| 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) 1999-2001 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
|
|---|
| 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 | # $Date Date: header in GSDL format (eg: 19990924)
|
|---|
| 50 | #
|
|---|
| 51 | #
|
|---|
| 52 | # John McPherson - June/July 2001
|
|---|
| 53 | # added (basic) MIME support and quoted-printable and base64 decodings.
|
|---|
| 54 | # Minor fixes for names that are actually email addresses (ie <...> was lost)
|
|---|
| 55 | #
|
|---|
| 56 | # See: * RFC 822 - ARPA Internet Text Messages
|
|---|
| 57 | # * RFC 2045 - Multipurpose Internet Mail Extensions (MIME) -part1
|
|---|
| 58 | # * RFC 2046 - MIME (part 2) Media Types (and multipart messages)
|
|---|
| 59 | # * RFC 2047 - MIME (part 3) Message Header Extensions
|
|---|
| 60 | # * RFC 1806 - Content Dispositions (ie inline/attachment)
|
|---|
| 61 | package EMAILPlug;
|
|---|
| 62 |
|
|---|
| 63 | use SplitPlug;
|
|---|
| 64 |
|
|---|
| 65 | use unicode;
|
|---|
| 66 |
|
|---|
| 67 | use sorttools;
|
|---|
| 68 | use util;
|
|---|
| 69 |
|
|---|
| 70 |
|
|---|
| 71 | # EMAILPlug is a sub-class of SplitPlug.
|
|---|
| 72 |
|
|---|
| 73 | sub BEGIN {
|
|---|
| 74 | @ISA = ('SplitPlug');
|
|---|
| 75 | }
|
|---|
| 76 |
|
|---|
| 77 | # Create a new EMAILPlug object with which to parse a file.
|
|---|
| 78 | # Accomplished by creating a new BasPlug and using bless to
|
|---|
| 79 | # turn it into an EMAILPlug.
|
|---|
| 80 |
|
|---|
| 81 | sub new {
|
|---|
| 82 | my ($class) = @_;
|
|---|
| 83 | my $self = new BasPlug ("EMAILPlug", @_);
|
|---|
| 84 | # this might not actually be true at read-time, but after processing
|
|---|
| 85 | # it should all be utf8.
|
|---|
| 86 | $self->{'input_encoding'}="utf8";
|
|---|
| 87 | return bless $self, $class;
|
|---|
| 88 | }
|
|---|
| 89 |
|
|---|
| 90 | sub get_default_process_exp {
|
|---|
| 91 | my $self = shift (@_);
|
|---|
| 92 | # mbx/email for mailbox file format, \d+ for maildir (each message is
|
|---|
| 93 | # in a separate file, with a unique number for filename)
|
|---|
| 94 | return q@([\\/]\d+|\.(mbx|email))$@;
|
|---|
| 95 | }
|
|---|
| 96 |
|
|---|
| 97 | # This plugin splits the mbox mail files at lines starting with From<sp>
|
|---|
| 98 | sub get_default_split_exp {
|
|---|
| 99 | return q^\nFrom .*\n^;
|
|---|
| 100 | }
|
|---|
| 101 |
|
|---|
| 102 |
|
|---|
| 103 | # do plugin specific processing of doc_obj
|
|---|
| 104 | sub process {
|
|---|
| 105 |
|
|---|
| 106 | my $self = shift (@_);
|
|---|
| 107 | my ($textref, $pluginfo, $base_dir, $file, $metadata, $doc_obj) = @_;
|
|---|
| 108 | my $outhandle = $self->{'outhandle'};
|
|---|
| 109 |
|
|---|
| 110 | # Check that we're dealing with a valid mail file
|
|---|
| 111 | return undef unless (($$textref =~ /From:/m) || ($$textref =~ /To:/m));
|
|---|
| 112 |
|
|---|
| 113 | # slightly more strict validity check, to prevent us from matching
|
|---|
| 114 | # .so.x files ...
|
|---|
| 115 | return undef unless (($$textref =~ /^From /) ||
|
|---|
| 116 | ($$textref =~ /^[-A-Za-z]{2,100}:/m));
|
|---|
| 117 |
|
|---|
| 118 | print $outhandle "EMAILPlug: processing $file\n"
|
|---|
| 119 | if $self->{'verbosity'} > 1;
|
|---|
| 120 |
|
|---|
| 121 | my $cursection = $doc_obj->get_top_section();
|
|---|
| 122 |
|
|---|
| 123 | #
|
|---|
| 124 | # Parse the document's text and extract metadata
|
|---|
| 125 | #
|
|---|
| 126 |
|
|---|
| 127 | # Protect backslashes
|
|---|
| 128 | $$textref =~ s@\\@\\\\@g;
|
|---|
| 129 |
|
|---|
| 130 | # Separate header from body of message
|
|---|
| 131 | my $Headers = $$textref;
|
|---|
| 132 | $Headers =~ s/\r?\n\r?\n(.*)$//s;
|
|---|
| 133 | $$textref = $1;
|
|---|
| 134 |
|
|---|
| 135 | # See if headers include non-ascii - RFC says whole header should be ascii.
|
|---|
| 136 | # not yet implemented, as we don't know what character set is the
|
|---|
| 137 | # user's default... We can do textcat to guess, or we can just choose
|
|---|
| 138 | # one of the charset fields later in the document (if there are any...)
