From Python

The Python binding converts the five types to and from the standard datetime module, both as results and as params=, to the microsecond (the resolution of both sides):

Graphersal	Python
`datetime`	aware `datetime.datetime` with `tzinfo=datetime.timezone(timedelta(seconds=offset))`
`localdatetime`	naive `datetime.datetime`
`date`	`datetime.date`
`time`	naive `datetime.time`
`duration` without months	`datetime.timedelta`
`duration` with months	`graphersal.Interval(months, days, micros)`

import datetime as dt
import graphersal

graph = graphersal.Graph()
at = dt.datetime(2023, 8, 2, 9, 30, tzinfo=dt.timezone(dt.timedelta(hours=2)))
graph.query("g.inject(x).date_add(DT.day, 1).next()", params={"x": at})
# datetime.datetime(2023, 8, 3, 9, 30, tzinfo=datetime.timezone(datetime.timedelta(seconds=7200)))
graph.query('duration("P1Y2M")')
# Interval('P1Y2M')

An aware datetime with any tzinfo (a zoneinfo.ZoneInfo too) is passed with the UTC offset it has at that instant; the zone name is not kept. An offset with a fraction of a second is refused.
An aware datetime.time is refused: a time of day has no instant without a date.
A timedelta becomes whole days and the rest, both with the sign of the whole (timedelta(hours=-1) is -PT1H).
graphersal.Interval is immutable, equal when its three fields are equal (Interval("P1M") != Interval(0, 30), as in the engine), has months, days, micros and to_iso(), and is built from the fields or from ISO 8601 text (Interval("P1Y2M3DT4H")).
Python's datetime holds the years 1 to 9999; a date before year 1 cannot be returned as one (read it with as_string()).

The binding's README has the details and more examples.

The Graphersal Documentation

From Python