Re^4: RFC: I rewrote a custom sort function, but not sure if it is the best way.

[Lady_Aleena]

roboticus, I decided to give the name transform a try, and so far it looks pretty good.

#!/usr/bin/perl
use strict;
use warnings;

my @roman_numerals = qw(I II III IV V VI VII VIII IX X);
my $roman_numerals_string = join('|',@roman_numerals);

sub name_transform {
  my $name = shift;
  
  # If a suffix is so long it needs a comman, let's get it here.
  my ($base_name,$pre_suffix) = split(', ',$name);
  
  # Split the rest of the name by the spaces.
  my @name = split(' ',$base_name);
  
  # Now check the first array item to see if it is a common prefix.
  # Some are there for fun.
  my $prefix = $name[0] =~ /(?:Lady|Lord|[MD][rs]|Mrs|Miss|Pres|Gov|Se
+n|officer)(?:|\.)/ ? shift @name : '';
  
  # Now check the last item of the array to see if it matches some com
+mon suffixes.
  # More Roman numerals can be aded.
  my $suffix = $pre_suffix ? $pre_suffix : $name[-1] =~ /(?:Jr|Sr|Esq|
+$roman_numerals_string)(?:|\.)/ ? pop @name : '';
  
  # All which should be left is the bare name. Even if only the first 
+name is left,
  # it will be treated as the last name and maybe sorted accordingly.
  my $last_name   = pop   @name;
  my $first_name  = shift @name // '';
  
  # Every name left should be middle names.
  my $middle_name = @name ? join(' ',@name) : '';
  
  return [$last_name,$first_name,$middle_name,$prefix,$suffix];
}

local $\ = "\n";

my @names = ('President Barack Hussein Obama II','Mrs. Amanda King','D
+r. Feelsgood',
             'Miss America','Officer Andy','Henry VIII', "Dr. John Smi
+th",
             "Eucalyptus A. Tree, Esquire","Roboticus", "Lady Aleena, 
+Baker of cookies",
             'Aleena Zarahlinda ibn Robert al-Hajnal Chaoshi-Mnemosyni
+od I');

for my $person (@names) {
  my $names = name_transform($person);
  print join('|',@{$names});
}

returns

Obama|Barack|Hussein|President|II
King|Amanda||Mrs.|
Feelsgood|||Dr.|
America|||Miss|
Andy|Officer|||
Henry||||VIII
Smith|John||Dr.|
Tree|Eucalyptus|A.||Esquire
Roboticus||||
Aleena|||Lady|Baker of cookies
Chaoshi-Mnemosyniod|Aleena|Zarahlinda ibn Robert al-Hajnal||I

What do you think?

[choroba]

What about Chiang Kai-shek? See Chinese name for more fun.

[Lady_Aleena]

choroba, there are so many rules about names I do not think I can cover them all. I did not cover van/von either. Depending on the country one where the name is being filed, van Helsing could be filed under H (Netherlands and Suriname) or V (Belgium). Van is also a Vietnamese middle name. The Dutch have a lot of surname prefixes which could be added to the surname. I am sure there are other rules I have missed. There are a whole slew of nobiliary particles each with their own sort rules I imagine.

[roboticus]

Lady_Aleena:

It looks pretty good. I'm glad I was able to be helpful. Sometimes I suspect that I'm informative without being helpful. ;^)

...roboticus

When your only tool is a hammer, all problems look like your thumb.