| |
 | Posted by Peter Alexander |  Permalink Reply |
|
Peter Alexander 
| http://d.puremagic.com/issues/show_bug.cgi?id=11097
Summary: Add version of std.algorithm.group that returns group
ranges
Product: D
Version: D2
Platform: All
OS/Version: All
Status: NEW
Severity: enhancement
Priority: P2
Component: Phobos
AssignedTo: nobody@puremagic.com
ReportedBy: peter.alexander.au@gmail.com
--- Comment #0 from Peter Alexander <peter.alexander.au@gmail.com> 2013-09-22 05:58:50 PDT ---
std.algorithm.group returns a range of (elem, count) tuples:
int[] arr = [ 1, 2, 2, 2, 2, 3, 4, 4, 4, 5 ];
assert(equal(group(arr), [ tuple(1, 1u), tuple(2, 4u), tuple(3, 1u),
tuple(4, 3u), tuple(5, 1u) ][]));
This is fine when the predicate is equality, but when the predicate is something else, the tuple is less useful. Here's an example of grouping strings by first character:
string[] arr = [ "Alice", "Andrew", "Ben", "Bob" ];
assert(equal(group!("a.front == b.front")(arr),
[ tuple("Alice", 2), tuple("Ben", 2) ]));
This isn't very useful because there aren't two Alice's and two Ben's. Alice and Ben are just one element from the group (btw, the documentation doesn't indicate that it is always the first element in the group that is returned).
It would be nice if there was a version of the algorithm that returned the groups themselves, working like this:
string[] arr = [ "Alice", "Andrew", "Ben", "Bob" ];
assert(equal(groups!("a.front == b.front")(arr),
[ ["Alice", "Andrew"], ["Ben", "Bob"] ]));
I have used the identifier "groups" here. I'm not bothered what it is called.
Once implemented, group may be elegantly implemented in terms of groups:
auto group(alias f, R)(R r)
{
return groups!(f)(r).map!(g => tuple(g.front, g.walkLength));
}
--
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
|
