RFC 822 mentions backslash-quoted CRLF in sections 3.4.3 (comments)
and 3.4.5 (quoted-strings). While the EBNF for quoted-pair in 822
permits only a single octet after the backslash, it is clear from
the normative text in the above cited sections that the interpretation
of a backslash followed by CR followed by LF in the respective
contexts is interpreted as a quoted CRLF pair.
RFC 2822 does not mention backslash-quoted CRLF at all. The ABNF
for quoted-pair and obs-qp each allow for only a single octet
following the backslash. Note that 2822 doesn't mention backslash-
quoted CRLF in section 4 (obsolete syntax) or in Appendix B (summary
of changes).
It appears that a backslash-quoted CRLF would be interpreted by an
RFC 2822 parser as a (deprecated) backslash-quoted CR followed by
a (deprecated) lone LF.
That's a different semantic interpretation from 822.
Given the lack of mention in section 4 and Appendix B, my first
question is: is the intent of 2822 to prohibit generation of
backslash-quoted CRLF as described in RFC 822?
Second question: is the different semantic interpretation
by 2822 parsers intentional?
Best regards,
Bruce Lilly