| 1 | ###########################################################################
|
|---|
| 2 | #
|
|---|
| 3 | # EMAILPlug.pm - a plugin for parsing email files
|
|---|
| 4 | #
|
|---|
| 5 | # A component of the Greenstone digital library software
|
|---|
| 6 | # from the New Zealand Digital Library Project at the
|
|---|
| 7 | # University of Waikato, New Zealand.
|
|---|
| 8 | #
|
|---|
| 9 | # Copyright (C) 1999-2002 New Zealand Digital Library Project
|
|---|
| 10 | #
|
|---|
| 11 | # This program is free software; you can redistribute it and/or modify
|
|---|
| 12 | # it under the terms of the GNU General Public License as published by
|
|---|
| 13 | # the Free Software Foundation; either version 2 of the License, or
|
|---|
| 14 | # (at your option) any later version.
|
|---|
| 15 | #
|
|---|
| 16 | # This program is distributed in the hope that it will be useful,
|
|---|
| 17 | # but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|---|
| 18 | # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|---|
| 19 | # GNU General Public License for more details.
|
|---|
| 20 | #
|
|---|
| 21 | # You should have received a copy of the GNU General Public License
|
|---|
| 22 | # along with this program; if not, write to the Free Software
|
|---|
| 23 | # Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
|
|---|
| 24 | #
|
|---|
| 25 | ###########################################################################
|
|---|
| 26 |
|
|---|
| 27 |
|
|---|
| 28 |
|
|---|
| 29 | # EMAILPlug
|
|---|
| 30 | #
|
|---|
| 31 | # by Gordon Paynter ([email protected])
|
|---|
| 32 | #
|
|---|
| 33 | # Email plug reads email files. These are named with a simple
|
|---|
| 34 | # number (i.e. as they appear in maildir folders) or with the
|
|---|
| 35 | # extension .mbx (for mbox mail file format)
|
|---|
| 36 | #
|
|---|
| 37 | # Document text:
|
|---|
| 38 | # The document text consists of all the text
|
|---|
| 39 | # after the first blank line in the document.
|
|---|
| 40 | #
|
|---|
| 41 | # Metadata (not Dublin Core!):
|
|---|
| 42 | # $Headers All the header content (optional, not stored by default)
|
|---|
| 43 | # $Subject Subject: header
|
|---|
| 44 | # $To To: header
|
|---|
| 45 | # $From From: header
|
|---|
| 46 | # $FromName Name of sender (where available)
|
|---|
| 47 | # $FromAddr E-mail address of sender
|
|---|
| 48 | # $DateText Date: header
|
|---|
| 49 | # $Date Date: header in GSDL format (eg: 19990924)
|
|---|
| 50 | #
|
|---|
| 51 | # $Title made up of Subject, Date and Sender (for default formatting)
|
|---|
| 52 | #
|
|---|
| 53 | #
|
|---|
| 54 | # John McPherson - June/July 2001
|
|---|
| 55 | # added (basic) MIME support and quoted-printable and base64 decodings.
|
|---|
| 56 | # Minor fixes for names that are actually email addresses (ie <...> was lost)
|
|---|
| 57 | #
|
|---|
| 58 | # See: * RFC 822 - ARPA Internet Text Messages
|
|---|
| 59 | # * RFC 2045 - Multipurpose Internet Mail Extensions (MIME) -part1
|
|---|
| 60 | # * RFC 2046 - MIME (part 2) Media Types (and multipart messages)
|
|---|
| 61 | # * RFC 2047 - MIME (part 3) Message Header Extensions
|
|---|
| 62 | # * RFC 1806 - Content Dispositions (ie inline/attachment)
|
|---|
| 63 |
|
|---|
| 64 | # 12/05/02 Added usage datastructure - John Thompson
|
|---|
| 65 |
|
|---|
| 66 | use strict;
|
|---|
| 67 | no strict "refs"; # so we can use a variable as a filehandle for print $out
|
|---|
| 68 |
|
|---|
| 69 | package EMAILPlug;
|
|---|
| 70 |
|
|---|
| 71 | use SplitPlug;
|
|---|
| 72 | # EMAILPlug is a sub-class of SplitPlug.
|
|---|
| 73 | @EMAILPlug::ISA = ('SplitPlug');
|
|---|
| 74 |
|
|---|
| 75 | use unicode; # gs conv functions
|
|---|
| 76 | use gsprintf 'gsprintf'; # translations
|
|---|
| 77 |
|
|---|
| 78 | use sorttools;
|
|---|
| 79 | use util;
|
|---|
| 80 |
|
|---|
| 81 |
|
|---|
| 82 |
|
|---|
| 83 | my $arguments =
|
|---|
| 84 | [ { 'name' => "process_exp",
|
|---|
| 85 | 'desc' => "{BasPlug.process_exp}",
|
|---|
| 86 | 'type' => "regexp",
|
|---|
| 87 | 'reqd' => "no",
|
|---|
| 88 | 'deft' => &get_default_process_exp() },
|
|---|
| 89 | { 'name' => "no_attachments",
|
|---|
| 90 | 'desc' => "{EMAILPlug.no_attachments}",
|
|---|
| 91 | 'type' => "flag",
|
|---|
| 92 | 'reqd' => "no" },
|
|---|
| 93 | { 'name' => "headers",
|
|---|
| 94 | 'desc' => "{EMAILPlug.headers}",
|
|---|
| 95 | 'type' => "flag",
|
|---|
| 96 | 'reqd' => "no" },
|
|---|
| 97 | { 'name' => "split_exp",
|
|---|
| 98 | 'desc' => "{EMAILPlug.split_exp}",
|
|---|
| 99 | 'type' => "regexp",
|
|---|
| 100 | 'reqd' => "no",
|
|---|
| 101 | 'deft' => &get_default_split_exp() }
|
|---|
| 102 | ];
|
|---|
| 103 |
|
|---|
| 104 | my $options = { 'name' => "EMAILPlug",
|
|---|
| 105 | 'desc' => "{EMAILPlug.desc}",
|
|---|
| 106 | 'abstract' => "no",
|
|---|
| 107 | 'inherits' => "yes",
|
|---|
| 108 | 'args' => $arguments };
|
|---|
| 109 |
|
|---|
| 110 | # Create a new EMAILPlug object with which to parse a file.
|
|---|
| 111 | # Accomplished by creating a new BasPlug and using bless to
|
|---|
| 112 | # turn it into an EMAILPlug.
|
|---|
| 113 |
|
|---|
| 114 | sub new {
|
|---|
| 115 | my ($class) = @_;
|
|---|
| 116 | my $self = new BasPlug ($class, @_);
|
|---|
| 117 | $self->{'plugin_type'} = "EMAILPlug";
|
|---|
| 118 | # 14-05-02 To allow for proper inheritance of arguments - John Thompson
|
|---|
| 119 | my $option_list = $self->{'option_list'};
|
|---|
| 120 | push( @{$option_list}, $options );
|
|---|
| 121 |
|
|---|
| 122 | if (!parsargv::parse(\@_,
|
|---|
| 123 | q^split_exp/.*/^, \$self->{'split_exp'},
|
|---|
| 124 | q^no_attachments^, \$self->{'ignore_attachments'},
|
|---|
| 125 | q^headers^, \$self->{'header_metadata'},
|
|---|
| 126 | "allow_extra_options")) {
|
|---|
| 127 | print STDERR "\nIncorrect options passed to $class.";
|
|---|
| 128 | print STDERR "\nCheck your collect.cfg configuration file\n";
|
|---|
| 129 | $self->print_txt_usage(""); # Use default resource bundle
|
|---|
| 130 | die "\n";
|
|---|
| 131 | }
|
|---|
| 132 | $self->{'assoc_filenames'} = {}; # to save attach names so we don't clobber
|
|---|
| 133 |
|
|---|
| 134 | # this might not actually be true at read-time, but after processing
|
|---|
| 135 | # it should all be utf8.
|
|---|
| 136 | $self->{'input_encoding'}="utf8";
|
|---|
| 137 | return bless $self, $class;
|
|---|
| 138 | }
|
|---|
| 139 |
|
|---|
| 140 | sub get_default_process_exp {
|
|---|
| 141 | my $self = shift (@_);
|
|---|
| 142 | # mbx/email for mailbox file format, \d+ for maildir (each message is
|
|---|
| 143 | # in a separate file, with a unique number for filename)
|
|---|
| 144 | # mozilla and IE will save individual mbx format files with a ".eml" ext.
