1#! PERL_COMMAND 2 3# This is a Perl script to demonstrate the possibilities of on-the-fly 4# delivery filtering in Exim. It is presented with a message on its standard 5# input, and must copy it to the standard output, transforming it as it 6# pleases, but of course it must keep to the syntax of RFC 822 for the headers. 7 8# The filter is run before any SMTP-specific processing, such as turning 9# \n into \r\n and escaping lines beginning with a dot. 10# 11# Philip Hazel, May 1997 12############################################################################# 13 14use warnings; 15BEGIN { pop @INC if $INC[-1] eq '.' }; 16use File::Basename; 17 18if ($ARGV[0] eq '--version') { 19 print basename($0) . ": $0\n", 20 "build: EXIM_RELEASE_VERSIONEXIM_VARIANT_VERSION\n", 21 "perl(runtime): $]\n"; 22 exit 0; 23} 24 25# If the filter is called with any arguments, insert them into the message 26# as X-Arg headers, just to verify what they are. 27 28for ($ac = 0; $ac < @ARGV; $ac++) 29 { 30 printf("X-Arg%d: %s\n", $ac, $ARGV[$ac]); 31 } 32 33# Now read the header portion of the message; this is easy to do in Perl 34 35$/ = ""; # set paragraph mode 36chomp($headers = <STDIN>); # read a paragraph, remove trailing newlines 37$/ = "\n"; # unset paragraph mode 38 39# Splitting up a sequence of unique headers is easy to do in Perl, but a 40# message may contain duplicate headers of various kinds. It is better 41# to extract the headers one wants from the whole paragraph, do any appropriate 42# munging, and then put them back (unless removing them altogether). Messing 43# with "Received:" headers is not in any case to be encouraged. 44 45# As a demonstration, we extract the "From:" header, add a textual comment 46# to it, and put it back. 47 48($pre, $from, $post) = 49 $headers =~ /^(|(?:.|\n)+\n) (?# Stuff preceding the From header, 50 which is either null, or any number 51 of characters, including \n, ending 52 with \n.) 53 From:[\s\t]* (?# Header name, with optional space or tab.) 54 ((?:.|\n)*?) (?# Header body, which contains any chars, 55 including \n, but we want to make it as 56 short as possible so as not to include 57 following headers by mistake.) 58 (|\n\S(?:.|\n)*)$ (?# Header terminates at end or at \n followed 59 by a non-whitespace character and 60 remaining headers.) 61 /ix; # case independent, regular expression, 62 # use extended features (ignore whitespace) 63 64# Only do something if there was a From: header, of course. It has been 65# extracted without the final \n, which is on the front of the $post 66# variable. 67 68if ($pre) 69 { 70 $headers = $pre . "From: $from (this is an added comment)" . $post; 71 } 72 73# Add a new header to the end of the headers; remember that the final 74# \n isn't there. 75 76$headers .= "\nX-Comment: Message munged"; 77 78# Write out the processed headers, plus a blank line to separate them from 79# the body. 80 81printf(STDOUT "%s\n\n", $headers); 82 83# As a demonstration of munging the body of a message, reverse all the 84# characters in each line. 85 86while (<STDIN>) 87 { 88 chomp; 89 $_ = reverse($_); 90 printf(STDOUT "%s\n", $_); 91 } 92 93# End 94