[pycrypto] Crypto.PublicKey.RSA and number.py
mads at kiilerich.com
Thu Mar 12 20:33:59 CST 2009
Anthony Honstain wrote, On 03/12/2009 06:47 AM:
> I am working towards using the Crypto.PublicKey.RSA encryption and
> eventually padding like PKCS#1. I have observed that the leading
> '\000' are removed. Is this the desired behavior of bytes_to_long and
> long_to_bytes? Or is this not an issue once a good padding scheme is
Yes, kind of.
RSA basically uses big numbers. The numbers are often encoded as byte
strings, and any (not too long) string can be interpreted as a number.
But just like in decimal numbers the leading zeroes doesn't matter; you
can't see from the value how many leading zeroes it was written with,
and an arbitrary number of leading zeroes can be added without changing
Yes, a good padding scheme will "fix" that; not so much in order to
preserve a string length, but more in order to ensure that the number is
big enough to preserve the assumed computational complexity of RSA.
-------------- next part --------------
A non-text attachment was scrubbed...
Size: 3435 bytes
Desc: S/MIME Cryptographic Signature
Url : http://lists.dlitz.net/pipermail/pycrypto/attachments/20090313/a0d0891b/attachment.bin
More information about the pycrypto