ph-grants-org-py

#!/usr/bin/python
# -*- coding: UTF-8 -*-

# Script to process the grant data per organization.
# The data should be provided in CSV format (processed by ph-xls2csv.sh). 
# The script combines all the grants for one particular organization in
# a unique CSV file.
#
# @param string file_in
# The csv file containing the data

import csv
import sys
import os

file_in = sys.argv[1]
folder_out = 'csv-data/orgs/'
csv_header = ['grant_id','school_id','org_id','year','district','school_name','grant_type','grant_amount','org_name','grant_summary_edited','grant_impact_area']

def generate_unique_ids(fn, col):
  "Generate a set of unique ids."
  ids = set()
  with open(fn, 'rb') as ifile:
    reader = csv.reader(ifile)
    header = reader.next()
    for row in reader:
      ids.add(row[col])
    return ids 

def write_csv(method,data,org):
  file_out = str(folder_out) + org + '.csv'
  with open(file_out,method) as ofile:
    writer = csv.writer(ofile)
    writer.writerow(data)

def main():
  # Check if folder exists, otherwise create it
  if not os.path.exists(folder_out):
    os.makedirs(folder_out)

  # Create a CSV file per organization and add the header
  unique_ids = generate_unique_ids(file_in,2)
  for org in unique_ids:
    write_csv('w',csv_header,org)

  # Loop over all the grants and add them to the proper organization csv
  with open(file_in, 'rb') as ifile:
    reader = csv.reader(ifile)
    header = reader.next()
    for row in reader:
      write_csv('a',row,row[2])


if __name__ == "__main__":
    main()