v2.1 uses a different folding mechanism to v3.0 and v4.0.
Currently the same mechanism is used as for v3.0 and v4.0.
The docs (vCard 2.1 specs (section 2.1.3), RFC 822#3.1.1) are somewhat confusing (notably, the examples do not match the description), but they say "Unfolding is accomplished by regarding CRLF immediately followed by a LWSP-char as equivalent to the LWSP-char." (LWSP-char is SPACE or HTAB.)
Thus, what I would expect is that unfolding translates \r\n[ \t] to a single SPACE or HTAB, respectively.
W.r.t. folding, the best bet is likely not to fold at all, given that the v2.1 specs do not seem to say anywhere that folding SHOULD or MUST be done. Otherwise, folding should probably only happen on SPACE or HTAB.
v2.1 uses a different folding mechanism to v3.0 and v4.0.
Currently the same mechanism is used as for v3.0 and v4.0.
The docs (vCard 2.1 specs (section 2.1.3), RFC 822#3.1.1) are somewhat confusing (notably, the examples do not match the description), but they say "Unfolding is accomplished by regarding CRLF immediately followed by a LWSP-char as equivalent to the LWSP-char." (
LWSP-charisSPACEorHTAB.)Thus, what I would expect is that unfolding translates
\r\n[ \t]to a singleSPACEorHTAB, respectively.W.r.t. folding, the best bet is likely not to fold at all, given that the v2.1 specs do not seem to say anywhere that folding SHOULD or MUST be done. Otherwise, folding should probably only happen on
SPACEorHTAB.