Hidden fields shouldn't need to implement Display

[zhiburt]

Hidden fields shouldn't need to implement Display

[zhiburt]

Understood.

Let me know if you'll encounter any difficulties on the second shot :)

If you don't mind I'll close the issue then.

And Yes.
Happy New Year.

[zhiburt]

Hi @nrdxp

Good observation,
I never thought about it.

Thank you for bringing this up and making the proposal.

I'll try to work on this tomorrow.

[zhiburt]

Hey @nrdxp

Could you show the exact type you're using?

Because I as see hiding fields in structs and tuples works well in master branch already. (surprisingly 😄)

The only issue I've spotted is when it's used in enum with #[field(inline)].
Is this your case?

Here's a bunch of tests

tabled/tests/derive_test.rs

Lines 593 to 715 in 061ccd7

	#[test]
	fn hidden_fields_may_not_implement_display() {
	{
	struct Something;
	#[derive(Tabled)]
	struct TupleStruct(#[header(hidden = true)] Something, &'static str);
	let st = TupleStruct(Something, "nrdxp");
	assert_eq!(vec!["nrdxp".to_owned()], st.fields());
	assert_eq!(vec!["1".to_owned()], TupleStruct::headers());
	}
	{
	struct Something;
	#[derive(Tabled)]
	struct Struct {
	#[header(hidden = true)]
	_gem: Something,
	name: &'static str,
	}
	let st = Struct {
	_gem: Something,
	name: "nrdxp",
	};
	assert_eq!(vec!["nrdxp".to_owned()], st.fields());
	assert_eq!(vec!["name".to_owned()], Struct::headers());
	}
	{
	struct Something;
	#[derive(Tabled)]
	enum Enum {
	#[field(inline("A::"))]
	A {
	name: &'static str,
	},
	#[field(inline("B::"))]
	B {
	issue: usize,
	#[header(hidden = true)]
	_gem: Something,
	name: &'static str,
	},
	#[field(inline("C::"))]
	C(usize, #[header(hidden = true)] Something, &'static str),
	D,
	}
	assert_eq!(
	vec![
	"A::name".to_owned(),
	"B::issue".to_owned(),
	"B::name".to_owned(),
	"C::0".to_owned(),
	"C::2".to_owned(),
	"D".to_owned()
	],
	Enum::headers()
	);
	let st = Enum::A { name: "nrdxp" };
	assert_eq!(
	vec![
	"nrdxp".to_owned(),
	"".to_owned(),
	"".to_owned(),
	"".to_owned(),
	"".to_owned(),
	"".to_owned()
	],
	st.fields()
	);
	let st = Enum::B {
	_gem: Something,
	issue: 32,
	name: "nrdxp",
	};
	assert_eq!(
	vec![
	"".to_owned(),
	"32".to_owned(),
	"nrdxp".to_owned(),
	"".to_owned(),
	"".to_owned(),
	"".to_owned()
	],
	st.fields()
	);
	let st = Enum::C(32, Something, "nrdxp");
	assert_eq!(
	vec![
	"".to_owned(),
	"".to_owned(),
	"".to_owned(),
	"32".to_owned(),
	"nrdxp".to_owned(),
	"".to_owned()
	],
	st.fields()
	);
	let st = Enum::D;
	assert_eq!(
	vec![
	"".to_owned(),
	"".to_owned(),
	"".to_owned(),
	"".to_owned(),
	"".to_owned(),
	"+".to_owned()
	],
	st.fields()
	);
	}
	}

[zhiburt]

By they how do you think maybe its worth to rename hidden to ignore?

[zhiburt]

*Coincidentally was closed after #34 merge 😥

[zhiburt]

@nrdxp ping

[zhiburt]

ping

[nrdxp]

Hey sorry, my notifications got flooded with other things and I haven't seen any of this til now. Unfortunately I don't have any of the code I was using at the time left over, I switched to using prettytable-rs for the time being, although I would much prefer printing the tables directly from structs as this crate provides, rather than manually building the tables as I'm doing with prettytable, so I'd eventually like to give this another shot.