#!/usr/bin/perl # $Header: /mhub4/sources/imap-tools/imapcopy.pl,v 1.14 2010/02/24 00:04:27 rick Exp $ ####################################################################### # Program name imapcopy.pl # # Written by Rick Sanders # # # # Description # # # # imapcopy is a utility for copying a user's messages from one # # IMAP server to another. # # # # imapcopy is called like this: # # ./imapcopy -S host1/user1/password1 -D host2/user2/password2 # # # # Optional arguments: # # -d debug # # -I show IMAP protocol exchanges # # -L logfile # # -m mailbox list (copy only certain mailboxes,see usage notes) # # -r reset the \DELETE flag on copied messages # # -p put copied mailboxes under a root mbx # # -M mailbox mapping (eg, src:inbox -> dst:inbox_copied) # ####################################################################### use Socket; use FileHandle; use Fcntl; use Getopt::Std; use IO::Socket; ################################################################# # Main program. # ################################################################# init(); # Get list of all messages on the source host # connectToHost($sourceHost, \$src) or exit; login($sourceUser,$sourcePwd, $src) or exit; namespace( $src, \$srcPrefix, \$srcDelim, $opt_x ); connectToHost( $destHost, \$dst ) or exit; login( $destUser,$destPwd, $dst ) or exit; namespace( $dst, \$dstPrefix, \$dstDelim, $opt_y ); @mbxs = getMailboxList( $src ); map_mbx_names( \%mbx_map, $srcDelim, $dstDelim ); $total=0; foreach $srcmbx ( @mbxs ) { $dstmbx = mailboxName( $srcmbx,$srcPrefix,$srcDelim,$dstPrefix,$dstDelim ); createMbx( $dstmbx, $dst ) unless mbxExists( $dstmbx, $dst); selectMbx( $dstmbx, $dst ); Log("Copying messages in $srcmbx mailbox") if $debug; getMsgList( $srcmbx, \@msgs, $src ); if ( $#msgs == -1 ) { Log("$srcmbx mailbox is empty"); next; } $copied=0; foreach $_ ( @msgs ) { ($msgnum,$date,$flags) = split(/\|/, $_); $message = fetchMsg( $msgnum, $srcmbx, $src ); next unless $message; $copied++ if insertMsg( $dst, $dstmbx, *message, $flags, $date ); } $total += $copied; if ( $use_utf7 ) { $dstmbx = Unicode::IMAPUtf7::imap_utf7_decode( $dstmbx ); } Log("Copied $copied messages to $dstmbx"); } Log("Copied $total total messages"); logout( $src ); logout( $dst ); exit; sub init { $os = $ENV{'OS'}; processArgs(); if ($timeout eq '') { $timeout = 60; } # Open the logFile # if ( $logfile ) { if ( !open(LOG, ">> $logfile")) { print STDOUT "Can't open $logfile: $!\n"; } select(LOG); $| = 1; } Log("$0 starting"); # Determine whether we have SSL support via openSSL and IO::Socket::SSL $ssl_installed = 1; eval 'use IO::Socket::SSL'; if ( $@ ) { $ssl_installed = 0; } } # # sendCommand # # This subroutine formats and sends an IMAP protocol command to an # IMAP server on a specified connection. # sub sendCommand { my $fd = shift; my $cmd = shift; print $fd "$cmd\r\n"; Log (">> $cmd") if $showIMAP; } # # readResponse # # This subroutine reads and formats an IMAP protocol response from an # IMAP server on a specified connection. # sub readResponse { my $fd = shift; $response = <$fd>; chop $response; $response =~ s/\r//g; push (@response,$response); Log ("<< $response") if $showIMAP;1 } # # Log # # This subroutine formats and writes a log message to STDERR. # sub Log { my $str = shift; # If a logfile has been specified then write the output to it # Otherwise write it to STDOUT if ( $logfile ) { ($sec,$min,$hour,$mday,$mon,$year,$wday,$yday,$isdst) = localtime; if ($year < 99) { $yr = 2000; } else { $yr = 1900; } $line = sprintf ("%.2d-%.2d-%d.