Changeset 16341

Timestamp:

2008-07-10 14:26:51 (4 months ago)

Author:

kjdon

Message:

save attachments in binary mode so they work on windows. Use filename_cat instead of hard coding forward slash in paths. added code for deleting tmp files. decode the filename header value - may be encoded like any other header value

Files:

• gsdl/trunk/perllib/plugins/EmailPlugin.pm (modified) (17 diffs)

gsdl/trunk/perllib/plugins/EmailPlugin.pm

@@ -62 +62 @@
-#       * RFC 1806 - Content Dispositions (ie inline/attachment)
-
-
+
-package EmailPlugin;
-
@@ -108 +108 @@
-                'args'     => $arguments };
-
-# Create a new EmailPlugin object with which to parse a file.
-# Accomplished by creating a new SplitTextFile and using bless to 
-# turn it into an EmailPlugin.
-
-sub new {
-    my ($class) = shift (@_);
@@ -123 +119 @@
-
-    $self->{'assoc_filenames'} = {}; # to save attach names so we don't clobber
+    $self->{'tmp_file_paths'} = (); # list of tmp files to delete after processing is finished
-
-    # this might not actually be true at read-time, but after processing
@@ -226 +223 @@
-        $value =~ s/^\s+//;
-        $value =~ s/\s+$//;
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
+
+
-        
-        # Store the metadata
@@ -407 +364 @@
-    } elsif ($mimetype ne "text/plain") {
-        $self->{'doc_obj'} = $doc_obj; # in case we need to associate files...
-
+default_header_encoding,$
-    } else { # mimetype eq text/plain
-
@@ -457 +414 @@
-}
-
+# delete any temp files that we have created
+sub clean_up_after_doc_obj_processing {
+    my $self = shift(@_);
+    
+    foreach my $tmp_file_path (@{$self->{'tmp_file_paths'}}) {
+        if (-e $tmp_file_path) {
+            &util::rm($tmp_file_path);
+        }
+    }
+    
+}
-
-# Convert a text string into HTML.
@@ -507 +475 @@
-sub text_from_mime_message {
-    my $self = shift(@_);
-
+default_header_encoding,$
-    my $outhandle=$self->{'outhandle'};
-    # Check for multiparts - $mimeinfo will be a boundary
@@ -597 +565 @@
-                }
-
-
+
+
-                                                       $is_first_part);
-            } # foreach message part.
@@ -618 +587 @@
-            my $msg_text;
-            if ($mimetype =~ m@multipart/@) {
-
+
+
-                                                   $text);
-            } else {
@@ -655 +625 @@
-sub process_multipart_part {
-    my $self = shift;
+    my $default_header_encoding = shift;
-    my $message_part = shift;
-    my $is_first_part = shift;
@@ -691 +662 @@
-        $filename=$1;
-        $filename =~ s@\r?\s*$@@; # remove trailing space, if any
+        # decode the filename
+        $self->decode_header_value($default_header_encoding, \$filename);
+
-    }
-    
@@ -745 +719 @@
-                    $self->text_from_mime_message($message_content_type,
-                                                  $message_content_info,
+                                                  $default_header_encoding,
-                                                  $message_part_body);
-            } else {
@@ -770 +745 @@
-        my $tmptext= $self->text_from_mime_message($part_content_type,
-                                                   $part_content_info,
+                                                   $default_header_encoding,
-                                                   $part_body);
-        $return_text.=$tmptext;
@@ -792 +768 @@
-                $encoding=$1; $encoding =~ tr/A-Z/a-z/;
-            }
-$ENV{'GSDLHOME'} . "/tmp"
+&util::filename_cat($ENV{'GSDLHOME'}, "tmp")
-            my $save_filename=$filename;
-            
@@ -807 +783 @@
-                $assoc_files->{$filename}=1;
-            }
-
-
+
+
+
+
-            my $part_text = $message_part;
-            $part_text =~ s/(.*?)\r?\n\r?\n//s; # remove header
@@ -820 +798 @@
-            close SAVE;
-            my $doc_obj=$self->{'doc_obj'};
-dir/$save
+
-                                     "$save_filename",
-                                     $part_content_type # mimetype
-                                     );
-clean up tmp area...
-# Can't do this as it hasn't been copied/linked yet!!!
-
+add this file to the list of tmp files for deleting later
+push(@{$self->{'tmp_file_paths'}}, $tmp_filename);
+
-            my $outhandle=$self->{'outhandle'};
-            print $outhandle "EmailPlugin: saving attachment \"$filename\"\n"; #
@@ -1002 +980 @@
-    }
-    return 0;
+}
+
+# words with non ascii characters in header values must be encoded in the 
+# following manner =?<charset>?[BQ]?<data>?= (rfc2047)
+
+sub decode_header_value {
+    my $self = shift(@_);
+    my ($default_header_encoding, $textref) = @_;
+    
+    if (!$$textref) {
+        # nothing to do!
+        return;
+    }
+    my $value = $$textref;
+    # decode headers if stored using =?<charset>?[BQ]?<data>?= (rfc2047)
+    if ($value =~ /=\?.*\?[BbQq]\?.*\?=/) {
+        my $original_value=$value;
+        my $encoded=$value;
+        $value="";
+        # we should ignore spaces between consecutive encoded-texts
+        $encoded =~ s@\?=\s+=\?@\?==\?@g;
+        while ($encoded =~ s/(.*?)=\?([^\?]*)\?([bq])\?([^\?]+)\?=//i) {
+            my ($charset, $encoding, $data)=($2,$3,$4);
+            my ($decoded_data);
+            my $leading_chars = "$1";
+            $self->convert2unicode($default_header_encoding, \$leading_chars);
+            $value.=$leading_chars; 
+            
+            $data=~s/^\s*//; $data=~s/\s*$//; # strip whitespace from ends
+            chomp $data;
+            $encoding =~ tr/BQ/bq/;
+            if ($encoding eq "q") { # quoted printable
+                $data =~ s/_/\ /g;  # from rfc2047 (sec 4.2.2)
+                $decoded_data=qp_decode($data);
+                # qp_decode adds \n, which is default for body text
+                chomp($decoded_data);
+            } else { # base 64
+                $decoded_data=base64_decode($data);
+            }
+            $self->convert2unicode($charset, \$decoded_data);
+            $value .= $decoded_data;
+        } # end of while loop
+        
+        # get any trailing characters
+        $self->convert2unicode($default_header_encoding, \$encoded);
+        $value.=$encoded;
+        
+        if ($value =~ /^\s*$/) { # we couldn't extract anything...
+            $self->convert2unicode($default_header_encoding,
+                                   \$original_value);
+            $value=$original_value;
+        }
+        $$textref = $value;
+    } # end of if =?...?=
+    
+    # In the absense of other charset information, assume the
+    # header is the default (usually "iso_8859_1") and convert to unicode.
+    else {
+        $self->convert2unicode($default_header_encoding, $textref);
+    }   
+    
-}
-