|
|---|
| 145 | return q@([\\/]\d+|\.(mbx|email|eml))$@;
|
|---|
| 146 | }
|
|---|
| 147 |
|
|---|
| 148 | # This plugin splits the mbox mail files at lines starting with From<sp>
|
|---|
| 149 | # It is supposed to be "\n\nFrom ", but this isn't always used.
|
|---|
| 150 | # add \d{4} so that the line ends in a year (in case the text has an
|
|---|
| 151 | # unescaped "From " at the start of a line).
|
|---|
| 152 | sub get_default_split_exp {
|
|---|
| 153 | return q^\nFrom .*\d{4}\n^;
|
|---|
| 154 |
|
|---|
| 155 | }
|
|---|
| 156 |
|
|---|
| 157 |
|
|---|
| 158 | # do plugin specific processing of doc_obj
|
|---|
| 159 | sub process {
|
|---|
| 160 |
|
|---|
| 161 | my $self = shift (@_);
|
|---|
| 162 | my ($textref, $pluginfo, $base_dir, $file, $metadata, $doc_obj, $gli) = @_;
|
|---|
| 163 | my $outhandle = $self->{'outhandle'};
|
|---|
| 164 |
|
|---|
| 165 | # Check that we're dealing with a valid mail file
|
|---|
| 166 | # mbox message files start with "From "
|
|---|
| 167 | # maildir messages usually start with Return-Path and Delivered-To
|
|---|
| 168 | # mh is very similar to maildir
|
|---|
| 169 | my $startoffile=substr($$textref,0,256);
|
|---|
| 170 | if (($startoffile !~ /^(From )/) &&
|
|---|
| 171 | ($startoffile !~ /^(From|To|Envelope.*|Received|Return-Path|Date|Subject|Content\-.*|MIME-Version|Forwarded):/im)) {
|
|---|
| 172 | return undef;
|
|---|
| 173 | }
|
|---|
| 174 |
|
|---|
| 175 |
|
|---|
| 176 | print STDERR "<Processing n='$file' p='EMAILPlug'>\n" if ($gli);
|
|---|
| 177 |
|
|---|
| 178 | gsprintf($outhandle, "EMAILPlug: {common.processing} $file\n")
|
|---|
| 179 | if $self->{'verbosity'} > 1;
|
|---|
| 180 |
|
|---|
| 181 | my $cursection = $doc_obj->get_top_section();
|
|---|
| 182 |
|
|---|
| 183 | #
|
|---|
| 184 | # Parse the document's text and extract metadata
|
|---|
| 185 | #
|
|---|
| 186 |
|
|---|
| 187 | # Protect backslashes
|
|---|
| 188 | $$textref =~ s@\\@\\\\@g;
|
|---|
| 189 |
|
|---|
| 190 | # Separate header from body of message
|
|---|
| 191 | my $Headers = $$textref;
|
|---|
| 192 | $Headers =~ s/\r?\n\r?\n(.*)$//s;
|
|---|
| 193 | $$textref = $1;
|
|---|
| 194 | $Headers .= "\n";
|
|---|
| 195 |
|
|---|
| 196 | # Unfold headers - see rfc822
|
|---|
| 197 | $Headers =~ s/\r?\n[\t\ ]+/ /gs;
|
|---|
| 198 | # Extract basic metadata from header
|
|---|
| 199 | my @headers = ("From", "To", "Subject", "Date");
|
|---|
| 200 | my %raw;
|
|---|
| 201 | foreach my $name (@headers) {
|
|---|
| 202 | $raw{$name} = "No $name value";
|
|---|
| 203 | }
|
|---|
| 204 |
|
|---|
| 205 | # Get a default encoding for the header - RFC says should be ascii...
|
|---|
| 206 | my $default_header_encoding="iso_8859_1";
|
|---|
| 207 |
|
|---|
| 208 | # We don't know what character set is the user's default...
|
|---|
| 209 | # We could use textcat to guess... for now we'll look at mime content-type
|
|---|
| 210 | # if ($Headers =~ /([[:^ascii:]])/) {
|
|---|
| 211 | # }
|
|---|
| 212 | if ($Headers =~ /^Content\-type:.*charset=\"?([a-z0-9\-_]+)/mi) {
|
|---|
| 213 | $default_header_encoding=$1;
|
|---|
| 214 | $default_header_encoding =~ s@\-@_@g;
|
|---|
| 215 | $default_header_encoding =~ tr/A-Z/a-z/;
|
|---|
| 216 | }
|
|---|
| 217 |
|
|---|
| 218 |
|
|---|
| 219 | # Examine each line of the headers
|
|---|
| 220 | my ($line, $name, $value);
|
|---|
| 221 | my @parts;
|
|---|
| 222 | foreach $line (split(/\n/, $Headers)) {
|
|---|
| 223 |
|
|---|
| 224 | # Ignore lines with no content or which begin with whitespace
|
|---|
| 225 | next unless ($line =~ /:/);
|
|---|
| 226 | next if ($line =~ /^\s/);
|
|---|
| 227 |
|
|---|
| 228 | # Find out what metadata is on this line
|
|---|
| 229 | @parts = split(/:/, $line);
|
|---|
| 230 | $name = shift @parts;
|
|---|
| 231 | # get fieldname in canonical form - first cap, then lower case.
|
|---|
| 232 | $name =~ tr/A-Z/a-z/;
|
|---|
| 233 | # uppercase the first character according to the current locale
|
|---|
| 234 | $name=~s/(.+)/\u$1/;
|
|---|
| 235 | next unless $name;
|
|---|
| 236 | next unless ($raw{$name});
|
|---|
| 237 |
|
|---|
| 238 | # Find the value of that metadata
|
|---|
| 239 | $value = join(":", @parts);
|
|---|
| 240 | $value =~ s/^\s+//;
|
|---|
| 241 | $value =~ s/\s+$//;
|
|---|
| 242 | # decode headers if stored using =?<charset>?[BQ]?<data>?= (rfc2047)
|
|---|
| 243 | if ($value =~ /=\?.*\?[BbQq]\?.*\?=/) {
|
|---|
| 244 | my $original_value=$value;
|
|---|
| 245 | my $encoded=$value;
|
|---|
| 246 | $value="";
|
|---|
| 247 | # we should ignore spaces between consecutive encoded-texts
|
|---|
| 248 | $encoded =~ s@\?=\s+=\?@\?==\?@g;
|
|---|
| 249 | while ($encoded =~ s/(.*?)=\?([^\?]*)\?([bq])\?([^\?]+)\?=//i) {
|
|---|
| 250 | my ($charset, $encoding, $data)=($2,$3,$4);
|
|---|
| 251 | my ($decoded_data);
|
|---|
| 252 | $value.="$1"; # any leading chars
|
|---|
| 253 | $data=~s/^\s*//; $data=~s/\s*$//; # strip whitespace from ends
|
|---|
| 254 | chomp $data;
|
|---|
| 255 | $encoding =~ tr/BQ/bq/;
|
|---|
| 256 | if ($encoding eq "q") { # quoted printable
|
|---|
| 257 | $data =~ s/_/\ /g; # from rfc2047 (sec 4.2.2)
|
|---|
| 258 | $decoded_data=qp_decode($data);
|
|---|
| 259 | # qp_decode adds \n, which is default for body text
|
|---|
| 260 | chomp($decoded_data);
|
|---|
| 261 | } else { # base 64
|
|---|
| 262 | $decoded_data=base64_decode($data);
|
|---|
| 263 | }
|
|---|
| 264 | $self->convert2unicode($charset, \$decoded_data);
|
|---|
| 265 | $value .= $decoded_data;
|
|---|
| 266 | } # end of while loop
|
|---|
| 267 |
|
|---|
| 268 | # get any trailing characters
|
|---|
| 269 | $self->convert2unicode($default_header_encoding, \$encoded);
|
|---|
| 270 | $value.=$encoded;
|
|---|
| 271 |
|
|---|
| 272 | if ($value =~ /^\s*$/) { # we couldn't extract anything...
|
|---|
| 273 | $self->convert2unicode($default_header_encoding,
|
|---|
| 274 | \$original_value);
|
|---|
| 275 | $value=$original_value;
|
|---|
| 276 | }
|
|---|
| 277 | } # end of if =?...?=
|
|---|
| 278 |
|
|---|
| 279 | # In the absense of other charset information, assume the
|
|---|
| 280 | # header is the default (usually "iso_8859_1") and convert to unicode.