%.2d:%.2d:%.2d %s\n", $mon + 1, $mday, $year + $yr, $hour, $min, $sec,$str); print LOG "$line"; } print STDOUT "$line" unless $quiet_mode; } sub createMbx { my $mbx = shift; my $conn = shift; # Create the mailbox if necessary sendCommand ($conn, "1 CREATE \"$mbx\""); while ( 1 ) { readResponse ($conn); last if $response =~ /^1 OK/i; last if $response =~ /already exists/i; if ( $response =~ /^1 NO|^1 BAD/ ) { Log ("Error creating $mbx: $response"); last; } } } # insertMsg # # This routine inserts a message into a user's mailbox # sub insertMsg { local ($conn, $mbx, *message, $flags, $date) = @_; local ($lenx); Log(" Inserting message") if $debug; $lenx = length($message); $totalBytes = $totalBytes + $lenx; $totalMsgs++; $flags = flags( $flags ); sendCommand ($conn, "1 APPEND \"$mbx\" ($flags) \"$date\" \{$lenx\}"); readResponse ($conn); if ( $response !~ /^\+/ ) { Log ("unexpected APPEND response: $response"); # next; push(@errors,"Error appending message to $mbx for $user"); return 0; } print $conn "$message\r\n"; undef @response; while ( 1 ) { readResponse ($conn); if ( $response =~ /^1 OK/i ) { last; } elsif ( $response !~ /^\*/ ) { Log ("unexpected APPEND response: $response"); # next; return 0; } } return 1; } # Make a connection to a IMAP host sub connectToHost { my $host = shift; my $conn = shift; Log("Connecting to $host") if $debug; ($host,$port) = split(/:/, $host); $port = 143 unless $port; # We know whether to use SSL for ports 143 and 993. For any # other ones we'll have to figure it out. $mode = sslmode( $host, $port ); if ( $mode eq 'SSL' ) { unless( $ssl_installed == 1 ) { warn("You must have openSSL and IO::Socket::SSL installed to use an SSL connection"); Log("You must have openSSL and IO::Socket::SSL installed to use an SSL connection"); exit; } Log("Attempting an SSL connection") if $debug; $$conn = IO::Socket::SSL->new( Proto => "tcp", SSL_verify_mode => 0x00, PeerAddr => $host, PeerPort => $port, ); unless ( $$conn ) { $error = IO::Socket::SSL::errstr(); Log("Error connecting to $host: $error"); exit; } } else { # Non-SSL connection Log("Attempting a non-SSL connection") if $debug; $$conn = IO::Socket::INET->new( Proto => "tcp", PeerAddr => $host, PeerPort => $port, ); unless ( $$conn ) { Log("Error connecting to $host:$port: $@"); warn "Error connecting to $host:$port: $@"; exit; } } Log("Connected to $host on port $port"); } sub sslmode { my $host = shift; my $port = shift; my $mode; # Determine whether to make an SSL connection # to the host. Return 'SSL' if so. if ( $port == 143 ) { # Standard non-SSL port return ''; } elsif ( $port == 993 ) { # Standard SSL port return 'SSL'; } unless ( $ssl_installed ) { # We don't have SSL installed on this machine return ''; } # For any other port we need to determine whether it supports SSL my $conn = IO::Socket::SSL->new( Proto => "tcp", SSL_verify_mode => 0x00, PeerAddr => $host, PeerPort => $port, ); if ( $conn ) { close( $conn ); $mode = 'SSL'; } else { $mode = ''; } return $mode; } # trim # # remove leading and trailing spaces from a string sub trim { local (*string) = @_; $string =~ s/^\s+//; $string =~ s/\s+$//; return; } # login # # login in at the source host with the user's name and password # sub login { my $user = shift; my $pwd = shift; my $conn = shift; sendCommand ($conn, "1 LOGIN $user $pwd"); while (1) { readResponse ( $conn ); last if $response =~ /^1 OK/i; if ($response =~ /^1 NO|^1 BAD/i) { Log ("unexpected LOGIN response: $response"); return 0; } } Log("Logged in as $user") if $debug; return 1; } # logout # # log out from the host # sub logout { my $conn = shift; undef @response; sendCommand ($conn, "1 LOGOUT"); while ( 1 ) { readResponse ($conn); if ( $response =~ /^1 OK/i ) { last; } elsif ( $response !