Zimbra

dutchie · #1 (**permalink**) 07-09-2009, 03:11 AM

The title says it all really. I'd prefer to write it in Perl, as there are Lucene bindings in the Ubuntu repositories, but if there is a simple shell command to do this then that would be perfect.

Basically, I need to get all emails that are to/from a particular address, regardless of who sent/received them.

Thanks in advance.

dutchie · #2 (**permalink**) 07-09-2009, 04:01 AM

OK, I found a solution. I can run "zmaccts", extract the account information using a regex, and run "zmmailbox -z -m <each one in turn> search ...", but this is very slow. Obviously a lot of this stuff can be cached, but as I'm going to be serving this online I can't be hanging around for a few seconds (edit MINUTES) while all this goes on.

Code:

#!/usr/bin/perl

use strict;
use warnings;

my $target = $ARGV[0];

die "Usage: $0 <target to search for>\n" unless $target;

die "Must run as user zimbra\n" if $> != 1001;

my @emails;

open ACCTS, "zmaccts|";
for (<ACCTS>) {
    last if /domain summary/o;
    next if /^-|account|^$/o;
    my @words = split /\s+/;
    push @emails, $words[0];
}

my @lines;
for (@emails) {
    my $found = `zmmailbox -z -m $_ s "from:$target OR to:$target"`;
    for my $line (split /\n/, $found) {
        next unless $line =~ /^\d+\./o;
        # get rid of meaningless rubbish at start of line
        $line =~ s/\d+\.\s+(-?\d+)\s+conv\s+//o;
        push @lines, $line;
        my $conv = `zmmailbox -z -m $_ gc $1`;
    }
}

# remove dupes from @lines
my %temp = map { $_, 1 } @lines;
@lines = sort keys %temp;

open OUTCACHE, "> $target.cache";
print OUTCACHE <<EOF;
Involving             Subject                                             Time
----------------------------------------------------------------------------------------
EOF
print OUTCACHE "$_\n" foreach @lines;

system "cat $target.cache";

wdimmit · #3 (**permalink**) 07-15-2009, 11:16 AM

That script, while painfully slow, works well - thanks for putting it out here. Have you had any success finding methods to increase performance?

dutchie · #4 (**permalink**) 07-22-2009, 12:14 PM

This script was written in a week of work experience, and this is as far as it got. It works, but is horrifically slow.

Code:

#!/usr/bin/perl

use strict;
use warnings;
use Text::Wrap;

my $target = $ARGV[0];

die "Usage: $0 <target to search for>\n" unless $target;

die "Must run as user zimbra\n" if $> != 1001;

mkdir $target;
chmod 0755, $target;

my @emails;
open ACCTS, "zmaccts|";
for (<ACCTS>) {
    last if /domain summary/;
    next if /^-|account|^$/;
    my @words = split /\s+/;
    push @emails, $words[0];
}

my @msgs;
for my $addr (@emails) {
    open SEARCH, "zmmailbox -z -m $addr s \"from:$target OR to:$target\"|";
    while (my $line = <SEARCH>) {
        next unless $line =~ /^\d+\./;
        # get rid of meaningless rubbish at start of line
        $line =~ s/\d+\.\s+(-?\d+)\s+conv\s+//;
        open CONV, "zmmailbox -z -m $addr gc $1|";
        while (my $msgline = <CONV>) {
            next unless $msgline =~ /^\d+\.\s+(\d+)/;
            my $mailmsg = `zmmailbox -z -m $addr gm $1`;
            push @msgs, $mailmsg;
        }
    }
}

# remove dupes from @msgs
my %temp = map { $_, 1 } @msgs;
@msgs = sort keys %temp;

open OUTCACHE, "> $target.cache";
print OUTCACHE <<EOF;
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN"
    "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
<html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en">
<head>
    <title>Emails involving $target</target>
    <meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
</head>

<body>
    <h1>Emails sent/received by $target</h1>
    <table border="1">
        <tr>
            <th>From</th>
            <th>To</th>
            <th>Subject</th>
            <th>Time</th>
        </tr>
EOF

my %dates_seen;
for (@msgs) {
    next if exists $dates_seen{$_};
    $dates_seen{$_}++;
    my ($date, $from, $to, $subject, $id);
    for (lines($_)) {
        $subject = $1 and last if /^Subject: (.*)/;
    }
    for (lines($_)) {
        $date    = $1 and last if /^Sent: (.*)/;
    }
    for (lines($_)) {
        $from    = $1 and last if /^From: (.*)/;
    }
    for (lines($_)) {
        $to      = $1 and last if /^To: (.*)/;
    }
    for (lines($_)) {
        $id      = $1 and last if /^Id: (.*)/;
    }
    print OUTCACHE <<EOF;
        <tr>
            <td>$from</td>
            <td>$to</td>
            <td><a href="$target/$id">$subject</a></td>
            <td>$date</td>
        </tr>
EOF
    s/^Id: |^Conversation-Id: |^Flags: |^Folder: |^Size//;
    open MSG, "> $target/$id";
    print MSG wrap("", "", $_);
}

print OUTCACHE <<EOF;
    </table>
</body>

</html>
EOF

sub lines {
    local $_ = shift;
    return split /\n/;
}

I also wrote a bit of PHP to display the results:

Code:

<?php
    $addr = $_REQUEST['email'];
    if (!$addr) {
        exit(1);
    }
    $filename = $addr . ".cache";
    if (file_exists($filename)) {
        $handle = fopen($filename, "r");
        $contents = fread($handle, filesize($filename));
        fclose($handle);
        echo $contents;
    }
?>

Zimbra

Downloads

Product Information

Toolbox

Why Join?