The Perl Toolchain Summit needs more sponsors. If your company depends on Perl, please support this very important event.

NAME

Geo::Postcodes::Selection - How to use the selection procedure/method of the Geo::Postcodes::* modules.

INTRODUCTION

This document uses the fictious country Utopia (with country code 'U2'), and the rest of this document will refer to the non-existent module Geo::Postcodes::U2.

SELECTION

selection procedure

 my @postcodes = Geo::Postcodes::U2::selection($field => $value);

This simple form will give a list of postcodes matching the specified field and value. Substitute 'U2' by a valid country subclass. The fields can be anyone given by the Geo::Postcodes::U2::get_fields() call, and the value either a literal text or a regular expression. The resulting list of postcodes is sorted. An empty list is also returned if one ore more of the arguments are invalid, but this can be checked by using the verify_selectionlist procedure described later on.

It is possible to specify more than one field/value pair, but then the mode should be given (but it will default to 'and' otherwise). Use as many field/value-pairs as required. The mode can be specified initially, between the field/value pairs, or not at all.

The following examples are equivalent:

 Geo::Postcodes::U2::selection('and', $field  => $value, 
                                      $field2 => $value2);
 Geo::Postcodes::U2::selection(       $field  => $value,
                               'and', $field2 => $value2);
 Geo::Postcodes::U2::selection(       $field  => $value,
                                      $field2 => $value2);

The field/value pairs are evaluated in the specified order, and the modes can be mixed freely:

 Geo::Postcodes::U2::selection($field1 => $value1,
                     'and',    $field2 => $value2,
                     'or',     $field3 => $value3,
                     'or not', $field3 => $value3,
                     'and not, $field4 => $value4,
                     'xor',    $field5 => $value5);

Modes

The modes (except 'all', 'none', 'not', and 'one') can be summarised by the following truth table:

 A  B | and and nand nand nor nor  or or  xnor xnor xor xor
      |     not      not      not     not      not      not
 -----+----------------------------------------------------
 0  0 |  0   0   1    1    1   0   0   1   1    0    0   1
 0  1 |  0   0   1    1    0   1   1   0   0    1    1   0
 1  0 |  0   1   1    0    0   0   1   1   0    1    1   0
 1  1 |  1   0   0    1    0   0   1   1   1    0    0   1

Using 'not' after the mode negates the second argument, and the 'nxxx' variants negate the result.

all

All the postcodes. This mode is only legal as the first argument, and any additional arguments are silently ignored.

 my @postcodes = Geo::Postcodes::U2::selection('all');

This will return all the postcodes, as a sorted list.

This is the same as sort get_postcodes(). The object oriented version (see below for syntax) will return a postcode object for each postcode, and may be handy in some circumstances - if time and memory usage is of no concern. Otherwise create the postcode objects only when needed, inside a foreach-loop on the procedure version - or use selection_loop.

and

The postcode is included in the result if it is included in all the expressions.

 my @postcodes = Geo::Postcodes::U2::selection(
    $field1 => $value1, 'and', $field2 => $value2);

Return postcodes matching all the field/value pairs.

The computation will work faster if the field/value pairs are given with the one with the most matches first, and the one with the least matches last. given first

and not

The postcode is included in the result if it is included in the first expression, but not the second one.

 my @postcodes = Geo::Postcodes::U2::selection(
    $field1 => $value1, 'and not', $field2 => $value2);

Return the postcodes not matching any of the field/value pairs. (This is the same as all - or, on sets of postcodes.)

nand

The postcode is included in the result if it is included in none or only one of the expressions.

 my @postcodes = Geo::Postcodes::U2::selection(
    $field1 => $value1, 'nand', $field2 => $value2);
nand not

The postcode is included in the result unless it is included in the first but not the second expression only.

 my @postcodes = Geo::Postcodes::U2::selection(
    $field1 => $value1, 'nand not', $field2 => $value2);
nor

The postcode is included in the result if it is included in none of the expressions.

 my @postcodes = Geo::Postcodes::U2::selection(
    $field1 => $value1, 'nor', $field2 => $value2);
nor not

The postcode is included in the result if it is included in the second expression only.

 my @postcodes = Geo::Postcodes::U2::selection(
    $field1 => $value1, 'nor not', $field2 => $value2);
none

This will return absolutely nothing.

