi everyone, great to see that this discussion gains momentum. Let me answer your questions below: @Hao 1. SEARCH_KEY planner representation and options Yes, internally nothing in the planner should change. The PTFs are just syntactic sugar to rules and nodes that already exist. Regarding t

Re: [DISCUSS] FLIP-517: Better Handling of Dynamic Table Primitives with PTFs

2025-04-14 Thread Lincoln Lee
+1 for the flip, the snapshot, changelog related ptfs are nice addition to flink sql! I have a question about the naming of the SEARCH_KEY, considering that it is a simplified alternative to lookup join, and the word 'search' may be more likely to remind users of keyword searching which is a bit d

Re: [DISCUSS] FLIP-517: Better Handling of Dynamic Table Primitives with PTFs

2025-04-14 Thread Shengkai Fang
Thanks for the FLIP, it helps a lot for us to develop features like VECTOR_SEARCH. But I have some questions about the FLIP: 1. Timing of Option Consumption for SEARCH_KEY Parameters For the FOR SYSTEM_TIME AS OF syntax, the planner leverages hints and catalog tables to load the table scan during

Re: [DISCUSS] FLIP-517: Better Handling of Dynamic Table Primitives with PTFs

2025-04-02 Thread Hao Li
Hi Timo, Any question I have is what's the SEARCH_KEY result schema you have in mind? Can it output multiple rows for every row in the left table or it needs to pack the result in a single row as an array? Thanks, Hao On Mon, Mar 24, 2025 at 10:20 AM Hao Li wrote: > Thanks Timo for the FLIP! T

RE: [DISCUSS] FLIP-517: Better Handling of Dynamic Table Primitives with PTFs

2025-03-27 Thread Gustavo de Morais
+1 That's a great proposal, Timo. I like the proposed simplified syntax. Also, being able to convert from or to a changelog is a really nice to have. Gustavo On 2025/03/21 13:23:37 Timo Walther wrote: > Hi everyone, > > I would like to start a discussion about FLIP-517: Better Handling of > D

RE: [DISCUSS] FLIP-517: Better Handling of Dynamic Table Primitives with PTFs

2025-03-27 Thread Gustavo de Morais
+1 That's a great proposal, Timo. I like the proposed simplified syntax. Also, being able to convert from or to a changelog is a really nice to have. Gustavo On 2025/03/21 13:23:37 Timo Walther wrote: > Hi everyone, > > I would like to start a discussion about FLIP-517: Better Handling of > Dyna

Re: [DISCUSS] FLIP-517: Better Handling of Dynamic Table Primitives with PTFs

2025-03-24 Thread Hao Li
Thanks Timo for the FLIP! This is a great improvement to the FLINK sql syntax around tables. I have two clarification questions: 1. For SEARCH_KEY ``` SELECT * FROM t_other, LATERAL SEARCH_KEY( input => t, on_key => DESCRIPTOR(k), lookup => t_other.name, options => MAP[ '