timeless 0.1.1

A datetime toolkit for people in a hurry.

Timeless: datetime for people in a hurry

A datetime toolkit for people in a hurry. It provides a simple API, heavily inspired by Pendulum.

Timeless is a work in progress.

🧠 Features

• ✔️ very simple API
• ✔️ easy to extend and use with other packages
• ✔️ built on top of standard packages and python-dateutil

📝 Why use Timeless?

To be quiet honest, Timeless is a weekend project. I love Pendulum, but since last year it doesnt seems to be actively maintained. For example, I missed a better type annotation suport. Subtely then I started to write Timeless for my own day-to-day use.

If this package is useful for you, you can say thanks staring this repo.

💻 Examples of usage

Timeless use two main concepts: Datetime and Period. A datetime is a point in Time, and a Period is a duration.

Timeless does`t differentiate between datetime and date objects.

All datetimes are assumed to be in the UTC+00:00 timezone if any other timezone is specified.

---

Datetime

import timeless

start = timeless.datetime(1900, 1, 1, zone="UTC")
end = start.add(years=1)

end.subtract(months=1)

start.set(year=2099, month=2, day=26, hour=5, zone="America/Sao_Paulo")

start.is_past()  # True
start.is_future()  # False
start.set(year=2099).is_future()  # True

Surely, you can get the current time:

timeless.today()
timeless.now()

Timeless heavily uses dateutil. The difference between two dates gives you a relativedelta object:

start.dif(end)

Periods

Create a date range with Period:

period_1 = timeless.period(start, end, freq="days")
period_2 = period_1.compute()

Periods are always yielded. Use compute or just list(period) to get the list of datetimes.

To easily change the frequency of a date range, use to:

period_1.to("hours")

Using Timeless and Pandas

Who else loves Pandas? The ones who says doens`t like Pandas probabily never used it for real. Timeless has some basic Pandas Timestamp compatibility methods.

import pandas as pd

pd_timestamp = pd.Timestamp('1900-01-01 00:00:00', tz=None)
pd_daterange = pd.date_range(pd_timestamp, periods=2, freq="MS")

timeless.from_pandas(pd_timestamp)
timeless.from_pandas(pd_daterange)

timeless.to_pandas(period_1)

Note that only the main Pandas freqs are implemented: D, W, M, A/ Y, H, T/min, S and U/ US. Freqs like MS, AS and YS are coerced to months and years, respectively.

Utilitaries

Parse some strings to datetime:

fill_date = timeless.datetime(2099, 2, 26, zone="UTC")
timeless.parse("1900", zone="America/Sao_Paulo", fill=fill_date)

Or get it as a ISO 8601 string:

timeless.now().to_iso()

Or find the next friday!

timeless.now().get_next("friday")

Timeless tries to be as flexible as possible. You can create some quite cool chain operations easily:

timeless.period(start, end.add(days=7), freq="days").to("weeks").compute()

🏗️ Development

Botree relies on Poetry.

Install the Python dependencies with:

poetry install

⚗️ Testing

poetry run pytest --cov=botree tests/