CSV files must contain at least three columns. One of the columns has the column header "id", and contains a unique ID for each CSV line. Another column has the column header "source" and contains the source language. All other columns with column headers are treated as the target language columns and imported. In this case, the column header name is arbitrary. Columns without a filled column header in the CSV file are ignored during import.
Column separator of CSV is a comma "," text delimiter is a quotation marks (").
Line separator of CSV is a Linux, Mac or Windows newline.
Encoding of the CSV file can utf-8, iso-8859-1 or windows-1252. The CSV file is exported en the same encoding as it existed during import (Note: For a translation into non-Western European languages therefore utf-8 should already be selected as the import encoding).
Overwriting of format specifications
The column header name of the ID column (by default mid) and the source language column (by default quelle) as well as the column separator and the text delimiter can be overridden in the installation-specific system configuration translate5. Search in your SQL database of translate5 with the following statement to find the csv-related configs:
SELECT * FROM `Zf_configuration` WHERE `name` LIKE '%csv%'