This mode is only legal as the first argument, and any additional arguments are silentliy ignored.

 my @postcodes = Geo::Postcodes::U2::selection('none');

This will return undef.

not

This mode can be used initially (as the first argument) to negate the first field/value pair. It is also possible to use 'and not' or any other 'xxx not'-mode initially.

Note that 'not' is not a valid mode, and it will default to 'and' for any additional field/value pairs if no mode is given.

The following examples are equivalent:

 Geo::Postcodes::U2::selection('not',     $field  => $value,
                               'and',     $field2 => $value2);
 Geo::Postcodes::U2::selection('not',     $field  => $value, 
                                          $field2 => $value2);
 Geo::Postcodes::U2::selection('or not',  $field  => $value,
                               'and',     $field2 => $value2);
 Geo::Postcodes::U2::selection('and not', $field  => $value,
                               'and',     $field2 => $value2);
 Geo::Postcodes::U2::selection('and not', $field  => $value,
                                          $field2 => $value2);

The following examples are equivalent:

 Geo::Postcodes::U2::selection('or not', $field  => $value,
                                         $field2 => $value2);
 Geo::Postcodes::U2::selection('not',    $field  => $value,
                               'or not', $field2 => $value2);
one

This mode can be used initially to limit the returned list of postcodes to just one (or zero). The returned postcode is chosen randomly from the result list.

 Geo::Postcodes::U2::selection('one', $field => $value);

It can also be used on its own, just to get a random postcode.

 Geo::Postcodes::U2::selection('one');
or

The postcode is included in the result if it is included in at least one of the expressions.

 my @postcodes = Geo::Postcodes::U2::selection(
    $field1 => $value1, 'or', $field2 => $value2);

Return postcodes matching one or more of the field/value pairs.

The computation will work faster if the field/value pairs are given with the one with the least matches first, and the one with the most matches last. given first

or not

The postcode is included in the result unless it is included in the second expression only.

 my @postcodes = Geo::Postcodes::U2::selection(
    $field1 => $value1, 'or not', $field2 => $value2);

It is also possible to achieved this by using 'or' and a reversed regular expression.

xnor

The postcode is included in the result if it is included in none or both expressions.

 my @postcodes = Geo::Postcodes::U2::selection(
    $field1 => $value1, 'xnor', $field2 => $value2);
xnor not

The postcode is included in the result if it is included in only one of the expressions. This mode is the same as 'xor'.

 my @postcodes = Geo::Postcodes::U2::selection(
    $field1 => $value1, 'xnor not', $field2 => $value2);
xor (exlusive or)

The postcode is included in the result if it is included in only one of the expressions.

 my @postcodes = Geo::Postcodes::U2::selection(
    $field1 => $value1, 'xor', $field2 => $value2);
xor not

The postcode is included in the result if it is included in both or none of the expressions.

The Search Value

The search value is parsed as the regular expression m{^$value$}i. This has the following implications:

m{...}i

The trailing i means that the search is done case insensitive. This does not work for the special norwegian and danish characters 'Æ', 'Ø' and 'Å' (as used in the subclasses 'NO' and 'DK') unless a use locale is used in the program, and the current locale supports these characters. (This will probably not work when the code is running with mod_perl on a web server.)

'As' will match 'AS', 'As', 'aS', and 'as'.

m{^...$}

The first (^; caret) and last ($; dollar sign) character inside the expression force a matcth to the whole expression.

'AS' will match exactly the four variations mentioned above, and nothing else (so that 'CHAS' or 'ASIMOV' will not match).

Use "Wildcards" or "Regular expressions" to match several characters at once.

Wildcards

The character % (the percent character) will match zero or more arbitrary characters (and is borrowed from standard SQL).

'%12' will match '1112' but not '1201'. 'O%D' will match all strings starting with an 'O' and ending with a 'D'. '%A%' will match all strings with an 'A' somewhere.

(The character % is changed to the regular expression '.*' (dot star) by the module.)

Regular expressions

.

The character . (a single dot) will match exactly one character.

'..11' will match a four character string, where the first two can be anything, followed by '11'.

?

The character ? (a question mark) will match the previous character zero or one time.

'%ØYA?' will match strings ending with 'ØY' or 'ØYA'.

'*'

The character * (a star) will match the previous character zero or more times.

[]

The expression '[AB]' will match one of 'A' or'B'.

%'[AB]' will match all names ending with an 'A' or 'B'.

()

The expression '(..)' will remember the part inside the paranthesis. See the next item for usage.