|
|---|
| 281 | else {
|
|---|
| 282 | $self->convert2unicode($default_header_encoding, \$value);
|
|---|
| 283 | }
|
|---|
| 284 |
|
|---|
| 285 | # Store the metadata
|
|---|
| 286 | $value =~ s@_@\\_@g; # protect against GS macro language
|
|---|
| 287 | $raw{$name} = $value;
|
|---|
| 288 | }
|
|---|
| 289 |
|
|---|
| 290 | # Extract the name and e-mail address from the From metadata
|
|---|
| 291 | my $frommeta = $raw{"From"};
|
|---|
| 292 | my $fromnamemeta;
|
|---|
| 293 | my $fromaddrmeta;
|
|---|
| 294 |
|
|---|
| 295 | $frommeta =~ s/\s*$//; # Remove trailing space, if any
|
|---|
| 296 |
|
|---|
| 297 | if ($frommeta =~ m/(.+)\s*<(.+)>/) {
|
|---|
| 298 | $fromnamemeta=$1;
|
|---|
| 299 | $fromaddrmeta=$2;
|
|---|
| 300 | } elsif ($frommeta =~ m/(.+@.+)\s+\((.*)\)/) {
|
|---|
| 301 | $fromnamemeta=$2;
|
|---|
| 302 | $fromaddrmeta=$1;
|
|---|
| 303 | }
|
|---|
| 304 | if (!defined($fromaddrmeta)) {
|
|---|
| 305 | $fromaddrmeta=$frommeta;
|
|---|
| 306 | }
|
|---|
| 307 | $fromaddrmeta=~s/<//; $fromaddrmeta=~s/>//;
|
|---|
| 308 | # minor attempt to prevent spam-bots from harvesting addresses...
|
|---|
| 309 | $fromaddrmeta=~s/@/@/;
|
|---|
| 310 |
|
|---|
| 311 | $doc_obj->add_utf8_metadata ($cursection, "FromAddr", $fromaddrmeta);
|
|---|
| 312 |
|
|---|
| 313 | if (defined($fromnamemeta) && $fromnamemeta) { # must be > 0 long
|
|---|
| 314 | $fromnamemeta =~ s/\"//g; # remove quotes
|
|---|
| 315 | $fromnamemeta =~ s/\s+$//; # remove trailing whitespace
|
|---|
| 316 | }
|
|---|
| 317 | else {
|
|---|
| 318 | $fromnamemeta = $fromaddrmeta;
|
|---|
| 319 | }
|
|---|
| 320 | # if name is an address
|
|---|
| 321 | $fromnamemeta =~ s/<//g; $fromnamemeta =~ s/>//g;
|
|---|
| 322 | $fromnamemeta=~s/@/@\;/;
|
|---|
| 323 | $doc_obj->add_utf8_metadata ($cursection, "FromName", $fromnamemeta);
|
|---|
| 324 |
|
|---|
| 325 | $raw{"From"}=$frommeta;
|
|---|
| 326 |
|
|---|
| 327 | # Process Date information
|
|---|
| 328 | if ($raw{"Date"} !~ /No Date/) {
|
|---|
| 329 | $raw{"DateText"} = $raw{"Date"};
|
|---|
| 330 |
|
|---|
| 331 | # Convert the date text to internal date format
|
|---|
| 332 | $value = $raw{"Date"};
|
|---|
| 333 | my ($day, $month, $year) = $value =~ /(\d?\d)\s([A-Z][a-z][a-z])\s(\d\d\d?\d?)/;
|
|---|
| 334 | # make some assumptions about the year formatting...
|
|---|
| 335 | # some (old) software thinks 2001 is 101, some think 2001 is 01
|
|---|
| 336 | if ($year < 20) { $year += 2000; } # assume not really 1920...
|
|---|
| 337 | elsif ($year < 150) { $year += 1900; } # assume not really 2150...
|
|---|
| 338 | $raw{"Date"} = &sorttools::format_date($day, $month, $year);
|
|---|
| 339 |
|
|---|
| 340 | } else {
|
|---|
| 341 | # We have not extracted a date
|
|---|
| 342 | $raw{"DateText"} = "Unknown.";
|
|---|
| 343 | $raw{"Date"} = "19000000";
|
|---|
| 344 | }
|
|---|
| 345 |
|
|---|
| 346 | # Add extracted metadata to document object
|
|---|
| 347 | foreach my $name (keys %raw) {
|
|---|
| 348 | $value = $raw{$name};
|
|---|
| 349 | if ($value) {
|
|---|
| 350 | # assume subject, etc headers have no special HTML meaning.
|
|---|
| 351 | $value = &text_into_html($value);
|
|---|
| 352 | # escape [] so it isn't re-interpreted as metadata
|
|---|
| 353 | $value =~ s/\[/[/g; $value =~ s/\]/]/g;
|
|---|
| 354 | } else {
|
|---|
| 355 | $value = "No $name field";
|
|---|
| 356 | }
|
|---|
| 357 | $doc_obj->add_utf8_metadata ($cursection, $name, $value);
|
|---|
| 358 | }
|
|---|
| 359 |
|
|---|
| 360 |
|
|---|
| 361 | # extract a message ID from the headers, if there is one, and we'll use
|
|---|
| 362 | # that as the greenstone doc ID. Having a predictable ID means we can
|
|---|
| 363 | # link to other messages, eg from In-Reply-To or References headers...
|
|---|
| 364 | if ($Headers =~ m@^Message-ID:(.+)$@mi) {
|
|---|
| 365 | my $id=escape_msg_id($1);
|
|---|
| 366 | $doc_obj->{'msgid'}=$id;
|
|---|
| 367 | }
|
|---|
| 368 | # link to another message, if this is a reply
|
|---|
| 369 | if ($Headers =~ m@^In-Reply-To:(.+)$@mi) {
|
|---|
| 370 | my $id=escape_msg_id($1);
|
|---|
| 371 | $doc_obj->add_utf8_metadata ($cursection, 'InReplyTo', $id);
|
|---|
| 372 | } elsif ($Headers =~ m@^References:.*\s([^\s]+)$@mi) {
|
|---|
| 373 | # References can have multiple, get the last one
|
|---|
| 374 | my $id=escape_msg_id($1);
|
|---|
| 375 | # not necessarily in-reply-to, but same thread...
|
|---|
| 376 | $doc_obj->add_utf8_metadata ($cursection, 'InReplyTo', $id);
|
|---|
| 377 | }
|
|---|
| 378 |
|
|---|
| 379 |
|
|---|
| 380 |
|
|---|
| 381 | my $mimetype="text/plain";
|
|---|
| 382 | my $mimeinfo="";
|
|---|
| 383 | my $charset = $default_header_encoding;
|
|---|
| 384 | # Do MIME and encoding stuff. Allow \s in mimeinfo in case there is
|
|---|
| 385 | # more than one parameter given to Content-type.
|
|---|
| 386 | # eg: Content-type: text/plain; charset="us-ascii"; format="flowed"
|
|---|
| 387 | if ($Headers =~ m@^content\-type:\s*([\w\.\-/]+)\s*(\;\s*.+)?\s*$@mi)
|
|---|
| 388 | {
|
|---|
| 389 | $mimetype=$1;
|
|---|
| 390 | $mimetype =~ tr/[A-Z]/[a-z]/;
|
|---|
| 391 |
|
|---|
| 392 | if ($mimetype eq "text") { # for pre-RFC2045 messages (c. 1996)
|
|---|
| 393 | $mimetype = "text/plain";
|
|---|
| 394 | }
|
|---|
| 395 |
|
|---|
| 396 | $mimeinfo=$2;
|
|---|
| 397 | if (!defined $mimeinfo) {
|
|---|
| 398 | $mimeinfo="";
|
|---|
| 399 | } else { # strip leading and trailing stuff
|
|---|
| 400 | $mimeinfo =~ s/^\;\s*//;
|
|---|
| 401 | $mimeinfo =~ s/\s*$//;
|
|---|
| 402 | }
|
|---|
| 403 | if ($mimeinfo =~ /charset=\"([^\"]+)\"/i) {
|
|---|
| 404 | $charset = $1;
|
|---|
| 405 | }
|
|---|
| 406 | }
|
|---|
| 407 |
|
|---|
| 408 | my $transfer_encoding="7bit";
|
|---|
| 409 | if ($Headers =~ /^content-transfer-encoding:\s*([^\s]+)\s*$/mi) {
|
|---|
| 410 | $transfer_encoding=$1;
|
|---|
| 411 | }
|
|---|
| 412 |
|
|---|
| 413 | if ($mimetype eq "text/html") {
|
|---|
| 414 | $$textref= $self->text_from_part($$textref, $Headers);
|
|---|
| 415 | } elsif ($mimetype ne "text/plain") {
|
|---|
| 416 | $self->{'doc_obj'} = $doc_obj; # in case we need to associate files...
