Thanks for the answer, I din’t know yet Transifex was also offering this feature.
Anyway we use CSV files that contains only one language and may have more than one columns to be translated, aside some other not-translatable columns (data):
1 | Acute Flaccid Paralysis .| Characterized by weakness or paralysis | ND | 0 | 1 | 1
2 | Cholera . . . . . . . . . . . . . | Infection of the small intestine . . . . . . . . | ND | 0 | 1 | 1
3 | Diarrhoea-Dysentry . . . .| Inflammatory disease of the colon . . . . . | ND | 2 | 1 | 1
In this case the “key” will be the first column, and the columns to be translated are second and third one, following ones are data needed to the software.
Another problem is that in the CVS we may have duplicated records, and this is because the application works like that (actually the “key” is the first column, not the English translation itself) so the XLSX has to be prepared before to upload on Transifex.
We could prepare one spreadsheet for each CVS with only one column to be translated, and then after the translations split the downloaded file into the original ones, but of course this would be a much longer procedure.
Anyway I will give it a try!