package Text::vFile; use strict; =head1 NAME Text::vFile - Generic module which can read and write "vFile" files such as vCard (RFC 2426) and vCalendar (RFC 2445). The result of loading this data is a collection of objects which will grant you easy access to the properties. Then the module can write your objects back to a data file. =head1 SYNOPIS use Text::vFile; my $objects = Text::vCard->load( "foo.vCard", "blort.vCard", "whee.vCard" ); foreach my $card (@$objects) { spam ( $card->email('pref') ); } # OR my $reader = Text::vFile->new( source_file => "foo.vCard" ); while ( my $object = $reader->next ) { spam ( $object ); } # OR my $reader = Text::vFile->new( source_text => $vcard_data ); while ( my $vcard = <$reader> ) { spam ( $vcard ); } =head1 DETAILS The way this processor works is that it reads the vFile line by line. 1 - BEGIN:(.*) tag $1 is looked up in classMap; class is loaded; new object of this class is created ie/ $Text::vFile::classMap{'VCARD'}="Text::vCard"; $object=$classMap{'VCARD'}->new; n.b. classMap is a package variable for Text::vFile 2 - All lines are read and stored until a BEGIN tag (goto 1) or END tag (goto 3) is reached 3 - END:(.*) tag Signals that all entry data has been obtained and now the rows of data are processed 4 - Data is concatenated - thanks to Text::iCal for the strategy; the tag label and data are obtained 5 - The data handler is identified via $object->varHandler->{$label} There are some generic handlers for common data types such as simple strings, dates, etc. More elaborate data types such as N, ADR, etc. need special treatment and are declared explititly in classes as "load_XXX" such as "load_N" You should be able to override and extend the processing by taking Text::vFile::Base.pm as your example and adjusting as necessary. The resulting data structure is a bit bulky - but is such that it can express vFile data completely and reliably $VAR1 = bless( { 'EMAIL' => [ { 'attr' => { 'email' => [ 'HOME' ], 'type' => [] }, 'sequence' => 1, 'type' => { 'internet' => 1 }, 'value' => 'email\\@domain.com' } ], 'TITLE' => { 'value' => 'Job Title' }, 'X-ICQ' => [ { 'attr' => { 'type' => [ 'WORK', 'pref' ] }, 'sequence' => 11, 'type' => { 'pref' => 1, 'work' => 1 }, 'value' => '12341234' } ], '_lines' => [ 'VERSION:2.1', 'N:Person;Test,Given;;;', 'FN:Test Person', .... ] }, "Text::vCard"); =head1 METHODS =over 4 =item \@objects = load( filename [, filename ... ] ) Loads the vFiles found in filenames supplied and returns all found items an array of objects. =cut sub load { my $self=shift; $self=$self->new unless ref($self); my @objects=(); foreach my $fn (@_) { $self->source_file( $fn ); while ( my $object = $self->next ) { push @objects, $object; } } return wantarray ? @objects : \@objects; } =item \@objects = parse( string [, string ... ] ) Loads the vFiles found in the strings passed in and returns all found items as objects. =cut sub parse { my $self=shift; $self=$self->new unless ref($self); my @objects=(); foreach my $text (@_) { $self->source_text( $text ); until ( $self->eod ) { push @objects, $self->next; } } return wantarray ? @objects : \@objects; } sub _open { my $self=shift; warn "No filename supplied" && return unless $self->{'source_file'}; open ($self->{'fh'}, $self->{'source_file'}) or warn "Cannot open $self->{'source_file'}\n"; } =item $loader->source_file( name ) Sets this filename to be the source of vfile data. Only one filename, can contain many vfile entries. =cut sub source_file { my $self=shift; if (@_) { $self->{'source_file'} = shift; delete $self->{'fh'}; delete $self->{'source_text'}; } return $self->{'source_file'}; } =item $loader->source_text( $scalar ) Sets this scalar to be the source of vfile data. Can contain many vfile. =cut sub source_text { my $self=shift; if (@_) { $self->{'source_text'} = shift; delete $self->{'fh'}; delete $self->{'source_file'}; } return $self->{'source_text'}; } # Classes inject their desired mappings use vars qw(%classMap); %classMap=( VFILE => "Text::vFile::Base", # VCALENDAR => "Text::vCalendar", # VALARM => "Text::vCalendar::vAlarm", # VEVENT => "Text::vCalendar::vEvent", # VTODO => "Text::vCalendar::vTodo", ); =item $object = class->new( options ) Create a new vfile loader. You will need to set its source to either a source_file or