|
|---|
| 417 | $$textref=$self->text_from_mime_message($mimetype,$mimeinfo,$$textref);
|
|---|
| 418 | } else { # mimetype eq text/plain
|
|---|
| 419 |
|
|---|
| 420 | if ($transfer_encoding =~ /quoted\-printable/) {
|
|---|
| 421 | $$textref=qp_decode($$textref);
|
|---|
| 422 | } elsif ($transfer_encoding =~ /base64/) {
|
|---|
| 423 | $$textref=base64_decode($$textref);
|
|---|
| 424 | }
|
|---|
| 425 | $self->convert2unicode($charset, $textref);
|
|---|
| 426 |
|
|---|
| 427 | $$textref = &text_into_html($$textref);
|
|---|
| 428 | $$textref =~ s@_@\\_@g; # protect against GS macro language
|
|---|
| 429 |
|
|---|
| 430 | }
|
|---|
| 431 |
|
|---|
| 432 |
|
|---|
| 433 | if ($self->{'header_metadata'} && $self->{'header_metadata'} == 1) {
|
|---|
| 434 | # Add "All headers" metadata
|
|---|
| 435 | $Headers = &text_into_html($Headers);
|
|---|
| 436 |
|
|---|
| 437 | $Headers = "No headers" unless ($Headers =~ /\w/);
|
|---|
| 438 | $Headers =~ s/@/@\;/g;
|
|---|
| 439 | # escape [] so it isn't re-interpreted as metadata
|
|---|
| 440 | $Headers =~ s/\[/[/g; $Headers =~ s/\]/]/g;
|
|---|
| 441 | $self->convert2unicode($charset, \$Headers);
|
|---|
| 442 |
|
|---|
| 443 | $Headers =~ s@_@\\_@g; # protect against GS macro language
|
|---|
| 444 | $doc_obj->add_utf8_metadata ($cursection, "Headers", $Headers);
|
|---|
| 445 | }
|
|---|
| 446 |
|
|---|
| 447 |
|
|---|
| 448 | # Add Title metadata
|
|---|
| 449 | my $Title = text_into_html($raw{'Subject'});
|
|---|
| 450 | $Title .= "<br>From: " . text_into_html($fromnamemeta);
|
|---|
| 451 | $Title .= "<br>Date: " . text_into_html($raw{'DateText'});
|
|---|
| 452 | $Title =~ s/\[/[/g; $Title =~ s/\]/]/g;
|
|---|
| 453 |
|
|---|
| 454 | $doc_obj->add_utf8_metadata ($cursection, "Title", $Title);
|
|---|
| 455 |
|
|---|
| 456 | # Add FileFormat metadata
|
|---|
| 457 | $doc_obj->add_metadata($cursection, "FileFormat", "EMAIL");
|
|---|
| 458 |
|
|---|
| 459 | # Add text to document object
|
|---|
| 460 | $$textref = "No message" unless ($$textref =~ /\w/);
|
|---|
| 461 |
|
|---|
| 462 | $doc_obj->add_utf8_text($cursection, $$textref);
|
|---|
| 463 |
|
|---|
| 464 | return 1;
|
|---|
| 465 | }
|
|---|
| 466 |
|
|---|
| 467 |
|
|---|
| 468 | # Convert a text string into HTML.
|
|---|
| 469 | #
|
|---|
| 470 | # The HTML is going to be inserted into a GML file, so
|
|---|
| 471 | # we have to be careful not to use symbols like ">",
|
|---|
| 472 | # which ocurs frequently in email messages (and use
|
|---|
| 473 | # > instead.
|
|---|
| 474 | #
|
|---|
| 475 | # This function also turns links and email addresses into hyperlinks,
|
|---|
| 476 | # and replaces carriage returns with <BR> tags (and multiple carriage
|
|---|
| 477 | # returns with <P> tags).
|
|---|
| 478 |
|
|---|
| 479 |
|
|---|
| 480 | sub text_into_html {
|
|---|
| 481 | my ($text) = @_;
|
|---|
| 482 |
|
|---|
| 483 | # Convert problem characters into HTML symbols
|
|---|
| 484 | $text =~ s/&/&/g;
|
|---|
| 485 | $text =~ s/</</g;
|
|---|
| 486 | $text =~ s/>/>/g;
|
|---|
| 487 | $text =~ s/\"/"/g;
|
|---|
| 488 |
|
|---|
| 489 | # convert email addresses and URIs into links
|
|---|
| 490 | # don't markup email addresses for now
|
|---|
| 491 | # $text =~ s/([\w\d\.\-]+@[\w\d\.\-]+)/<a href=\"mailto:$1\">$1<\/a>/g;
|
|---|
| 492 |
|
|---|
| 493 | # try to munge email addresses a little bit...
|
|---|
| 494 | $text =~ s/@/@/;
|
|---|
| 495 | # assume hostnames are \.\w\- only, then might have a trailing '/.*'
|
|---|
| 496 | # assume URI doesn't finish with a '.'
|
|---|
| 497 | $text =~ s@((http|ftp|https)://[\w\-]+(\.[\w\-]+)*/?((&|\.)?[\w\?\=\-_/~]+)*(\#[\w\.\-_]*)?)@<a href=\"$1\">$1<\/a>@g;
|
|---|
| 498 |
|
|---|
| 499 |
|
|---|
| 500 | # Clean up whitespace and convert \n charaters to <BR> or <P>
|
|---|
| 501 | $text =~ s/ +/ /g;
|
|---|
| 502 | $text =~ s/\s*$//g;
|
|---|
| 503 | $text =~ s/^\s*//g;
|
|---|
| 504 | $text =~ s/\n/\n<br>/g;
|
|---|
| 505 | $text =~ s/<br>\s*<br>/<p>/gi;
|
|---|
| 506 |
|
|---|
| 507 | return $text;
|
|---|
| 508 | }
|
|---|
| 509 |
|
|---|
| 510 |
|
|---|
| 511 |
|
|---|
| 512 |
|
|---|
| 513 | #Process a MIME message.
|
|---|
| 514 | # the textref we are given DOES NOT include the header.
|
|---|
| 515 | sub text_from_mime_message {
|
|---|
| 516 | my $self = shift(@_);
|
|---|
| 517 | my ($mimetype,$mimeinfo,$text)=(@_);
|
|---|
| 518 | my $outhandle=$self->{'outhandle'};
|
|---|
| 519 | # Check for multiparts - $mimeinfo will be a boundary
|
|---|
| 520 | if ($mimetype =~ /multipart/) {
|
|---|
| 521 | my $boundary="";
|
|---|
| 522 | if ($mimeinfo =~ m@boundary=(\"[^\"]+\"|[^\s]+)\s*$@im) {
|
|---|
| 523 | $boundary=$1;
|
|---|
| 524 | if ($boundary =~ m@^\"@) {
|
|---|
| 525 | $boundary =~ s@^\"@@; $boundary =~ s@\"$@@;
|
|---|
| 526 | }
|
|---|
| 527 | } else {
|
|---|
| 528 | print $outhandle "EMAILPlug: (warning) couldn't parse MIME boundary\n";
|
|---|
| 529 | }
|
|---|
| 530 | # parts start with "--$boundary"
|
|---|
| 531 | # message ends with "--$boundary--"
|
|---|
| 532 | # RFC says boundary is <70 chars, [A-Za-z'()+_,-./:=?], so escape any
|
|---|
| 533 | # that perl might want to interpolate. Also allows spaces...
|
|---|
| 534 | $boundary=~s/\\/\\\\/g;
|
|---|
| 535 | $boundary=~s/([\?\+\.\(\)\:\/\'])/\\$1/g;
|
|---|
| 536 | my @message_parts = split("\r?\n\-\-$boundary", "\n$text");
|
|---|
| 537 | # remove first "part" and last "part" (final --)
|
|---|
| 538 | shift @message_parts;
|
|---|
| 539 | my $last=pop @message_parts;
|
|---|
| 540 | # if our boundaries are a bit dodgy and we only found 1 part...
|
|---|
| 541 | if (!defined($last)) {$last="";}
|
|---|
| 542 | # make sure it is only -- and whitespace
|
|---|
| 543 | if ($last !~ /^\-\-\s*$/ms) {
|
|---|
| 544 | print $outhandle "EMAILPlug: (warning) last part of MIME message isn't empty\n";
|
|---|
| 545 | }
|
|---|
| 546 | foreach my $message_part (@message_parts) {
|
|---|
| 547 | # remove the leading newline left from split.