~ /^\*/ ) { Log ("unexpected LOGOUT response: $response"); last; } } close $conn; return; } # getMailboxList # # get a list of the user's mailboxes from the source host # sub getMailboxList { my $conn = shift; my @mbxs; my @mailboxes; # Get a list of the user's mailboxes # if ( $mbxList ) { # The user has supplied a list of mailboxes so only processes # the ones in that list @mbxs = split(/,/, $mbxList); foreach $mbx ( @mbxs ) { trim( *mbx ); push( @mailboxes, $mbx ); } return @mailboxes; } if ($debug) { Log("Get list of user's mailboxes",2); } sendCommand ($conn, "1 LIST \"\" *"); undef @response; while ( 1 ) { readResponse ($conn); if ( $response =~ /^1 OK/i ) { last; } elsif ( $response !~ /^\*/ ) { Log ("unexpected response: $response"); return 0; } } @mbxs = (); for $i (0 .. $#response) { $response[$i] =~ s/\s+/ /; if ( $response[$i] =~ /"$/ ) { $response[$i] =~ /\* LIST \((.*)\) "(.+)" "(.+)"/i; $mbx = $3; } elsif ( $response[$i] =~ /\* LIST \((.*)\) NIL (.+)/i ) { $mbx = $2; } else { $response[$i] =~ /\* LIST \((.*)\) "(.+)" (.+)/i; $mbx = $3; } $mbx =~ s/^\s+//; $mbx =~ s/\s+$//; if ($response[$i] =~ /NOSELECT/i) { if ($debug) { Log("$mbx is set NOSELECT,skip it",2); } next; } if ($mbx =~ /^\#/) { # Skip public mbxs next; } if ($mbx =~ /^\./) { # Skip mailboxes starting with a dot next; } push ( @mbxs, $mbx ) if $mbx ne ''; } return @mbxs; } # getMsgList # # Get a list of the user's messages in the indicated mailbox on # the source host # sub getMsgList { my $mailbox = shift; my $msgs = shift; my $conn = shift; my $seen; my $empty; my $msgnum; my $from; my $flags; trim( *mailbox ); sendCommand ($conn, "1 EXAMINE \"$mailbox\""); undef @response; $empty=0; while ( 1 ) { readResponse ( $conn ); if ( $response =~ / 0 EXISTS/i ) { $empty=1; } if ( $response =~ /^1 OK/i ) { last; } elsif ( $response !~ /^\*/ ) { Log ("unexpected response: $response"); return 0; } } sendCommand ( $conn, "1 FETCH 1:* (uid flags internaldate body[header.fields (From Date)])"); undef @response; while ( 1 ) { readResponse ( $conn ); if ( $response =~ /^1 OK/i ) { last; } last if $response =~ /^1 NO|^1 BAD/; } @msgs = (); $flags = ''; for $i (0 .. $#response) { last if $response[$i] =~ /^1 OK FETCH complete/i; if ( $response[$i] =~ /^From: (.+)/i ) { $from = $1; } if ($response[$i] =~ /FLAGS/) { # Get the list of flags $response[$i] =~ /FLAGS \(([^\)]*)/; $flags = $1; $flags =~ s/\\Recent//; } if ( $response[$i] =~ /INTERNALDATE/) { $response[$i] =~ /INTERNALDATE (.+) BODY/i; # $response[$i] =~ /INTERNALDATE "(.+)" BODY/; $date = $1; $date =~ /"(.+)"/; $date = $1; $date =~ s/"//g; } # if ( $response[$i] =~ /\* (.+) [^FETCH]/ ) { if ( $response[$i] =~ /\* (.