rPickle 0.5.0

A safe and efficient Python serialization library

rPickle

A safe and efficient Python serialization library.

Features

• 🔒 Safe - No arbitrary code execution, unlike pickle
• ⚡ Fast - Optimized binary format
• 📦 Compact - Small serialized size
• 🔄 Circular references - Handles self-referential structures
• 🎨 Extensible - Custom type support via extensions

🔒 Safety

rPickle is safe in the sense that it does not execute arbitrary code during deserialization.

Unlike pickle (which can call arbitrary functions during loads()), rPickle only reconstructs data structures.
It does not use __reduce__(), no implicit imports, and no function calls.

✅ rPickle.loads(data) does not execute any code — it only restores data.
❌ pickle.loads(data) may execute arbitrary code defined in __reduce__.

However, safety is not absolute:

• If you use custom extensions (Extension), you are responsible for the code you provide — the library author is not liable.
• rPickle does not encrypt data or prevent tampering. If you need integrity or confidentiality, take additional measures.

Requirements

• Python 3.15+ (preview)
• Python 3.11+ (full)

Installation

pip install rPickle

Quick Start

import rPickle

# Serialize
data = {'name': 'Alice', 'scores': [95, 87, 92]}
packed = rPickle.dumps(data)

# Deserialize
restored = rPickle.loads(packed)
print(restored)  # {'name': 'Alice', 'scores': [95, 87, 92]}

API

Core Function

Function	Description
dumps(obj)	Serialize object to bytes
loads(data)	Deserialize from bytes
dump(obj, file)	Serialize to file
load(file)	Deserialize from file

Extensions

from datetime import datetime

data = {'created': datetime.now()}
packed = rPickle.dumps(data, extensions=rPickle.ext.datetime_ext)
restored = rPickle.loads(packed, extensions=rPickle.ext.datetime_ext)

Custom Extensions

Add support for your own types using the extensions parameter.

from datetime import datetime

# 1. Define dump function (type → bytes)
def dump_datetime(dt: datetime) -> bytes:
    return dt.timestamp().to_bytes(8, 'little')

# 2. Define load function (bytes → type)
def load_datetime(data: bytes) -> datetime:
    timestamp = int.from_bytes(data, 'little')
    return datetime.fromtimestamp(timestamp)

# 3. Register your extension
my_extensions = rPickle.ext.Extension(typ=datetime, load_func=load_datetime, dump_func=dump_datetime)

# 4. Use it
data = {'created': datetime.now()}
packed = rPickle.dumps(data, extensions=my_extensions)
restored = rPickle.loads(packed, extensions=my_extensions)

Supported Types

• None, bool, int, float, complex, str
• bytes, bytearray
• list, tuple, set, frozenset
• dict, frozendict (Python 3.15+)
• range, slice
• Ellipsis, NotImplemented
• sentinel (Python 3.15+)
• Custom types

Built-in Extensions

rPickle comes with ready-to-use extensions for common types: see the list below.

Name	Type
datetime_ext	datetime.datetime
Decimal_ext	decimal.Decimal
UUID_ext	uuid.UUID
Fraction_ext	fractions.Fraction
Path_ext	pathlib.Path
defaultdict_ext	collections.defaultdict

License

MIT