|
|---|
| 548 | $message_part=~s/^\r?\n//;
|
|---|
| 549 | }
|
|---|
| 550 | if ($mimetype eq "multipart/alternative") {
|
|---|
| 551 | # check for an HTML version first, then TEXT, otherwise use first.
|
|---|
| 552 | my $part_text="";
|
|---|
| 553 | foreach my $message_part (@message_parts) {
|
|---|
| 554 | if ($message_part =~ m@\s*content\-type:\s*text/html@mis)
|
|---|
| 555 | {
|
|---|
| 556 | # Use the HTML version
|
|---|
| 557 | $part_text= $self->text_from_part($message_part);
|
|---|
| 558 | $mimetype="text/html";
|
|---|
| 559 | last;
|
|---|
| 560 | }
|
|---|
| 561 | }
|
|---|
| 562 | if ($part_text eq "") { # try getting a text part instead
|
|---|
| 563 | foreach my $message_part (@message_parts) {
|
|---|
| 564 | if ($message_part =~ m@^content\-type:\s*text/plain@mis)
|
|---|
| 565 | {
|
|---|
| 566 | # Use the plain version
|
|---|
| 567 | $part_text= $self->text_from_part($message_part);
|
|---|
| 568 | if ($part_text =~/[^\s]/) {
|
|---|
| 569 | $part_text = text_into_html($part_text);
|
|---|
| 570 | }
|
|---|
| 571 | $mimetype="text/plain";
|
|---|
| 572 | last;
|
|---|
| 573 | }
|
|---|
| 574 | }
|
|---|
| 575 | }
|
|---|
| 576 | if ($part_text eq "") { #use first part (no html/text part found)
|
|---|
| 577 | $part_text = $self->text_from_part(shift @message_parts);
|
|---|
| 578 | $part_text = text_into_html($part_text);
|
|---|
| 579 | }
|
|---|
| 580 | if ($part_text eq "") { # we couldn't get anything!!!
|
|---|
| 581 | # or it was an empty message...
|
|---|
| 582 | # do nothing...
|
|---|
| 583 | gsprintf($outhandle, "{BasPlug.empty_file} - empty body?\n");
|
|---|
| 584 | } else {
|
|---|
| 585 | $text = $part_text;
|
|---|
| 586 | }
|
|---|
| 587 | } elsif ($mimetype =~ m@multipart/(mixed|digest|related|signed)@) {
|
|---|
| 588 | $text = "";
|
|---|
| 589 | # signed is for PGP/GPG messages... the last part is a hash
|
|---|
| 590 | if ($mimetype =~ m@multipart/signed@) {
|
|---|
| 591 | pop @message_parts;
|
|---|
| 592 | }
|
|---|
| 593 | my $is_first_part=1;
|
|---|
| 594 | foreach my $message_part (@message_parts) {
|
|---|
| 595 | if ($is_first_part && $text ne "") {$is_first_part=0;}
|
|---|
| 596 |
|
|---|
| 597 | if ($mimetype eq "multipart/digest") {
|
|---|
| 598 | # default type - RTFRFC!! Set if not already set
|
|---|
| 599 | $message_part =~ m@^(.*)\n\r?\n@s;
|
|---|
| 600 | my $part_header=$1;
|
|---|
| 601 | if ($part_header !~ m@^content-type@mi) {
|
|---|
| 602 | $message_part="Content-type: message/rfc822\n"
|
|---|
| 603 | . $message_part; # prepend default type
|
|---|
| 604 | }
|
|---|
| 605 | }
|
|---|
| 606 |
|
|---|
| 607 | $text .= $self->process_multipart_part($message_part,
|
|---|
| 608 | $is_first_part);
|
|---|
| 609 | } # foreach message part.
|
|---|
| 610 | } else {
|
|---|
| 611 | # we can't handle this multipart type (not mixed or alternative)
|
|---|
| 612 | # the RFC also mentions "parallel".
|
|---|
| 613 | }
|
|---|
| 614 | } # end of ($mimetype =~ multipart)
|
|---|
| 615 | elsif ($mimetype =~ m@message/rfc822@) {
|
|---|
| 616 | my $msg_header = $text;
|
|---|
| 617 | $msg_header =~ s/\r?\n\r?\n(.*)$//s;
|
|---|
| 618 | $text = $1;
|
|---|
| 619 |
|
|---|
| 620 | if ($msg_header =~ /^content\-type:\s*([\w\.\-\/]+)\s*\;?\s*(.+?)\s*$/mi)
|
|---|
| 621 | {
|
|---|
| 622 | $mimetype=$1;
|
|---|
| 623 | $mimeinfo=$2;
|
|---|
| 624 | $mimetype =~ tr/[A-Z]/[a-z]/;
|
|---|
| 625 |
|
|---|
| 626 | my $msg_text;
|
|---|
| 627 | if ($mimetype =~ m@multipart/@) {
|
|---|
| 628 | $msg_text = text_from_mime_message($self, $mimetype, $mimeinfo,
|
|---|
| 629 | $text);
|
|---|
| 630 | } else {
|
|---|
| 631 | $msg_text=$self->text_from_part($text,$msg_header);
|
|---|
| 632 | }
|
|---|
| 633 |
|
|---|
| 634 | my $brief_header=text_into_html(get_brief_headers($msg_header));
|
|---|
| 635 | $text= "\n<b><<attached message>></b><br>";
|
|---|
| 636 | $text.= "<table><tr><td width=\"5%\"> </td>\n";
|
|---|
| 637 | $text.="<td>" . $brief_header . "\n</p>" . $msg_text
|
|---|
| 638 | . "</td></tr></table>";
|
|---|
| 639 | }
|
|---|
| 640 | } else {
|
|---|
| 641 | # we don't do any processing of the content.
|
|---|
| 642 | }
|
|---|
| 643 |
|
|---|
| 644 | return $text;
|
|---|
| 645 | }
|
|---|
| 646 |
|
|---|
| 647 |
|
|---|
| 648 |
|
|---|
| 649 | # used for turning a message id into a more friendly string for greenstone
|
|---|
| 650 | sub escape_msg_id {
|
|---|
| 651 | #msgid
|
|---|
| 652 | my $id=shift;
|
|---|
| 653 | chomp $id; $id =~ s!\s!!g; # remove spaces
|
|---|
| 654 | $id =~ s![<>\[\]]!!g; # remove [ ] < and >
|
|---|
| 655 | $id =~ s![_&]!-!g; # replace symbols that might cause problems
|
|---|
| 656 | $id =~ s!@!-!g; # replace @ symbol, to avoid spambots
|
|---|
| 657 | return $id;
|
|---|
| 658 | }
|
|---|
| 659 |
|
|---|
| 660 |
|
|---|
| 661 |
|
|---|
| 662 | sub process_multipart_part {
|
|---|
| 663 | my $self = shift;
|
|---|
| 664 | my $message_part = shift;
|
|---|
| 665 | my $is_first_part = shift;
|
|---|
| 666 |
|
|---|
| 667 | my $return_text="";
|
|---|
| 668 | my $part_header=$message_part;
|
|---|
| 669 | my $part_body;
|
|---|
| 670 | if ($message_part=~ /^\s*\n/) {
|
|---|
| 671 | # no header... use defaults
|
|---|
| 672 | $part_body=$message_part;
|
|---|
| 673 | $part_header="Content-type: text/plain; charset=us-ascii";
|
|---|
| 674 | } elsif ($part_header=~s/\r?\n\r?\n(.*)$//s) {
|
|---|
| 675 | $part_body=$1;
|
|---|
| 676 | } else {
|
|---|
| 677 | # something's gone wrong...
|
|---|
| 678 | $part_header="";
|
|---|
| 679 | $part_body=$message_part;
|
|---|
| 680 | }
|
|---|
| 681 |
|
|---|
| 682 | $part_header =~ s/\r?\n[\t\ ]+/ /gs; #unfold
|
|---|
| 683 | my $part_content_type="";
|
|---|
| 684 | my $part_content_info="";
|
|---|
| 685 |
|
|---|
| 686 | if ($part_header =~ m@^content\-type:\s*([\w\.\-/]+)\s*(\;.*)?$@mi) {
|
|---|
| 687 | $part_content_type=$1; $part_content_type =~ tr/A-Z/a-z/;
|
|---|
| 688 | $part_content_info=$2;
|
|---|
| 689 | if (!defined($part_content_info)) {
|
|---|
| 690 | $part_content_info="";
|
|---|
| 691 | } else {
|
|---|
| 692 | $part_content_info =~ s/^\;\s*//;
|
|---|
| 693 | $part_content_info =~ s/\s*$//;
|
|---|
| 694 | }
|
|---|
| 695 | }
|
|---|
| 696 | my $filename="";
|
|---|
| 697 | if ($part_header =~ m@name=\"?([^\"\n]+)\"?@mis) {
|
|---|
| 698 | $filename=$1;
|
|---|
| 699 | $filename =~ s@\r?\s*$@@; # remove trailing space, if any
|
|---|
| 700 | }
|
|---|
| 701 |
|
|---|
| 702 | # disposition - either inline or attachment.