+) FETCH/ ) { ($msgnum) = split(/\s+/, $1); } if ( $msgnum && $from && $date ) { push (@$msgs,"$msgnum|$date|$flags"); $msgnum = $from = $date = ''; } } } sub mbxExists { my $mbx = shift; my $conn = shift; my $status = 1; # Determine whether a mailbox exists sendCommand ($conn, "1 EXAMINE \"$mbx\""); while (1) { readResponse ($conn); last if $response =~ /^1 OK/i; if ( $response =~ /^1 NO|^1 BAD/ ) { $status = 0; last; } } return $status; } sub fetchMsg { my $msgnum = shift; my $mbx = shift; my $conn = shift; my $message; Log(" Fetching msg $msgnum...") if $debug; # sendCommand ($conn, "1 EXAMINE \"$mbx\""); # while (1) { # readResponse ($conn); # last if ( $response =~ /^1 OK/i ); # } sendCommand( $conn, "1 FETCH $msgnum (rfc822)"); while (1) { readResponse ($conn); last if $response =~ /^1 NO|^1 BAD/; if ( $response =~ /^1 OK/i ) { $size = length($message); last; } elsif ($response =~ /message number out of range/i) { Log ("Error fetching uid $uid: out of range",2); $stat=0; last; } elsif ($response =~ /Bogus sequence in FETCH/i) { Log ("Error fetching uid $uid: Bogus sequence in FETCH",2); $stat=0; last; } elsif ( $response =~ /message could not be processed/i ) { Log("Message could not be processed, skipping it ($user,msgnum $msgnum,$destMbx)"); push(@errors,"Message could not be processed, skipping it ($user,msgnum $msgnum,$destMbx)"); $stat=0; last; } elsif ($response =~ /^\*\s+$msgnum\s+FETCH\s+\(.*RFC822\s+\{[0-9]+\}/i) { ($len) = ($response =~ /^\*\s+$msgnum\s+FETCH\s+\(.*RFC822\s+\{([0-9]+)\}/i); $cc = 0; $message = ""; while ( $cc < $len ) { $n = 0; $n = read ($conn, $segment, $len - $cc); if ( $n == 0 ) { Log ("unable to read $len bytes"); return 0; } $message .= $segment; $cc += $n; } } } return $message; } sub usage { print STDOUT "usage:\n"; print STDOUT " imapcopy -S sourceHost/sourceUser/sourcePassword\n"; print STDOUT " -D destHost/destUser/destPassword\n"; print STDOUT " -d debug\n"; print STDOUT " -I show IMAP protocol exchanges\n"; print STDOUT " -L logfile\n"; print STDOUT " -m mailbox list (eg \"Inbox, Drafts, Notes\". Default is all mailboxes)\n"; print STDOUT " -r clear \\DELETE flag from deleted messages\n"; print STDOUT " -p put copied mailboxes under a root mailbox\n"; print STDOUT " -x source (eg, -x '. INBOX.'\n"; print STDOUT " -y destination\n"; print STDOUT " -M mailbox map file. Maps src mbxs to dst mbxs. "; print STDOUT " -q quiet mode (still writes to the logfile)\n"; print STDOUT "Each line in the file should be 'src mbx:dst mbx'\n"; exit; } sub processArgs { if ( !getopts( "dS:D:L:m:hIp:M:rqx:y:" ) ) { usage(); } ($sourceHost,$sourceUser,$sourcePwd) = split(/\//, $opt_S); ($destHost, $destUser, $destPwd) = split(/\//, $opt_D); $mbxList = $opt_m; $logfile = $opt_L; $root_mbx = $opt_p; $debug = 1 if $opt_d; $showIMAP = 1 if $opt_I; $quiet_mode = 1 if $opt_q; $reset_dflag = $opt_r; $mbx_map_fn = $opt_M; # $src_mbx_delim = $opt_x; # $dst_mbx_delim = $opt_y; usage() if $opt_h; } sub selectMbx { my $mbx = shift; my $conn = shift; # Some IMAP clients such as Outlook and Netscape) do not automatically list # all mailboxes. The user must manually subscribe to them. This routine # does that for the user by marking the mailbox as 'subscribed'. sendCommand( $conn, "1 SUBSCRIBE \"$mbx\""); while ( 1 ) { readResponse( $conn ); if ( $response =~ /^1 OK/i ) { Log("Mailbox $mbx has been subscribed") if $debug; last; } elsif ( $response =~ /^1 NO|^1 BAD/i ) { Log("Unexpected response to subscribe $mbx command: $response"); last; } } # Now select the mailbox sendCommand( $conn, "1 SELECT \"$mbx\""); while ( 1 ) { readResponse( $conn ); if ( $response =~ /^1 OK/i ) { last; } elsif ( $response =~ /^1 NO|^1 BAD/i ) { Log("Unexpected