ietf-nntp new draft of the NNTP spec released

[Russ Allbery]

Charles Lindsey <chl at clw.cs.man.ac.uk> writes:

> It seems to me that INN only allows the '!' at the start of a wildmat,
> and not, for example, comp.!sys.*. Is that correct?

Yes.

> However, the wording you quote above seems vague. I guess it means:
> 	"Start from the right. Stop when you match the group. If the place
> 	where you stopped has a '!', reject the group, otherwise accept."
> But that does not deal with the case where no match is found at all. If the
> pattern given is
> 	!comp.*
> then I would expect alt.foo to get through, but from what you say it seems
> not.

This varies for different parts of INN.  Some places prepend a logical *
to the beginning of the pattern; other places don't.

>>    See wildmat(5) for full details on wildmat patterns.

> Please could you post the text of that man page here.

Should have been wildmat(3).

                                  WILDMAT(3)
                             C Library Functions

NAME
     wildmat - perform shell-style wildcard matching

SYNOPSIS
     int
     wildmat(text, pattern)
         char       *text;
         char       *pattern;

DESCRIPTION
     Wildmat is part of libinn (3).  Wildmat compares the text against the
     pattern and returns non-zero if the pattern matches the text.  The
     pattern is interpreted according to rules similar to shell filename
     wildcards, and not as a full regular expression such as those handled by
     the grep(1) family of programs or the regex(3) or regexp(3) set of
     routines.

     The pattern is interpreted as follows:

     \x   Turns off the special meaning of x and matches it directly; this is
          used mostly before a question mark or asterisk, and is not special
          inside square brackets.

     ?    Matches any single character.

     *    Matches any sequence of zero or more characters.

     [x...y]
          Matches any single character specified by the set x...y.  A minus
          sign may be used to indicate a range of characters.  That is,
          [0-5abc] is a shorthand for [012345abc].  More than one range may
          appear inside a character set; [0-9a-zA-Z._] matches almost all of
          the legal characters for a host name.  The close bracket, ], may be
          used if it is the first character in the set.  The minus sign, -,
          may be used if it is either the first or last character in the set.

     [^x...y]
          This matches any character not in the set x...y, which is
          interpreted as described above.  For example, [^]-] matches any
          character other than a close bracket or minus sign.

HISTORY
     Written by Rich $alz <rsalz at uunet.uu.net> in 1986, and posted to Usenet
     several times since then, most notably in comp.sources.misc in March,
     1991.

     Lars Mathiesen <thorinn at diku.dk> enhanced the multi-asterisk failure mode
     in early 1991.

     Rich and Lars increased the efficiency of star patterns and reposted it
     to comp.sources.misc in April, 1991.

     Robert Elz <kre at munnari.oz.au> added minus sign and close bracket
     handling in June, 1991.

     This is revision 1.2, dated 1998/04/16.

SEE ALSO
     grep(1), regex(3), regexp(3).

-- 
Russ Allbery (rra at stanford.edu)         <URL:http://www.eyrie.org/~eagle/>