|
|---|
| 703 | # NOT CURRENTLY USED - we display all text types instead...
|
|---|
| 704 | # $part_header =~ /^content\-disposition:\s*([\w+])/mis;
|
|---|
| 705 |
|
|---|
| 706 | # add <<attachment>> to each part except the first...
|
|---|
| 707 | if (!$is_first_part) {
|
|---|
| 708 | $return_text.="\n<p><hr><strong><<attachment>>";
|
|---|
| 709 | # add part info header
|
|---|
| 710 | my $header_text="<br>Type: $part_content_type<br>\n";
|
|---|
| 711 | if ($filename ne "") {
|
|---|
| 712 | $header_text.="Filename: $filename\n";
|
|---|
| 713 | }
|
|---|
| 714 | $header_text =~ s@_@\\_@g;
|
|---|
| 715 | $return_text.=$header_text . "</strong></p>\n<p>\n";
|
|---|
| 716 | }
|
|---|
| 717 |
|
|---|
| 718 | if ($part_content_type =~ m@text/@)
|
|---|
| 719 | {
|
|---|
| 720 | my $part_text= $self->text_from_part($message_part);
|
|---|
| 721 | if ($part_content_type !~ m@text/(ht|x)ml@) {
|
|---|
| 722 | $part_text = text_into_html($part_text);
|
|---|
| 723 | }
|
|---|
| 724 | if ($part_text eq "") {
|
|---|
| 725 | $part_text = ' ';
|
|---|
| 726 | }
|
|---|
| 727 | $return_text .= $part_text;
|
|---|
| 728 | } elsif ($part_content_type =~ m@message/rfc822@) {
|
|---|
| 729 | # This is a forwarded message
|
|---|
| 730 | my $message_part_headers=$part_body;
|
|---|
| 731 | $message_part_headers=~s/\r?\n\r?\n(.*)$//s;
|
|---|
| 732 | my $message_part_body=$1;
|
|---|
| 733 | $message_part_headers =~ s/\r?\n[\t\ ]+/ /gs; #unfold
|
|---|
| 734 |
|
|---|
| 735 | my $rfc822_formatted_body=""; # put result in here
|
|---|
| 736 | if ($message_part_headers =~
|
|---|
| 737 | /^content\-type:\s*([\w\.\-\/]+)\s*(\;.*)?$/ims)
|
|---|
| 738 | {
|
|---|
| 739 | # The message header uses MIME flags
|
|---|
| 740 | my $message_content_type=$1;
|
|---|
| 741 | my $message_content_info=$2;
|
|---|
| 742 | if (!defined($message_content_info)) {
|
|---|
| 743 | $message_content_info="";
|
|---|
| 744 | } else {
|
|---|
| 745 | $message_content_info =~ s/^\;\s*//;
|
|---|
| 746 | $message_content_info =~ s/\s*$//;
|
|---|
| 747 | }
|
|---|
| 748 | $message_content_type =~ tr/A-Z/a-z/;
|
|---|
| 749 | if ($message_content_type =~ /multipart/) {
|
|---|
| 750 | $rfc822_formatted_body=
|
|---|
| 751 | $self->text_from_mime_message($message_content_type,
|
|---|
| 752 | $message_content_info,
|
|---|
| 753 | $message_part_body);
|
|---|
| 754 | } else {
|
|---|
| 755 | $message_part_body=$self->text_from_part($part_body,
|
|---|
| 756 | $message_part_headers);
|
|---|
| 757 | $rfc822_formatted_body=text_into_html($message_part_body);
|
|---|
| 758 | }
|
|---|
| 759 | } else {
|
|---|
| 760 | # message doesn't use MIME flags
|
|---|
| 761 | $rfc822_formatted_body=text_into_html($message_part_body);
|
|---|
| 762 | $rfc822_formatted_body =~ s@_@\\_@g;
|
|---|
| 763 | }
|
|---|
| 764 | # Add the returned text to the output
|
|---|
| 765 | # don't put all the headers...
|
|---|
| 766 | # $message_part_headers =~ s/^(X\-.*|received|message\-id|return\-path):.*\n//img;
|
|---|
| 767 | my $brief_headers=get_brief_headers($message_part_headers);
|
|---|
| 768 | $return_text.=text_into_html($brief_headers);
|
|---|
| 769 | $return_text.="</p><p>\n";
|
|---|
| 770 | $return_text.=$rfc822_formatted_body;
|
|---|
| 771 | $return_text.="</p>\n";
|
|---|
| 772 | # end of message/rfc822
|
|---|
| 773 | } elsif ($part_content_type =~ /multipart/) {
|
|---|
| 774 | # recurse again
|
|---|
| 775 |
|
|---|
| 776 | my $tmptext= $self->text_from_mime_message($part_content_type,
|
|---|
| 777 | $part_content_info,
|
|---|
| 778 | $part_body);
|
|---|
| 779 | $return_text.=$tmptext;
|
|---|
| 780 | } else {
|
|---|
| 781 | # this part isn't text/* or another message...
|
|---|
| 782 | if ($is_first_part) {
|
|---|
| 783 | # this is the first part of a multipart, or only part!
|
|---|
| 784 | $return_text="\n<p><hr><strong><<attachment>>";
|
|---|
| 785 | # add part info header
|
|---|
| 786 | my $header_text="<br>Type: $part_content_type<br>\n";
|
|---|
| 787 | $header_text.="Filename: $filename</strong></p>\n<p>\n";
|
|---|
| 788 | $header_text =~ s@_@\\_@g;
|
|---|
| 789 | $return_text.=$header_text;
|
|---|
| 790 | }
|
|---|
| 791 |
|
|---|
| 792 | # save attachment by default
|
|---|
| 793 | if (!$self->{'ignore_attachments'}
|
|---|
| 794 | && $filename ne "") { # this part has a file...
|
|---|
| 795 | my $encoding="8bit";
|
|---|
| 796 | if ($part_header =~
|
|---|
| 797 | /^content-transfer-encoding:\s*(\w+)/mi ) {
|
|---|
| 798 | $encoding=$1; $encoding =~ tr/A-Z/a-z/;
|
|---|
| 799 | }
|
|---|
| 800 | my $tmpdir=$ENV{'GSDLHOME'} . "/tmp";
|
|---|
| 801 | my $save_filename=$filename;
|
|---|
| 802 |
|
|---|
| 803 | # make sure we don't clobber files with same name;
|
|---|
| 804 | # need to keep state between .mbx files
|
|---|
| 805 | my $assoc_files=$self->{'assoc_filenames'};
|
|---|
| 806 | if ($assoc_files->{$filename}) { # it's been set...
|
|---|
| 807 | $assoc_files->{$filename}++;
|
|---|
| 808 | $filename =~ m/(.+)\.(\w+)$/;
|
|---|
| 809 | my ($filestem, $ext)=($1,$2);
|
|---|
| 810 | $save_filename="${filestem}_"
|
|---|
| 811 | . $assoc_files->{$filename} . ".$ext";
|
|---|
| 812 | } else { # first file with this name
|
|---|
| 813 | $assoc_files->{$filename}=1;
|
|---|
| 814 | }
|
|---|
| 815 | open (SAVE, ">$tmpdir/$save_filename") ||
|
|---|
| 816 | warn "EMAILPlug: Can't save attachment as $tmpdir/$save_filename: $!";
|
|---|
| 817 | my $part_text = $message_part;
|
|---|
| 818 | $part_text =~ s/(.*?)\r?\n\r?\n//s; # remove header
|
|---|
| 819 | if ($encoding eq "base64") {
|
|---|
| 820 | print SAVE base64_decode($part_text);
|
|---|
| 821 | } elsif ($encoding eq "quoted-printable") {
|
|---|
| 822 | print SAVE qp_decode($part_text);
|
|---|
| 823 | } else { # 7bit, 8bit, binary, etc...
|
|---|
| 824 | print SAVE $part_text;
|
|---|
| 825 | }
|
|---|
| 826 | close SAVE;
|
|---|
| 827 | my $doc_obj=$self->{'doc_obj'};
|
|---|
| 828 | $doc_obj->associate_file("$tmpdir/$save_filename",
|
|---|
| 829 | "$save_filename",
|
|---|
| 830 | $part_content_type # mimetype
|
|---|
| 831 | );
|
|---|
| 832 | # clean up tmp area...
