"Number::RecordLocator" encodes integers into a 32 character "alphabet" designed to be short and easy to
read and pronounce. The encoding maps:
0 to O
1 to I
S to F
B to P
With a 32 bit encoding, you can map 33.5 million unique ids into a 5 character code.
This certainly isn't an exact science and I'm not yet 100% sure of the encoding. Feedback is much
appreciated.
new
Instantiate a new "Number::RecordLocator" object. Right now, we don't actually store any object-specific
data, but in the future, we might.
init
Initializes our integer to character and character to integer mapping tables.
encodeINTEGER
Takes an integer. Returns a Record Locator string.
decodeSTRING
Takes a record locator string and returns an integer. If you pass in a string containing an invalid
character, it returns undef.
canonicalizeSTRING
To compare a Record Locator string with another you can do:
print "ALWAYS TRUE\n" if $generator->decode("B0") == $generator->decode("PO");
However, this method provides an alternative:
my $rl_string = $generator->encode(725);
print "ALWAYS TRUE\n" if $generator->canonicalize("b0") eq $rl_string;
print "ALWAYS TRUE\n" if $generator->canonicalize("BO") eq $rl_string;
print "ALWAYS TRUE\n" if $generator->canonicalize("P0") eq $rl_string;
print "ALWAYS TRUE\n" if $generator->canonicalize("po") eq $rl_string;
This is primarily useful if you store the record locator rather than just the original integer and don't
want to have to decode your strings to do comparisons.
Takes a general Record Locator string and returns one with character mappings listed in "DESCRIPTION"
applied to it. This allows string comparisons to work. This returns "undef" if a non-alphanumeric
character is found in the string.