python-dateutil-rs 0.0.9

A Rust-backed port of python-dateutil

python-dateutil-rs

A high-performance Rust-backed port of python-dateutil (v2.9.0).

Status: All core modules (easter, relativedelta, parser, rrule, tz) are rewritten in Rust via PyO3/maturin, delivering 1.3x–94x speedups over python-dateutil.

Features

• Drop-in replacement for python-dateutil — same API, same behavior
• Rust-accelerated: easter, relativedelta, parser (parse / isoparse), rrule, tz, weekday
• Full module coverage: parser, relativedelta, rrule, tz, easter, utils
• Comprehensive test suite inherited from the original project
• Benchmark infrastructure for side-by-side performance comparison (original vs Rust)
• Python 3.10–3.14 supported on Linux and macOS

Installation

pip install python-dateutil-rs

Usage

from dateutil_rs.parser import parse
from dateutil_rs.relativedelta import relativedelta
from dateutil_rs.rrule import rrule, MONTHLY
from dateutil_rs.tz import gettz, tzutc
from dateutil_rs.easter import easter

# Parse date strings
dt = parse("2024-01-15T10:30:00+09:00")

# Relative deltas
next_month = dt + relativedelta(months=+1)

# Recurrence rules
monthly = rrule(MONTHLY, count=5, dtstart=parse("2024-01-01"))

# Timezones
tokyo = gettz("Asia/Tokyo")
utc = tzutc()

# Easter
easter_date = easter(2024)

Development

Prerequisites

• Python 3.10+
• uv (recommended) or pip

Setup

git clone https://github.com/wakita181009/dateutil-rs.git
cd dateutil-rs
uv sync --extra dev

Running Tests

# Run the test suite
uv run pytest tests/ -x -q

# Run with coverage
uv run pytest tests/ --cov=dateutil

Linting

uv run ruff check tests/ python/
uv run ruff format --check tests/ python/

Benchmarks

Benchmarks compare the original python-dateutil (PyPI) and the Rust extension (dateutil_rs) using pytest-benchmark.

Easter

Benchmark	python-dateutil	dateutil-rs (Rust)	Speedup
single call (Western)	0.51 µs	0.13 µs	3.9x
single call (Orthodox)	0.35 µs	0.11 µs	3.2x
single call (Julian)	0.29 µs	0.06 µs	4.9x
1000 years (Western)	437.88 µs	70.46 µs	6.2x
500 years × 3 methods	567.99 µs	110.33 µs	5.2x

RelativeDelta

Benchmark	python-dateutil	dateutil-rs (Rust)	Speedup
create simple	0.94 µs	0.19 µs	4.9x
add months to datetime	1.62 µs	0.13 µs	12.7x
subtract from datetime	3.03 µs	0.18 µs	16.5x
multiply by scalar	1.49 µs	0.08 µs	18.7x
diff between datetimes	2.81 µs	0.33 µs	8.6x
sequential add ×12	19.41 µs	1.70 µs	11.4x

Parser — parse()

Benchmark	python-dateutil	dateutil-rs (Rust)	Speedup
simple date	8.78 µs	5.96 µs	1.5x
datetime with tz	17.99 µs	6.52 µs	2.8x
fuzzy parsing	29.42 µs	8.45 µs	3.5x
10 various formats	165.30 µs	63.13 µs	2.6x

Parser — isoparse()

Benchmark	python-dateutil	dateutil-rs (Rust)	Speedup
isoparse date	0.81 µs	0.08 µs	10.7x
isoparse datetime+tz	3.11 µs	0.61 µs	5.1x
isoparse with µs	2.67 µs	0.11 µs	23.5x

RRule

Benchmark	python-dateutil	dateutil-rs (Rust)	Speedup
daily 100	105.63 µs	63.11 µs	1.7x
weekly 52	105.33 µs	34.44 µs	3.1x
monthly 120	448.46 µs	114.12 µs	3.9x
yearly 100	2,144.58 µs	235.02 µs	9.1x
rrulestr complex	6.60 µs	0.90 µs	7.4x

Timezone

Benchmark	python-dateutil	dateutil-rs (Rust)	Speedup
gettz various (×10)	714.93 µs	7.59 µs	94.3x
gettz offset	20.20 µs	5.26 µs	3.8x
resolve_imaginary	6.54 µs	3.19 µs	2.0x
datetime_exists	3.15 µs	1.62 µs	1.9x
convert chain	6.74 µs	4.08 µs	1.7x

Measured on Apple Silicon (M-series), Python 3.13, release build. Full results: benchmarks/RESULTS.md

# Install the original python-dateutil for comparison
uv pip install python-dateutil

# Run benchmarks
make bench

# Run and save results as JSON
make bench-save

Project Structure

dateutil-rs/
├── crates/dateutil-rs/    # Rust implementation (PyO3 extension)
│   └── src/
│       ├── lib.rs         # Crate root + #[pymodule] definition
│       ├── common.rs      # Weekday (MO–SU with N-th occurrence)
│       ├── easter.rs      # Easter date calculations
│       ├── relativedelta.rs # Relative date arithmetic
│       ├── parser/        # Date/time string parsing + ISO-8601
│       ├── rrule/         # Recurrence rules (RFC 5545)
│       ├── tz/            # Timezone support (TZif, POSIX TZ, gettz cache)
│       └── utils.rs       # Utility functions
├── python/dateutil_rs/    # Python package (maturin mixed layout)
│   ├── __init__.py        # Re-exports from Rust native module
│   ├── parser.py          # Rust parse/isoparse + fallback for custom parserinfo
│   ├── relativedelta.py   # Rust RelativeDelta
│   ├── easter.py          # Rust easter
│   ├── rrule.py           # Rust rrule/rruleset/rrulestr
│   ├── tz.py              # Rust timezone classes + gettz (cached)
│   ├── common.py          # Rust weekday constants
│   └── utils.py           # Rust within_delta + python-dateutil fallback
├── tests/                 # Test suite (~13k lines)
├── benchmarks/            # pytest-benchmark comparisons
├── .github/workflows/     # CI (lint + test matrix)
├── pyproject.toml
├── Makefile
└── LICENSE

Implementation Status

Module	Rust	Notes
easter	✅	3.2x–6.2x faster
relativedelta	✅	3.5x–18.7x faster
parser (parse)	✅	1.3x–3.5x faster; custom parserinfo tables forwarded to Rust
parser (isoparse)	✅	5.1x–23.5x faster
rrule	✅	1.7x–9.1x faster
tz	✅	1.0x–94.3x faster (gettz cached)
common (Weekday)	✅
utils	✅	today(), default_tzinfo(), within_delta() all Rust-native

Roadmap

1. Python-only phase — Pure Python port with full test coverage ✅
2. Rust core + PyO3 bindings — easter, relativedelta, parser, weekday, utils ✅
3. Rust rrule — Rewrite recurrence rules in Rust ✅
4. Rust tz — Rewrite timezone support in Rust (with gettz cache) ✅
5. Release — Publish to crates.io and PyPI with pre-built wheels (manylinux, macOS, Windows)

License

MIT