|
|---|
| 833 | # Can't do this as it hasn't been copied/linked yet!!!
|
|---|
| 834 | # &util::rm("$tmpdir/$save_filename");
|
|---|
| 835 | my $outhandle=$self->{'outhandle'};
|
|---|
| 836 | print $outhandle "EMAILPlug: saving attachment \"$filename\"\n"; #
|
|---|
| 837 |
|
|---|
| 838 | # be nice if "download" was a translatable macro :(
|
|---|
| 839 | $return_text .="<a href=\"_httpdocimg_/$save_filename\">download</a>";
|
|---|
| 840 | } # end of save attachment
|
|---|
| 841 | } # end of !text/message part
|
|---|
| 842 |
|
|---|
| 843 |
|
|---|
| 844 | return $return_text;
|
|---|
| 845 | }
|
|---|
| 846 |
|
|---|
| 847 |
|
|---|
| 848 | # Return only the "important" headers from a set of message headers
|
|---|
| 849 | sub get_brief_headers {
|
|---|
| 850 | my $msg_header = shift;
|
|---|
| 851 | my $brief_header = "";
|
|---|
| 852 |
|
|---|
| 853 | # Order matters!
|
|---|
| 854 | if ($msg_header =~ /^(From:.*)$/im) {$brief_header.="$1\n";}
|
|---|
| 855 | if ($msg_header =~ /^(To:.*)$/im) {$brief_header.="$1\n";}
|
|---|
| 856 | if ($msg_header =~ /^(Cc:.*)$/im) {$brief_header.="$1\n";}
|
|---|
| 857 | if ($msg_header =~ /^(Subject:.*)$/im) {$brief_header.="$1\n";}
|
|---|
| 858 | if ($msg_header =~ /^(Date:.*)$/im) {$brief_header.="$1\n";}
|
|---|
| 859 |
|
|---|
| 860 | return $brief_header;
|
|---|
| 861 | }
|
|---|
| 862 |
|
|---|
| 863 |
|
|---|
| 864 | # Process a MIME part. Return "" if we can't decode it.
|
|---|
| 865 | # should only be called for parts with type "text/*" ?
|
|---|
| 866 | # Either pass the entire mime part (including the part's header),
|
|---|
| 867 | # or pass the mime part's text and optionally the part's header.
|
|---|
| 868 | sub text_from_part {
|
|---|
| 869 | my $self = shift;
|
|---|
| 870 | my $text = shift || '';
|
|---|
| 871 | my $part_header = shift;
|
|---|
| 872 |
|
|---|
| 873 | my $type="text/plain"; # default, overridden from part header
|
|---|
| 874 | my $charset="ascii"; # default, overridden from part header
|
|---|
| 875 |
|
|---|
| 876 | if (! $part_header) { # no header argument was given. check the body
|
|---|
| 877 | $part_header = $text;
|
|---|
| 878 | # check for empty part header (leading blank line)
|
|---|
| 879 | if ($text =~ /^\s*\r?\n/) {
|
|---|
| 880 | $part_header="Content-type: text/plain; charset=us-ascii";
|
|---|
| 881 | } else {
|
|---|
| 882 | $part_header =~ s/\r?\n\r?\n(.*)$//s;
|
|---|
| 883 | $text=$1; if (!defined($text)) {$text="";}
|
|---|
| 884 | }
|
|---|
| 885 | $part_header =~ s/\r?\n[\t ]+/ /gs; #unfold
|
|---|
| 886 | }
|
|---|
| 887 |
|
|---|
| 888 | if ($part_header =~
|
|---|
| 889 | /content\-type:\s*([\w\.\-\/]+).*?charset=\"?([^\;\"\s]+)\"?/is) {
|
|---|
| 890 | $type=$1;
|
|---|
| 891 | $charset=$2;
|
|---|
| 892 | }
|
|---|
| 893 | my $encoding="";
|
|---|
| 894 | if ($part_header =~ /^content\-transfer\-encoding:\s*([^\s]+)/mis) {
|
|---|
| 895 | $encoding=$1; $encoding=~tr/A-Z/a-z/;
|
|---|
| 896 | }
|
|---|
| 897 | # Content-Transfer-Encoding is per-part
|
|---|
| 898 | if ($encoding ne "") {
|
|---|
| 899 | if ($encoding =~ /quoted\-printable/) {
|
|---|
| 900 | $text=qp_decode($text);
|
|---|
| 901 | } elsif ($encoding =~ /base64/) {
|
|---|
| 902 | $text=base64_decode($text);
|
|---|
| 903 | } elsif ($encoding !~ /[78]bit/) { # leave 7/8 bit as is.
|
|---|
| 904 | # rfc2045 also allows binary, which we ignore (for now).
|
|---|
| 905 | my $outhandle=$self->{'outhandle'};
|
|---|
| 906 | print $outhandle "EMAILPlug: unknown transfer encoding: $encoding\n";
|
|---|
| 907 | return "";
|
|---|
| 908 | }
|
|---|
| 909 | }
|
|---|
| 910 | if ($type eq "text/html") {
|
|---|
| 911 | # only get stuff between <body> tags, or <html> tags.
|
|---|
| 912 | $text =~ s@^.*<html[^>]*>@@is;
|
|---|
| 913 | $text =~ s@</html>.*$@@is;
|
|---|
| 914 | $text =~ s/^.*?<body[^>]*>//si;
|
|---|
| 915 | $text =~ s/<\/body>.*$//si;
|
|---|
| 916 | }
|
|---|
| 917 | elsif ($type eq "text/xml") {
|
|---|
| 918 | $text=~s/</</g;$text=~s/>/>/g;
|
|---|
| 919 | $text="<pre>\n$text\n</pre>\n";
|
|---|
| 920 | }
|
|---|
| 921 | # convert to unicode
|
|---|
| 922 | $self->convert2unicode($charset, \$text);
|
|---|
| 923 |
|
|---|
| 924 | $text =~ s@_@\\_@g; # protect against GS macro language
|
|---|
| 925 | return $text;
|
|---|
| 926 | }
|
|---|
| 927 |
|
|---|
| 928 |
|
|---|
| 929 |
|
|---|
| 930 |
|
|---|
| 931 | # decode quoted-printable text
|
|---|
| 932 | sub qp_decode {
|
|---|
| 933 | my $text=shift;
|
|---|
| 934 |
|
|---|
| 935 | # if a line ends with "=\s*", it is a soft line break, otherwise
|
|---|
| 936 | # keep in any newline characters.
|
|---|
| 937 |
|
|---|
| 938 | $text =~ s/=\s*\r?\n//mg;
|
|---|
| 939 | $text =~ s/=([0-9A-Fa-f]{2})/chr (hex "0x$1")/eg;
|
|---|
| 940 | return $text;
|
|---|
| 941 | }
|
|---|
| 942 |
|
|---|
| 943 | # decode base64 text. This is fairly slow (since it's interpreted perl rather
|
|---|
| 944 | # than compiled XS stuff like in the ::MIME modules, but this is more portable
|
|---|
| 945 | # for us at least).
|
|---|
| 946 | # see rfc2045 for description, but basically, bits 7 and 8 are set to zero;
|
|---|
| 947 | # 4 bytes of encoded text become 3 bytes of binary - remove 2 highest bits
|
|---|
| 948 | # from each encoded byte.
|
|---|
| 949 |
|
|---|
| 950 |
|
|---|
| 951 | sub base64_decode {
|
|---|
| 952 | my $enc_text = shift;
|
|---|
| 953 | # A=>0, B=>1, ..., '+'=>62, '/'=>63
|
|---|
| 954 | # also '=' is used for padding at the end, but we remove it anyway.
|
|---|
| 955 | my $mimechars="ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/";
|
|---|
| 956 | # map each MIME char into it's value, for more efficient lookup.
|
|---|
| 957 | my %index;
|
|---|
| 958 | map { $index{$_} = index ($mimechars, $_) } (split ('', $mimechars));
|
|---|
| 959 | # remove all non-base64 chars. eval to get variable in transliteration...