|
|---|
| 139 | # if ($Headers =~ /([[:^ascii:]])/) {
|
|---|
| 140 | # }
|
|---|
| 141 |
|
|---|
| 142 | # Unfold headers - see rfc822
|
|---|
| 143 | $Headers =~ s/\r?\n[\t\ ]+/ /gs;
|
|---|
| 144 | # Extract basic metadata from header
|
|---|
| 145 | my @headers = ("From", "To", "Subject", "Date");
|
|---|
| 146 | my %raw;
|
|---|
| 147 | foreach my $name (@headers) {
|
|---|
| 148 | $raw{$name} = "No $name value";
|
|---|
| 149 | }
|
|---|
| 150 |
|
|---|
| 151 | # Examine each line of the headers
|
|---|
| 152 | my ($line, $name, $value);
|
|---|
| 153 | my @parts;
|
|---|
| 154 | foreach $line (split(/\n/, $Headers)) {
|
|---|
| 155 |
|
|---|
| 156 | # Ignore lines with no content or which begin with whitespace
|
|---|
| 157 | next unless ($line =~ /:/);
|
|---|
| 158 | next if ($line =~ /^\s/);
|
|---|
| 159 |
|
|---|
| 160 | # Find out what metadata is on this line
|
|---|
| 161 | @parts = split(/:/, $line);
|
|---|
| 162 | $name = shift @parts;
|
|---|
| 163 | # uppercase the first character according to the current locale
|
|---|
| 164 | $name=~s/(.+)/\u$1/;
|
|---|
| 165 | next unless $name;
|
|---|
| 166 | next unless ($raw{$name});
|
|---|
| 167 |
|
|---|
| 168 | # Find the value of that metadata
|
|---|
| 169 | $value = join(":", @parts);
|
|---|
| 170 | $value =~ s/^\s+//;
|
|---|
| 171 | $value =~ s/\s+$//;
|
|---|
| 172 |
|
|---|
| 173 | # decode headers if stored using =?<charset>?[BQ]?<data>?= (rfc2047)
|
|---|
| 174 | if ($value =~ /=\?/) {
|
|---|
| 175 | my $original_value=$value;
|
|---|
| 176 | my $encoded=$value;
|
|---|
| 177 | $value="";
|
|---|
| 178 | # this isn't quite right yet regarding spaces between encoded-texts
|
|---|
| 179 | # (see examples, section 8. of rfc).
|
|---|
| 180 | while ($encoded =~ s/(.*?)=\?([^\?]*)\?([bq])\?([^\?]+)\?=\s*//i) {
|
|---|
| 181 | my ($charset, $encoding, $data)=($2,$3,$4);
|
|---|
| 182 | my ($decoded_data);
|
|---|
| 183 | $value.="$1"; # any leading chars
|
|---|
| 184 | $data=~s/^\s*//; $data=~s/\s*$//; # strip whitespace from ends
|
|---|
| 185 | chomp $data;
|
|---|
| 186 | $encoding =~ tr/BQ/bq/;
|
|---|
| 187 | if ($encoding eq "q") { # quoted printable
|
|---|
| 188 | $data =~ s/_/\ /g; # from rfc2047 (sec 4.2.2)
|
|---|
| 189 | $decoded_data=qp_decode($data);
|
|---|
| 190 | } else { # base 64
|
|---|
| 191 | $decoded_data=base64_decode($data);
|
|---|
| 192 | }
|
|---|
| 193 | $self->convert2unicode($charset, \$decoded_data);
|
|---|
| 194 | $value .= $decoded_data;
|
|---|
| 195 | } # end of while loop
|
|---|
| 196 |
|
|---|
| 197 | # get any trailing characters
|
|---|
| 198 | $self->convert2unicode("iso_8859_1", \$encoded);
|
|---|
| 199 | $value.=$encoded;
|
|---|
| 200 |
|
|---|
| 201 | if ($value =~ /^\s*$/) { # we couldn't extract anything...
|
|---|
| 202 | $self->convert2unicode("iso_8859_1", \$original_value);
|
|---|
| 203 | $value=original_value;
|
|---|
| 204 | }
|
|---|
| 205 | } # end of if =?...?=
|
|---|
| 206 |
|
|---|
| 207 | # Store the metadata
|
|---|
| 208 | $raw{$name} = $value;
|
|---|
| 209 | }
|
|---|
| 210 |
|
|---|
| 211 | # Extract the name and e-mail address from the From metadata
|
|---|
| 212 | $frommeta = $raw{"From"};
|
|---|
| 213 | my $fromnamemeta;
|
|---|
| 214 | my $fromaddrmeta;
|
|---|
| 215 |
|
|---|
| 216 | $frommeta =~ s/\s*$//; # Remove trailing space, if any
|
|---|
| 217 |
|
|---|
| 218 | if ($frommeta =~ m/(.+)\s*<(.+)>/) {
|
|---|
| 219 | $fromnamemeta=$1;
|
|---|
| 220 | $fromaddrmeta=$2;
|
|---|
| 221 | } elsif ($frommeta =~ m/(.+@.+)\s+\((.*)\)/) {
|
|---|
| 222 | $fromnamemeta=$2;
|
|---|
| 223 | $fromaddrmeta=$1;
|
|---|
| 224 | }
|
|---|
| 225 | if (!defined($fromaddrmeta)) {
|
|---|
| 226 | $fromaddrmeta=$frommeta;
|
|---|
| 227 | }
|
|---|
| 228 | $fromaddrmeta=~s/<//; $fromaddrmeta=~s/>//;
|
|---|
| 229 | # minor attempt to prevent spam-bots from harvesting addresses...
