main.py

import sample_ReBac as sr
import pattern_tracker as pt
import dict_to_csv as d2c
import networkx_to_csv as n2c
import rule_finder as rf
import miscellaneous_sorted.hueristic_miner as hm
import hueristic_miner2 as hm2
import lla_csv

print('starting')
# Iterations is for hueristic_miner only
iterations = 1000

max_rule_length = 5
num_nodes = 5000
num_edges = 15000
relationships = ['a', 'b', 'c', 'd', 'e']
rules = [
    # ['a', 'b', 'c', 'd', 'e'],
    # ['a', 'a', 'a', 'a', 'a'],
    # ['e', 'd', 'c', 'b', 'a']
]
# Generating our graph
graph = sr.generate_graph(num_nodes, num_edges, relationships)
lla = sr.create_lla(graph)
lla = sr.grant_access(rules, lla, graph)

# Tracking the present and missing relationship patterns
tracker = pt.Tracker(relations= relationships.copy(), max_depth=max_rule_length) # YOU MUST MANUALLY INPUT A THE RELATIONSHIPS, YOU CANNOT USE A LIST OF RELATIONSHIPS!!!! 
tracker.detect_pattern(graph)
missing = tracker.show_missing()


# If a pattern is missing from the graph, we add nodes to represent them. 
# We could potentially add nodes to represent all patterns.
if missing:
    sr.update_graph(graph, missing, rules)
    lla = sr.create_lla(graph)
    lla = sr.grant_access(rules, lla, graph)

# Implement algorithims to find policy under this line
print('\n', '1st run')
new_lla = hm2.hueristic_miner(lla, max_rule_length, relationships.copy(), graph, iterations) 
lla_csv.save_lla(new_lla)
# print(rf.find_policy(graph, max_rule_length, lla, relationships.copy()))
print(lla_csv.load_lla())