Untitled

import networkx as nx
import pandas as pd
import sqlite3
import re
import matplotlib.pyplot as plt
import numpy as np
import ast
from functions import fill_conditional_seq, program_to_graph

pgms_csv = pd.read_csv("ground_truth.csv")
list_of_programs = pgms_csv['Program_Name'].to_list()
list_of_identifiers = pgms_csv['Identifier'].to_list()

conn    = sqlite3.connect(r"app_db_pc.db")
data    = pd.read_sql_query(con=conn, sql=f"""SELECT * FROM BRE_LOGIC_TBL WHERE Program_Name IN ({','.join([f"'{p}'" for p in list_of_programs])})""")

data['Source_Value'] = data['Source'].apply(lambda x: ast.literal_eval(x)[0] if isinstance(x,str) else x)

data['Identifier'] = data[['Program_Name', 'Target_Value', 'Source_Value', 'Children', 'Parents']].astype(str).agg('/'.join, axis=1)

data = data[data['Identifier'].isin(list_of_identifiers)]

data = data.drop_duplicates(subset=['Identifier'], keep='first')
data['Actuals'] = data[['Program_Name','Sequence']].astype(str).agg('_'.join,axis=1)

data['Program_Snippet'] = data['Program_Name'].map(pgms_csv.set_index('Program_Name')['Program_Snippet'])

data.to_csv("ground_truth.csv")