|
|---|
| 230 | $fromaddrmeta=~s/@/@/;
|
|---|
| 231 | $doc_obj->add_utf8_metadata ($cursection, "FromAddr", $fromaddrmeta);
|
|---|
| 232 |
|
|---|
| 233 | if (defined($fromnamemeta)) {
|
|---|
| 234 | $fromnamemeta =~ s/\"//g;
|
|---|
| 235 | }
|
|---|
| 236 | else {
|
|---|
| 237 | $fromnamemeta = $fromaddrmeta;
|
|---|
| 238 | }
|
|---|
| 239 | # if name is an address
|
|---|
| 240 | $fromnamemeta =~ s/<//g; $fromnamemeta =~ s/>//g;
|
|---|
| 241 | $fromnamemeta=~s/@/@\;/;
|
|---|
| 242 | $doc_obj->add_utf8_metadata ($cursection, "FromName", $fromnamemeta);
|
|---|
| 243 |
|
|---|
| 244 | # Escape < and > in the whole From field;
|
|---|
| 245 | $raw{"From"}=$frommeta;
|
|---|
| 246 |
|
|---|
| 247 | # Process Date information
|
|---|
| 248 | if ($raw{"Date"} !~ /No Date/) {
|
|---|
| 249 | $raw{"DateText"} = $raw{"Date"};
|
|---|
| 250 |
|
|---|
| 251 | # Convert the date text to internal date format
|
|---|
| 252 | $value = $raw{"Date"};
|
|---|
| 253 | my ($day, $month, $year) = $value =~ /(\d?\d)\s([A-Z][a-z][a-z])\s(\d\d\d?\d?)/;
|
|---|
| 254 | if ($year < 100) { $year += 1900; }
|
|---|
| 255 | $raw{"Date"} = &sorttools::format_date($day, $month, $year);
|
|---|
| 256 |
|
|---|
| 257 | } else {
|
|---|
| 258 | # We have not extracted a date
|
|---|
| 259 | $raw{"DateText"} = "Unknown.";
|
|---|
| 260 | $raw{"Date"} = "19000000";
|
|---|
| 261 | }
|
|---|
| 262 |
|
|---|
| 263 | # Add extracted metadata to document object
|
|---|
| 264 | foreach my $name (keys %raw) {
|
|---|
| 265 | $value = $raw{$name};
|
|---|
| 266 | if ($value) {
|
|---|
| 267 | # assume subject, etc headers have no special HTML meaning.
|
|---|
| 268 | $value = &text_into_html($value);
|
|---|
| 269 | # escape [] so it isn't re-interpreted as metadata
|
|---|
| 270 | $value =~ s/\[/[/g; $value =~ s/\]/]/g;
|
|---|
| 271 | } else {
|
|---|
| 272 | $value = "No $name field";
|
|---|
| 273 | }
|
|---|
| 274 | $doc_obj->add_utf8_metadata ($cursection, $name, $value);
|
|---|
| 275 | }
|
|---|
| 276 |
|
|---|
| 277 | my $mimetype="text/plain";
|
|---|
| 278 | my $mimeinfo="";
|
|---|
| 279 | my $charset = "iso_8859_1";
|
|---|
| 280 | # Do MIME and encoding stuff
|
|---|
| 281 | if ($Headers =~ /^content\-type:\s*([\w\/\-]+)\s*\;?\s*(.+?)\s*$/mi)
|
|---|
| 282 | {
|
|---|
| 283 | $mimetype=$1;
|
|---|
| 284 | $mimetype =~ tr/[A-Z]/[a-z]/;
|
|---|
| 285 | $mimeinfo=$2;
|
|---|
| 286 | if ($mimeinfo =~ /charset=\"([^\"]+)\"/) {
|
|---|
| 287 | $charset = $1;
|
|---|
| 288 | }
|
|---|
| 289 | }
|
|---|
| 290 |
|
|---|
| 291 | my $transfer_encoding="7bit";
|
|---|
| 292 | if ($Headers =~ /^content-transfer-encoding:\s*([^\s]+)\s*$/mi) {
|
|---|
| 293 | $transfer_encoding=$1;
|
|---|
| 294 | }
|
|---|
| 295 | if ($mimetype eq "text/html") {
|
|---|
| 296 | $$textref= $self->text_from_part("$Headers\n$$textref");
|
|---|
| 297 | } elsif ($mimetype ne "text/plain") {
|
|---|
| 298 | $$textref= $self->text_from_mime_message($mimetype,$mimeinfo,$$textref,
|
|---|
| 299 | $outhandle);
|
|---|
| 300 | } elsif ($transfer_encoding =~ /quoted\-printable/) {
|
|---|
| 301 | $$textref=qp_decode($$textref);
|
|---|
| 302 | $self->convert2unicode($charset, $textref);
|
|---|
| 303 | } elsif ($transfer_encoding =~ /base64/) {
|
|---|
| 304 | $$textref=base64_decode($$textref);
|
|---|
| 305 | $self->convert2unicode($charset, $textref);
|
|---|
| 306 | } else {
|
|---|
| 307 | $self->convert2unicode($charset, $textref);
|
|---|
| 308 | }
|
|---|
| 309 |
|
|---|
| 310 |
|
|---|
| 311 | # Add "All headers" metadata
|
|---|
| 312 | $Headers = &text_into_html($Headers);