|
|---|
| 960 | # also remove '=' - we'll assume (!!) that there are no errors in the encoding
|
|---|
| 961 | eval "\$enc_text =~ tr|$mimechars||cd";
|
|---|
| 962 | my $decoded="";
|
|---|
| 963 | while (length ($enc_text)>3)
|
|---|
| 964 | {
|
|---|
| 965 | my $fourchars=substr($enc_text,0,4,"");
|
|---|
| 966 | my @chars=(split '',$fourchars);
|
|---|
| 967 | $decoded.=chr( $index{$chars[0]} << 2 | $index{$chars[1]} >> 4);
|
|---|
| 968 | $decoded.=chr( ($index{$chars[1]} & 15) << 4 | $index{$chars[2]} >> 2);
|
|---|
| 969 | $decoded.=chr( ($index{$chars[2]} & 3 ) << 6 | $index{$chars[3]});
|
|---|
| 970 | }
|
|---|
| 971 | # if there are any input chars left, there are either
|
|---|
| 972 | # 2 encoded bytes (-> 1 raw byte) left or 3 encoded (-> 2 raw) bytes left.
|
|---|
| 973 | my @chars=(split '',$enc_text);
|
|---|
| 974 | if (length($enc_text)) {
|
|---|
| 975 | $decoded.=chr($index{$chars[0]} << 2 | (int $index{$chars[1]} >> 4));
|
|---|
| 976 | }
|
|---|
| 977 | if (length($enc_text)==3) {
|
|---|
| 978 | $decoded.=chr( ($index{$chars[1]} & 15) << 4 | $index{$chars[2]} >> 2);
|
|---|
| 979 | }
|
|---|
| 980 | return $decoded;
|
|---|
| 981 | }
|
|---|
| 982 |
|
|---|
| 983 | sub convert2unicode {
|
|---|
| 984 | my $self = shift(@_);
|
|---|
| 985 | my ($charset, $textref) = @_;
|
|---|
| 986 |
|
|---|
| 987 | if (!$$textref) {
|
|---|
| 988 | # nothing to do!
|
|---|
| 989 | return;
|
|---|
| 990 | }
|
|---|
| 991 |
|
|---|
| 992 | # first get our character encoding name in the right form.
|
|---|
| 993 | $charset = "iso_8859_1" unless defined $charset;
|
|---|
| 994 | $charset =~ tr/A-Z/a-z/; # lowercase
|
|---|
| 995 | $charset =~ s/\-/_/g;
|
|---|
| 996 | if ($charset =~ /gb_?2312/) { $charset="gb" }
|
|---|
| 997 | # assumes EUC-KR, not ISO-2022 !?
|
|---|
| 998 | $charset =~ s/^ks_c_5601_1987/korean/;
|
|---|
| 999 | if ($charset eq 'utf_8') {$charset='utf8'}
|
|---|
| 1000 |
|
|---|
| 1001 | my $outhandle = $self->{'outhandle'};
|
|---|
| 1002 |
|
|---|
| 1003 | if ($charset eq "utf8") {
|
|---|
| 1004 | # no conversion needed, but lets check that it's valid utf8
|
|---|
| 1005 | # see utf-8 manpage for valid ranges
|
|---|
| 1006 | $$textref =~ m/^/g; # to set \G
|
|---|
| 1007 | my $badbytesfound=0;
|
|---|
| 1008 | while ($$textref =~ m!\G.*?([\x80-\xff]+)!sg) {
|
|---|
| 1009 | my $highbytes=$1;
|
|---|
| 1010 | my $highbyteslength=length($highbytes);
|
|---|
| 1011 | # replace any non utf8 complaint bytes
|
|---|
| 1012 | $highbytes =~ /^/g; # set pos()
|
|---|
| 1013 | while ($highbytes =~
|
|---|
| 1014 | m!\G (?: [\xc0-\xdf][\x80-\xbf] | # 2 byte utf-8
|
|---|
| 1015 | [\xe0-\xef][\x80-\xbf]{2} | # 3 byte
|
|---|
| 1016 | [\xf0-\xf7][\x80-\xbf]{3} | # 4 byte
|
|---|
| 1017 | [\xf8-\xfb][\x80-\xbf]{4} | # 5 byte
|
|---|
| 1018 | [\xfc-\xfd][\x80-\xbf]{5} # 6 byte
|
|---|
| 1019 | )*([\x80-\xff])? !xg
|
|---|
| 1020 | ) {
|
|---|
| 1021 | my $badbyte=$1;
|
|---|
| 1022 | if (!defined $badbyte) {next} # hit end of string
|
|---|
| 1023 | my $pos=pos($highbytes);
|
|---|
| 1024 | substr($highbytes, $pos-1, 1, "\xc2\x80");
|
|---|
| 1025 | # update the position to continue searching (for \G)
|
|---|
| 1026 | pos($highbytes) = $pos+1; # set to just after the \x80
|
|---|
| 1027 | $badbytesfound=1;
|
|---|
| 1028 | }
|
|---|
| 1029 | if ($badbytesfound==1) {
|
|---|
| 1030 | # claims to be utf8, but it isn't!
|
|---|
| 1031 | print $outhandle "EMAILPlug: Headers claim utf-8 but bad bytes "
|
|---|
| 1032 | . "detected and removed.\n";
|
|---|
| 1033 |
|
|---|
| 1034 | my $replength=length($highbytes);
|
|---|
| 1035 | my $textpos=pos($$textref);
|
|---|
| 1036 | # replace bad bytes with good bytes
|
|---|
| 1037 | substr( $$textref, $textpos-$replength,
|
|---|
| 1038 | $replength, $highbytes);
|
|---|
| 1039 | # update the position to continue searching (for \G)
|
|---|
| 1040 | pos($$textref)=$textpos+($replength-$highbyteslength);
|
|---|
| 1041 | }
|
|---|
| 1042 | }
|
|---|
| 1043 | return;
|
|---|
| 1044 | }
|
|---|
| 1045 |
|
|---|
| 1046 | # It appears that we can't always trust ascii text so we'll treat it
|
|---|
| 1047 | # as iso-8859-1 (letting characters above 0x80 through without
|
|---|
| 1048 | # converting them to utf-8 will result in invalid XML documents
|
|---|
| 1049 | # which can't be parsed at build time).
|
|---|
| 1050 | $charset = "iso_8859_1" if ($charset eq "us_ascii" || $charset eq "ascii");
|
|---|
| 1051 |
|
|---|
| 1052 | if ($charset eq "iso_8859_1") {
|
|---|
| 1053 | # test if the mailer lied, and it has win1252 chars in it...
|
|---|
| 1054 | # 1252 has characters between 0x80 and 0x9f, 8859-1 doesn't
|
|---|
| 1055 | if ($$textref =~ m/[\x80-\x9f]/) {
|
|---|
| 1056 | print $outhandle "EMAILPlug: Headers claim ISO charset but MS ";
|
|---|
| 1057 | print $outhandle "codepage 1252 detected.\n";
|
|---|
| 1058 | $charset = "windows_1252";
|
|---|
| 1059 | }
|
|---|
| 1060 | }
|
|---|
| 1061 | my $utf8_text=&unicode::unicode2utf8(&unicode::convert2unicode($charset,$textref));
|
|---|
| 1062 |
|
|---|
| 1063 | if ($utf8_text ne "") {
|
|---|
| 1064 | $$textref=$utf8_text;
|
|---|
| 1065 | } else {
|
|---|
| 1066 | # we didn't get any text... unsupported encoding perhaps? Or it is
|
|---|
| 1067 | # empty anyway. We'll try to continue, assuming 8859-1. We could strip
|
|---|
| 1068 | # characters out here if this causes problems...
|
|---|
| 1069 | my $outhandle=$self->{'outhandle'};
|
|---|
| 1070 | print $outhandle "EMAILPlug: falling back to iso-8859-1\n";
|
|---|
| 1071 | $$textref=&unicode::unicode2utf8(&unicode::convert2unicode("iso_8859_1",$textref));
|
|---|
| 1072 |
|
|---|
| 1073 | }
|
|---|
| 1074 | }
|
|---|
| 1075 |
|
|---|
| 1076 |
|
|---|
| 1077 | sub set_OID {
|
|---|
| 1078 | my $self = shift (@_);
|
|---|
| 1079 | my ($doc_obj, $id, $segment_number) = @_;
|
|---|
| 1080 |
|
|---|
| 1081 | if ( exists $doc_obj->{'msgid'} ) {
|
|---|
| 1082 | $doc_obj->set_OID($doc_obj->{'msgid'});
|
|---|
| 1083 | } else {
|
|---|
| 1084 | $doc_obj->set_OID("$id\_$segment_number");
|
|---|
| 1085 | }
|
|---|
| 1086 | }
|
|---|
| 1087 |
|
|---|
| 1088 |
|
|---|
| 1089 | # Perl packages have to return true if they are run.
|
|---|
| 1090 | 1;
|
|---|