It can also be used in combination with back references; \1, \2 and so on. (..)\1 will match postcodes starting with two caharcters, and ending with the same ones (e.g. '1919', 7272', but not '1221'). (.)(.)\2\1 will match postcodes where the first and fourth digit is the same, and the second and third digit is the same.

|

The expression 'A|BBB' will match either 'A' or 'BBB'.

Be careful with this construct, as '%ÅS|%SKOG' will not match '%ÅS' or '%SKOG', but rather everything starting with 'ÅS' or ending with 'SKOG' - caused by the '^...$' that the expression is wrapped in. Use '%(ÅS|SKOG)' to get the desired result.

External Procedures

It is also possible to call external procedures, and have them decide if the postcode should be included. Specify 'procedure', \&procedure_to_do_the_job instead of a field/value pair.

The procedure is passed the postcode, and must return true or false.

 sub procedure_to_do_the_job
 {
   my postcode = shift;
   return 1 if ...
   return 0;
 }

selection method

 my @postcodobjects = Geo::Postcodes::U2->selection(xxxx);

This works just as the procedure version (see above), but will return a list of postcode objects (instead of just a list of postcodes).

When not to use Selection

Do not use selection when you are after a single postcode, but use the constructor (the new call) or the xxx_of procedures to access the values of the fields directly.

  my @obj_list = Geo::Postcodes::U2->selection('postcode' -> '1178');
  my $obj      = Geo::Postcodes::U2->new('1178');

Using selection causes a loop that iterates over all the postcodes, and this is not a good idea when you have the answer already - a single postcode.

The valid procedure can be used to check if the postcode actually exist. See the tutorial.

SELECTION_LOOP

It is possible, even adviceable, to use external procedures in the argument list to the selection call itself, if the procedure is only used to select from the postcodes. The procedure shown in this section have a print statement, and is okay - if somewhat contrived.

selection_loop procedure

The first argument is a pointer to a procedure which will be called for each postcode returned by the selection call, with the rest of the arguments.

 sub post_check
 {
   my $postcode = shift;
   print "$postcode\n" if something($postcode) > 3.14;
 }

 Geo::Postcodes::U2::selection_loop(\&post_check, xxx, yyy);

selection_loop method

As above, but the value passed to the specified procedure is a postcode object.

 sub post_check
 {
   my $object = shift;
   print $object->postcode() . "\n" if something($object->postcode()) > 3.14;
 }

 Geo::Postcodes::U2->selection_loop(\&post_check, xxx, yyy);

SUPPORTING PROCEDURES

verify_selectionlist

The selection procedure/method will return an empty list, if there are errors in the argument list, or if the selection did not find any matches. Use this procedure (from the child class) to verify that the arguments are valid for use by the selection procedure/method.

 my($status, @list) = Geo::Postcodes::U2::verify_selectionlist(@arguments);

A status value of true (1) is followed by a modified version of the original arguments. This will replace things as 'and' 'not' by 'and not', but the selection procedure/method copes with both. (The returned list may be optimised in future versions of the module.)

A status value of false (0) is followed by a list of diagnostic messages, up to the point where the verification failed.

External procedures are recognised, and must actually exist for the test to accept them.

Returns true if the mode is one of the list returned by get_selectionmodes, documented below.

Returns true if the mode is one of the list returned by get_initial_selectionmodes, documented below.

Geo::Postcodes::get_selectionmodes

A sorted list of legal selection modes; 'and', 'and not', 'nand', 'nand not', 'nor', 'nor not', 'or', 'or not', 'xnor', 'xnor not', 'xor' and 'xor not'.

Geo::Postcodes::get_initial_selectionmodes

As Geo::Postcodes::get_selectionmode, with the addition of 'all', none', 'not' and 'one'. The list is sorted.

SEE ALSO

The tutorial perldoc Geo::Postcodes::Tutorial or man Geo::Postcodes::Tutorial, and the documentation for the individual country modules; e.g. perldoc Geo::Postcodes::NO or man Geo::Postcodes::NO.

COPYRIGHT AND LICENCE

Copyright (C) 2006 by Arne Sommer - perl@bbop.org

This library is free software; you can redistribute them and/or modify it under the same terms as Perl itself.

1 POD Error

The following errors were encountered while parsing the POD:

Around line 253:

Non-ASCII character seen before =encoding in ''Æ','. Assuming CP1252