So I have a setup whereby content is generated from an Excel sheet, which generates a CSV files in various languages, which then lets me select language specific files from within the ID file itself. The CSV generator basically removes any problematic content for the ID labels, and then just spits out the various content flavours.
This process has been working fine for the past couple of years, with the exception of a couple of files which started acting up back in the summer. They worked fine for a brief while and then stopped working alltogether. The designers managed to create merged versions from the file, but if I try to select a CSV and remap the content it causes ID to crash out.
Have tried exporting the file to an idml and then resaving, but still does the same.
Other automated ID files are able to use the CSV content associated with the problem file, and any CSV selected from the problem file will cause the same behaviour.
The file lives on a network drive & is accessed by colleagues via a VPN, but moving the problem file locally makes no difference (and the other 140+ work fine).
Is the csv file saved with UTF-8 encoding?
Saved as Unicode (format 42) from Excel, which uses UCS2-Little Endian encoding.
While I do not have programing knowledge of font encoding methodology, here is some basic information about UTF-8 and UTF-16. Each of these encodings can encode the same amount of information, just different procedures to identify characters. Endianness then adds an additional layer of complexity in regard to assigning and accessing byte order.
My experience with ID data merge using (csv and txt) and (UTF-8 and UTF-16) specifically with accommodating cyrillic and arabic characters, UTF-8 has worked more consistently over UTF-16.