|
|---|
| 313 |
|
|---|
| 314 | $Headers = "No headers" unless ($Headers =~ /\w/);
|
|---|
| 315 | $Headers =~ s/@/@\;/g;
|
|---|
| 316 | # escape [] so it isn't re-interpreted as metadata
|
|---|
| 317 | $Headers =~ s/\[/[/g; $Headers =~ s/\]/]/g;
|
|---|
| 318 |
|
|---|
| 319 | $doc_obj->add_utf8_metadata ($cursection, "Headers", $Headers);
|
|---|
| 320 |
|
|---|
| 321 | # Add text to document object
|
|---|
| 322 | if ($mimetype eq "text/plain") {
|
|---|
| 323 | $$textref = &text_into_html($$textref);
|
|---|
| 324 | }
|
|---|
| 325 | $$textref = "No message" unless ($$textref =~ /\w/);
|
|---|
| 326 | $doc_obj->add_utf8_text($cursection, $$textref);
|
|---|
| 327 |
|
|---|
| 328 | return 1;
|
|---|
| 329 | }
|
|---|
| 330 |
|
|---|
| 331 |
|
|---|
| 332 | # Convert a text string into HTML.
|
|---|
| 333 | #
|
|---|
| 334 | # The HTML is going to be inserted into a GML file, so
|
|---|
| 335 | # we have to be careful not to use symbols like ">",
|
|---|
| 336 | # which ocurs frequently in email messages (and use
|
|---|
| 337 | # > instead.
|
|---|
| 338 | #
|
|---|
| 339 | # This function also turns links and email addresses into hyperlinks,
|
|---|
| 340 | # and replaces carriage returns with <BR> tags (and multiple carriage
|
|---|
| 341 | # returns with <P> tags).
|
|---|
| 342 |
|
|---|
| 343 |
|
|---|
| 344 | sub text_into_html {
|
|---|
| 345 | my ($text) = @_;
|
|---|
| 346 |
|
|---|
| 347 | # Convert problem characters into HTML symbols
|
|---|
| 348 | $text =~ s/&/&/go;
|
|---|
| 349 | $text =~ s/</</go;
|
|---|
| 350 | $text =~ s/>/>/go;
|
|---|
| 351 | $text =~ s/\"/"/go;
|
|---|
| 352 |
|
|---|
| 353 | # convert email addresses and URIs into links
|
|---|
| 354 | # don't markup email addresses for now
|
|---|
| 355 | # $text =~ s/([\w\d\.\-]+@[\w\d\.\-]+)/<a href=\"mailto:$1\">$1<\/a>/g;
|
|---|
| 356 |
|
|---|
| 357 | # assume hostnames are \.\w\- only, then might have a trailing '/.*'
|
|---|
| 358 | # assume URI doesn't finish with a '.'
|
|---|
| 359 | $text =~ s@((http|ftp|https)://[\w\-]+(\.[\w\-]+)*/?((&|\.)?[\w\?\=\-_/~]+)*)@<a href=\"$1\">$1<\/a>@g;
|
|---|
| 360 |
|
|---|
| 361 |
|
|---|
| 362 | # Clean up whitespace and convert \n charaters to <BR> or <P>
|
|---|
| 363 | $text =~ s/ +/ /go;
|
|---|
| 364 | $text =~ s/\s*$//o;
|
|---|
| 365 | $text =~ s/^\s*//o;
|
|---|
| 366 | $text =~ s/\n/\n<BR>/go;
|
|---|
| 367 | $text =~ s/<BR>\s*<BR>/<P>/go;
|
|---|
| 368 |
|
|---|
| 369 | return $text;
|
|---|
| 370 | }
|
|---|
| 371 |
|
|---|
| 372 |
|
|---|
| 373 |
|
|---|
| 374 |
|
|---|
| 375 | #Process a MIME message.
|
|---|
| 376 | # the textref we are given DOES NOT include the header.
|
|---|
| 377 | sub text_from_mime_message {
|
|---|
| 378 | my $self = shift(@_);
|
|---|
| 379 | my ($mimetype,$mimeinfo,$text,$outhandle)=(@_);
|
|---|
| 380 |
|
|---|
| 381 | # Check for multiparts - $mimeinfo will be a boundary
|
|---|
| 382 | if ($mimetype =~ /multipart/) {
|
|---|
| 383 | $boundary="";
|
|---|
| 384 | if ($mimeinfo =~ m@boundary=(\"[^\"]+\"|[^\s]+)\s*$@im) {
|
|---|
| 385 | $boundary=$1;
|
|---|
| 386 | if ($boundary =~ m@^\"@) {
|
|---|
| 387 | $boundary =~ s@^\"@@; $boundary =~ s@\"$@@;
|
|---|
| 388 | }
|
|---|
| 389 | } else {
|
|---|
| 390 | print $outhandle "EMAILPlug: (warning) couldn't parse MIME boundary\n";
|
|---|
| 391 | }
|
|---|
| 392 | # parts start with "--$boundary"
|
|---|
| 393 | # message ends with "--$boundary--"
|
|---|
| 394 | # RFC says boundary is <70 chars, [A-Za-z'()+_,-./:=?], so escape any
|
|---|
| 395 | # that perl might want to interpolate. Also allows spaces...
