Deduplication — unique / squish / repeated

Drop duplicates globally (unique), only when adjacent (squish), or find them (repeated).

Three methods handle repeated elements differently: unique removes all duplicates, squish collapses only adjacent runs, and repeated returns the duplicates themselves.

squish — collapse adjacent runs #

squish removes a duplicate only when it directly follows its twin, so a value can reappear later.

say (1, 1, 2, 2, 3, 1).squish;
say (1, 2, 2, 3, 3, 3).repeated;

(1 2 3 1)
(2 3 3)

The trailing 1 survives squish (it isn't adjacent to the first 1); repeated returns each element that has appeared before.

unique by a key #

unique drops all repeats; :as compares by a computed key instead of the value itself — here, by absolute value.

say (1, -1, 2, -2, 3).unique(:as(*.abs));

(1 2 3)

-1 is dropped as a duplicate of 1 because their .abs match.

Notes #

• unique is O(n) with a seen-set and keeps first-seen order; squish is streaming and only remembers the previous element — cheaper, but only for sorted or grouped data.
• repeated is the complement of unique: together they partition a list into first-occurrences and later-occurrences.
• Sort first if you want squish to behave like unique: @a.sort.squish.