source_text. Then use the next method to get each next object. =cut sub new { my $class = ref($_[0]) ? ref(shift) : shift; my $opts = ref($_[0]) ? $_[0] : {@_}; my $self = bless {}, $class; map { $self->$_( $opts->{$_} ) } keys %$opts; return $self; } =item \@objects = Class->next Gets next object from vfile =cut use overload '<>' => \&next, fallback => 1, ; sub next { my $self=shift; if ($self->{'source_file'}) { $self->_open unless $self->{'fh'}; } my $fh=$self->{'fh'}; if ($self->{'source_text'}) { $self->{'text'} = [ split (/[\r\n]+/, $self->{'source_text'}) ] unless $self->{'text'}; } return () unless $fh || $self->{'text'}; # my $parent=shift; # $self->{'_parent'}=$parent if ref $parent; my $line = $fh ? <$fh> : shift @{$self->{'text'}}; return if $self->eod; my $decoder; # UTF-16/32 detection if ( $line =~ /\000/ ) { eval "use Encode;"; die "Cannot decode this file - need the Encode module; $@\n" if $@; if ($line =~ /\000\000\000/) { if ($line =~ /^\000/) { $decoder=Encode::find_encoding("UTF-32BE"); } else { $decoder=Encode::find_encoding("UTF-32LE"); } } else { if ($line =~ /^\000/) { $decoder=Encode::find_encoding("UTF-16BE"); } else { $decoder=Encode::find_encoding("UTF-16LE"); } } } $line = $decoder->decode( $line ) if $decoder; # VFILE class detection # - see BEGIN until found or return at EOD contition until ( $line =~ /^BEGIN:/i ) { $line = $fh ? <$fh> : shift @{$self->{'text'}}; return if $self->eod; $line = $decoder->decode( $line ) if $decoder; } $line =~ /^BEGIN:(.*)/i; my $kind=uc $1; my $class=$classMap{ $kind }; die "In parseable begin tag $line - unknown class\n" unless $class; eval "use $class"; die "Cannot load $class\n" if $@; my $varHandler=$class->varHandler; my $thing=$class->new; my @lines=(); my $ended=0; until ( $self->eod ) { $line = $fh ? <$fh> : shift @{$self->{'text'}}; $line = $decoder->decode($line) if $decoder; $line =~ s/[\r\n]+$//; # Sub object - like EVENT, etc. if ($line =~ /^BEGIN:(.+)/) { # $thing=$1; # my $subclass= $classMap{uc $thing} || die "Don't know how to load ${thing}s\n"; # eval "use $subclass"; die $@ if $@; # push @{$self->{$thing}}, $subclass->new->load($fh, $self); # next; die "sub object loading not done\n"; } if ($line =~ /^END:(.*)/) { warn "bad end of data block - found END:$1 instead of END:" . uc $kind . "\n" unless uc $1 eq $kind; $ended=1; last; } push @lines, $line; } warn "premature end of data block - missing end tag\n" unless $ended; $thing->{'_lines'}= [ @lines ]; while ( @lines ) { my $line = shift @lines; while ( @lines && $lines[0] =~ /^\s(.*)/ ) { $line .= $1; shift @lines; } # Non-typed line data if ( $line =~ /^([\w\-]+):(.*)/ ) { my ($var, $data)=(uc $1, $2); my $h; if (UNIVERSAL::can( $thing, "load_$var")) { $h="load_$var"; } else { $h="load_singleText"; } $h = "load_$varHandler->{$var}" if exists $varHandler->{$var}; $thing->$h($var, undef, $data); next; } # ATTR OR Typed line data if ( $line =~ /^([\w\-]+);([^:]*):(.*)/ ) { my ($var, $attr_dat, $data)=(uc $1, $2, $3); my %attr=(); map { /([^=]+)=(.*)/; push @{$attr{lc $1}}, $2 } split (/(?{$var}" if exists $varHandler->{$var}; $thing->$h($var, \%attr, $data); next; } $self->error( $line ); } return $thing; } =item $loader->eod Returns true if loader is at end of data for current source. =cut sub eod { if ( $_[0]->{'fh'} ) { return eof $_[0]->{'fh'}; } return 0 if exists $_[0]->{'text'} && @{$_[0]->{'text'}}; return 1; } =item $object->error Called when a line cannot be successfully decoded =back =cut sub error { warn ref($_[0]) . " ERRORLINE: $_[1]\n"; } =head1 SUPPORT For technical support please email to jlawrenc@cpan.org ... for faster service please include "Text::vFile" and "help" in your subject line. =head1 AUTHOR Jay J. Lawrence - jlawrenc@cpan.org Infonium Inc., Canada http://www.infonium.ca/ =head1 COPYRIGHT Copyright (c) 2003 Jay J. Lawrence, Infonium Inc. All rights reserved. This program is free software; you can redistribute it and/or modify it under the same terms as Perl itself. The full text of the license can be found in the LICENSE file included with this module. =head1 ACKNOWLEDGEMENTS Leo - for a very productive exchange on how this should work plus suffering through a few growing pains. Net::iCal - whose loading code inspired me for mine =head1 SEE ALSO RFC 2425, 2426, 2445 =cut 1;