◇ spreadsheet in, TMX out

Convert Excel or CSV to TMX

Clients send glossaries and legacy translations as spreadsheets; CAT tools want TMX. Converting by hand loses accents to the wrong encoding, and most converters drop the extra columns — client, product, notes — that you'd want to filter on later.

━━ the problem

which column is which language?

A first row reading "English (US)" and "French (Canada)" is obvious to you and opaque to a converter. Without real codes, the TMX imports as the wrong pair or not at all.

the accents break

A CSV saved from Excel on Windows is often not UTF-8. Read as UTF-8, café becomes café; read the other way, it's the other columns that break.

the context disappears

Columns like Client, Product or Notes are what make a TM filterable later. Most quick converters keep two columns and throw the rest away.

━━ how TM Cleaner handles it

Drop the .xlsx, .csv or .tsv. Languages come from the header row (codes like en-US, or names like English (US)), or you type them in the upload form. Other named columns become <prop> entries, the encoding is detected, and the result goes through the same repair and preview as any TMX. “Convert only” keeps every row.

━━ frequently asked

what layout does my spreadsheet need?+

one row per translation pair and one column per language. put the language codes or names in the first row, or leave them out and choose the two languages in the upload form (column A is then the source).

can I convert several spreadsheets at once?+

yes: drop them together, or in a zip. they're merged into one TMX.

will empty or duplicate rows be removed?+

with “Convert only”, no: every row with source text is kept (a row with no source can't be a translation unit). pick “Clean it” instead to remove empty translations and duplicates while converting.

━━ related pages

clean your tmx now →

free · no account up to 50 MB · files up to 2 GB · try the demo