On Thu, 16 Jan 2020 at 15:37, Wietse Venema <wie...@porcupine.org> wrote:
> Dominic Raferd:
> > On Thu, 16 Jan 2020 at 14:34, Wietse Venema <wie...@porcupine.org>
> wrote:
> >
> > > Dominic Raferd:
> > > > Thanks Christian that was very helpful. I have it working now for
> > > > postscreen and I think (but am waiting for an incoming instance) for
> > > > smtpd. Weird
> > > > that they have such different approaches (postscreen_dnsbl_reply_map
> and
> > > > rbl_reply_maps). And I could not find a way to use pcre with
> > > rbl_reply_maps
> > > > because it throws a warning if I reference any variables such as
> > > $rbl_code
> > > > - but such variables do seem to work in a hash file.
> > >
> > > Use $$name instead of $name.
> > >
> > > As documented:
> > >
> > > TEXT SUBSTITUTION
> > > Substitution of substrings (text that matches patterns inside
> > > "()")
> > > from the matched expression into the result string is
> requested
> > > with
> > > $1, $2, etc.; specify $$ to produce a $ character as
> output.
> > > The
> > > macros in the result string may need to be written as ${n} or
> > > $(n) if
> > > they aren't followed by whitespace.
> > >
> >
> > Thanks Wietse. I had read that but I interpreted the text 'specify $$ to
> > produce a $ character as output' as meaning that $$ would produce a
> > hard-coded dollar sign, not a sign that could then be re-interpreted as
> the
> > start of a variable. Perhaps the text could clarify this?
>
> Oh ye of little faith. You can test my suggestion with the postmap command.
>
> postmap -q 'search strinng here' pcre:/path/to/file
>
Before I saw through a glass darkly. Now I'm a believer.