response to SELECT $mbx command: $response"); last; } } } sub namespace { my $conn = shift; my $prefix = shift; my $delimiter = shift; my $mbx_delim = shift; # Query the server with NAMESPACE so we can determine its # mailbox prefix (if any) and hierachy delimiter. if ( $mbx_delim ) { # The user has supplied a mbx delimiter and optionally a prefix. Log("Using user-supplied mailbox hierarchy delimiter $mbx_delim"); ($$delimiter,$$prefix) = split(/\s+/, $mbx_delim); return; } @response = (); sendCommand( $conn, "1 NAMESPACE"); while ( 1 ) { readResponse( $conn ); if ( $response =~ /^1 OK/i ) { last; } elsif ( $response =~ /^1 NO|^1 BAD/i ) { Log("Unexpected response to NAMESPACE command: $response"); last; } } foreach $_ ( @response ) { if ( /NAMESPACE/i ) { my $i = index( $_, '((' ); my $j = index( $_, '))' ); my $val = substr($_,$i+2,$j-$i-3); ($val) = split(/\)/, $val); ($$prefix,$$delimiter) = split( / /, $val ); $$prefix =~ s/"//g; $$delimiter =~ s/"//g; last; } last if /^1 NO|^1 BAD/; } unless ( $$delimiter ) { # NAMESPACE command is not supported by the server # so we will have to figure it out another way. $delim = getDelimiter( $conn ); $$delimiter = $delim; $$prefix = ''; } if ( $debug ) { Log("prefix >$$prefix<"); Log("delim >$$delimiter<"); } } sub mailboxName { my $srcmbx = shift; my $srcPrefix = shift; my $srcDelim = shift; my $dstPrefix = shift; my $dstDelim = shift; my $dstmbx; my $substChar = '_'; # Adjust the mailbox name if the source and destination server # have different mailbox prefixes or hierarchy delimiters. if ( ($srcPrefix eq $dstPrefix) and ( $srcDelim eq $dstDelim ) ) { # They are the same on both, no need to adjust anything $dstmbx = $srcmbx; return $dstmbx; } if ( ($srcmbx =~ /[$dstDelim]/) and ($dstDelim ne $srcDelim) ) { # The mailbox name has a character that is used on the destination # as a mailbox hierarchy delimiter. We have to replace it. $srcmbx =~ s^[$dstDelim]^$substChar^g; } if ( $debug ) { Log("src mbx $srcmbx"); Log("src prefix $srcPrefix"); Log("src delim $srcDelim"); Log("dst prefix $dstPrefix"); Log("dst delim $dstDelim"); } $srcmbx =~ s/^$srcPrefix//; $srcmbx =~ s/\\$srcDelim/\//g; # Change the mailbox name if the user has supplied mapping rules. if ( $mbx_map{"$srcmbx"} ) { $srcmbx = $mbx_map{"$srcmbx"} } if ( ($srcPrefix eq $dstPrefix) and ($srcDelim eq $dstDelim) ) { # No adjustments necessary $dstmbx = $srcmbx; if ( $root_mbx ) { # Put folders under a 'root' folder on the dst $dstmbx =~ s/^$dstPrefix//; $dstDelim =~ s/\./\\./g; $dstmbx =~ s/^$dstDelim//; $dstmbx = $dstPrefix . $root_mbx . $dstDelim . $dstmbx; if ( uc($srcmbx) eq 'INBOX' ) { # Special case for the INBOX $dstmbx =~ s/INBOX$//i; $dstmbx =~ s/$dstDelim$//; } $dstmbx =~ s/\\//g; } return $dstmbx; } $srcmbx =~ s#^$srcPrefix##; $dstmbx = $srcmbx; if ( $srcDelim ne $dstDelim ) { # Need to substitute the dst's hierarchy delimiter for the src's one $srcDelim = '\\' . $srcDelim if $srcDelim eq '.'; $dstDelim = "\\" . $dstDelim if $dstDelim eq '.'; $dstmbx =~ s#$srcDelim#$dstDelim#g; $dstmbx =~ s/\\//g; } if ( $srcPrefix ne $dstPrefix ) { # Replace the source prefix with the dest prefix $dstmbx =~ s#^$srcPrefix## if $srcPrefix; if ( $dstPrefix ) { $dstmbx = "$dstPrefix$dstmbx" unless uc($srcmbx) eq 'INBOX'; } $dstDelim = "\\$dstDelim" if $dstDelim eq '.'; $dstmbx =~ s#^$dstDelim##; } if ( $root_mbx ) { # Put folders under a 'root' folder on the dst $dstDelim =~ s/\./\\./g; $dstmbx =~ s/^$dstPrefix//; $dstmbx =~ s/^$dstDelim//; $dstmbx = $dstPrefix . $root_mbx . $dstDelim . $dstmbx; if ( uc($srcmbx) eq 'INBOX' ) { # Special case for the INBOX $dstmbx =~ s/INBOX$//i; $dstmbx =~ s/$dstDelim$//; } $dstmbx =~ s/\\//g; } return $dstmbx; } sub flags { my $flags = shift; my @newflags; my $newflags; # Make sure the flags list contains only standard # IMAP flags. return unless $flags; $flags =~ s/\\Recent//i; foreach $_ ( split(/\s+/, $flags) ) { next unless substr($_,0,1) eq '\\'; push( @newflags, $_ ); } $newflags = join( ' ', @newflags ); $newflags =~ s/\\Deleted//ig if $opt_r; $newflags =~ s/^\s+|\s+$//g; return $newflags; } sub map_mbx_names { my $mbx_map = shift; my $srcDelim = shift; my $dstDelim = shift; # The -M argument causes imapcopy to read the # contents of a file with mappings between source and # destination mailbox names. This permits the user to # to change the name of a mailbox when copying messages. # # The lines in the file should be formatted as: # : # For example: # Drafts/2008/Save: Draft_Messages/2008/Save # Action Items: Inbox # # Note that if the names contain non-ASCII characters such # as accents or diacritical marks then the Perl module # Unicode::IMAPUtf7 module must be installed. return unless $mbx_map_fn; unless ( open(MAP, "<$mbx_map_fn") ) { Log("Error opening mbx map file $mbx_map_fn: $!"); exit; } $use_utf7 = 0; while( ) { chomp; s/[\r\n]$//; # In case we're on Windows s/^\s+//; next if /^#/; next unless $_; ($srcmbx,$dstmbx) = split(/\s*:\s*/, $_); # Unless the mailbox name is entirely ASCII we'll have to use # the Modified UTF-7 character set. $use_utf7 = 1 unless isAscii( $srcmbx ); $use_utf7 = 1 unless isAscii( $dstmbx ); $srcmbx =~ s/\//$srcDelim/g; $dstmbx =~ s/\//$dstDelim/g; $$mbx_map{"$srcmbx"} = $dstmbx; } close MAP; if ( $use_utf7 ) { eval 'use Unicode::IMAPUtf7'; if ( $@ ) { Log("At least one mailbox map contains non-ASCII characters. This means you"); Log("have to install the Perl Unicode::IMAPUtf7 module in order to map mailbox "); Log("names between the source and destination servers."); print "At least one mailbox map contains non-ASCII characters. This means you\n"; print "have to install the Perl Unicode::IMAPUtf7 module in order to map mailbox\n"; print "names between the source and destination servers.\n"; exit; } } my %temp; foreach $srcmbx ( keys %$mbx_map ) { $dstmbx = $$mbx_map{"$srcmbx"}; Log("Mapping src:$srcmbx to dst:$dstmbx"); if ( $use_utf7 ){ # Encode the name in Modified UTF-7 charset $srcmbx = Unicode::IMAPUtf7::imap_utf7_encode( $srcmbx ); $dstmbx = Unicode::IMAPUtf7::imap_utf7_encode( $dstmbx ); } $temp{"$srcmbx"} = $dstmbx; } %$mbx_map = %temp; %temp = (); } sub isAscii { my $str = shift; my $ascii = 1; # Determine whether a string contains non-ASCII characters my $test = $str; $test=~s/\P{IsASCII}/?/g; $ascii = 0 unless $test eq $str; return $ascii; } sub getDelimiter { my $conn = shift; my $delimiter; # Issue a 'LIST "" ""' command to find out what the # mailbox hierarchy delimiter is. sendCommand ($conn, '1 LIST "" ""'); @response = ''; while ( 1 ) { readResponse ($conn); if ( $response =~ /^1 OK/i ) { last; } elsif ( $response !~ /^\*/ ) { Log ("unexpected response: $response"); return 0; } } for $i (0 .. $#response) { $response[$i] =~ s/\s+/ /; if ( $response[$i] =~ /\* LIST \((.*)\) "(.*)" "(.*)"/i ) { $delimiter = $2; } } return $delimiter; }