Re: empty quoted strings and other oddities

[Bruce Lilly <[e-mail]>]

Jacob Palme wrote:

> Should you for example accept
> From: John F. Kennedy <[e-mail]>
> 
> Standards have a long time accepted the "K." even
> if the user-friendly name is not quoted, but implementors
> have not been told that accepting such a "." followed
> by space is in the liberal category of what you should
> accept but not produce.

That is in the obs-phrase syntax of RFC 2822.  It is a
bit unusual compared to most of the rest of the obs-
syntax; it was *not* allowed under 822's rules and is
still a MUST NOT generate case, but there's a note in
section 4.1 that implies that it may be formally permitted
in the future.  And of course, 2822-conforming implementations
of receivers MUST recognize it as a valid phrase.

IMO, formally permitting generation would be good.  While

   Wilt "The Stilt" Chamberlain

is a valid phrase and is as would normally be written in
text not neccessarily specific to header formats,

   Bullwinkle J"." Moose
   Bullwinkle "J." Moose
   Bullwinkle "J, Moose"
   "Bullwinkle J." Moose
   "Bullwinkle J. Moose"
   "Bullwinkle" "J""." "Moose"
   ...
   Bullwinkle =?us-ascii*en-us?Q?J=2E?= Moose

are valid phrases under the RFC 2822 generate syntax rules
but are unlikely to be typed as such by a user. In the
first case the double quotes are part of the name, whereas
in the variants of the second example which use double quotes
they are mere syntactic sugar coating [the 2047/2231 encoded
variant is more like syntactic scaffolding in the Gothic
tradition].

I.e. permitting a phrase with an unquoted dot would be
a recognition of existing practice which serves user desires.
Unlike some of the relaxations of syntax permitted by 2822,
such as the oxymoronic but 2822-valid (under obs- rules)
time-stamp line

   Received:

That was not legal under 821/822 syntax, and there's no point
in a time-stamp line with no time stamp.