bar chart race for discord messages sent

README.md

Instructions

1: Enable devtools in Discord

2: Download DiscordChatExporter and use it to download every channel as JSON

3: Setup a virtualenv if you want

4: Install dependancies with pip install simdjson pandas git+https://github.com/programiz/bar_chart_race.git@master

5: Run load.py modifying it to to use the path of your json files, it'll save a sorted and concatenated version in the current directory

6: Run main.py modifying it to use the path of that json file

load.py

import simdjson
import json
from datetime import datetime
import glob

messages = []
for message_file in glob.glob('C:\\Users\\SachK\\Documents\\csesoc\\*.json'):
	cur_messages = simdjson.load(open(message_file, encoding="utf-8"), strict=False)
	messages.extend(cur_messages['messages'])

def parse_time(date_string):
    try:
        timestamp = datetime.strptime(date_string[:len(date_string) - 6], '%Y-%m-%dT%H:%M:%S.%f')
    except ValueError:
        timestamp = datetime.strptime(date_string[:len(date_string) - 6], '%Y-%m-%dT%H:%M:%S')
    return timestamp

sorted_messages = sorted(
    messages,
    key=lambda x: parse_time(x['timestamp'])
)

print(sorted_messages[0])
print(sorted_messages[-1])

with open('cse.json', 'w') as f:
    json.dump(sorted_messages, f)

main.py

import bar_chart_race as bcr
import pandas as pd
import simdjson as json
from datetime import datetime
import glob

def parse_time(date_string):
    try:
        timestamp = datetime.strptime(date_string[:len(date_string) - 6], '%Y-%m-%dT%H:%M:%S.%f')
    except ValueError:
        timestamp = datetime.strptime(date_string[:len(date_string) - 6], '%Y-%m-%dT%H:%M:%S')
    return timestamp

message_file = "C:\\Users\\SachK\\Documents\\cse.json"
messages = json.load(open(message_file, encoding="utf-8"), strict=False)

print(len(messages))
cur_day = parse_time(messages[0]['timestamp']).date()
user_names = []
for message in messages:
	if not message['author']['name'] in user_names:
		user_names.append(message['author']['name'])

print(user_names)
user_temp = [int(0)] * len(user_names)
df = pd.DataFrame(columns=['date'].append(user_names))
for message in messages:
	if message['type'] == 'Default':
		index = user_names.index(message['author']['name'])
		user_temp[index] += 1
		if parse_time(message['timestamp']).date() != cur_day:
			s = {'date': cur_day.strftime('%d/%m/%Y')}
			for user in user_names:
				index = user_names.index(user)
				s[user] = int(user_temp[index])
			df = df.append(s, ignore_index=True)
			print(df)
			cur_day = parse_time(message['timestamp']).date()

df = df.set_index('date')
df = df.astype(int)
df.fillna(0, inplace=True)
print(df)
bcr.bar_chart_race(
        df=df, 
        filename='csesoc1.mp4', 
        orientation='h', 
        sort='desc', 
        n_bars=16, 
        period_length=200,
		steps_per_period=6, 
        end_period_pause=0,
        interpolate_period=False, 
        filter_column_colors=False,
		title='Messages sent on CSESoc by user')