Clean and validate CSV data
Work from a safe copy, validate the structure, resolve repeated keys and review the result before conversion.
Begin with CSV Validator →1. Check the structure
Keep an untouched original. Open CSV Validator and run it on the input sample below. Expect 5 data rows and 3 columns. A structurally valid file can still contain repeated IDs or incorrect values.
2. Choose the duplicate policy
- Keep an untouched copy of your original CSV. Start with the synthetic sample below.
- Open CSV Duplicate Key Resolver and paste the sample into CSV.
- Set Key columns (comma-separated) to
id. Choose Keep first, then press Run. - Expect 2 duplicate groups, 2 removed rows and 3 remaining data rows. Compare the output with the example.
- Try Keep last and Remove all duplicate groups to see how the policy changes the result. Review the names before downloading.
Try a duplicate-key example
Five data rows contain two repeated IDs. The two rows for 101 have different names, so choosing a policy changes the information you keep.
Input CSV
id,name,team 101,Ada,Ops 101,Ada Lee,Ops 102,Bo,Data 103,Chen,Data 103,Chen,Data
| Policy, using key id | Rows kept | What happens |
|---|---|---|
| Keep first | 3 | 101 keeps Ada; 102 keeps Bo; 103 keeps Chen. |
| Keep last | 3 | 101 keeps Ada Lee; 102 keeps Bo; 103 keeps Chen. |
| Remove all duplicate groups | 1 | Only 102, Bo, Data remains. Both repeated IDs are removed completely. |
Choose a business key deliberately. Adding name to the key would make Ada and Ada Lee different keys. Blank values, spaces and letter case can also change which rows match. Review those fields before removing data.
3. Check the cleaned result
With Keep first, expect 3 data rows and 3 columns. Confirm that retaining Ada rather than Ada Lee matches your intended rule. Save the result separately from the original.
4. Convert only after review
Open CSV to JSON and paste the cleaned CSV. Expect an array of 3 objects with id, name and team properties. CSV values remain strings; confirm what the receiving system expects.
For your own file, inspect blank keys, whitespace and case before resolving duplicates. Review untrusted cells that spreadsheet software might interpret as formulas before opening a downloaded CSV.
Continue your workflow
Privacy
The linked tools process their working input locally in your browser. The copy buttons on this page copy only the displayed synthetic sample; they do not send it to a server. Normal website and network infrastructure may still process ordinary connection metadata needed to deliver and protect the site.