help with new program logic

[fkany@xxxxxxxxxxxxxxxxxx]

I'm trying to develop the program logic flow for a Canadian zip code
ranging program.  I've written an easy zip code ranging progam for US zip
codes.  It was easy to write because all I had to do was check to see if
the current_zipcode was equal to the previous_zipcode plus 1.

Below is an example of how the program ranged the input file.
Input file:
CITY1       13760
CITY1       13761
CITY1       13762
CITY2       13855
CITY2       13856
CITY3       12345
CITY3       12348
CITY4       23567
CITY5       65432

Output file:
CITY1       13760 - 13762
CITY2       13855 - 13856
CITY3       12345 - 00000
CITY3       12348 - 00000
CITY4       23567 - 00000
CITY5       65432 - 00000

Here's how I want to range the Canadian zip codes:
Input file:
CITY1       T0B0Z0
CITY2       T0A0A0
CITY3       T0A0C0
CITY4       T4A1A1
CITY4       T4A1A2
CITY4       T4A1A3
CITY4       T4A1B1
CITY4       T4A1B2
CITY4       T4A1B3
CITY4       T4A1D1
CITY4       T4A1D2
CITY4       T4A1D3

Output file:
CITY1       T0B0Z0 - 000000
CITY2       T0A0A0 - 000000
CITY3       T0A0C0 - 000000
CITY4       T4A1A1 - T4A1B3
CITY4       T4A1D1 - T4A1D3

Here's some ideas of what I was trying to make something out of:
> if first 5 positions from current_zipcode is equal to first 5 positions
of previous_zipcode, they match, now check to make sure the 6th position is
in numerical sequence with the previous_zipcode
      -if not in sequence, start a new range
      -if in sequence, get next record to see if it's also in sequence

> if first 5 positions from current_zipcode do not match first 5 positions
of previous_zipcode
      -if first 4 positions from current_zipcode is equal to first 4
positions of previous_zipcode, they match, now check to make sure the
      5th positions is in alphabetical sequence with the previous_zipcode
            *if not in sequence, start a new range
            *if in sequence, get next record to see if it's also in
sequence

The above might be a good way to go, but if anyone else has a better/easier
way to get this done, please let me know.

Thanks,

Frank