name	python-datetime
description	Work with dates, times, and timezones to this project's standards. Use when creating, parsing, serialising, comparing, or storing datetimes, or handling timezones.

Python Datetime

Standards for dates, times, and timezones, using the standard library datetime and zoneinfo.

Always use timezone-aware datetimes; never naive ones. Attach a zone at the boundary where a value enters the system.
Use the standard library zoneinfo for timezones, not pytz (whose legacy localize() dance is a frequent source of bugs).

Serialise with ISO 8601 via isoformat() and parse with fromisoformat() — not strftime/strptime with a hand-written format, and not all-digit numeric formats.
When parsing datetimes in pandas, pass an explicit format= to pd.to_datetime rather than relying on inference.

Be careful using datetimes as dict keys or set members: two values for the same instant can carry different tzinfo objects and hash differently, so a lookup silently misses. Normalise to one zone (e.g. UTC) first.

See python-pandas for typing DataFrame timestamp columns as pd.Timestamp.

name	python-datetime
description	Work with dates, times, and timezones to this project's standards. Use when creating, parsing, serialising, comparing, or storing datetimes, or handling timezones.

Standards for dates, times, and timezones, using the standard library datetime and zoneinfo.

Always use timezone-aware datetimes; never naive ones. Attach a zone at the boundary where a value enters the system.
Use the standard library zoneinfo for timezones, not pytz (whose legacy localize() dance is a frequent source of bugs).

Serialise with ISO 8601 via isoformat() and parse with fromisoformat() — not strftime/strptime with a hand-written format, and not all-digit numeric formats.
When parsing datetimes in pandas, pass an explicit format= to pd.to_datetime rather than relying on inference.

Be careful using datetimes as dict keys or set members: two values for the same instant can carry different tzinfo objects and hash differently, so a lookup silently misses. Normalise to one zone (e.g. UTC) first.

See python-pandas for typing DataFrame timestamp columns as pd.Timestamp.

python-datetime