From a messy CSV to clean data

An export from one system needs to go into another, and the import fails, or the names come out as Müller, or the file is too big for the upload form. These tools check, fix and reshape CSV files in your browser, so the data isn’t uploaded anywhere.

1. Find the broken rows

Drop the file on the Data Validator. It checks the CSV against the standard (RFC 4180) and lists every problem with its line: a row with more or fewer fields than the header, or a quote that isn’t closed. Click a problem to jump to it. A stray comma inside an unquoted field is the usual cause.

2. Fix the encoding

If you have the spreadsheet itself (.xlsx), skip this step: open it in CSV ⇄ JSON, which reads the sheet as text with nothing garbled. If accented names are garbled, or Excel shows them wrongly, drop the file on the Encoding Fixer. It detects the file’s encoding, repairs text that was garbled on the way, and saves a UTF-8 copy. Tick Start saved files with a UTF-8 BOM for a file that Excel will open. The line endings can be made all LF or all CRLF there too, if the importing system is fussy.

3. Look through it

Paste it into CSV ⇄ JSON and choose Show as a table. Sort by a column to bring blanks and odd values to the top or bottom, and use Filter the table to find particular rows. The delimiter (comma, tab, semicolon or pipe) is detected for you.

For an API or a script, CSV → JSON turns each row into an object keyed by the column names. Values stay as text, so 34 becomes "34". To SQL writes the statements that create a table and load the rows into a database, each column typed from its values.

4. Split or join large files

When an import takes only so many rows at a time, set Rows per file in CSV Split & Join and Split: you get a ZIP of numbered files, each with the header. Join goes the other way, and refuses files whose headers don’t match, naming the one that differs.

5. Chart it

For a quick picture of the result, paste two or more columns (a label and a value, or a header row and several series) into the Chart Maker. Choose the chart type and colours, and download it as PNG or SVG.