'''

VERSION 0.8

Created on October 2012

@author: pbradley

CHANGE LOG

-- 12/7/2012 0.2 (pbradley) : Updated kvals list to reflect NDI fields vs CDC fields. Updated MATCH_CRITERIA_x dictionaries.

-- 12/17/2012 0.3 (pbradley) : Updated OutputMatchData() method to remove hardcoded input field header dependency. Changed PatientID to ID.

Added kvals as an argument.

-- 12/21/2012 0.4 (pbradley) : Changed output file to include M or P for Match or Possible match indicator. Added possible match criteria.

Changed match value output to be just the value of the matched fields, not the entire record.

-- 01/03/2013 0.5 (pbradley) : Added support for New field within input file that will be used to prevent match comparisons

between "old" (i.e. existing records).

-- 01/08/2013 0.6 (pbradley) : removed trailing space in sqlfile.write() from ', ' to ','

-- 01/24/2013 0.7 (pbradley) : added scoreRec() method to system to score the completeness of a record and return the score in the output

file. The last two fiels in the output file are the scores and the correspond to the IDs in the first two fields

-- 01/25/2013 0.8 (pbradley) : Updated MATCH_CRITERIA_1 to include LastName

Design Notes:

-- Match order dependencies; once a match is found the base record is no longer used in subsequent searches

TODO:

-- Need to consider the difference between NULL/missing fields vs differing data. For example if there are two records where

the FNAME, LNAME, and DOB match, where one has a SSN and the other's SSN field is NULL -- how should this be handled compared

to a scenario where the other's SSN field is a DIFFERENT value.

'''

import sys

sys.path = ['c:/work/SDRmatch2/febrl-0.4.2'] + ['c:/work/SDRmatch2/libsvm'] + sys.path

import time

import os

import errno

import glob

import datetime

import argparse

import logging

import csv

from os.path import join as pjoin, isdir, isfile

from stringcmp import do_stringcmp

import random

'''

Dictionaries containing keys and minimum match values

REF kvals = ['LastName', 'FirstName', 'MiddleName','Suffix','DOB','Sex','Surname','SSN']

'''

MATCH_CRITERIA_1 = {'SSN': 1,

'LastName': 0.9}

MATCH_CRITERIA_2 = {'LastName': 1,

'FirstName': 1,

'DOB': 0.9,

'Surname': 1 }

MATCH_CRITERIA_3 = {'LastName': 0.85,

'MiddleName': 0.85 }

MATCH_CRITERIA_4 = { 'LastName': 0.85,

'DOB': 1,

'Sex': 1,

'SSN': 0.85 }

MATCH_CRITERIA_5 = { 'FirstName': 0.85,

'SSN': 0.95 }

MATCH_CRITERIA_6 = { 'MiddleName': 0.85,

'SSN': 0.95 }

CRITERIA_LIST = [MATCH_CRITERIA_1, MATCH_CRITERIA_2,

MATCH_CRITERIA_3, MATCH_CRITERIA_4,

MATCH_CRITERIA_5, MATCH_CRITERIA_6]

POSMATCH_CRITERIA_1 = {'LastName': 1,

'FirstName': 0.85,

'DOB': 1,

'Surname': 0.85 }

POSMATCH_CRITERIA_2 = {'LastName': 1,

'FirstName': 1,

'DOB': 1 }

POSBL_LIST = [POSMATCH_CRITERIA_1, POSMATCH_CRITERIA_2]

REC_VAL = {'SSN':10, 'LastName':10, 'DOB':5, 'FirstName':5, 'MiddleName':1,'Suffix':1,'Sex':1,'Surname':1}

class DataFile(object):

return line[kval]

def timeStamped(fname, fmt='%Y-%m-%d-%H-%M-%S_{fname}'):

return datetime.datetime.now().strftime(fmt).format(fname=fname)

class DbFile(DataFile):

return self._keys

class InFile(DataFile):

output.close()

class Output(DataFile):

self.ofile.close()

class Directory(object):

yield name

def process(in_file, kvals, out_dir, log_dir):

inf2.close()

LOGGING_LEVELS = {'critical': logging.CRITICAL,

'debug': logging.DEBUG}

def main():

logging.error(str(e))

if __name__ == "__main__":

main()