Rheel Data Management

Strictly-typed, human-readable data management for Python.

Rheel Data Management (RDM) provides a clean .rdm file format and Python API for structured, section-based data storage with enforced types.

Since version 2.0 RDM features Rheel Temporary Data (RTD), which provides clean .rtd files for storing temporary data that expires at a given time. RTD is not section-based but features all other functions that RDM does.

It is designed for developers who want more structure and type safety than JSON or TOML --- without the complexity of a database.

✨ Features

• Strict type enforcement
• Native Python type support
• Nested generics (list[int], dict[int, str])
• Union types (str | int)
• Mixed generics (list[str | int])
• Dynamic operations (add(), extend(), multiply())
• Key deletion support (delete())
• Section-based structure
• Human-readable aligned format
• Atomic file saves (corruption protection)
• Custom type registry
• Supports datetime, date, time, Path and Optional natively
• Safe variable handling with deep copy

NEW in v2:

• Temporary data system with expiration support (.rtd)
• Per-key TTL (time-to-live) expiration
• Automatic expired key cleanup on save
• Expiration extension and inspection API

📦 Installation

pip install Rheel-Data-Management

📥 Import

import rheelDM

📄 Example .rdm File

# Rheel Data Management 2.0

[user123]
name  : str           = "Steve"
score : int           = 42
tags  : list[str]     = ["admin", "tester"]
prefs : dict[int,str] = {1: "dark", 2: "light"}

📄 Example .rtd File

# Rheel Temporary Data @ Rheel Data Management 2.0

[2026-03-04T18:14:37.891803+00:00]
effect : str = "burning"

[2026-03-04T18:35:07.172342+00:00]
premium_member : bool = True

Clean. Typed. Readable.

🚀 Basic Usage

For each heading there are 2 examples (when required). The first is RDM and the second is RTD

Create and Save

import rheelDM

data = rheelDM.Obj()
user = data.section("user123") # for simplicity user always refers to this section in RDM files

user.set("name", str, "Steve")
user.set("score", int, 42)
user.set("roles", list[str], ["admin", "tester"])
user.set("prefs", dict[int, str], {1: "dark", 2: "light"})

data.save("botdata.rdm")

import rheelDM, datetime

temp = rheelDM.TempObj()

temp.set("name", str, "Steve", ttl=30) # Expires 30 seconds from now (also works with datetime.timedelta)
temp.set("score", int, 42, expires_at=datetime.datetime(2026,3,3,18,0,tzinfo=datetime.timezone.utc)) # Expires at given timestamp

temp.save("tempbotdata.rtd")

Load Data

from datetime import datetime

default_data = rheelDM.Obj()
default_data.section("user123").set("name", str, "noName")
default_data.section("user123").set("score", int, 0)

default_items = rheelDM.Obj.load("default_items.rdm")

loaded = rheelDM.Obj.load("botdata.rdm", default_data) # uses a copy of default if file not found (this can also be a boolean for catching missing files)
user = loaded.section("user123")

print(user.get("name"))  # "Steve"
print(user.get("score")) # 42
print(user.get("items", default_items)) # [1] (uses a copy of default because "items" is not defined in botdata.rdm)

default = rheel.TempObj.load("default.rtd")

loaded = rheel.TempObj.load("tempbotdata.rtd", default)

print(loaded.get("score")) # 42 or <ExpiredKey key="score" expired_at='2026-03-03T18:00:00.000000+00:00'>
print(loaded.get("name", default)) # "Steve" or default
print(loaded.get_expiration("name")) # timestamp of expiration

Modify Data

Overwrite

# same with RTD
user.set("name", str, "Alex")
print(user.get("name")) # "Alex"

# How to prevent overwrites:
user.set("name", str, "Steve", overwrite=False) # -> raise KeyError
print(user.get("name"))                         # "Alex"

Add or Multiply (int, float, datetime)

# same with RTD
from datetime import datetime

user.set("score", int, 10)   # 10
user.add("score", 5)         # 15
user.add("score", -3)        # 12
user.multiply("score", 2)    # 24
user.multiply("score", 0.25) # 6

user.set("last_login", datetime, datetime.now()) # 2026-02-21T21:39:18.398038
user.add("last_login", 3605)                     # 2026-02-21T22:39:23.398038

Extend (str, list, set, dict, Path)

# same with RTD
from pathlib import Path

user.set("name", str, "Steve")   # "Steve"
user.extend("name", " the Hero") # "Steve the Hero"

user.set("items", list[int], [1]) # [1]
user.extend("items", 2)           # [1, 2]
user.extend("items", [3, 4])      # [1, 2, 3, 4]

user.set("roles", set[str], {"a"}) # {"a"}
user.extend("roles", "b")          # {"a", "b"}

user.set("settings", dict[str,int], {"a":1}) # {"a": 1}
user.extend("settings", {"b":2})             # {"a": 1, "b": 2}

config_dir = Path("/settings")
user.set("config_file", Path, config_dir)   # Path(/settings)
user.extend("config_file", "steve3828.rdm") # Path(/settings/steve3828.rdm)

Extend Expiration:

# not featured in RDM
temp.extend_expiration("name", seconds=30)                                                                   # adds 30 seconds
temp.extend_expiration("name", new_expires_at=datetime.datetime(2026,3,3,18,0,tzinfo=datetime.timezone.utc)) # set new expiration timestamp

Convert non-RDM Files

# not featured in RTD
rheelDM.Obj.convert_file("/settings/data.json", overwrite=True)
rheelDM.Obj.load("/settings/data.rdm")

Delete Data

user.delete("score") # key, value and type of "score" will be deleted entirely from the RDM or RTD file.

🧠 Supported Types

• str
• int
• float
• bool
• None
• Optional[T]
• list[T]
• set[T]
• tuple[T]
• dict[K, V]
• Nested generics (e.g. list[dict[int, str]])
• Union types (e.g. str | int)
• Mixed generics (e.g. list[str | int])
• datetime:
  • datetime
  • date
  • time
• pathlib:
  • Path

Custom Type Registry

You can register your own types globally.

import rheelDM

class Color:
    def __init__(self, hex_code: str):
        self.hex = hex_code

#using lambda, where 'v' is the stored value
rheelDM.TypeRegistry.register(
    "Color",
    Color,
    lambda v: f'#{v.hex}',
    lambda v: Color(v[1:])
)

#using callables (functions)
def serialize_color(v): ...
def deserialize_color(v): ...

rheelDM.TypeRegistry.register(
    "Color",
    Color,
    serialize_color,  # do not call them here,
    deserialize_color # just pass the callable
)

Now you can use it like any native type:

# same with TempObj and set()
data.section("settings").set("theme", Color, Color("ff8800"))

🛡 Why RDM Instead of JSON or TOML?

Strict Typing

JSON and TOML do not enforce types.
RDM validates everything on write and load.

Python-Native Types

JSON cannot store:

• datetime
• Path
• set
• tuple
• Union types
• Nested generics
• int as dict-keys
• NoneType (stores it as null)
• Optional

TOML cannot store:

• Path
• set
• NoneType
• Optional
• int as dict-keys

RDM can store all of these, even the most complex and custom types.

Cleaner Structure

Section-based (not RTD) and auto-aligned format keeps large files organized and readable.

Human Editable

Minimal syntax makes manual edits easier than ever:

key : type = value

🎯 Ideal Use Cases

• Game save systems
• Discord bot data
• Typed configuration systems
• CLI tool configs
• Small to medium persistent data storage