NAME

       Biblio::LCC - parse and normalize LC-style call numbers

SYNOPSIS

           use Biblio::LCC;
           $normalized = Biblio::LCC->normalize('PS3573.A472242 A88 1998');
           ($begin, $limit) = Biblio::LCC->normalize_range('E184.5-E185');
           @parts = Biblio::LCC->parse($call_number);
           $call_number = Biblio::LCC->add($class, $offset);

DESCRIPTION

       Biblio::LCC parses Library of Congress classification ranges and call numbers and
       normalizes them into a form suitable for a straight ASCII sort.

PUBLIC METHODS

       normalize($call_number)
               $normalized = Biblio::LCC->normalize('PS3573.A472242 A88 1998');

           Convert an LC-style class (e.g., 'PS' or 'E184.5') or call number (e.g.,
           'PS3573.A472242 A88 1998') into a string that may be compared to other normalized call
           numbers (see normalize_range below).

       normalize_range($call_number_range)
               ($begin, $limit) = Biblio::LCC->normalize_range('E184.5-E185');

           Parse a call number range, producing a pair of strings B and L such that a call number
           falls within the range if and only if its normalized form, in a straight lexicographic
           ASCII comparison, is greater than or equal to B and strictly less than L.

           The range may be specified in one of three ways:

           call number - call number
               A pair of call numbers; the range includes the beginning call number, the ending
               call number, and any call numbers that have the ending call number as a prefix.

               For example, the (unnormalized) range "AS131-AS142" encompasses any class or call
               number from "AS131" up to but not including "AS143"

               In this form, the alphabetic string that begins the second call number may be
               omitted, so (for example) "E184.5-185" is equivalent to "E184.5-E185".

               Space is optional around the hyphen.

           call number
               A single class or call number, in unnormalized form.  This is equivalent to a pair
               in which each call number is the same.  For example, the unnormalized range
               "QA141.5.A1" encompasses call numbers from "QA141.5.A1" up to but not including
               "QA141.5.A2".

           call number < call number
               A pair of call numbers; the range includes the first call number and any call
               number up and not including the ending call number.

               For example, the unnormalized range "DT6.7<DT7" includes everything greater than
               "DT6.7" and less than "DT7".

               In this form, the alphabetic string that begins the second call number may be
               omitted, as in the form of range that uses a hyphen to separate the parts.

               Space is optional around the less-than sign (<).

       parse($call_number)
               ($alpha, $int, $frac, $rmdr) = Biblio::LCC->parse($call_number);

           Split an LC call number into alphabetic, integer, decimal fraction, and remainder
           (i.e., everything else).

       add($class, $offset)
               $call_number = Biblio::LCC->add($class, $offset);

           Add an offset (e.g., '180.3') to a base LC class (e.g., 'GN1600') to produce another
           LC class (e.g., 'GN1780.3').

           The base class may have only alphabetic and integer parts; an exception will be thrown
           if it has a fractional part (e.g., as in 'GN1600.1') or a remainder (e.g., as in
           'GN1600 R5').

HOW IT WORKS

       Call numbers are first analyzed into four parts.  For example, take the call number
       GB1001.72.M32 E73 1988.

       alpha
               GB

           The one to three alphabetic characters that begin the call number.

       integer
               1001

           An integer from 1 to 9999 that follows.

       fraction
               72

           Digits that follow a decimal point after the integer part.

       remainder
               M32 E73 1988

           Everything that follows.

       The LC Classification allows for a wide range of possible call numbers that do not fall
       into the simple (alpha, integer, fraction, remainder) model that this module implements.
       For example, the following are all valid call numbers:

       E514.5 17th .S76 1986 =item G3804.N4:3B7 K142t =item G3824.Y6C5 s50 .W5

       It may be that in some cases further analysis, and fully correct sorting, are not possible
       without hardcoded knowledge of the LC classification.  In many cases, however, a more
       sophisticated parsing model, while more complex, would result in better normalization.

BUGS

       There are no known bugs.  Please report bugs on this module's RT page:
       <http://rt.cpan.org/NoAuth/ReportBug.html?Queue=Biblio-LCC>.

TO DO

       Implement a "new" method and rewrite other methods so they may be used as class or
       instance methods.

       Special handling of "special" call numbers (e.g., in the Gs).

       Allow caller to specify prefixes to strip (e.g., "Folio").

       Parse straight from the 050 or 090 field of a MARC record.

       Better error reporting.

AUTHOR

       Paul Hoffman (nkuitse AT cpan DOT org)

COPYRIGHT

       Copyright 2007-2008 Paul M. Hoffman.

       This is free software, and is made available under the same terms as Perl itself.