Hi Ron, Thanks for the suggestion.
At the end, reading the pod file and capturing the key from head1 is easy (below). I could use a Pod::Simple::SimpleTree , but I would still to have the output as raw pod. Probably using a parser to select pars of pod is not a good idea ? François my $fname = shift; my $mod; open $mod, "<$fname" or die "Can't open $fname $! \n"; my %data; my $data_ar; my $key; my $ispod; for my $line (<$mod>) { chomp $line; if ( $line =~ /^=cut/ ) { $ispod = 0; } if ( $line =~ /^=head1\s*(.+)$/i ) { $ispod = 1; if ($key) { #$key read from the previous =head1 $data{$key} = $data_ar; #store the array of lines as an array ref $data_ar = undef; # and prepare a new array } $key = $1; #now fetch the new value for $key } push @{$data_ar}, $line if ($ispod) ; #store the line in the array if we are in a pod section } close $mod; $data{$key} = $data_ar; #fetch the lines from the last =head1 read =for comment for $key ( keys %data ) { print "*$key*\n"; print join( "\n", @{ $data{$key} } ), "\n"; } =cut return %data; -----Original Message----- From: Ron Savage [mailto:r...@savage.net.au] Sent: 28 March 2017 23:36 To: pod-people@perl.org Subject: Re: raw_data with Pod::Simple::PullParser Hi François Can't help specifically but I do suggest you use a tree rather than a hash. It just makes much more sense to me to store the pod that way. -- Ron Savage - savage.net.au