|
|---|
| 396 | $boundary=~s/\\/\\\\/g;
|
|---|
| 397 | $boundary=~s/([\?\+\.\(\)\:\/\'])/\\$1/g;
|
|---|
| 398 | my @message_parts = split("\r?\n\-\-$boundary", "\n$text");
|
|---|
| 399 | # remove first "part" and last "part" (final --)
|
|---|
| 400 | shift @message_parts;
|
|---|
| 401 | my $last=pop @message_parts;
|
|---|
| 402 | # if our boundaries are a bit dodgy and we only found 1 part...
|
|---|
| 403 | if (!defined($last)) {$last="";}
|
|---|
| 404 | # make sure it is only -- and whitespace
|
|---|
| 405 | if ($last !~ /^\-\-\s*$/ms) {
|
|---|
| 406 | print $outhandle "EMAILPlug: (warning) last part of MIME message isn't empty\n";
|
|---|
| 407 | }
|
|---|
| 408 | foreach my $message_part (@message_parts) {
|
|---|
| 409 | # remove the leading newline left from split.
|
|---|
| 410 | $message_part=~s/^\r?\n//;
|
|---|
| 411 | }
|
|---|
| 412 | if ($mimetype eq "multipart/alternative") {
|
|---|
| 413 | # check for an HTML version first, then TEXT, otherwise use first.
|
|---|
| 414 | my $part_text="";
|
|---|
| 415 | foreach my $message_part (@message_parts) {
|
|---|
| 416 | if ($message_part =~ m@\s*content\-type:\s*text/html@mis)
|
|---|
| 417 | {
|
|---|
| 418 | # Use the HTML version
|
|---|
| 419 | $part_text= $self->text_from_part($message_part);
|
|---|
| 420 | $mimetype="text/html";
|
|---|
| 421 | last;
|
|---|
| 422 | }
|
|---|
| 423 | }
|
|---|
| 424 | if ($part_text eq "") { # try getting a text part instead
|
|---|
| 425 | foreach my $message_part (@message_parts) {
|
|---|
| 426 | if ($message_part =~ m@^content\-type:\s*text/plain@mis)
|
|---|
| 427 | {
|
|---|
| 428 | # Use the plain version
|
|---|
| 429 | $part_text= $self->text_from_part($message_part);
|
|---|
| 430 | if ($part_text =~/[^\s]/) {
|
|---|
| 431 | $part_text="<pre>".$part_text."</pre>";
|
|---|
| 432 | }
|
|---|
| 433 | $mimetype="text/plain";
|
|---|
| 434 | last;
|
|---|
| 435 | }
|
|---|
| 436 | }
|
|---|
| 437 | }
|
|---|
| 438 | if ($part_text eq "") { # use first part
|
|---|
| 439 | $part_text= $self->text_from_part(shift @message_parts);
|
|---|
| 440 | }
|
|---|
| 441 | if ($part_text eq "") { # we couldn't get anything!!!
|
|---|
| 442 | # or it was an empty message...
|
|---|
| 443 | # do nothing...
|
|---|
| 444 | print $outhandle "EMAILPlug: no text - empty body?\n";
|
|---|
| 445 | } else {
|
|---|
| 446 | $text=$part_text;
|
|---|
| 447 | }
|
|---|
| 448 | } elsif ($mimetype =~ m@multipart/(mixed|digest|related)@) {
|
|---|
| 449 | $text="";
|
|---|
| 450 | foreach my $message_part (@message_parts) {
|
|---|
| 451 | my $part_header=$message_part;
|
|---|
| 452 | my $part_body;
|
|---|
| 453 | if ($message_part=~ /^\s*\n/) {
|
|---|
| 454 | # no header... use defaults
|
|---|
| 455 | $part_body=$message_part;
|
|---|
| 456 | $part_header="Content-type: text/plain; charset=us-ascii";
|
|---|
| 457 | } elsif ($part_header=~s/\r?\n\r?\n(.*)$//sg) {
|
|---|
| 458 | $part_body=$1;
|
|---|
| 459 | } else {
|
|---|
| 460 | # something's gone wrong...
|
|---|
| 461 | $part_header="";
|
|---|
| 462 | $part_body=$message_part;
|
|---|
| 463 | }
|
|---|
| 464 |
|
|---|
| 465 | $part_header =~ s/\r?\n[\t\ ]+/ /gs; #unfold
|
|---|
| 466 | my $part_content_type="";
|
|---|
| 467 | my $part_content_info="";
|
|---|
| 468 | if ($mimetype eq "multipart/digest") {
|
|---|
| 469 | # default type - RTFRFC!!
