wishful thinking #csv

csv_stream_example.pl

Example 1

	use CSV::Stream;
	use CSV::Stream::Column;

	my $planets_stream = CSV::Stream->new(
	    columns => [
	        CSV::Stream::Column->new( name => 'Name' ),
	        CSV::Stream::Column->new( name => 'Distance' ),
	        CSV::Stream::Column->new( 
	            name           => 'Temperature' 
	            transformation => sub {
	            	my $self = shift;

	            	print $self->value * 10;
	            }
	        ),
	    ]
	    filename  => $filename,
	    separator => q{;},
	    # ... usual Text::CSV_XS configuration ...
	);

	my $planets_stream_out = CSV::Stream->new(
	    filename  => $output_filename
	);

	# Loops over every columns, applying transformations if any
	# Columns names and ordering are the same
	#
	CSV::Stream->stream( $planets_stream => $planets_stream_out );

-----------------------------------------------------------------------

Example 2 (combining columns into new one, keeping the others)

	use CSV::Stream;
	use CSV::Stream::Column;

	my $planets_stream = CSV::Stream->new(
	    columns => [
	        CSV::Stream::Column->new( name => 'Name' ),
	        CSV::Stream::Column->new( name => 'Distance' ),
	        CSV::Stream::Column->new( name => 'Temperature' ),
	    ]
	    filename  => $filename,
	    separator => q{;},
	    # ...
	);

	my $planets_stream_out = CSV::Stream->new(
	    filename  => $output_filename
	    columns_add => [
	        # 1. Come fare per specificare che voglio le stesse colonne?
	        #    Ad esempio usando columns_add anziche` colums. 
	        #    Il che semplificherebbe anche la semantica di value
	        # 2. Ha senso passare l'intera riga di partenza?

	        CSV::Stream::Column->new( 
	            name  => 'Orbit' 
	            value => sub {
	            	my $self         = shift;
	            	my $original_row = shift;

	            	return $original_row->distance * 3.14;
	            }
	        )
	    ]
	);

	# Loops over every columns, applying transformations if any
	CSV::Stream->stream( $planets_stream => $planets_stream_out );

-----------------------------------------------------------------------

Example 3 (combining columns into new one, throwing away the others)

	use CSV::Stream;
	use CSV::Stream::Column;

	my $planets_stream = CSV::Stream->new(
	    columns => [
	        CSV::Stream::Column->new( name => 'Name' ),
	        CSV::Stream::Column->new( name => 'Distance' ),
	        CSV::Stream::Column->new( name => 'Temperature' ),
	    ]
	    filename  => $filename,
	    separator => q{;},
	    # ...
	);

	my $planets_stream_out = CSV::Stream->new(
	    filename  => $output_filename
	    columns => [ ], # No original columns

	    columns_add => [
	        # 1. Come fare per specificare che voglio le stesse colonne?
	        #    Ad esempio usando columns_add anziche` colums. 
	        #    Il che semplificherebbe anche la semantica di value
	        # 2. Ha senso passare l'intera riga di partenza?

	        CSV::Stream::Column->new( 
	            name           => 'Temperature' 
	            transformation => sub {
	            	my $self = shift;

	            	print $self->value * 10;
	            }
	        ),

	        CSV::Stream::Column->new( 
	            name  => 'Orbit' 
	            value => sub {
	            	my $self         = shift;
	            	my $original_row = shift;

	            	return $original_row->distance * 3.14;
	            }
	        )
	    ]
	);

	# Loops over every columns, applying transformations if any
	CSV::Stream->stream( $planets_stream => $planets_stream_out );

-----------------------------------------------------------------------

Example 4 (filter the original)
TODO

-----------------------------------------------------------------------

Example 5 (filter and aggregation, # of lines in output differs from
the original)

	use CSV::Stream;
	use CSV::Stream::Column;

	my $planets_stream = CSV::Stream->new(
	    columns => [
	        CSV::Stream::Column->new( name => 'Name' ),
	        CSV::Stream::Column->new( name => 'Distance' ),
	        CSV::Stream::Column->new( name => 'Temperature' ),
	    ]
	    filename  => $filename,
	    separator => q{;},
	    # ...
	);

	my $planets_stream_out = CSV::Stream->new(
	    filename  => $output_filename
	    columns => [ 

	        # Ho bisogno di un modo per mantenere un valore
	        # globale alla computazione. In questo caso sono
	        # Praticamente due accumulatori (anzi, un accumulatore
	        # e una serie di etichette)

	        # Direi che questa colonna potrebbe essere 
	        # un iteratore
	        CSV::Stream::Column->new( 
	            name => 'Proximity',
	            value => Class::Stream::Value::Iterator->new(
	                values => [ qw/ 
	                    Near 
	                    Medium 
	                    Far 
	                    Outer 
	                / ]
	            )
	        ),
	        CSV::Stream::Column->new( 
	            name => 'Count',
	            value_accumulator => sub {
	            	my $self = shift; 
	            	my $context = shift; # questo dovrebbe essere un oggetto
	            	                     # che rappresenta la riga che sta
	            	                     # leggendo

	            	my $distance = $context->current_line->distance;

	            	given( $distance ) {
	            		when ( $_ < 100               ) { $_stash->{ Near }++   }
	            		when ( $_ >= 100 and $_ < 300 ) { $_stash->{ Medium }++ }
	            		when ( $_ >= 300 and $_ < 900 ) { $_stash->{ Far }++    }
	            		when ( $_ >= 900              ) { $_stash->{ Outer }++  }
	                }
	            }

	            value => sub {
	            	my $self    = shift;
	            	my $context = shift; # questo invece la riga che 
	            	                     # sta costruendo

	            	my $proximity_label_value = 
	            	    $context->proximity->value;

	            	return $_stash->{ ucfirst $proximity_label_value };
	            }
	        ),
	    ]
	);