3Sum dedup.

Why the duplicate-skip works — step through both algorithms side by side.

INTERACTIVE · CLICK NEXT

LC 15

Sorted input

Anchor (this round)

i = 0 · value -1

Inner two-pointer searches L+R = 1 (to cancel the -1 anchor).

Goal

find all unique triplets

anchor (i)

L pointer

R pointer

match found

duplicate

skip in progress

Step 0 / 0

Without skip.

naive · finds duplicates

step

—

L · R

—

sum

—

Press Next → to begin.

triplets recorded

— none yet —

With skip.

canonical · no duplicates

step

—

L · R

—

sum

—

Press Next → to begin.

triplets recorded

— none yet —

The takeaway.

Without skip: when the two pointers land on duplicate values after a match, they emit the same triplet again. The set silently drops it, but the CPU still did the work.

With skip: after a match, we walk L past any equal neighbors (look-ahead: nums[L] == nums[L+1]) and walk R past its equal neighbors (look-behind: nums[R] == nums[R-1]). Then we advance both into fresh territory.

Net effect: no duplicate triplet is ever even reached. No set needed. Result is deterministic in order.

      SEE ALSO →
      Flashcards · two-pointers category ·
      Cheatsheet · pattern 2 ·
      Practice script · LC 15 ·
      LC 16
    