Results 32 comments of dongwang

Convert to draft as this PR depending on #21959. So it should be reviewed after #21959 is merged.

> Is this a cherry pick? At first glance, I don't see substantial similarities with the referenced PR. I referenced the implementation of trino, which may evolve gradually through a...

> Based on my reading of the code, this wouldn't qualify as a cherry pick as it's substantially different, and even the way it's performed is different between the two...

> Instead of making changes directly to `Type`, can we just have a new utility method in the Iceberg connector which takes in the Type, the Object, and from those...

> I think we can start in the Iceberg connector, and extract if or when an additional use case comes along. Good idea! Have done this, please take a look,...

Hi @ZacBlanco, I have combined `getPreviousValue()` and `getNextValue()` to a single function `getAdjacentValue()`, and used `anyNot(...)` to extract functions `thoroughlyPushdown(...)` and `notThoroughlyPushdown(...)` which could have a more clearer meaning. Then...

@tdcmeehan @ZacBlanco just resolved the conflict in `IcebergPlanOptimizer`.

Hi @evanvdia, just wanted to confirm that are you still working on this?