Table cleaner workspace
Your input is processed only in this browser. See the privacy policy.
How to use it
- Copy a table with headers from Excel or Google Sheets and paste it. CSV, Markdown tables and tables copied from web pages work too, and the format is detected automatically.
- Check the grid below to confirm every cell landed in the right column. Use the ▾ menu on each header to rename, move, delete, trim, change case or split a column, and pick a phone, date, address, pseudonymization or duplicate rule if needed.
- Switch between Original and Cleaned to review the highlighted changed cells and warnings, then copy as TSV for Excel, CSV, Markdown or JSON, or download the CSV and change log. Press Undo to step back.
Examples
Find equivalent phone numbers
Normalize 010 1234 5678 and +82 10-1234-5678 to one Korean format and flag the second as a duplicate candidate.
- Input
name,phone\nMin,010 1234 5678\nJin,+82 10-1234-5678
- Result
Both become 010-1234-5678; the second row is flagged
Keep the newest customer row
Use email as the key, keep the row with the newest date, and record every excluded row in the audit CSV.
Details
Table Cleaner combines phone, date, address, pseudonymization and conditional duplicate rules in one workspace. It uses the same quote-aware CSV parser as the CSV fixer, so delimiters and line breaks inside quoted cells remain cells and identifier-like values keep their leading zeros.
A date such as 03/04/2026 is inherently ambiguous when both numbers are at most 12. The tool follows your selected month/day order and leaves a warning for review. Address cleanup only normalizes spaces and punctuation or proposes a comma-based split; it does not verify that an address exists or is deliverable.
Pseudonymization creates consistent replacements or masks in this session, but it is not a guarantee of anonymization. People may still be identifiable from other columns or rare combinations. Review the whole dataset before sharing it. Input and output are neither uploaded nor saved to browser storage.
Frequently asked questions
Is my file uploaded?
No. Reading, transformation, preview and downloads happen only in the current browser tab.
Can I paste a table from a web page or a Markdown document?
Yes. A table copied from a web page is read by its HTML cell boundaries, including merged cells, and a Markdown table is parsed with its alignment row and \| escapes. Only the text is extracted from pasted HTML; it is never inserted into the page or run.
Does address cleanup validate addresses?
No. It applies spacing and punctuation rules and can split on commas; it does not check existence, postal codes or deliverability.
Does pseudonymization fully anonymize data?
No. It masks selected direct identifiers.
Can I audit removed duplicates?
Yes. The change log CSV contains the original row number, values and removal reason.