G
Glenn Parker
Gavin said:
Instead of "one", why not "lone", e.g. The Lone Ranger?
Instead of "one", why not "lone", e.g. The Lone Ranger?
J
Joel VanderWerf
Glenn said:
Let's call it #singleton
T
trans. (T. Onoma)
| > Gavin Sinclair wrote:
| >> I find #one quite uncompelling,
| >> because it sounds like #any: it doesn't express its one_and_only_one
| >> meaning.
| >
| > Instead of "one", why not "lone", e.g. The Lone Ranger?
Array#tonto
| Let's call it #singleton
LOL
| >> I find #one quite uncompelling,
| >> because it sounds like #any: it doesn't express its one_and_only_one
| >> meaning.
| >
| > Instead of "one", why not "lone", e.g. The Lone Ranger?
Array#tonto
| Let's call it #singleton
LOL
B
Brian Schröder
Why not #scalar as it returns the scalar value of the array if the array is a 1-tuple.
Regards,
Brian
B
Brian Schröder
Sorry for the noise, I read the rejection further down only after sending this. (Though I still think #to_scalar more natural than #one, because #one does not convey a clear meaning to me)
e.g.
[].one => 1 (It returns one?)
[].one => true (It is only one array?)
[:a,:b,:c,:d].one => [1,1,1,1] (Fill with ones?)
Regards,
Brian
T
trans. (T. Onoma)
On Tuesday 30 November 2004 05:25 am, Brian Schröder wrote:
| Sorry for the noise, I read the rejection further down only after sending
| this. (Though I still think #to_scalar more natural than #one, because #one
| does not convey a clear meaning to me)
|
| e.g.
|
| [].one => 1 (It returns one?)
| [].one => true (It is only one array?)
| [:a,:b,:c,:d].one => [1,1,1,1] (Fill with ones?)
At first it occurred to me that this is nothing more than Array#first, except
that an error is desired if it isn't single. But this led me to another
oddity about #first, you can;t distinguish the results of these two forms:
[].first
[nil].first
To improve upon this leads me to three possibilities:
1) Live with it. It's manegable.
2) Introduce the concept of nack --which is an encapsulated
and diverted error. This has certain advantages, but
it is a "bigger" change.
[].first #=> nack
(f = [].first).nack_raise #=> (Exception)
where #nack_raise is defined in kernel to raise the nack if
it is one other wise return self.
3) Add a parameter to #first.
[].firstempty) #=> (Exception)
[1].firstempty) #=> 1
[1,2].firstempty) #=> 1
[].firstonly) #=> (Exception)
[1].firstonly) #=> 1
[1,2].firstonly) #=> (Exception)
T.
| Sorry for the noise, I read the rejection further down only after sending
| this. (Though I still think #to_scalar more natural than #one, because #one
| does not convey a clear meaning to me)
|
| e.g.
|
| [].one => 1 (It returns one?)
| [].one => true (It is only one array?)
| [:a,:b,:c,:d].one => [1,1,1,1] (Fill with ones?)
At first it occurred to me that this is nothing more than Array#first, except
that an error is desired if it isn't single. But this led me to another
oddity about #first, you can;t distinguish the results of these two forms:
[].first
[nil].first
To improve upon this leads me to three possibilities:
1) Live with it. It's manegable.
2) Introduce the concept of nack --which is an encapsulated
and diverted error. This has certain advantages, but
it is a "bigger" change.
[].first #=> nack
(f = [].first).nack_raise #=> (Exception)
where #nack_raise is defined in kernel to raise the nack if
it is one other wise return self.
3) Add a parameter to #first.
[].first
empty) #=> (Exception)[1].first
empty) #=> 1[1,2].first
empty) #=> 1[].first
only) #=> (Exception)[1].first
only) #=> 1[1,2].first
only) #=> (Exception)T.