NAME Parse::CSV::Colnames - Highly flexible CSV parser including column names (field names) manipulation NOTE This Module derives from Parse::CSV by Adam Kennedy inheriting its methods. The main extensions are methods for column names manipulation and some simple method-fixes. SYNOPSIS Column names manipulation makes only sense if the fields-parameter is auto, i.e. column names are in the first line. # Parse a colon-separated variables file from a handle as a hash # based on headers from the first line. my $objects = Parse::CSV::Colnames->new( handle => $io_handle, sep_char => ';', fields => 'auto', # select only rows where column name fieldname is "value" filter => sub { if($_->{fieldname} eq "value") {$_} else {undef} } ); # get column names my @fn=$objects->colnames # you want lower case field names @fn=map {lc} @fn; # you want field names without blanks @fn=map { s/\s+//g} @fn; # set column names $objects->colnames(@fn); while ( my $object = $objects->fetch ) { $object->do_something; } DESCRIPTION This module is only an extension of Parse::CSV For a detailed description of all methods see Parse::CSV For a detailed description of the underlying csv-parser see Text::CSV_XS Fixed METHODS These methods have not work in the parent module Parse::CSV yet, because Adam Kennedy is very busy. combine $status = $csv->combine(@columns); The "combine" method is passed through to the underlying Text::CSV_XS object. See example 3. It sets the fields and constructs the corresponding csv string from the arguments. You can read this array with the "fields" method. string $line = $csv->string; The "string" method is passed through to the underlying Text::CSV_XS object. See example 3 and example 4. It returns the parsed string or the corresponding combine-setting. print $status = $csv->print($io, $colref); The "print" method is passed through to the underlying Text::CSV_XS object. See example 1. It prints the string of the corresponding @$colref directly to an IO handle. Added METHODS fields @fields = $csv->fields; The "fields" method is passed through to the underlying Text::CSV_XS object. It returns the input to "combine"-method or the actual row as an array. colnames @colnames = $csv->colnames("fn1","fn2") # sets colnames or @colnames = $csv->colnames; # gets colnames The "colnames" method sets or gets colnames (="fields"-param). So you can rename the colnames (hash-keys in Parse::CSV::Colnames object). pushcolnames @colnames = $csv->pushcolnames("fn1","fn2") The "pushcolnames" method adds colnames at the end of $csv->colnames (="fields"-param). You can do that if the "filter"-method adds some new fields at the end of fields-array in Parse::CSV::Colnames object . Please consider that these colnames or fields are not in the underlying Text::CSV_XS object. See example 1 and example 4. pushcombine @colnames = $csv->pushcombine("fn1","fn2") The "pushcombine" method adds fields at the end of the actual row (="fields"-method) and constructs the corresponding csv string. You can read the result with the "fields"-method. The pushcombine and pushcolnames belong together. See example 4. EXAMPLES You can test these examples with copy and paste Example 1 Using "csv->print", "csv->pushcolnames" #!/usr/bin/perl use strict; use warnings; use Parse::CSV::Colnames; my $fh=\*DATA; my $fhout=\*STDOUT; # only for demo my $csv = Parse::CSV::Colnames->new( #file => "testnamen.csv", handle => $fh, sep_char => ';', fields => 'auto', binary => 1, # for german umlauts and utf filter => sub { $_->{country}="Germany"; $_->{product}=$_->{factor1}*$_->{factor2}; # select only rows where column name product>0 if($_->{product}>0) { $_; } else { undef } } ); # add colnames at the end $csv->pushcolnames(qw(product country)); # get column names my @fn=$csv->colnames; # you want lower case field names @fn=map {lc} @fn; # you want field names without blanks map { s/\s+//g} @fn; # set column names $csv->colnames(@fn); # headerline for direct output $csv->print($fhout,[$csv->colnames]); # print header-line print "\n"; while(my $line=$csv->fetch) { # csv direct output $csv->print($fhout,[$csv->fields,$line->{product},$line->{country}]); # only input-fields are printed with method fields print "\n"; } __DATA__ Name;Given Name;factor1;factor2 Hurtig;Hugo;5.4;4.6 Schnallnichts;Carlo;6.4;4.6 Weissnich;Carola;7.4;4.6 Leer;Hinnerk;0;4.6 Keine Ahnung;Maximilian;8.4;4.6 Example 2 Building new fields by hand with map #!