Champion "deconstruction in from and let"

[gafter]

Summary

Allow declaring multiple range at once with let. For example, ... let (x, y) = deconstructable ....

• Proposal added
• Discussed in LDM
• Decision in LDM
• Finalized (done, rejected, inactive)
• Spec'ed

See also dotnet/roslyn#13964

https://github.com/dotnet/csharplang/blob/master/meetings/2016/LDM-2016-12-07-14.md#deconstruction

[jhgbrt]

Is there any progress on this? The more one gets used to deconstruction, the more annoying it becomes that from and let clauses in LINQ queries don't support this :-)

[333fred]

Is there any progress on this?

This feature is in the Any Time bucket, which means we are unlikely to spend time on it ourselves, but if a motivated community member would like to contribute, we would be happy to review. Currently, this issue needs a specification for LDM to review and approve.

[jhgbrt]

Any chance this could make it in C# 12?

[CyrusNajmabadi]

@jhgbrt see the comment above yours.

[sab39]

Does this cover things like

from person in People
where person.Age is int age && age >= 18 // person.Age is a nullable int
select (person, age)

or would that be a separate ticket?

[333fred]

I don't think that would be covered by this. That would be changing the scoping of existing code, which would be concerning.

[sab39]

Oof. I see what you mean - my code example would compile today if something called age existed in an outer scope. That seems to me like a footgun in the status quo, though: if someone declares a variable in a pattern expression in a where clause and then uses the same identifier in the next line in a select clause, it's unlikely that they intended to pull in a different value from outside, and even more unlikely that someone reading the code would recognize what's happening.

Of course, breaking existing code is a big deal and should only be done in extenuating circumstances. But does "the code is probably not doing what they intended in the first place" count as an extenuating circumstance?