Re: [Technical Errata Reported] RFC7230 (4189)

[Willy Tarreau <[e-mail]>]

On Tue, Apr 21, 2015 at 10:24:47AM -0500, Zhong Yu wrote:
> Another question about obs-fold before we proceed with the formal
> definitions. Consider the following example
> 
> foo: bar<CRLF>
> <SP><CRLF>
> ...
> 
> It won't be surprising if some parser mistakes the 2nd line as an
> "empty line" that terminates the headers. Visually it *is* an empty
> line.
> 
> In spirit, obs-fold should be followed by visible chars, otherwise
> it's very confusing and problematic.

I disagree, a parser doesn't "see" characters, it consumes them. Here
you have a space after a CRLF, so it's a continuation of a folded header,
that's as simple as that. And it's important that it's properly defined
so that it's not abused by senders trying to put parsers in a situation
which is not well defined.

> RFC 822 $3.2 appears to suggest the same thing, that obs-fold can only
> appear between two non-empty segments.

And what is the parser supposed to do if it receives something which does
not match this rule ? That's always the problem when adding exceptions to
well-defined rules, it requires more work on the recipient side to properly
handle the situation. In short, it *adds* more risks of confusion.

Regards,
Willy