main.py

import os
import gzip
import re
import matplotlib.pyplot as plt
import pandas as pd
import calmap
from datetime import datetime, timedelta
from collections import defaultdict

# --- CONFIGURE THESE ---
LOG_DIRS = [
    os.path.expanduser("~/AppData/Roaming/.minecraft/logs"),
    # os.path.expanduser("~/AppData/Roaming/.minecraft/instances/SomeModpack/logs"),
    # Add as many folders as you like
]
CUTOFF_DATE = datetime.strptime("2000-01-01", "%Y-%m-%d")

START_PATTERN = re.compile(r"\[(\d{2}:\d{2}:\d{2})\]")
END_PATTERN   = re.compile(r"\[(\d{2}:\d{2}:\d{2})\]")

def extract_date_from_filename(filename):
    match = re.match(r"(\d{4}-\d{2}-\d{2})", filename)
    if match:
        return datetime.strptime(match.group(1), "%Y-%m-%d")
    return None

def read_session_times(filepath, is_gz, inferred_date):
    open_func = gzip.open if is_gz else open
    start_time = end_time = None
    with open_func(filepath, 'rt', errors='ignore') as f:
        for line in f:
            if not start_time:
                m = START_PATTERN.search(line)
                if m:
                    start_time = m.group(1)
            m = END_PATTERN.search(line)
            if m:
                end_time = m.group(1)
    if start_time and end_time and inferred_date:
        try:
            base = inferred_date.strftime('%Y-%m-%d')
            start_dt = datetime.strptime(f"{base} {start_time}", "%Y-%m-%d %H:%M:%S")
            end_dt   = datetime.strptime(f"{base} {end_time}",   "%Y-%m-%d %H:%M:%S")
            if end_dt > start_dt:
                return start_dt.date(), end_dt - start_dt
        except Exception:
            pass
    return None, None

playtime_per_day = defaultdict(timedelta)

for log_dir in LOG_DIRS:
    if not os.path.isdir(log_dir):
        print(f"Skipping missing folder: {log_dir}")
        continue
    for filename in sorted(os.listdir(log_dir)):
        if not (filename.endswith(".log") or filename.endswith(".log.gz")):
            continue
        filepath  = os.path.join(log_dir, filename)
        is_gz     = filename.endswith(".gz")
        file_date = extract_date_from_filename(filename)
        if file_date:
            if file_date < CUTOFF_DATE:
                continue
        else:
            file_date = datetime.today()
        print(f"Found a log from: {file_date}  [{log_dir}]")
        day, duration = read_session_times(filepath, is_gz, file_date)
        if day and duration:
            playtime_per_day[day] += duration

TOTAL = 0
print("\n=== Playtime per Day ===")
for day in sorted(playtime_per_day.keys()):
    total = playtime_per_day[day]
    hours, remainder = divmod(total.total_seconds(), 3600)
    minutes, seconds = divmod(remainder, 60)
    print(f"{day}: {int(hours)}h {int(minutes)}m {int(seconds)}s")
    TOTAL += total.total_seconds()

series = pd.Series({
    pd.Timestamp(date): duration.total_seconds() / 3600
    for date, duration in playtime_per_day.items()
})

calmap.calendarplot(
    series,
    how='sum',
    cmap='YlGnBu',
    fillcolor='lightgray',
    linewidth=0.5,
    daylabels='MTWTFSS',
    dayticks=[0, 1, 2, 3, 4, 5, 6],
    monthticks=1,
    yearlabels=True,
    yearlabel_kws={'color': 'black', 'fontsize': 14},
    fig_kws={'figsize': (16, 6)},
)

hours, remainder = divmod(TOTAL, 3600)
days, hours = divmod(hours, 24)
minutes, seconds = divmod(remainder, 60)
plt.suptitle(
    f"Total playtime: {int(days)}d {int(hours)}h {int(minutes)}m {int(seconds)}s",
    fontsize=18
)
plt.tight_layout()
plt.show()