|
|---|
| 470 | $part_content_type="message/rfc822";
|
|---|
| 471 | }
|
|---|
| 472 | if ($part_header =~ m@^content\-type:\s*([\w+/\-]+)\s*\;?\s*(.*?)\s*$@mi) {
|
|---|
| 473 | $part_content_type=$1; $part_content_type =~ tr/A-Z/a-z/;
|
|---|
| 474 | $part_content_info=$2;
|
|---|
| 475 | }
|
|---|
| 476 | my $filename="";
|
|---|
| 477 | if ($part_header =~ m@name=\"?([\w\.\-\\/]+)\"?@mis) {
|
|---|
| 478 | $filename=$1;
|
|---|
| 479 | }
|
|---|
| 480 |
|
|---|
| 481 | # disposition - either inline or attachment.
|
|---|
| 482 | # NOT CURRENTLY USED - we display all text types instead...
|
|---|
| 483 | # $part_header =~ /^content\-disposition:\s*([\w+])/mis;
|
|---|
| 484 |
|
|---|
| 485 | # add <<attachment>> to each part except the first...
|
|---|
| 486 | if ($text ne "") {
|
|---|
| 487 | $text.="\n<p><hr><strong><<attachment>></>";
|
|---|
| 488 | # add part info header
|
|---|
| 489 | $text.="<br>Type: $part_content_type<br>\n";
|
|---|
| 490 | if ($filename ne "") {
|
|---|
| 491 | $text.="Filename: $filename\n";
|
|---|
| 492 | }
|
|---|
| 493 | $text.="</strong></p>\n";
|
|---|
| 494 | }
|
|---|
| 495 |
|
|---|
| 496 | if ($part_content_type =~ m@text/@)
|
|---|
| 497 | {
|
|---|
| 498 | my $part_text= $self->text_from_part($message_part);
|
|---|
| 499 | if ($part_content_type !~ m@text/(ht|x)ml@) {
|
|---|
| 500 | $part_text=text_into_html($part_text);
|
|---|
| 501 | }
|
|---|
| 502 | if ($part_text eq "") {
|
|---|
| 503 | $part_text='<<empty message>>';
|
|---|
| 504 | }
|
|---|
| 505 | $text.=$part_text;
|
|---|
| 506 | } elsif ($part_content_type =~ m@message/rfc822@) {
|
|---|
| 507 | # This is a forwarded message
|
|---|
| 508 | my $message_part_headers=$part_body;
|
|---|
| 509 | $message_part_headers =~ s/\r?\n[\t\ ]+/ /gs; #unfold
|
|---|
| 510 | $message_part_headers=~s/\r?\n\r?\n(.*)$//s;
|
|---|
| 511 | my $message_part_body=$1;
|
|---|
| 512 |
|
|---|
| 513 | my $rfc822_formatted_body=""; # put result in here
|
|---|
| 514 | if ($message_part_headers =~
|
|---|
| 515 | /^content\-type:\s*([\w\/\-]+)\s*\;?\s*(.*?)\s*$/ims)
|
|---|
| 516 | {
|
|---|
| 517 | # The message header uses MIME flags
|
|---|
| 518 | my $message_content_type=$1;
|
|---|
| 519 | my $message_content_info=$2;
|
|---|
| 520 | if (!defined($message_content_info)) {
|
|---|
| 521 | $message_content_info="";
|
|---|
| 522 | }
|
|---|
| 523 | $message_content_type =~ tr/A-Z/a-z/;
|
|---|
| 524 | if ($message_content_type =~ /multipart/) {
|
|---|
| 525 | $rfc822_formatted_body=
|
|---|
| 526 | $self->text_from_mime_message($message_content_type,
|
|---|
| 527 | $message_content_info,
|
|---|
| 528 | $message_part_body,
|
|---|
| 529 | $outhandle);
|
|---|
| 530 | } else {
|
|---|
| 531 | $message_part_body= $self->text_from_part($part_body);
|
|---|
| 532 | $rfc822_formatted_body=text_into_html($message_part_body);
|
|---|
| 533 | }
|
|---|
| 534 | } else {
|
|---|
| 535 | # message doesn't use MIME flags
|
|---|
| 536 | $rfc822_formatted_body=text_into_html($message_part_body);
|
|---|
| 537 | }
|
|---|
| 538 | # Add the returned text to the output
|
|---|
| 539 | # don't put all the headers...
|
|---|
| 540 | $message_part_headers =~ s/^(X\-.*|received|message\-id|return\-path):.*\n//img;
|
|---|
| 541 | $text.=text_into_html($message_part_headers);
|
|---|
| 542 | $text.="<p>\n";
|
|---|
| 543 | $text.=$rfc822_formatted_body;
|
|---|
| 544 | # end of message/rfc822
|
|---|
| 545 | } elsif ($part_content_type =~ /multipart/) {
|
|---|
| 546 | # recurse again
|
|---|
| 547 |
|
|---|
| 548 | $tmptext= $self->text_from_mime_message($part_content_type,
|
|---|
| 549 | $part_content_info,
|
|---|
| 550 | $part_body,
|
|---|
| 551 | $outhandle);
|
|---|
| 552 | $text.=$tmptext;
|
|---|
| 553 | } elsif ($text eq "") {
|
|---|
| 554 | # we can't do anything with this part, but if it's the first
|
|---|
| 555 | # part then make sure it is mentioned..
