> > For now the ASN.1 string is not an opaque structure.
>
> Considering how OpenSSL design evolved, I suspect that the time may come
> when this string would become opaque. Therefore, I suggest that
> getter/setter functions should be added.
Any thoughts about the interface?
Alas, since I don't have any experience actually using these flags, I don't consider myself competent suggesting the API here.
Some flag bits look internal, and not necessarily ideal for applications to modify directly, so
while these might do:
long ASN1_STRING_get_flags(ASN1_STRING *);
void ASN1_STRING_set_flags(ASN1_STRING *, long);
This makes sense. Also, if there are flags that users are likely to need - perhaps explicit getter/setter calls for those individual flags? This would lower the likelihood of erroneously affecting unintended flags by a "ricochet"...
Perhaps all that's needed is:
int ASN1_BIT_STRING_set_unused(ASN1_STRING *, int);
which sets the unused bit count, checking that the ASN1_STRING type
is V_ASN1_BIT_STRING and the count is in the range [0,7].
That I don't know. I can't figure how or why I would use ..._set_unused()... But as I said, my competence in this particular thing is low.
Attachment:
smime.p7s
Description: S/MIME cryptographic signature