/usr/bin/perl use strict; use warnings; use Parse::CSV::Colnames; my $fh=\*DATA; my $csv = Parse::CSV::Colnames->new( #file => "testnamen.csv", handle => $fh, sep_char => ';', fields => 'auto', binary => 1, # for german umlauts filter => sub { $_->{country}="Germany"; $_->{product}=$_->{factor1}*$_->{factor2}; # select only rows where column name product>0 if($_->{product}>0) { $_; } else { undef } } ); #add new fieldname at the end $csv->pushcolnames(qw(product)); # get column names my @fn=$csv->colnames; # you want lower case field names @fn=map {lc} @fn; # you want field names without blanks map { s/\s+//g} @fn; # set column names $csv->colnames(@fn); # headerline with only 2 fields my @outcolnames1=(qw(givenname product)); print join(";",@outcolnames1) . "\n"; while(my $line=$csv->fetch) { print join(";",map {$line->{$_}} @outcolnames1) . "\n"; } __DATA__ Name;Given Name;factor1;factor2 Hurtig;Hugo;5.4;4.6 Schnallnichts;Carlo;6.4;4.6 Weissnich;Carola;7.4;4.6 Leer;Hinnerk;0;4.6 Keine Ahnung;Maximilian;8.4;4.6 Example 3 Using "csv->combine" and "csv->string" #!/usr/bin/perl use strict; use warnings; use Parse::CSV::Colnames; my $fh=\*DATA; my $csv = Parse::CSV::Colnames->new( #file => "testnamen.csv", handle => $fh, sep_char => ';', fields => 'auto', binary => 1, # for german umlauts filter => sub { $_->{country}="Germany"; $_->{product}=$_->{factor1}*$_->{factor2}; # select only rows where column name product>0 if($_->{product}>0) { $_; } else { undef } } ); $csv->pushcolnames(qw(product country)); # get column names my @fn=$csv->colnames; # you want lower case field names @fn=map {lc} @fn; # you want field names without blanks map { s/\s+//g} @fn; # set column names $csv->colnames(@fn); # headerline my @outcolnames2=(qw(givenname product country)); $csv->combine(@outcolnames2); print $csv->string . "\n"; while(my $line=$csv->fetch) { # csv output $csv->combine(map {$line->{$_}} @outcolnames2); print $csv->string . "\n"; } __DATA__ Name;Given Name;factor1;factor2 Hurtig;Hugo;5.4;4.6 Schnallnichts;Carlo;6.4;4.6 Weissnich;Carola;7.4;4.6 Leer;Hinnerk;0;4.6 Keine Ahnung;Maximilian;8.4;4.6 Example 4 Using "csv->pushcombine" , "csv->pushcolnames" and "csv->string" #!/usr/bin/perl use strict; use warnings; use Parse::CSV::Colnames; my $fh=\*DATA; my $csv = Parse::CSV::Colnames->new( #file => "testnamen.csv", handle => $fh, sep_char => ';', fields => 'auto', binary => 1, # for german umlauts filter => sub { $_->{country}="Germany"; $_->{product}=$_->{factor1}*$_->{factor2}; # select only rows where column name product>0 if($_->{product}>0) { $_; } else { undef } } ); $csv->pushcolnames(qw(product country)); # get column names my @fn=$csv->colnames; # you want lower case field names @fn=map {lc} @fn; # you want field names without blanks map { s/\s+//g} @fn; # set column names $csv->colnames(@fn); # headerline $csv->combine($csv->colnames); print $csv->string . "\n"; while(my $line=$csv->fetch) { # csv output $csv->pushcombine(map {$line->{$_}} qw(product country)); # is like $csv->pushcombine($line->{product},$line->{country}); print $csv->string . "\n"; } __DATA__ Name;Given Name;factor1;factor2 Hurtig;Hugo;5.4;4.6 Schnallnichts;Carlo;6.4;4.6 Weissnich;Carola;7.4;4.6 Leer;Hinnerk;0;4.6 Keine Ahnung;Maximilian;8.4;4.6 TODO Creating Methods "popcolnames" and "popcombine". These methods delete the last fieldnames (column names) or fields. (I will add these methods if anybody wants this) Creating Methods "(un)shiftcolnames" and "(un)shiftcombine". These methods add/delete the first fieldnames (column names) or fields. (I will add these methods if anybody wants this) Integrating methods "getline_hr" and "column_names" of the underlying object Text::CSV_XS. SUPPORT Bugs should always be reported via the CPAN bug tracker at AUTHORS Uwe Sarnowski Author of the parent modul Parse::CSV : Adam Kennedy SEE ALSO Parse::CSV, Text::CSV_XS COPYRIGHT Copyright 2011 Uwe Sarnowski 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.