|
|---|
| 556 |
|
|---|
| 557 | $text.="\n<p><hr><strong><<attachment>></>";
|
|---|
| 558 | # add part info header
|
|---|
| 559 | $text.="<br>Type: $part_content_type<br>\n";
|
|---|
| 560 | if ($filename ne "") {
|
|---|
| 561 | $text.="Filename: $filename\n";
|
|---|
| 562 | }
|
|---|
| 563 | $text.="</strong></p>\n";
|
|---|
| 564 | }
|
|---|
| 565 | } # foreach message part.
|
|---|
| 566 | } else {
|
|---|
| 567 | # we can't handle this multipart type (not mixed or alternative)
|
|---|
| 568 | # the RFC also mentions "parallel".
|
|---|
| 569 | }
|
|---|
| 570 | } # end of ($mimetype !~ multipart)
|
|---|
| 571 | else {
|
|---|
| 572 | # we don't do any processing of the content.
|
|---|
| 573 | }
|
|---|
| 574 |
|
|---|
| 575 | return $text;
|
|---|
| 576 | }
|
|---|
| 577 |
|
|---|
| 578 |
|
|---|
| 579 |
|
|---|
| 580 |
|
|---|
| 581 |
|
|---|
| 582 |
|
|---|
| 583 | # Process a MIME part. Return "" if we can't decode it.
|
|---|
| 584 | sub text_from_part {
|
|---|
| 585 | my $self = shift(@_);
|
|---|
| 586 | my $text=shift;
|
|---|
| 587 | my $part_header=$text;
|
|---|
| 588 | # check for empty part header (leading blank line)
|
|---|
| 589 | if ($text =~ /^\s*\r?\n/) {
|
|---|
| 590 | $part_header="Content-type: text/plain; charset=us-ascii";
|
|---|
| 591 | } else {
|
|---|
| 592 | $part_header =~ s/\r?\n\r?\n(.*)$//s;
|
|---|
| 593 | $text=$1; if (!defined($text)) {$text="";}
|
|---|
| 594 | }
|
|---|
| 595 | $part_header =~ s/\r?\n[\t ]+/ /gs; #unfold
|
|---|
| 596 | $part_header =~ /content\-type:\s*([\w\/]+).*?charset=\"?([^\;\"\s]+)\"?/is;
|
|---|
| 597 | my $type=$1;
|
|---|
| 598 | my $charset=$2;
|
|---|
| 599 | if (!defined($type)) {$type="";}
|
|---|
| 600 | if (!defined($charset)) {$charset="ascii";}
|
|---|
| 601 | my $encoding="";
|
|---|
| 602 | if ($part_header =~ /^content\-transfer\-encoding:\s*([^\s]+)/mis) {
|
|---|
| 603 | $encoding=$1; $encoding=~tr/A-Z/a-z/;
|
|---|
| 604 | }
|
|---|
| 605 | # Content-Transfer-Encoding is per-part
|
|---|
| 606 | if ($encoding ne "") {
|
|---|
| 607 | if ($encoding =~ /quoted\-printable/) {
|
|---|
| 608 | $text=qp_decode($text);
|
|---|
| 609 | } elsif ($encoding =~ /base64/) {
|
|---|
| 610 | $text=base64_decode($text);
|
|---|
| 611 | } elsif ($encoding !~ /[78]bit/) { # leave 7/8 bit as is.
|
|---|
| 612 | # rfc2045 also allows binary, which we ignore (for now).
|
|---|
| 613 | # maybe this shouldn't go to stderr, but anyway...
|
|---|
| 614 | print STDERR "EMAILPlug: unknown encoding: $encoding\n";
|
|---|
| 615 | return "";
|
|---|
| 616 | }
|
|---|
| 617 | }
|
|---|
| 618 | if ($type eq "text/html") {
|
|---|
| 619 | # only get stuff between <body> tags, or <html> tags.
|
|---|
| 620 | $text =~ s@^.*<html[^>]*>@@is;
|
|---|
| 621 | $text =~ s@</html>.*$@@is;
|
|---|
| 622 | $text =~ s/^.*?<body[^>]*>//si;
|
|---|
| 623 | $text =~ s/<\/body>.*$//si;
|
|---|
| 624 | }
|
|---|
| 625 | elsif ($type eq "text/xml") {
|
|---|
| 626 | $text=~s/</</g;$text=~s/>/>/g;
|
|---|
| 627 | $text="<pre>\n$text\n</pre>\n";
|
|---|
| 628 | }
|
|---|
| 629 | # convert to unicode
|
|---|
| 630 | $self->convert2unicode($charset, \$text);
|
|---|
| 631 | return $text;
|
|---|
| 632 | }
|
|---|
| 633 |
|
|---|
| 634 |
|
|---|
| 635 | # decode quoted-printable text
|
|---|
| 636 | sub qp_decode {
|
|---|
| 637 | my $text=shift;
|
|---|
| 638 |
|
|---|
| 639 | my @lines=split('\n', $text);
|
|---|
| 640 |
|
|---|
| 641 | # if a line ends with "=\s*", it is a soft line break, otherwise
|
|---|
| 642 | # keep in any newline characters.
|
|---|
| 643 | foreach my $line (@lines) {
|
|---|
| 644 | if ($line !~ s/=\s*$//) {$line.="\n";}
|
|---|
| 645 |
|
|---|
| 646 | if ($line =~ /=[0-9A-Fa-f]{2}/) { # it contains an escaped char
|
|---|
| 647 | my @hexcode_segments=split('=',$line);
|
|---|
| 648 | shift @hexcode_segments;
|
|---|
| 649 | my @hexcodes;
|
|---|
| 650 | foreach my $hexcode (@hexcode_segments) {
|
|---|
| 651 | $hexcode =~ s/^(..).*$/$1/; # only need first 2 chars
|
|---|
| 652 | chomp($hexcode); # just in case...
