ietf-822
[Top] [All Lists]

Re: Revisiting RFC 2822 grammar (obs-utext and unstructured)

2004-02-05 07:18:56

Charles Lindsey wrote:

Oh Dear! You misread what I wrote. When I said "*text", I didn't meant
"*text", I meant the set of rules named 'utext', 'ctext', 'dtext', etc"
which all have the property that they produce just a single character.
Your latest offering seemed to be breaking with that convention.

ctext, dtext, qtext, and text are single-octet productions (or should be).
In 2822, obs-text and obs-utext are not.

obs-text only appears in text (where obs-char makes more sense) and
obs-utext (which is defined as obs-text). obs-utext appears only in utext,
and utext appears only in unstructured.

I'll take another look at the unstructured and related productions.


What I would like him to do would be
    Subject: Re: foo
but the proper way to address that issue would be to establish some rule
or convention about whether subjects could be refolded or have WSP
collapsed in the course of generating a followup/reply.

But following a strict reading of 3.6.5 in RFC 2822, I would argue that
the only compliant way would be
    Subject:Re:           foo
from which it is evident that every MUA implementation known to me is
non-compliant with RFC 2822 :-( . Another bug for Pete to worry over...

The main point is that any such "rule or convention" imposes structure
and that is clearly not compatible with the description and history of
the Subject field being unstructured.