Prepare CSV encoding and delimiters for Tableau Prep

Encoding and delimiter are separate decisions

A Tableau Prep source can be text-delimited even when its extension is .csv. First identify the list separator used by the producing system; then choose the character set that renders the values correctly. Changing only one can leave the flow with shifted columns or mojibake.

• Columns are all in one field: select the actual delimiter, such as semicolon or tab.
• Names are garbled: set the source encoding explicitly before choosing UTF-8 output.
• Rows split unexpectedly: diagnose quoted commas and embedded line breaks before conversion.

• Convert the CSV encoding or delimiter →Set source and output formats explicitly and create a separate CSV.
• Diagnose CSV column counts and quoted fields →Review structural findings before changing a value or header.
• Check the CSV character encoding →Choose the source encoding from the preview instead of guessing from the country.

Use the source application and Tableau preview together

Record the producing application's locale, separator, and encoding. In Tableau Prep, confirm the resulting field names and data types; automatic type inference is a separate decision from the bytes in the CSV.

• A German or French export may use semicolons because commas represent decimals.
• A tab-delimited text file should be treated as structured text, not repaired by replacing tabs inside values.
• Do not overwrite a trusted source workbook while testing a new encoding.

• Diagnose CSV column counts and quoted fields →Review structural findings before changing a value or header.
• Check the CSV character encoding →Choose the source encoding from the preview instead of guessing from the country.
• Split a large CSV into traceable batches →Keep the header and reconcile every output row.

Convert the file for a repeatable Prep flow

LocaCSV changes the selected representation, not the meaning of measures, dates, or dimensions. Keep data type decisions visible in the Prep flow.

1. 1Open a copy of the export in LocaCSV's format conversion workspace.
2. 2Select the source encoding and delimiter that produce the expected preview.
3. 3Choose the UTF-8 or destination encoding and delimiter required by the flow, then save a separate CSV.
4. 4Load the converted file in Tableau Prep and save the input step only after its fields and types look correct.

Verify the first refresh before scaling the flow

• The field count and header names match the previous Prep input.
• Japanese, accents, and identifiers render correctly in the input preview.
• The row count before and after conversion is the same.
• A sample flow output has the expected dates, numeric values, and null handling.

• Convert the CSV encoding or delimiter →Set source and output formats explicitly and create a separate CSV.
• Diagnose CSV column counts and quoted fields →Review structural findings before changing a value or header.
• Append monthly CSV files with the same header →Check column order and file order before a vertical merge.

If the flow still changes the data, inspect types

A valid CSV can still be interpreted as a date, number, or null by Tableau Prep. Compare the input step's type settings and locale before changing the file again. If the source is monthly, keep identical headers before appending it.

• Append monthly CSV files with the same header →Check column order and file order before a vertical merge.
• Split a large CSV into traceable batches →Keep the header and reconcile every output row.
• Match a Snowflake file format after conversion →

Give Tableau Prep a stable source

Store the source settings beside the converted file so the next export can follow the same encoding and delimiter contract.