download_rescuetime_data.py

"""
Download data from RescueTime.

Inspired by https://gist.github.com/veekas/2fe3cf30fe6375f5d121c6372a550000
Also see https://www.rescuetime.com/anapi/setup/documentation
"""

import urllib.parse
import urllib.request
import requests
from tqdm import tqdm


def download(date_start, date_end, api_key):
    """Download data in an interval (inclusive).
    Format: YYYY-MM-DD
    Saves to {date_start}-{date_end}.csv
    """

    params = urllib.parse.urlencode(
        {
            "key": api_key,
            "perspective": "interval",
            "resolution_time": "hour",
            "format": "csv",
            "restrict_begin": date_start,
            "restrict_end": date_end,
        }
    )
    url = f"https://www.rescuetime.com/anapi/data?{params}"
    print(f"Downloading for interval {date_start} to {date_end} from\n{url}")
    save_path = f"{date_start}-{date_end}.csv"

    # Copied from
    # https://stackoverflow.com/questions/37573483/progress-bar-while-download-file-over-http-with-requests
    # Streaming, so we can iterate over the response.
    r = requests.get(url, stream=True)

    # Total size in bytes.
    total_size = int(r.headers.get("content-length", 0))
    block_size = 1024  # 1 Kibibyte
    progress = tqdm(total=total_size, unit="iB", unit_scale=True)
    with open(save_path, "wb") as f:
        for data in r.iter_content(block_size):
            progress.update(len(data))
            f.write(data)
    progress.close()

    if total_size != 0 and progress.n != total_size:
        print("ERROR, something went wrong")
    else:
        print(f"Data saved to {save_path}")


def get_year_month_start(year_end, month_end):
    if month_end - 3 <= 0:
        year_start = year_end - 1
    else:
        year_start = year_end
    month_start = (month_end - 3) % 12
    if month_start == 0:
        month_start = 12
    return year_start, month_start


def main():
    # get from https://www.rescuetime.com/anapi/manage
    api_key = "FILL ME"

    year_end = 2020
    month_end = 5
    day = 11
    for _ in range(20):
        year_start, month_start = get_year_month_start(year_end, month_end)
        download(
            f"{year_start}-{month_start:02}-{day + 1:02}",
            f"{year_end}-{month_end:02}-{day:02}",
            api_key,
        )
        year_end, month_end = year_start, month_start


if __name__ == "__main__":
    main()