CS3723 Pgm 6 Python (40 pts)

This is a continuation of program #5.  In this program, you will compare addresses to determine how close they match.  This isn't just a simple comparison.

For each customer:

Show the customer.

In program #5, you retrieved multiple addresses for a customer.  Each address should be represented by a dictionary, and you should have a list of addresses. 

For each address:

Show the address with a sequence number.  (This was done in program #5.)

Break the street line (which was concatenated from multiple LINE commands in program #5) into its parts (streetNr, streetName, streetType, direction, apartmentNumber).  This part of the assignment requires thinking.

Show the parts of a street line.

Show pairs of addresses (referencing the sequence number) and provide a score (0..100) of how closely these match.

Your business people have provided this guidance:

1. Your code should recognize equivalence of common abbreviations.  It is probably best to map them to the longer value. 

Street types (not an exhaustive list):

RD, ROAD

AVE, AVENUE

ST, STREET

Directions (not an exhaustive list):

SOUTH, S  (when appropriate)

SW, SOUTHWEST, SOUTH WEST (when appropriate)

2. There is usually at least one word of a street name before a street type, assuming the street type exists.  (e.g., 123 E ST)

3. The direction (e.g., SOUTH) might occur after the street number or after the street type.

Example addresses including name, street address, city, state, and zip   

BOB WIRE

1       123 DIRT RD

         SAN ANTONIO, TX 78210

2       123 DIRT

         SAN ANTONIO, TX 78210

3       123 DIRT LN

         SAN ANTONIO, TX 78210

PENNY LOAFER

1       111 SHOE LN 

         SAN ANTONIO, TX  78249-1234

2       111 SHOE ST 

         SAN ANTONIO, TX  78249-1234

3       111 BOOT ST SOUTH APT #5A 

         SAN ANTONIO, TX  78230

4       111 S BOOT STREET NR 5A

         SAN ANTONIO, TX  78230

FLO N WATER

1       45 S.W. VISTA RIO GRANDE RD

         SAN ANTONIO, TX 78210

2       45 S WEST VISTA RIO GRANDE RD

         SAN ANTONIO, TX 78210

3       45 SOUTHWEST VISTA RIO GRANDE ROAD

         SAN ANTONOI, TX 78210

HOLLY WOOD

1      12 WEST AVE APT 23D

LOS ANGELES, CA 90009

2   12 WEST AVENUE #23D

LOS ANGELES, CA 90009

3   12 WEST A ST NR 23

LOS ANGELES, CA 90009

BILL MELATER

1      36A E COMMERCE ST #5

       LOS ANGLES, CA 90009-2312

2      36-A COMMERCE ST EAST APT. NR. 5

LOS ANGELES, CA 90009

3      36A COMMERCE APT 5

       LOS ANGELES, CA 90009

4      36A E COMMERCE ST #8

       LOS ANGELES, CA 90009

5      36A EAST COMMEREC APT. 5

     LOS ANGELES, CA 9009

Scoring Matches:

Seq

Information

Scoring

1

Street number

+20 both not empty and they match

-20  both have values, but they don't match

0   both empty

-20  only one empty

2

Street Type

+10 both not empty and they match

-10  both have values, but they don't match

+10 both empty

+5 only one empty

3

Direction

+5 both not empty and they match

-10  both have values, but they don't match

+5 both empty

-5 only one empty

4

AptNum

+20 both not empty and they match

-20  both have values, but they don't match

+10 both empty

-10  only one empty

5   mostly match (scale) where both have values

5

City

+20 both not empty and they match

-20  both have values, but they don't match

+10 both empty

-10  only one empty

15 mostly match (scale) where both have values

6

State

+10 both not empty and they match

-20  both have values, but they don't match

0   both empty

0   only one empty

7

Street Name

+20 both not empty and they match

-5 both have values, but they don't match

-20  both empty

-20  only one empty

+10 mostly match (scale) where both have values

8

Zip Code

+80 both not empty, both len of 10, and they match (if the ZIP+4 values match, these are most likely the same address)

+5 both not empty, both len of 5, and they match

+0 same len, but they don't match.  Note that zip codes change a lot without all the addresses being updated

+5 lens not equal, the first 5 characters match

+0 lens not equal, the first 5 characters don't match.

Notes:

1.  If the computed score is less than 0, set it to 0.  If it is greater than 100, set it to 100.

2. For scoring where mostly match is provided:

Only use when the values are both not empty and they do not match. 

Use the SequenceMatcher's ratio to see if they mostly match.

If the ratio is greater than 0.6, add the ratio * the value in the table to the score.

Otherwise, score using the both have values, but they do not match value.

3. Apartment numbers come in a variety of formats.  They may be preceded with APT, NR, and/or #.  Frequently those have periods and run into the actual apartment number (which isn't necessarily a number).

4. Some punctuation isn't valuable.

5. Initially test with the data from program #5, adding the data in the examples above.You will be provided with additional data later.

Extra work:

1. (5 pts) Python can receive optional command line arguments.  Add a "-v" command line argument which when specified will show the following extra information when showing the scores.  (for the convenience of grading, it must be shown like this)

                                                                            Address Address Score  

     z=  5 Snum= 20 STy=  5 Dir=  5 Apt= 10.0 C= 20.0 State= 10 StNM= 20.0     1       2     95

     z=  5 Snum= 20 STy=-10 Dir=  5 Apt= 10.0 C= 20.0 State= 10 StNM= 20.0     1       3     80

     z=  5 Snum= 20 STy=  5 Dir=  5 Apt= 10.0 C= 20.0 State= 10 StNM= 20.0     2       3     95

     Python receives command line arguments in the sys.argv list. import sys and then sys.argv[1]will be the "-v" if it is provided

     python3 p6python.py -v <p6Input.txt >p6Out.txt 

      Please turn in output with and without the -v command line argument.

2. (max 10 pts) Your program must support the -v verbose option.  Any customers after PENNY LANE are for extra credit.