Sorting and deduping lines

Published: 2026-09-05

How to reorder multi-line lists and remove duplicates—locale sort, numeric digit order, case folding, keep-first vs sort-unique, reverse, and shuffle—without uploading the paste.

Pastes of emails, IDs, tags, log hosts, or spreadsheet columns often arrive as one value per line. Two jobs come up constantly: put those lines in a stable order, and drop repeats. Spreadsheet sorts and shell pipelines (sort, uniq) work, but for a quick browser pass—especially when the list is sensitive—local tools keep the text in the tab.

LocalTools ships Sort lines and Dedupe lines. Both split on JavaScript newline rules (LF, CR, or CRLF), preserve a trailing newline when present, and never upload the paste.

Sort vs dedupe (and when to chain them)

Goal	Prefer	Notes
Alphabetical or reverse order	Sort lines	Ascending / descending with locale rules
Flip current order only	Sort → Reverse	Does not re-alphabetize
Randomize for sampling	Sort → Shuffle	New permutation each shuffle
Drop repeats, keep encounter order	Dedupe → Keep first occurrence	First spelling wins
Unique set, then A→Z	Dedupe → Sort unique	Dedupe then locale sort

Typical cleanup: dedupe first (or use sort-unique), then sort if you still need a different order. Sorting alone does not remove duplicates—adjacent equals stay unless you dedupe.

What “sorted” means: locale, case, and digits

Browsers order strings with Unicode collation for the current locale, not a simple ASCII byte chart. That affects accents (é vs e) and language-specific letter order.

For A→Z / Z→A (and for dedupe’s sorted mode), LocalTools exposes:

Case-insensitive (accent-aware) — usual default for human lists; Apple and apple group together when sorting, and count as the same line when deduping (the first spelling is kept).
Case-sensitive — exact comparison; useful when Prod and prod must stay distinct.
Lexicographic vs numeric digits — lexicographic puts 10 before 2 ("1" < "2"). Numeric digit ordering puts 2 before 10, which matches how people expect version fragments and numbered labels to read.

Reverse and shuffle ignore comparison options: reverse flips the current sequence; shuffle uses Web Crypto randomness per action (use Shuffle again for another permutation).

Dedupe details that surprise people

Empty lines count like any other line—two blank rows are duplicates of each other.
Leading/trailing spaces are part of the line. "alpha " and "alpha" are different unless you normalize first with Whitespace normalizer.
Case-insensitive match folds for comparison only; output keeps the first representative’s casing and accents.
“Repeated occurrences removed” counts extras after the first keep—not how many distinct keys you had.

If you need fuzzy near-duplicates (typos), line equality is the wrong tool; see Levenshtein distance later, or edit with Find & replace (batch).

Practical workflow

Paste one item per line (exports from sheets, grep output, CMS tag lists).
If whitespace is noisy, normalize CRLF/tabs/trailing spaces first.
Dedupe: keep first occurrence for “stable list order,” or sort unique for a clean glossary.
Sort when you need A→Z, Z→A, reverse, or a random sample order.
Copy the output; check line counts under the boxes when you care how many rows dropped.

For measuring length after cleanup, use Text statistics. For turning cleaned titles into URL paths, see slug rules.

Why sort and dedupe locally

Customer emails, internal hostnames, and unreleased feature flags should not ride through a remote “list cleaner.” Both tools reorder and filter entirely in the browser—same local-only model as the rest of LocalTools.

Try it locally

Sort: open Sort lines, paste, choose A→Z / Z→A / reverse / shuffle, set case and digit options for sort modes, then copy.

Dedupe: open Dedupe lines, paste, choose keep-first or sort-unique, set match sensitivity (and digit order when sorting), then copy.