disable coercison for unmatched struct type

[Lordworms]

Which issue does this PR close?

• Closes Inconsistent and incorrect struct field coercion #14396

Rationale for this change

What changes are included in this PR?

Are these changes tested?

Are there any user-facing changes?

[alamb]

Thank you @Lordworms -- I hope to review this tomorrow

[alamb]

Hi @Lordworms -- thank you very much for this PR and for working on the issue

While reviewing this PR I noticed a few other tests that I think we should add (are you willing to do so?)

1. Adding a test for coalesece(column1, column2) and `coalesce(column2, column3) which is similar to union and case
2. Adding a test for comparison column1 = column and column2 = column3 which is also similar

Then in terms of code, As this PR demonstrates, there are at least three places that do versions of this struct comparision that need to be changed as they have different versions of code that does about the same thing by coercing structs.

I think @jayzhan211's suggestion on https://github.com/apache/datafusion/pull/14384/files#r1937492704

Yes, I think type union resolution is the correct on for CASE

Means that we should be aiming towards consolidating / using the same rules.

Maybe you could move the checks into struct_coercion and then update any other coercion that needed to coerce structs to use that one?

[Lordworms]

struct_coercion

Sure, I can do that

Hi @Lordworms -- thank you very much for this PR and for working on the issue

While reviewing this PR I noticed a few other tests that I think we should add (are you willing to do so?)

1. Adding a test for coalesece(column1, column2) and `coalesce(column2, column3) which is similar to union and case
2. Adding a test for comparison column1 = column and column2 = column3 which is also similar

Then in terms of code, As this PR demonstrates, there are at least three places that do versions of this struct comparision that need to be changed as they have different versions of code that does about the same thing by coercing structs.

I think @jayzhan211's suggestion on https://github.com/apache/datafusion/pull/14384/files#r1937492704

Yes, I think type union resolution is the correct on for CASE

Means that we should be aiming towards consolidating / using the same rules.

Maybe you could move the checks into struct_coercion and then update any other coercion that needed to coerce structs to use that one?

Ok, I'll add that