Clean names copied from a spreadsheet: dedupe, trim, sort
Paste a column of names copied from a spreadsheet or web page into the left pane, e.g. Zhang San
Li Si
Zhang San
Alice
alice . Spreadsheet cells often carry leading/trailing spaces, CRLF line endings and duplicates.
Step 1: choose an ordered pipeline. Enter trim-lines,deduplicate,sort-lines in "Step order". Trimming first lets " Zhang San " merge with "Zhang San"; deduplicating first keeps them as distinct lines. Changing the order changes the result predictably — that is by design.
Step 2: set dedup switches. Default comparison is case-sensitive and keeps the first occurrence, so Alice and alice both survive; enable "ignore case" to keep only the first Alice. Blank lines are kept (repeated blanks collapse to one) unless "drop blank lines" is checked.
Step 3: choose a sort mode. Lexicographic sorting defaults to locale=en with Intl.Collator string comparison; codepoint gives deterministic order independent of the host runtime. For all-numeric lines choose numeric mode: 2 / 10 / 1 becomes 1 / 2 / 10, while lexicographic mode gives 1 / 10 / 2. The result metadata names the mode and locale.
Step 4: review and export. The input’s dominant EOL (CRLF or LF) and trailing-newline state are preserved by default; emoji and combining marks are never silently rewritten. After Run, check the right pane, then Copy or Download. Downloads are UTF-8 with no BOM by default (Chinese stays intact) and are named text-deduplicate-<timestamp>.txt; an explicit BOM option exists for legacy Notepad. Parameter changes only refresh the preview — they never overwrite the left pane — and Clear Result clears only the right pane. Inputs over 2 MiB / 100 000 lines are rejected before processing with a readable error, and Cancel works mid-run.
Also useful for batch name lists and copied-table cleanup. This site does no AI rewriting and no subtitle timecodes; see the related-site configuration for those (config only until those sites are live).