package Data::JavaScript; use strict; use vars qw($VERSION @ISA @EXPORT $UNDEF); use subs qw(quotemeta); require Exporter; @ISA = qw(Exporter); @EXPORT = qw(jsdump hjsdump); $VERSION = 1.08; $UNDEF = q(''); sub import{ foreach( grep{ref($_)} @_ ){ if(ref($_) eq 'HASH'){ if(exists($_->{UNDEF})){ $UNDEF = $_->{UNDEF}; } } } Data::JavaScript->export_to_level(1, grep {!ref($_)} @_); } sub quotemeta { my $text = CORE::quotemeta(shift); $text =~ s/\\ / /g; $text =~ s/\\([^\x20-\x7E])/sprintf("\\%03o", ord($1))/ge; $text; } sub jsdump { my $sym = shift; return "var $sym;\n" unless (@_); # my $elem = $#_ ? [@_] : $_[0]; my $elem = shift; my $undef = shift; my %dict; my @res = __jsdump($sym, $elem, \%dict, $undef); $res[0] = "var " . $res[0]; wantarray ? @res : join("\n", @res, ""); } sub hjsdump { my @res = (''); wantarray ? @res : join("\n", @res, ""); } sub __jsdump { my ($sym, $elem, $dict, $undef) = @_; unless (ref($elem)) { if(! defined($elem) ){ return "$sym = @{[defined($undef) ? $undef : $UNDEF]};"; } elsif ($elem =~ /^-?(\d+\.?\d*|\.\d+)([eE]-?\d+)?$/) { return "$sym = " . eval($elem) . ";"; } return "$sym = '" . quotemeta($elem) . "';"; } if ($dict->{$elem}) { return "$sym = " . $dict->{$elem} . ";"; } $dict->{$elem} = $sym; if (UNIVERSAL::isa($elem, 'ARRAY')) { my @list = ("$sym = new Array;"); my $n = 0; foreach (@$elem) { my $newsym = "$sym\[$n]"; push(@list, __jsdump($newsym, $_, $dict, $undef)); $n++; } return @list; } if (UNIVERSAL::isa($elem, 'HASH')) { my @list = ("$sym = new Object;"); my ($k, $old_k, $v); foreach $k (keys %$elem) { $k = quotemeta($old_k=$k); my $newsym = (($k =~ /^[a-z_]\w+$/i) ? "$sym.$k" : "$sym\['$k']"); push(@list, __jsdump($newsym, $elem->{$old_k}, $dict, $undef)); } return @list; } } 1; __END__ # Below is the stub of documentation for your module. You better edit it! =head1 NAME Data::JavaScript - Perl extension for dumping structures into JavaScript code =head1 SYNOPSIS use Data::JavaScript; B use Data::JavaScript {UNDEF=>0}; @code = jsdump('my_array', $array_ref, 0); $code = jsdump('my_object', $hash_ref); $code = hjsdump('my_stuff', $array_ref B $hash_ref); =head1 DESCRIPTION This module is aimed mainly for CGI programming, when a perl script generates a page with client side JavaScript code that needs access to structures created on the server. It works by creating one line of JavaScript code per datum. Therefore, structures cannot be created anonymously and needed to be assigned to variables. This enables dumping big structures. You may define a default to be substitued in dumping of undef values at compile time by supplying the default value in anonymous hash like so use Data::JavaScript {UNDEF=>'null'}; =over =item jsdump('name', \$reference, [$undef]); The first argument is required, the name of JavaScript object to create. The second argument is required, a hashref or arrayref. Structures can be nested, circular referrencing is supported EXPERIMENTALLY. The third argument is optional, a scalar whose value is to be used en lieu of undefenied values when dumping a structure. If unspecified undef is output as C<''>. Other useful values might be C<0>, C and C When called in list context, the functions return a list of lines. In scalar context, it returns a string. =item hjsdump('name', \$reference, [$undef]); hjsdump is identical to jsdump except that it adds HTML tags to embed the script inside an HTML page. =back =head1 AUTHOR Maintained by Jerrad Pierce Ariel Brosh, schop@cpan.org. Inspired by WDDX.pm JavaScript support. =head1 CREDITS Garick Hamlin B, fixing of quoting bug. =head1 SEE ALSO perl(1), L. =cut