dately 3.3.0

Deterministic date, time, timezone, holiday, and natural-language temporal processing for Python.

Overview

dately parses, detects, converts, replaces, and normalizes date and time values. It accepts individual strings as well as lists, dictionaries, NumPy arrays, and pandas Series. Its rule-based NLP engine also resolves expressions such as:

• first Monday of next month
• last 5 weekends
• Q3 of last year
• 3 days ago starting from April 10

The library includes Cython extensions for frequently used parsing operations and compatibility handling for platform-specific date-format behavior.

Installation

Install the latest published release from PyPI:

python -m pip install dately

To test a local checkout in an isolated environment:

python -m venv .venv-test
source .venv-test/Scripts/activate
python -m pip install --upgrade pip setuptools wheel
python -m pip install -e .
python -m pytest -q

Release downloads provide Windows x64 wheels for supported CPython versions and a source distribution. See Distribution builds for the artifact matrix and release procedure.

Quick start

import dately as dtly

dtly.detect_date_format("2024-03-13T14:30:00")
# '%Y-%m-%dT%H:%M:%S'

dtly.replace_datestring(
    "2024-03-13T14:30:00",
    year=2025,
    month=12,
    day=25,
)
# '2025-12-25T14:30:00'

dtly.replace_timestring(
    "2024-03-13 T14:30:00",
    hour=9,
    minute=15,
    tzinfo="Europe/London",
)
# '2024-03-13 T09:15:00 Europe/London'

dtly.parse("first Monday of next month")
# datetime.date(...) relative to the current date

Capabilities

Format detection and conversion

• Detect common and custom date/time formats.
• Convert values to a requested output format.
• Extract individual year, month, day, time, and timezone components.
• Process scalar values and nested collections.

Date and time replacement

• Replace selected date or time components without rebuilding the input manually.
• Preserve the original date/time separator where supported.
• Work with ISO-8601 and non-ISO strings.

Natural-language parsing

• Resolve relative, ordinal, anchored, and range expressions.
• Interpret quarters, weeks, weekends, weekdays, months, and years.
• Configure the first day of the week with set_week_start().
• Return exact dates or date ranges through a deterministic rule-based pipeline.

Timezones and holidays

TimeZoner and Holidate are exposed as lazy-loaded public objects:

import dately as dtly

country_codes = dtly.TimeZoner.CountryCodes
zones = dtly.TimeZoner.Zones
holidays = dtly.Holidate

Bundled timezone and holiday properties work with local data. Operations that request current remote information still depend on their respective external services being available. See the issue and fix log for implementation details.

Cross-platform formatting

Windows and Unix-like platforms differ in their support for flags that suppress leading zeros in strftime directives. dately detects and normalizes these cases so formatting behavior remains consistent across supported platforms.

Examples

Runnable examples with expected output are available in the examples directory:

• Date conversion and formatting
• Replacing date and time components
• Timezone operations
• Natural-language parsing

Documentation

• Distribution builds
• Issue and fix log

License

dately is distributed under the MIT License. See LICENSE for details.