[Exim] String expansion: repeat a character sequence n times

Startseite
Nachricht löschen
Nachricht beantworten
Autor: Sheldon Hearn
Datum:  
To: exim-users
Betreff: [Exim] String expansion: repeat a character sequence n times
Hi folks,

The short question:
===================

I'm looking for a string expansion item in Exim that will allow me to
repeat a character sequence n times.

For example, I'd like something like this

    ${repeat{\\+}{3}}


to produce this:

    \+\+\+


"Does such a thing exist?" he asked, somewhat hopefully. I've read
"String expansions" through and can't find it, but perhaps it's a new
feature that's arrived since 4.20?

The reason:
===========

I have a transport that is only used for certain local recipients:

remote_smtp_subject_spam:
  driver = smtp
  no_delay_after_cutoff
  headers_remove = \
    ${if \
      match {$h_X-Spam-Score:}{\N\+\+\+\+\+\N} \
      {subject} {} \
    }
  headers_add = \
    ${if \
      match {$h_X-Spam-Score:}{\N\+\+\+\+\+\N} \
      {Subject: ***SPAM*** $h_Subject:} {} \
    }


This tags the Subject line for messages with a spam score of 5 or
greater. I have customers who only allow filtering for some recipients,
and whose staff can't filter on the X-Spam-Score header.

But now I want the threshold to be a per-user value, obtained from a
database.

Sure, I could store '\+\+\+' in the database, but that feels very, very
ugly. So I'd like to be able to do something like this:

MAILBOX_SPAMTAG_THRESH= \
  ${lookup \
    pgsql{ \
      SELECT SpamTagScore \
    FROM Mailbox \
       WHERE Name = '${quote_pgsql:$local_part}' \
    } \
    {$value} \
    {no} \
  }


remote_smtp_subject_spam:
  driver = smtp
  no_delay_after_cutoff
  headers_remove = \
    ${if \
      match {$h_X-Spam-Score:}{${repeat{\\\\+}{MAILBOX_SPAMTAG_THRESH}}} \
      {subject} {} \
    }
  headers_add = \
    ${if \
      match {$h_X-Spam-Score:}{${repeat{\\\\+}{MAILBOX_SPAMTAG_THRESH}}} \
      {Subject: ***SPAM*** $h_Subject:} {} \
    }


Hope springs eternal. :-)

Ciao,
Sheldon.