New Forum | Previous | Next | (P-PDF) Developers
Conf: (P-PDF) Developers, Msg: 94258
Date: 8/13/2003 07:22 PM
I develop and market a PDF generator tool which runs on an IBM iSeries. This presents some special challenges because the iSeries is an EBCDIC machine.
I am currently developing a version of my product which supports non-Latin alphabets (Greek, Cyrillic, also Chinese, Japanese, Korean). I have a question regarding PDF passwords.
I would like to make it possible for users to supply a password in their local language (Greek, Chinese etc.) when creating the PDF file using my tool. The password string they supply to my tool will be in EBCDIC. Normally I convert this to ASCII before including it in the PDF, but there are some extra complications when dealing with non-Latin alphabets.
Ideally therefore I would like to handle the password as a Unicode string. Is this possible? Can a PDF password be stored in the encryption dictionary as a Unicode string? When Acrobat checks the password that has been input by the user who is trying to open the file, will it convert the string to Unicode?
The PDF reference manual indicates that passwords should be a maximum of 32-bytes long. *If* the password can be in Unicode, does this mean that the password will be limited to 15 characters (i.e. 15 two-byte characters plus the 2-byte Unicode x'FEFF' marker at the front)?
Any help gratefully received by a currently rather confused developer?