affRo/affro_cluster.py

import sys 
##import functions
from functions_cluster import *
from matching_cluster import *
from create_input_cluster import *
import json

dix_org = load_json('dictionaries/dix_acad.json')
dix_mult = load_json('dictionaries/dix_mult.json')
dix_city = load_json('dictionaries/dix_city.json')
dix_country = load_json('dictionaries/dix_country.json')
dix_org_oaire = load_json('dictionaries/dix_acad_oaire.json')
dix_mult_oaire = load_json('dictionaries/dix_mult_oaire.json')
dix_country_oaire = load_json('dictionaries/dix_country_oaire.json')
dix_status = load_json('dictionaries/dix_status.json')
dix_grids = load_json('dictionaries/dix_grids_rors.json')

    
def find_ror(input, simU, simG):
    result = Aff_Ids(input, dix_org, dix_mult, dix_city, dix_country, simU, simG)
    result_oaire = Aff_Ids(input, dix_org_oaire, dix_mult_oaire, dix_country_oaire, dix_country_oaire, simU, simG)
    
    results_upd = []

    for r in result:
       
        if  dix_status[r[2]][0] == 'active':
            results_upd.append([r[1], 'ROR', r[2], 'active'])
        else:
            if dix_status[r[2]][1] == '':
                results_upd.append([r[1], 'ROR', r[2], dix_status[r[2]][0]])
            else:

                results_upd.append([r[1], 'ROR', r[2], dix_status[r[2]][0]])

                results_upd.append([r[1], 'ROR', dix_status[r[2]][1], 'active'])
    
    for r in result_oaire:
        results_upd.append([r[1],'OpenOrgs', r[2], None])
        
    if len(results_upd)>0:
        result_dict =  [{'Provenance': 'AffRo', 'PID':'ROR', 'Value':x[2], 'Confidence':x[0], 'Status':x[3]} if x[1] == 'ROR' else {'Provenance': 'AffRo', 'PID':'OpenOrgs', 'Value':x[2], 'Confidence':x[0], 'Status': 'active'} for x in results_upd]

    else:
        result_dict =  []

        
    return result_dict


def affro(raw_aff_string):
    try:
        result = find_ror(create_df_algorithm(raw_aff_string), 0.65, 0.82)
       
        return result
    except Exception as e:
        # Return some indication of an error, or log the row
        print(f"Error: {str(e)}")
        print(raw_aff_string)
        pass


if __name__ == "__main__":
    if len(sys.argv) != 2:
        print("Usage: python affro_spark.py <string> <float1> <float2>")
        sys.exit(1)

    string_arg = sys.argv[1]
   # float_arg1 = float(sys.argv[2])
   # float_arg2 = float(sys.argv[3])

    print(affro(string_arg))