from os import name
from tracemalloc import start
import pandas as pd
import numpy as np
from collections import defaultdict
import time
import os

pd.set_option('display.max_colwidth', 150)

def init_stock_df(cur_stock):
    if not os.path.exists(f"data/stocks/{cur_stock}.csv"):
        return pd.DataFrame()

    COLUMNS_DROP_STOCK = ['volume', 'high', 'low', 'adj close']

    df = pd.read_csv(f"data/stocks/{cur_stock}.csv")
    df = df.drop(columns=COLUMNS_DROP_STOCK)
    df.rename(columns={'date': 'Date'}, inplace=True)
    df['Date'] = pd.to_datetime(df.Date)
    df['Price Change %'] = ((df['close'] - df['open']) / df['open']) * 100    
    
    return df

def main():
    CHUNK_SIZE = 1000
    COLUMNS_DROP = ['Unnamed: 0', 'Url', 'Publisher', 'Author', 'Article', 
                    'Lsa_summary', 'Luhn_summary', 'Lexrank_summary']


    cur_stock = ""
    stock = pd.DataFrame()
    rows_to_add = [] 
    
    for chunk in pd.read_csv("data/news/nasdaq_exteral_data.csv", chunksize=CHUNK_SIZE, parse_dates=['Date']):
        chunk = chunk.drop(columns=COLUMNS_DROP)
        chunk['Date'] = chunk['Date'].dt.date
    
        if cur_stock == "":
            cur_stock = chunk.loc[0, 'Stock_symbol']
            stock = init_stock_df(cur_stock)
                        
        for index, row in chunk.iterrows():
            if row['Stock_symbol'] == cur_stock:
                rows_to_add.append(row)
                continue
                
            # Turn articles into its own data frame 
            # and aggregate articles published in the same day separated by "|"
            articles = pd.DataFrame(rows_to_add)
            articles = articles.dropna()
            
            if len(articles) != 0 and 'Date' in articles and len(stock) != 0:
                articles['Date'] = pd.to_datetime(articles.Date)
                
                aggregated_articles = articles.groupby('Date').agg({'Article_title': ' | '.join, 'Textrank_summary': ' | '.join}).reset_index()
                
                merged_data = pd.merge(stock, aggregated_articles, on='Date', how='left')
                stock_output_path = f"data/price_movement/{cur_stock}.csv"
                merged_data.to_csv(stock_output_path, index=False)

            rows_to_add = []
            cur_stock = row['Stock_symbol']
            stock = init_stock_df(cur_stock)




if __name__ == '__main__':
    main()