etl.py

import os
import glob
import psycopg2
import pandas as pd
from sql_queries import *
import json


def process_song_file(cur, filepath):
    """
    Reads the data in song data json files and inserts them into songs 
    table and artists table.
    
    Takes in a cursor object cur and a string of the location path of the 
    files that will be processed.
    """


    df = pd.read_json(filepath, lines = True)

    # insert song record
    song_data = list(df[['song_id', 'title', 'artist_id', 'year', 
                         'duration']].values[0])
    cur.execute(song_table_insert, song_data)
    
    # insert artist record
    artist_data = (df[['artist_id', 'artist_name', 'artist_location',
                       'artist_latitude', 'artist_longitude']]
                       .values[0].tolist())

    cur.execute(artist_table_insert, artist_data)


def process_log_file(cur, filepath):
    """
    Reads the data in log data json files, processes the data and 
    inserts it into the time table, users table and songplays table.
    
    Takes in a cursor object as cur and a string as the location path
    of the files that will be processed
    """

    # open log file
    data = [json.loads(line) for line in open (filepath, 'r')]

    df = pd.DataFrame(data)

    # filter by NextSong action
    df = df.loc[df['page'] == 'NextSong']

    # convert timestamp column to datetime
    t = pd.to_datetime(df['ts'], unit = 'ms')
    
    # insert time data records
    time_data = [df['ts'], t.dt.hour, t.dt.day, t.dt.isocalendar().week, 
                 t.dt.month, t.dt.year, t.dt.dayofweek ]
    column_labels = ('timestamp', 'hour', 'day', 'week', 'month', 'year', 
                     'weekday')
    time_df = pd.DataFrame(dict(zip(column_labels, time_data)))

    for i, row in time_df.iterrows():
        cur.execute(time_table_insert, list(row))

    # load user table
    user_df = df[['userId', 'firstName', 'lastName', 'gender', 'level']]

    # insert user records
    for i, row in user_df.iterrows():
        cur.execute(user_table_insert, row)

    # insert songplay records
    for index, row in df.iterrows():
        
        # get songid and artistid from song and artist tables
        cur.execute(song_select, (row.song, row.artist, row.length))
        results = cur.fetchone()
        
        if results:
            songid, artistid = results
        else:
            songid, artistid = None, None

        # insert songplay record
        songplay_data = (row.ts, row.userId, row.level, songid, artistid, 
                         row.sessionId, row.location, row.userAgent)

        cur.execute(songplay_table_insert, songplay_data)


def process_data(cur, conn, filepath, func):
    """
    Gets a all files from a certain path, passes the file to a function 
    and returns progress as a string.
    
    Takes in a cursor object cur, a connection with a database as conn, 
    a string as the location path of the files as a keyword argument filepath
    and the function that is used on the files as keyword argument func.
    """


    # get all files matching extension from directory
    all_files = []
    for root, dirs, files in os.walk(filepath):
        files = glob.glob(os.path.join(root,'*.json'))
        for f in files :
            all_files.append(os.path.abspath(f))

    # get total number of files found
    num_files = len(all_files)
    print(f'{num_files} files found in {filepath}')

    # iterate over files and process
    for i, datafile in enumerate(all_files, 1):
        func(cur, datafile)
        conn.commit()
        print(f'{i}/{num_files} files processed.'


def main():
    """
    Creates a database connection conn with a cursor object cur and calls 
    process_data on filepath of song data and process_data on 
    filepath of log data.
    """
    conn = psycopg2.connect("""host=127.0.0.1 dbname=sparkifydb user=student
                            password=student""")
    cur = conn.cursor()

    process_data(cur, conn, filepath='data/song_data', func=process_song_file)
    process_data(cur, conn, filepath='data/log_data', func=process_log_file)

    conn.close()


if __name__ == "__main__":
    main()