|
|---|
| 653 | my $char=chr (hex "0x$hexcode");
|
|---|
| 654 | $line =~ s/=$hexcode/$char/g;
|
|---|
| 655 | }
|
|---|
| 656 | }
|
|---|
| 657 | }
|
|---|
| 658 | $text= join('', @lines);
|
|---|
| 659 | return $text;
|
|---|
| 660 | }
|
|---|
| 661 |
|
|---|
| 662 | # decode base64 text. This is fairly slow (since it's interpreted perl rather
|
|---|
| 663 | # than compiled XS stuff like in the ::MIME modules, but this is more portable
|
|---|
| 664 | # for us at least).
|
|---|
| 665 | # see rfc2045 for description, but basically, bits 7 and 8 are set to zero;
|
|---|
| 666 | # 4 bytes of encoded text become 3 bytes of binary - remove 2 highest bits
|
|---|
| 667 | # from each encoded byte.
|
|---|
| 668 |
|
|---|
| 669 |
|
|---|
| 670 | sub base64_decode {
|
|---|
| 671 | my $enc_text = shift;
|
|---|
| 672 | # A=>0, B=>1, ..., '+'=>62, '/'=>63
|
|---|
| 673 | # also '=' is used for padding at the end, but we remove it anyway.
|
|---|
| 674 | my $mimechars="ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/";
|
|---|
| 675 | # map each MIME char into it's value, for more efficient lookup.
|
|---|
| 676 | my %index;
|
|---|
| 677 | map { $index{$_} = index ($mimechars, $_) } (split ('', $mimechars));
|
|---|
| 678 | # remove all non-base64 chars. eval to get variable in transliteration...
|
|---|
| 679 | # also remove '=' - we'll assume (!!) that there are no errors in the encoding
|
|---|
| 680 | eval "\$enc_text =~ tr|$mimechars||cd";
|
|---|
| 681 | my $decoded="";
|
|---|
| 682 | while (length ($enc_text)>3)
|
|---|
| 683 | {
|
|---|
| 684 | my $fourchars=substr($enc_text,0,4,"");
|
|---|
| 685 | my @chars=(split '',$fourchars);
|
|---|
| 686 | $decoded.=chr( $index{$chars[0]} << 2 | $index{$chars[1]} >> 4);
|
|---|
| 687 | $decoded.=chr( ($index{$chars[1]} & 15) << 4 | $index{$chars[2]} >> 2);
|
|---|
| 688 | $decoded.=chr( ($index{$chars[2]} & 3 ) << 6 | $index{$chars[3]});
|
|---|
| 689 | }
|
|---|
| 690 | # if there are any input chars left, there are either
|
|---|
| 691 | # 2 encoded bytes (-> 1 raw byte) left or 3 encoded (-> 2 raw) bytes left.
|
|---|
| 692 | my @chars=(split '',$enc_text);
|
|---|
| 693 | if (length($enc_text)) {
|
|---|
| 694 | $decoded.=chr($index{$chars[0]} << 2 | (int $index{$chars[1]} >> 4));
|
|---|
| 695 | }
|
|---|
| 696 | if (length($enc_text)==3) {
|
|---|
| 697 | $decoded.=chr( ($index{$chars[1]} & 15) << 4 | $index{$chars[2]} >> 2);
|
|---|
| 698 | }
|
|---|
| 699 | return $decoded;
|
|---|
| 700 | }
|
|---|
| 701 |
|
|---|
| 702 | sub convert2unicode {
|
|---|
| 703 | my $self = shift(@_);
|
|---|
| 704 | my ($charset, $textref) = @_;
|
|---|
| 705 |
|
|---|
| 706 | # first get our character encoding name in the right form.
|
|---|
| 707 | $charset = "iso_8859_1" unless defined $charset;
|
|---|
| 708 | $charset=~tr/A-Z/a-z/;
|
|---|
| 709 | $charset=~s/\-/_/g;
|
|---|
| 710 | $charset=~s/gb2312/gb/;
|
|---|
| 711 | # assumes EUC-KR, not ISO-2022 !?
|
|---|
| 712 | $charset=~s/ks_c_5601_1987/korean/;
|
|---|
| 713 |
|
|---|
| 714 | # It appears that we can't always trust ascii text so we'll treat it
|
|---|
| 715 | # as iso-8859-1 (letting characters above 0x80 through without
|
|---|
| 716 | # converting them to utf-8 will result in invalid XML documents
|
|---|
| 717 | # which can't be parsed at build time).
|
|---|
| 718 | $charset = "iso_8859_1" if ($charset eq "us_ascii" || $charset eq "ascii");
|
|---|
| 719 |
|
|---|
| 720 | $$textref=&unicode::unicode2utf8(&unicode::convert2unicode($charset,$textref));
|
|---|
| 721 | }
|
|---|
| 722 |
|
|---|
| 723 |
|
|---|
| 724 | # Perl packages have to return true if they are run.
|
|---|
| 725 | 1;
|
|---|
