How to use this tool
- Load a CSV and select its parsing options.
- Load the column controls, then uncheck unwanted fields, edit header names or use the move buttons.
- Apply the changes and inspect the output schema before downloading.
How the operation works
Column selection uses original field positions, not guesses based on content. Reordering moves every row’s matching field with its header. Output names must be nonempty and unique. The resulting file uses the selected input delimiter and preserves all accepted records.
Editing schema without touching values
Column selection uses original field positions, not guesses based on content. Unchecking a column drops that field from every record; values in remaining columns are unchanged. Renaming a header changes only the header string—amount_usd → amount_eur does not convert currency.
Reordering moves every row’s matching field with its header. Output names must be nonempty and unique; duplicate or blank names are rejected rather than silently overwritten.
Keep, rename and reorder
Input: id,name,city 00123,Ada,Lahore Keep name then id; rename name → full_name: full_name,id Ada,00123
The city column is dropped. Leading zeros in id survive because values are never parsed as numbers. Reload the column controls after changing the input file so the control list matches the new schema.
Before merge and import
Typical use: two exports have the same data in different column orders or with different header spellings. Align both with this tool, then merge. For a narrower import schema, uncheck unused fields after alignment. For structural checks first, open the file in CSV Viewer.
After editing, inspect the exported header order against the receiving application’s documented columns before importing a large batch.
Common mistakes
Renaming headers to “fix” types does not transform values. Dropping a key column before deduplication removes the ability to deduplicate on it later—reorder operations so keys are still present when needed. After changing the input, always reload controls; stale column lists apply edits to the wrong positions.
A table with no columns cannot produce useful CSV. Keep at least one field.
Limits
5 MiB per input, 50,000 records, 200 columns. The selected input delimiter is used for output. Processing is local; files are not uploaded to a conversion service.
A worked example
id,name,city 00123,Ada,Lahore
Keep name then id → name,id Ada,00123
Renaming a header does not transform the field values underneath it. For example, changing amount_usd to amount_eur does not convert currency. After changing the input, reload the column controls. Check the exported header order against the receiving application before importing.
Schema changes without changing values
Column selection uses the original field positions, not header text or content guesses. Keep or remove a column, rename its header, and move it left or right; every row matching field moves with the header. Output header names must be nonempty and unique. At least one column must remain, because a table with no columns cannot produce a useful CSV.
Renaming a header does not transform the values underneath it. Changing amount_usd to amount_eur does not convert currency, and changing id to customer_id does not create a new identifier. If the input text changes, reload the column controls before applying another change.
Preparing a file for another application
Match the destination expected header names, order and delimiter. A database import or API may require exact names and a fixed order; a spreadsheet is more forgiving but may re-type identifiers on open. Test with a small file before processing a full export, and keep the original so the schema change can be reversed.
Use the same delimiter as the receiving system and inspect the output header row and a few data rows. Column editing is structural: it does not validate emails, dates, currency or postal codes.
Processing and privacy
CSV Column Editor rewrites headers and column order locally; the table is not uploaded. Input stays in browser memory until you reset or leave the page. Copy and download run only when you choose them. No tool input is sent to a processing service. Website pages and scripts still make ordinary hosting requests; the privacy policy separates those from tool processing.
Frequently asked questions
Can I remove every column?
No. A table with no columns cannot produce a useful CSV output. Keep at least one.
Does renaming a header change the data?
No. Only the header string changes. Values underneath remain exactly as loaded.
Why must headers be unique?
Duplicate names would overwrite fields in JSON conversion and confuse importers. The tool rejects them.
Can I create new computed columns?
No. This tool selects, renames and reorders existing fields only.
Does it preserve leading zeros?
Yes. Field values are kept as text; no numeric coercion is applied.
Does renaming a column convert its values?
No. Only the header changes. Convert units or formats separately with the appropriate tool or script.
How this tool was checked
CSV Column Editor was checked with keep, rename and reorder operations, duplicate header names, every-column-removed guards, delimiter preservation and multi-row data. See the testing method or report a failure.